Spontaneous generation of diverse recombinant prion strains: sulfated glycanco factors facilitate strain emergence but do not determine specific strain properties

RESEARCH Open Access © The Author(s) 2025. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit h t t p : / / c r e a t i v e c o m m o n s . o r g / l i c e n s e s / b y - n c - n d / 4 . 0 /. Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 https://doi.org/10.1186/s40478-025-02175-w Acta Neuropathologica Communications †Nuria L. Lorenzo and Hasier Eraña have same contribution. *Correspondence: Jesús R. Requena [email protected] Joaquín Castilla [email protected] Full list of author information is available at the end of the article Abstract Significant advances over the last two decades in the study of in vitro prion formation and propagation have revealed that distinct cofactors can facilitate or induce spontaneous prion misfolding. This, in turn, has raised important questions about the role of cofactors and their potential significance in vivo in prion diseases. Key questions include whether cofactors are necessary for prion infectivity or whether they might play a crucial role in determining strain features without being essential for infection. Grounded in previous work that showed that polyanions such as RNA or dextran sulfate facilitate spontaneous prion misfolding in vitro, we have addressed whether other chemically similar cofactors could expand the diversity of PrPSc conformers and whether these would exhibit distinctive strain features. Using the Protein Misfolding Shaking Amplification (PMSA) and three different polyanionic cofactors (heparin, chondroitin sulfate and pentosan polysulfate), we obtained and characterized a total of nine conformers and compared them to previously generated strains obtained with dextran sulfate. All nine conformers proved infectious in transgenic mice, generating distinct prion strains and suggesting that different cofactors can indeed drive the formation of distinct conformers. However, the observed variations within conformers generated with the same cofactor indicate a degree of structural flexibility, likely resulting in related but distinguishable groups of conformers. Our study demonstrates that sulfated glycans not only facilitate in vitro spontaneous PrPSc generation but also enable the emergence of multiple distinct prion strains, providing insights into the molecular mechanisms underlying strain diversity and their potential relevance to spontaneous prion diseases. Keywords Prion diseases, PrPSc, Recombinant prions, Cofactors, Spontaneous misfolding, Spontaneous prions, Prion strains, PMSA, Sulfated glycans, Strain diversity Spontaneous generation of diverse recombinant prion strains: sulfated glycan cofactors facilitate strain emergence but do not determine specific strain properties Nuria L.Lorenzo1†, HasierEraña2,3,4†, EnricVidal5,6, Jorge M.Charco2,3,4, LucíaParga1, Carlos M.Díaz-Domínguez2, SoniaVeiga1, SusanaBravo7, SamantaGiler5,6, MarivíGeijo8, Jesús R.Requena1* and JoaquínCastilla2,3,9* Page 2 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 Introduction Transmissible spongiform encephalopathies (TSE) or prion diseases are a group of rare, fatal, and rapidly progressing neurodegenerative disorders that affect humans and other mammals. According to their etiology, prion diseases can be classified as acquired, genetic or idiopathic conditions (commonly referred to, as sporadic). Acquired cases are related to exogenous exposure to contaminated food, drugs or surgical material; genetic cases are associated with mutations in the prion protein (PrP) gene, and idiopathic prion diseases have no known cause [1, 2]. Idiopathic prion diseases, and specifically, sporadic Creutzfeldt-Jakob disease (sCJD), represent the most common cause of these disorders with 85–90% of diagnosed TSE cases [1, 2], and therefore, understanding the mechanism of spontaneous generation of PrPSc is of key relevance in the prion field. Independently of their origin, all prion diseases share a common physiopathological mechanism based on the accumulation of the aberrantly folded form of the cellular prion protein (PrPC), termed PrPSc, in the central nervous system (CNS) and in lymphoreticular tissues [2–4]. While PrPC is a monomeric, protease-sensitive and globular membrane-anchored glycoprotein, PrPSc is an insoluble, protease-resistant and β sheet-rich amyloid [1, 2]. The key feature of PrPSc is that it induces its abnormal conformation to the physiological counterpart through a self-templating process [5]. Although this templating mechanism, leading to propagation of the PrPSc conformation, is not known at an atomistic level, a general understanding of it has been possible after the groundbreaking cryo-EM-based elucidation of the structure of PrPSc [6–10]. PrPSc comprises a Parallel-In-Register-Beta Stack (PIRIBS), spanning PrP residues ∼90–230 (mouse numbering) with an intrinsically disordered ∼23–90 N-terminal tail. This tail is dispensable for the templating/infectious activity of PrPSc, which resides in the PIRIBS core [11]. PrPC, in turn, also comprises two domains: an intrinsically flexible/unfolded N-terminal domain that is longer than that of PrPSc, spanning residues ∼23–120, and an α helix-rich globular domain spanning ∼121–230 [12, 13]. The ∼90–121 flexible/unfolded stretch present in PrPC but not in PrPSc is very likely key in the templated conversion process given that it can easily dock and lock into the rigid, template-like corresponding isosequential stretch in the surface of PrPSc. Such spontaneous dock and lock mechanism is typical of seeded templated conversion, and propagation, of amyloid forming proteins that are intrinsically disordered such as Amyloid β [14, 15]. Conversion of the ∼90–120 stretch to the PIRIBS conformation is highly favored by the large decrease of enthalpy afforded by the extensive of newly formed cross-β hydrogen bonds, substantially higher than the decrease of entropy incurred in by transition from disorder to a highly ordered β sheet-rich conformation [14, 15]. It is well known, since the seminal studies by Dobson and collaborators, that the amyloid conformation is the most stable one for any protein, as long as the energy barrier to unfold it is provided [16]. Once the ∼90–120 stretch is locked in place, the remaining ∼121–230 folded domain of PrPC must unfold to dock and lock. The specific atomistic pathway of this part of the conversion remains to be resolved and likely exhibits some characteristics of a polymer-templated conversion as proposed by Griffith in his seminal paper suggesting a proteinic nature for the scrapie agent [17, 18]. But how is the initial PrPSc assembly formed in spontaneous prion diseases, in the absence of any pre-existing PrPSc template? How are PrPSc prions “born”? To address this question, in vitro models of spontaneous PrPSc generation are essential. While early experiments showed that small titers of infectivity can be generated by spontaneous conversion of unfolded PrP to an amyloid conformation under denaturing conditions [19], such PrP amyloid conformer is different from bona fide PrPSc, as revealed by atomistic structures resolved by cryo-EM of one type of PrP amyloid [20, 21], and the other [6–10]. Following the groundbreaking generation in vitro of the first bona fide PrPSc from a set of completely defined components including brainpurified PrPC [22], the first spontaneous generation of bona fide recombinant PrPSc was achieved by Wang et al.. using Protein Misfolding Cyclic Amplification (PMCA) adapted to a recombinant PrPC substrate [23]. In the absence of any seed, but in the presence of RNA isolated from normal mouse liver and the anionic phospholipid 1-palmitoyl-2-oleoylphosphatidylglycerol (POPG) as cofactors, these researchers generated a PrP conformer with a 100% attack rate to wild-type mice and a ∼90–230 PK-resistant core. In the absence of definitive cryo-EM evidence showing that such conformer is PrPSc, all the available biological and biochemical evidence suggests that it is. Subsequent studies identified other cofactors such as phosphatidylethanolamine [24–26] as alternative facilitators of spontaneous acquisition of the PrPSc conformation by recombinant PrPC. Of note, all these cofactors have in common that they are negatively charged/ zwitterionic. The introduction of Protein Misfolding Shaking Amplification (PMSA) [27] represented a substantial leap forward in the modelling of spontaneous generation of PrPSc in vitro. PMSA was originally developed as a robust alternative to PMCA to propagate recombinant PrPSc using a pre-formed PrPSc seed, dextran sulfate as cofactor, recombinant PrPC as substrate and shaking in the presence of small zirconia beads as the force to fragment growing recombinant PrPSc assemblies and thus multiply the templating surfaces [27]. Its efficiency and Page 3 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 scalability allow facile production of milligram quantities of fully infectious recombinant PrPSc. Furthermore, the use of dextran sulfate results in selection and reproducible propagation of only certain conformers/strains of the several that might be present in the initial seed [28], so that the final PrPSc product is homogeneous, with biochemical and biophysical properties corresponding to a single strain [27]. Shortly thereafter, the detection of infrequent, random occurrence of spontaneous recombinant PrPSc species in unseeded samples subjected to PMSA prompted a systematical search for conditions enhancing such events. We found that substitution of zirconia beads by glass beads, and some other minor adjustments led to a robust, reproducible PMSA variant capable of inducing facile, spontaneous generation of recombinant PrPSc in the absence of a pre-formed PrPSc seed [29]. Such modified PMSA method easily generates large quantities of spontaneous recombinant PrPSc in a short period of time, needing just 1–4 passages vs. 17 in the original PMCA-based method [23]. When recombinant bank vole (BV) PrPC with the 109I polymorphism was subjected to this version of PMSA, with dextran sulfate as the sole cofactor, four different conformers/ strains of spontaneously generated recombinant BVPrPSc were obtained, each one with different biochemical and biological properties. These four strains propagated true over many seeded passages [29]. Each strain presented a unique pattern of PK-resistant peptides, incubation time in first and second passages and lesion profile when inoculated to transgenic mice expressing ~ 1x BVPrP(109I) [29]. Subsequently, we applied this methodology to a very large collection of recombinant PrPC substrates from hundreds of species across most mammalian orders and families and achieved spontaneous recombinant PrPSc from the majority of them [30]. This study demonstrated the versatility of the PMSA method and the many possibilities it offers for the study of spontaneous generation of PrPSc at the molecular level. For example, given the easiness with which the combination of BVPrP(109I), the most spontaneously prionogenic PrP sequence that we have encountered (further details on its misfolding proneness can be found at the PrPdex webpage, previously published in Eraña et al., 2024 [30]), we wondered whether dextran sulfate was an essential ingredient for such easy conversion into a variety of BVPrPSc strains, forming a lucky combination with BVPrPC, or whether it could be substituted by other polyanionic glycans acting as cofactors. Furthermore, we became interested in knowing whether different cofactors might induce spontaneous PrPSc conformers with similar biochemical and biological characteristics, so as to begin to tackle the role of cofactors in determining the structure of different PrPSc strains and its underlying rules. In this study, we systematically explore how three different negatively charged polysaccharides—heparin, chondroitin sulfate, and pentosan polysulfate—influence spontaneous recombinant PrPSc generation and conformational diversity. Through biochemical characterization, electron microscopy, and in vivo assessment of infectivity and strain properties, we demonstrate that these different cofactors can generate unique PrPSc conformers with distinct biological features. Materials and methods Expression and purification of Recombinant bank vole PrP 109I Recombinant full-length bank vole PrP (109I) was expressed in E. coli Rosetta (DE3) competent cells (Millipore) harboring the pOPIN E expression vector as previously described [31]. Briefly, the wild-type I109 bank vole Prnp sequence (amino acids 23–231, based on GenBank accession number PQ327920) was cloned into the pOPIN E expression vector by homologous recombination after using the oligonucleotides 5’ AGGAGATATACCATGAAGAAGCGGCCAAAGCCTGG3’ and 5’ GTGATGGTGATGTTTGGAACTTCTCCCTTCGTAGTA3’ to amplify it from bank vole genomic DNA. Bacteria were transformed by heat-shock and grown in LB broth with 50 µg/mL ampicillin sodium salt at 37 °C with shaking. Protein expression was induced by adding 1 mM Isopropyl β-D-1-thiogalactopyranoside. Subsequently, the bacterial culture was centrifuged at 4,500 g for 15 min at 4 °C. Pellets were resuspended and lysed with lysis buffer containing 50 mM Tris-HCl, 5 mM EDTA, 1% Titon X-100, 1 mM PMSF (adjusted to pH 8), 100 µg/mL lysozyme, 100 U/mL DNase, and 20 mM MgCl2. The suspension was then incubated for 30 min with vigorous shaking at room temperature and subsequently centrifuged at 8,500 g at 4 °C for 1 h. Pellets containing bacterial inclusion bodies were resuspended in 50 mL of washing buffer (20 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% Srkosyl, pH 8), and after additional centrifugation at 8,500 g, at 4 °C for 1 h, the pellet was dissolved in inclusion buffer containing 20 mM Tris-HCl, 0.5 M NaCl, and 6 M GdnHCl (pH 8). The suspension was incubated overnight at 37 °C with vigorous shaking, followed by final centrifugation at 8500 g and 4 °C for 1 h. The supernatant was filtered through a 0.20 μm-pore syringe filter. Protein purification was carried out using a histidine affinity column (HisTrap FF Crude 5 mL chromatographic columns, GE Healthcare) coupled to an ÄKTA Start FPLC system (GE Healthcare). Although this protein does not contain a His-tag, purification was performed with a histidine affinity column, taking advantage of the natural histidine residues present in the octapeptide repeat region of PrP. After purification, the proteins were eluted in 30 mL of elution buffer containing 20 mM Tris-HCl, 500 mM NaCl, 500 Page 4 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 mM imidazole, and 2 M GdnHCl (pH 8). Subsequently, the quality and purity of protein batches were assessed by total protein staining after SDS-PAGE. Finally, GdnHCl was added to a final concentration of 6 M. The final protein concentration was adjusted to 25 mg/mL by concentrating using 10 kDa centrifugal filter units (Amicon Ultra-15, 10 kDa, Millipore). For long-term storage, purified protein preparations were aliquoted and frozen stored at −80 °C. De novo generation of PrPSc by PMSA with different cofactors Aliquots of recombinant protein stored at −80 °C in 6 M GdnHCl were thawed on ice, diluted 1:5 in PBS, and dialyzed against PBS (1:10,000) for 1 h at room temperature. After dialysis, the protein was centrifuged for 15 min at 19,000 g and 4 °C. Subsequently, the supernatant containing PrP folded into its native conformation was diluted to a final concentration of 2 µM with conversion buffer (CB) consisting of 0.15 M NaCl and 1% Triton-X-100 in PBS, supplemented with a cofactor: 0.1% (w/v) heparin sodium salt from porcine intestinal mucosa (Sigma, Cat No SRE00027, 15–19 kDa), 0.1% (w/v) chondroitin sulfate sodium salt from shark cartilage (Sigma, Cat No C4384, 50–80 kDa), or 0.1% (w/v) pentosan polysulfate (MedChem Express, 4–6 kDa). Note that heparin and chondroitin sulfate are chemically heterogeneous biopolymers with molecular weights that vary depending on sulfation degree and chain length, resulting in considerable batch-to-batch variability. The mixture containing CB, the recombinant protein, and the corresponding cofactor is referred to hereafter as “substrate”. PrP conversion was performed by PMSA as previously described [27]. Briefly, 500 µL of substrate was aliquoted into 2 mL screw-cap conical tubes containing ~ 180 mg of 1 mm or 0.1 mm glass beads (Sigma-Aldrich). Tubes were continuously shaken at 700 rpm at 39 °C for 24 h using thermoblocks (Digital Shaking Drybath, Thermo Scientific). To perform serial rounds of PMSA, a 1:10 dilution of the PMSA product was used as a seed and added to tubes containing fresh substrate supplemented during the first three passages with 0.1–1 mm glass beads, or with 1 mm zirconia-silica beads (BioSpec Products Inc.) for subsequent passages for stabilization. Some of the samples showing spontaneous rec-PrPres after the three serial rounds in the presence of glass beads, chosen for their differences in electrophoretic mobility patterns, were propagated further for stabilization, performing four additional serial PMSA rounds of 24 h with 1:10 dilution. These stabilized PMSA products were used for most biochemical and biological characterization experiments, and also for longterm storage and further propagation through coating of 2.3 mm diameter zirconia-silica beads (BioSpec Products Inc.), as previously explained [32]. Detection of PK-resistant recombinant PrP by total protein staining The PMSA product was analyzed by SDS-PAGE and stained after digestion with PK. For this purpose, the PMSA product was subjected to PK (Roche) digestion to a final concentration of 25µg/mL at 42°C for 1h. Subsequently, samples were centrifuged at 19,000 g at 4°C for 15min. The supernatant was carefully discarded, and the pellet was washed with 500 µL of PBS and centrifuged again at 19,000 g at 4°C for 5min. After removing the PBS, the pellet was resuspended in 15 µL of 1x loading buffer (NuPAGE 4x, Invitrogen Life Technologies) and boiled at 100°C for 10min. Subsequently, samples were loaded onto 4–12% acrylamide gels (NuPAGE Midi gel, Invitrogen Life Technologies). Electrophoresis was performed for 1h 20min (10min at 70V, 10min at 110V, and 1h at 150V). The gel was then stained using BlueSafe (NZYtech) for total protein staining at room temperature with gentle agitation. Animal housing and experimental procedures All animal experiments were conducted in accordance with European and national regulations on animal protection for experimental and other scientific purposes, and with approval from the respective institutional ethics committees. We used a transgenic mouse model expressing ~ 1x bank vole PrP, bearing isoleucine at position 109 (GenBank accession number PQ327920), hereinafter TgVole(I109)1x (FVB/N.129Ola-Tg (Prnp-Bvole109I) C594PRC/Cicb) [27, 29, 32]. Animals were housed in groups of 3–6 per cage in a controlled environment at a room temperature of 22 °C, with a 12-hour light/darkness cycle, and 60% relativ humidity in HEPA-filtered, individually ventilated cages. Animals had access to food and water ad libitum, and received environmental enrichment. For prion transmission experiments, the following protocols were employed. Inoculum preparation: Inocula from PMSA products were prepared by diluting the PMSA product 1:10 in PBS prior to inoculation. All PMSA-derived inocula were normalized based on equal volumes from reactions containing identical starting amounts of recombinant PrPC substrate, with reactions carried out to conversion plateau where available substrate was depleted. This approach was chosen over quantification of PK-resistant material to avoid potential artifacts from strain-specific differences in protease sensitivity. While this ensures comparable amounts of converted protein across preparations, modest differences in final yields between strains may still influence infectivity comparisons. Similarly, inocula from brain homogenates of terminally ill mice were prepared by diluting to 1% (w/v) in PBS. Intracerebral inoculation: Sixto eight-week-old male and female Page 5 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 TgVole(I109)1x mice were anesthetized using either isoflurane (IsoVet, Braun) or a combination of ketamine/ medetomidine (75/1 mg/kg) (Imalgene 1000, Boehringer Ingelheim/Domtor, Ecuphar). For the ketamine/medetomidine protocol, anesthesia was subsequently reversed with atipamezole hydrochloride (1 mg/kg) (Antisedan, Ecuphar). Following cranial exposure, a small perforation was created in the right parietal bone, through which 20 μL of inoculum was administered to the right cerebral hemisphere at approximately 3 mm depth using a precision syringe with a sterile 27-gauge hypodermic needle (Terumo). To prevent reflux along the injection tract, the needle remained in position for an additional 20 seconds before gradual withdrawal. Post-procedure, animals received subcutaneous buprenorphine (0.3 mg/kg) for analgesia and were maintained on a heating pad until complete recovery from anesthesia. Clinical monitoring post-inoculation: Following inoculation, mice were monitored daily for general health status. Detailed clinical assessment was performed twice weekly until the first appearance of neurological signs, after which monitoring was increased to daily observation. The presence of TSE-associated clinical signs was scored (0–3) based on the following parameters: kyphosis, gait abnormalities, altered coat state, depressed mental state, flattened back, eye discharge, hyperactivity, loss of body condition, and incontinence. Animals showing sustained clinical signs (score ≥ 2 in two or more categories) or severe neurological impairment that compromised welfare were humanely euthanized by cervical dislocation and subsequent decapitation. Survival time was calculated as the interval between inoculation and euthanasia, expressed as days post-inoculation (dpi). Attack rate was determined as the ratio between animals developing confirmed prion disease (by histopathological and/or biochemical analyses) and the total number of inoculated animals. Animals found dead due to intercurrent disease before 100 days post-inoculation were excluded from the study. Results are expressed as mean incubation period ± standard error of the mean (SEM) for each experimental group. Brains were removed and cut sagittally into two halves. One half was frozen for biochemical characterization, and the other half was fixed in 4% paraformaldehyde (PFA) in PBS for histopathological and immunohistochemical analysis. Histopathological and immunohistochemical analysis For histopathological analyses, transverse sections of the fixed brains were performed at the levels of the medulla oblongata, piriform cortex, and optic chiasm (Supplementary Fig.1). Samples were embedded in paraffin wax after dehydration through increasing alcohol concentrations and xylene. For morphological evaluation, 4 µm sections were mounted on glass microscope slides and stained with hematoxylin (Sigma-Aldrich) and eosin (Casa Álvarez). For immunohistochemical analysis, sections were mounted on 3triethoxysilylpropylaminecoated glass microscope slides (DAKO). Deparaffinized sections were subjected to epitope retrieval by immersion in formic acid and boiling at low pH (6.15) in a pressure cooker, followed by pre-treatment with proteinase K (Roche). Endogenous peroxidases were blocked by immersion in 3% H2O2 in methanol. Subsequently, sections were incubated overnight with anti-PrP 6C2 monoclonal antibody (1:1000) (CVI-Wageningen UR) and visualized using the goat anti-mouse EnVision system (DAKO) with 3,3’-diaminobenzidine (Sigma-Aldrich) as the chromogen substrate. For negative controls, incubation with the primary antibody was omitted. Spongiform lesion was evaluated following a semiquantitative approach previously described [33]. Briefly, fourteen different brain areas were examined (Supplementary Fig. 1): piriform cortex (Pfc), hippocampus (H), occipital cortex (Oc), temporal cortex (Tc), parietal cortex (Pc), frontal cortex (Fc), striatum (S), thalamus (T), hypothalamus (HT), mesencephalon (M), medulla oblongata (Mobl), cerebellar nuclei (Cm), cerebellar vermis (Cv), and cerebellar cortex (Cc). Spongiform lesions in each area were scored as follows: (0) absence of spongiosis, (1) mild, (2) moderate, (3) intense, and (4) maximum level of lesion. Each area was evaluated globally as a region for scoring. Brain lesion profiles were plotted using the anatomical regions and the mean of the semiquantitative spongiosis scores for each area. Error bars represent the standard error of the mean [33]. Western blot of brain homogenates Frozen brains were suspended at 10% (w/v) in PBS containing complete protease inhibitors cocktail (Roche) and were homogenized using a Dounce homogenizer until a homogeneous mixture was obtained. Brain homogenates were stored at −80°C. Fifteen microliters of brain homogenates was mixed with 5 µL of buffer containing 2% (v/v) NP40 and 5% (w/v) Sarkosyl in PBS, or alternatively, 10 µL of brain homogenate was mixed with 10 µL of the same digestion buffer. Subsequently, samples were digested with 85µg/mL PK for 1h at 56°C and 450rpm. The reaction was stopped by adding loading buffer (1:3 ratio) and heating at 99 °C for 10 min. For PNGase F digestion, the same 10% brain homogenates were digested with PK as indicated above and immediately after digestion, denaturing buffer [20% SDS (NZYTech) and 100 mM of DTT (Sigma-Aldrich)] was added to each sample for a final concentration of 4% SDS and 20 mM of DTT. The samples were incubated at 100°C for 30min and then precipitated with ice-cold methanol, by incubating the samples at −20°C for 30min and centrifugation at 20,000 g for another 30min. Supernatants were Page 6 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 discarded, and the pellets were air-dried prior to PNGase digestion. Dry pellets were resuspended in 1X PNGase F buffer and digested with 1.5µl of PNGase F (Gibco) for 1h at 50°C. Digestion was stopped by adding NuPage 4X Loading Buffer (Invitrogen Life Technologies) and boiling them for 10min. Samples were electrophoresed on 4–12% acrylamide gel (NuPAGE Midi gel, Invitrogen Life Technologies) and transferred to a PVDF membrane using the iBlot 3 (Invitrogen Life Technologies) transfer system. Subsequently, blocking, washing, and antibody incubations were performed with the iBind™ Flex Western Device (Invitrogen Life Technologies). The primary monoclonal antibody used was SAF83 (1:400) (recognizing epitope 130–160) (Vitro S.A.), and the secondary antibody was peroxidase-conjugated anti-mouse IgG (m-IgGκ BP-HRP, Santa Cruz Biotechnology). Images were acquired using the iBright CL750 imaging system (Invitrogen Life Technologies). Glycoform ratio quantification was performed over digitalized WB images through densitometric analysis, done using the AlphView software (Version 3.4.0, FluorChem Q). Mass spectrometry For mass spectrometry-based analysis, 2 mL of the PMSA product was digested by adding PK to a final concentration of 25µg/mL and incubating at 42°C for 1h. After digestion, samples were centrifuged for 30min at 19,000 g, the supernatant was discarded, and the pellet was resuspended and washed with 1mL of PBS. Subsequently, the sample was centrifuged again for 30min at 19,000 g, the supernatant was discarded, and the pellet was resuspended in 50 µL of 6M GdnHCl with 3 pulses of a tip sonicator and incubated for 1h at 37°C. After incubation, TFA was added to a final concentration of 1%. Samples (4 µL) were injected into a nanoLC 400 micro liquid chromatography system coupled to a Triple TOF 6600 high-speed mass spectrometer with a microflow source (SCIEX) and equipped with a silica-based reversed-phase ChromXP C18 column (150 × 0.30 mm, 3 μm particle size, 120 Å pore size). A YMC-TRIART C18 trap column was connected online prior to the separating column (3μm particle size, 120 Å pore size) (Tecnokroma). After sample loading and washing with 0.1% formic acid in water to remove GdnHCl and other non-peptide components, the flow was switched to the analytical column, and separation proceeded at a flow rate of 5µl/min with a solvent system consisting of 0.1% formic acid in water as mobile phase A, and 0.1% formic acid in acetonitrile as mobile phase B. Peptides were separated over 40min with a gradient ranging from 2% to 90% mobile phase B. Data acquisition was performed using a TripleTOF 6600 System (SCIEX) with a data-dependent workflow. Source and interface conditions were as follows: ionspray voltage floating (ISVF) 5500V, curtain gas (CUR) 25, collision energy (CE) 10, and ion source gas 1 (GS1) 25. The instrument was operated with Analyst TF 1.7.1 software (SCIEX). Switching criteria were set to ions with mass-to-charge ratio (m/z) greater than 350 and less than m/z 1400, with a charge state of 2–5, a mass tolerance 250 ppm, and an abundance threshold of more than 200 counts (cps). Previously targeted ions were excluded for 15s. The instrument was automatically calibrated every 4h using tryptic peptides from PepCalMix (SCIEX) as an external calibrant. For data analysis, the sample TIC was analyzed using the PeakView2.2 software, which allows protein reconstruction. The LC-MS Peptide Reconstruct feature uses a peak-finding algorithm to identify groups of peaks that form isotope series and charge series. Protein deconvolution was carried out in the range of 800 to 20,000 Da. PK-resistant fragments were identified using the GPMAW software tool, based on a ± 1 Da difference between the observed peak and the theoretical molecular mass. Transmission electron microscopy Zirconia-silica beads coated with each PMSA product were used as seeds to propagate the nine different conformers for their purification and visualization by transmission electron microscopy with negative staining. Using 2–3 prion-loaded beads for each PMSA product, two-step PMSA reactions were performed. First, 4 mL of substrate was seeded and submitted to PMSA for 4h, which was then transferred to fresh substrate (1:5 dilution) supplemented with clean zirconia-silica beads and subjected to an additional 20h of PMSA. Sodium dodecyl sulfate (SDS) was added to a final concentration of 0.1% (v/v) to the approximately 20 mL of each PMSA product, which was then incubated at room temperature for 2h on a rotating mixer. After centrifugation of the PMSA products at 100 g in a swinging-bucket rotor for 10min at RT. Subsequently, most of the supernatant was recovered and centrifuged again at 1,000 g for 90min at RT. In this case, most of the supernatant was discarded and the pellet was resuspended in the remaining volume (approximately 100 µL) and washed with 2 mL of 10 mM Tris-NaCl containing 0.1% of SDS. These concentrated samples were digested with 150µg/ml of PK at 37°C for 45min with mild shaking. After digestion, they were centrifuged at 1000 g for 10min at RT using a swinging-bucket rotor, and the supernatants were completely discarded. The pellets were washed three times using 10 mM TrisNaCl with 0.05% of SDS in the first wash and 10 mM Tris NaCl with 0.02% of SDS in the subsequent washes. After the final wash and centrifugation at 1000 g for 5min at RT, the supernatant was completely discarded and the final pellet for each PMSA product was resuspended in 20–50 µL of 10 mM Tris-NaCl with 0.02% SDS. After thorough resuspension, each sample was deposited onto freshly Page 7 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 glow-discharged carbon-coated copper grids (Carbon Film 400 Mesh, Cu, Electron Microscopy Sciences). After 1min at RT, the grids were washed with deionized water for 1min and stained with freshly prepared, filtered 5% uranyl acetate solution for 45s. Imaging was conducted using a JEM-1230 (JEOL) transmission electron microscope operated at 100kV and equipped with an Orius SC1000 CCD camera (GATAN). Quantitative morphological analysis of fibril heterogeneity To address the morphological diversity observed among different conformers, a systematic quantitative analysis was performed on transmission electron microscopy images. Eighteen representative TEM micrographs were selected for each of the nine conformers (Hep0103, CoS01-03, PPS01-03) and arranged in 6 × 3 grids for comprehensive visualization (Supplementary Fig.7. Images were acquired using identical microscopy conditions without pre-selection bias. Fields were captured sequentially as soon as fibrils became visible, with only overlapping fields containing identical structures being excluded. No selection was made based on fibril type, size, or density within each field. All image acquisition and subsequent quantitative analyses were performed by a single observer to ensure consistency. Preliminary assessment of fibril populations showed that eight conformers presented predominantly uniform morphological characteristics within each image set, whereas CoS02 displayed two morphologically distinct fibril subpopulations that were analyzed separately (CoS02a and CoS02b). Quantitative morphological classification was performed according to three parameters: Fibril Length (< 50nm, 50–200nm, 200–500nm, > 500nm), Curvature Measurements (straight, slightly curved, moderately curved, highly curved), and Bundling Frequency (isolated fibrils, small bundles, large bundles). Each parameter was assessed independently of total fibril number per field. To ensure comparability across images with varying fibril densities, raw counts were converted to percentages relative to the total fibril population within each image. Quantitative data are expressed as mean ± standard error of the mean (SEM) calculated from the 18 individual measurements per conformer. Conformer-specific morphological profiles were generated by averaging percentage values across all analyzed parameters. Statistical analysis Statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Descriptive statistics including mean, standard deviation, and standard error of the mean (SEM) were calculated for each experimental group of mice. Incubation periods were expressed as mean ± SEM and compared between groups using one-way ANOVA followed by GamesHowell’s multiple comparisons test when appropriate. Survival time was also plotted as Kaplan-Meier survival curves. Attack rates were calculated as the percentage of animals developing confirmed prion disease (determined by clinical signs, histopathology, and/or biochemical analysis) relative to the total number of inoculated animals. For semiquantitative histopathological scoring, lesion profiles were compared between conformers using non-parametric tests (Kruskal-Wallis followed by Dunn’s multiple comparisons test) due to the ordinal nature of the scoring system (0–4 scale). Statistical significance was set at p < 0.05 for all analyses. Data are presented as mean ± SEM unless otherwise indicated. Results Recombinant bank vole (I109) PrP spontaneously misfolded by PMSA exhibits the key biochemical hallmarks of bona fide prions As a logical next step from our previous work [29, 30] we decided to explore whether dextran sulfate could be replaced by a different polyanionic glycan, with a particular interest in the emergence of more than one PrPSc strain. We thus substituted dextran sulfate by two glycosaminoglycans (GAGs), heparin and chondroitin sulfate, and one GAG mimic, pentosan polysulfate. We chose them because, while they are different from dextran sulfate, they share with them common features such as high density of negative charges afforded by -SO3− moieties (Supplementary Fig. 2). Furthermore, GAGs are physiologically relevant to prion disease progression in vivo as shown before [34]. For our purpose, 6 independent, unseeded PMSA reactions were set up for each of the three cofactors, 3 with 0.1 mm and 3 with 1 mm glass beads (Supplementary Fig. 3). Each reaction was submitted to 3 serial PMSA rounds of 24 h. After each round a portion of the reaction product was added to the next reaction mixture. Another portion of the reaction product was used to assess the presence of PK-resistant PrP, a surrogate of PrPSc emergence. For this, the sample was subjected to PK treatment (25 µg/mL, 1 h, 42 °C, 1/0.43 molar ratio PrP/PK) and PK-resistant fragments assessed by SDS-PAGE with total protein staining [29, 30]. After three rounds, 3 of the 6 serial reaction products of each cofactor set, showing apparently distinct PK fragmentation patterns were chosen as distinct rec-PrPSc conformer candidates. We provisionally assumed these products to be PrPSc based on their PK-resistant pattern being overall similar -if with nuancesto those of previously generated and propagated recombinant prions, and to their in vitro propagative nature. Of note, PK-resistant PrP bands were already found in 15/18 tubes after the first round of PMSA, and in 18/18 after the third round. Thus, the frequency of potential spontaneous misfolding was Page 8 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 83% in the first, 89% i the second and 10% in the third PMSA rond. These results are comparable to those previously obtained with dextran sulfate [29], demonstrating that heparin, chondroitin sulfate and pentosan polysulfate are roughly equipotent to dextran sulfate as facilitators of spontaneous generation of recombinant bank vole PrPSc. The consistent emergence of PK-resistant products across different cofactors supports the hypothesis that polyanionic molecules share common mechanisms for facilitating spontaneous prion misfolding. These 9 distinct reaction products, hereafter referred to as “spontaneous rec-PrPSc conformers”, were subsequently propagated using standard PMSA with zirconia beads [27] (Supplementary Fig. 3) and PK-resistant patterns of stabilized products are shown in Fig. 1. Several conformers obtained with different cofactors appeared to be remarkably similar; for example, PPS02 is very similar to Ust09, and Hep01 to CoS03. However, careful semiquantitative assessment of band intensities showed every conformer to be unique (Supplementary Fig. 4). No obvious cofactor-specific trend could be discerned. Every conformer presented a ~ 16 kDa PKresistant band, corresponding to the classic ~ 90–231 PrPSc PK-resistant fragment (bank vole PrP numbering). This fragment, with added mass from glycans and GPI-anchor, is referred to as PrP27-30 in brain-derived PrPSc [2, 35–37]. This PK-resistant fragment is now known to correspond to the PrPSc PIRIBS core [6–10] and is also prominent in all bona fide rec-PrPSc generated to date, whether spontaneously or by seeded templating of rec-PrPC [23–25, 27, 29, 30, 37]. In addition to this fragment, the 9 different conformers featured combinations of 4 main additional PK-resistant fragments with apparent molecular masses of ~ 9, ~8, and ~ 6 kDa (Fig. 1). An exact identification of these fragments, carried out by means of mass spectrometry, will be described in the next section. However, we can anticipate that the ~ 9 and ~ 6 kDa bands are C-terminal and N-terminal fragments of the ~ 16 kDa band, respectively, resulting from a secondary PK nick at Asn153 or Met154, whereas the ~ 8 kDa fragment result from an alternative nick at Tyr163. No N-terminal fragment complementary to this fragment was detected. Other minor fragments were seen across the different conformers. Similar patterns have been obtained for other PMSA-derived rec-PrPSc conformers; their main difference from patterns obtained from brainderived PrPSc conformers such as RML, GPI-less RML, 263 K, or Drowsy is the larger abundance of fragments derived from internal nicks in the putative PIRIBS core. Fig. 1 Distinct electrophoretic mobility patterns following PK digestion of spontaneously misfolded recombinant PrP using PMSA and 3 different polyanionic cofactors. Three PMSA-treated samples with heparin displayed distinct patterns of PK-resistant fragments, with different proportions of specific fragments. For example, Hep02 features a dominant ~16 kDa fragment and much fainter ~9 kDa and ~6 kDa fragments, which are much more intense in Hep01 with respect to the ~16 kDa fragment. In some cases, differences in the patterns are subtle, as observed between Hep01 and Hep03. Similar differences exist between the 3 different patterns observed with chondroitin sulfate and those with pentosan polysulfate. For comparison, patterns from two conformers previously generated with dextran sulfate, Ust02 and Ust09 [29], are also shown. Molecular weight markers slightly overestimate the real MW of PK-resistant PrP fragments (see mass spectrometry-based data below). The arrows indicate ~9 kDa and ~8 kDa fragments, respectively Page 9 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 Fig. 2 (See legend on next page.) Page 16 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 Cross-cofactor comparison suggests potential strain convergence despite distinct origins Comparison of the lesion profiles and incubation periods after second passage among strains generated with different cofactors suggests that some may represent closely related entities despite their distinct origins and electrophoretic mobility patterns observed in the PMSA products. This convergence could be explained by differential protective effects exerted by each cofactor, which might alter PK cleavage sites during in vitro propagation. Alternatively, this phenomenon might reflect the inherent constraints of the bank vole PrP(I109) sequence, which may only be capable of adopting a limited repertoire of stable misfolded conformations, regardless of the initiating cofactor. Most notably, the striking similarity of lesion profiles between Hep01, Hep02, and PPS02, together with their incubation periods of approximately 80 dpi, suggest shared structural characteristics, at least after adaptation to propagation in vivo. Similarly, despite showing different PrPSc aggregate morphologies, CoS01 and CoS02 displayed almost identical spongiform lesions, which, albeit with greater intensity, were also shared by PPS01. Combined with incubation periods of approximately 100 dpi, these similarities could indicate the formation of highly related strains. These findings suggest that while different cofactors can initiate prion misfolding, the ultimate structural and biological properties may be constrained by the intrinsic folding landscape of the specific PrP sequence employed, as illustrated in Fig.5. Discussion Cryo-EM-based elucidation of several PrPSc strains [6– 10] has shown that it is a PIRIBS amyloid akin to other PrP amyloids purported to be non-infectious [20, 21]. The only difference, although not a trivial one, between these two broad categories of PrP amyloid conformers is that bona fide PrPSc, at least for all the handful of prion strains with atomic level structures available, features a large ~ 90–230 PIRIBS core composed of two lobes, roughly coincident with the two subdomains of the PrPC folded domain (β1-α1-β2 and α2-α3). The N-terminal lobe of PrPSc also incorporates the ~ 90–120 portion of the highly flexible N-terminal PrPC tail. In turn, amyloids with reduced infectivity feature only one C-terminal lobe [20, 21]. While these single-lobe PrP amyloids propagate easily in vitro like all amyloids, growing evidence shows they can also propagate in the brain, if inoculated intracerebrally, and cause prion disease [19, 43]. However, their capacity to provoke clinical disease is substantially reduced compared to bona fide PrPSc, often requiring two or even three clinically silent passages in suitable animal models [44–46]. This explains why the first spontaneous recombinant prion to be generated had an unusually long incubation time and required a 16x level of PrP overexpression attained in a transgenic model [19]. Therefore, paradoxically, while this prion definitively proved the prion paradigm, it was most likely not a bona fide PrPSc conformer. Thus, its mode of generation (facile incubation of recombinant PrP under denaturing conditions with no cofactors, with simple agitation), has no relevance to the mechanism of genesis of the bona fide prions that appear spontaneously in the brain of mammals, including humans. Such spontaneous prions are central to the vast majority of human prion diseases: they directly cause idiopathic cases, likely underlie the transmissible ones such as those seen in kuru and iatrogenic CJD, and may contribute to genetic cases where mutations facilitate spontaneous misfolding, most —if not all— of which can be traced back to an ancestral spontaneous PrPSc conformer. After the groundbreaking experiment of Legname et al.., many groups pursued, with little success, methods to generate spontaneous PrP conformers with closer resemblance to brain PrPSc in terms of infectivity and structure. Given that at the time the structure of PrPSc had not yet been elucidated, biochemical properties such as a large ~ 90–230 PK-resistant core were used as a structural surrogate. Wang et al. were the first to succeed in generating a spontaneous PrP conformer with the biological and biochemical characteristics of brain PrPSc. Adapting PMCA, they generated an infectious recombinant PrP product that exhibited an incubation time similar to that of a brain-derived, bona fide PrPSc in wild-type mice and yielded a ~ 90–230 PK-resistant fragment upon treatment with protease. Of note, Wang´s successful reaction mixture contained two anionic cofactors: POPG and RNA [23]. Unfortunately, this and other subsequent PMCA-based methods, all utilizing these and other negatively charged or zwitterionic cofactors such as phosphatidylethanolamine [24, 47–49] exhibited substantial variability, often producing PrP conformers with low or no infectivity and variable biochemical characteristics. In contrast, our PMSA-based method robustly and reproducibly generates large amounts PK-resistant PrP conformers that demonstrate infectivity in transgenic mice [29, 30]. However, establishing the specific infectivity of these conformers is challenging due to potential conformer mixtures, where highly infectious minor components might coexist with less infectious major populations. This limitation also applies to PrPSc isolated from infected brain tissue. While we cannot definitively determine specific infectivity values, the abundant production of infectious material and consistent attack rates (100%) suggest that our PMSA-generated conformers possess substantial biological activity. In its original formulation, the key elements of our PMSA method to generate spontaneous rec-PrPSc were shaking, glass beads, the detergent Triton X-100, and Page 17 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 dextran sulfate as a cofactor [29, 30]. Here we show that three additional sulfated glycans are equally adept at facilitating robust spontaneous generation of rec-PrPSc. Therefore, the original cofactor that we used when developing our method, dextran sulfate, is not a sine qua non element for success. Neither is the specific original bank vole 109I PrP sequence, which can be replaced by many other PrP sequences [30]. This suggests that, by exclusion, glass beads are the key element in our method. In fact, we already noted during the development of the method that substitution of glass beads with zirconia beads, which are effective at facilitating propagation of pre-formed PrPSc [27], or beads of other materials such as steel, has a profound impact on the performance of the method [29]. Furthermore, studies carried out by us in parallel to these, and published very recently, showed that complete elimination of any cofactor, as long as glass beads are maintained, allows spontaneous generation of infectious PrPSc. However, substantially lower efficiency was obtained. Unlike in the dextran-complemented reactions, generation of rec-PrPres was inconsistent across different experiments and replicates. In many cases, after the initial 24 h of PMSA, the protease-resistant products detected featured only low molecular weight fragments (~ 10 kDa and < 5 kDa), which evolved in subsequent rounds into the electrophoretic pattern associated with recombinant bona fide PrPSc prions, characterized by a 16 kDa fragment [32]. This means that while the glass beads are the only component that cannot be eliminated, robust and consistent spontaneous conversion of PrPC to PrPSc requires combining them with a sulfated glycan, which needs not be dextran sulfate. Our data show that heparin, chondroitin sulfate, and pentosan polysulfate also share with dextran sulfate the capacity to facilitate spontaneous generation of a diversity of PrPSc strains. This is an important feature of our method, not shared by earlier methods which generate only one PrPSc strain [25]. It is known that brains affected by spontaneous prion disease may harbor more than one prion strain, which must necessarily have emerged together [50, 51]. In fact, a notion widely held in the prion field is that pure prion strains rarely exist; rather, populations of slightly different PrPSc conformers termed “quasi-species” predominate, often with a main component and many minor ones [42]. Our method thus offers a unique tool to study the emergence of such quasi-species. Our results do not provide any evidence of correlations between a given cofactor and the biochemical (and therefore structural) and biological properties of the spontaneous PrPSc conformers whose emergence it facilitates. As shown in Fig. 1 and Supplementary Fig. 4, PK-resistant patterns seem randomly distributed across the PrPSc conformers obtained with the three cofactors (four, if we include the two strains previously obtained with dextran sulfate and used for comparative purposes). The pattern of a PrPSc strain obtained with one given cofactor is often more similar to the pattern of a strain obtained with a different cofactor than to the patterns of strains obtained with the same cofactor. The same is true for incubation times. Mean incubation times in the first passage did not vary significantly across cofactors, and within each cofactor group, very large differences were seen. Furthermore, strains showing relatively similar PK-resistant patterns, such as Hep01 and Hep03 or PPS02 and Hep02, exhibited very discordant incubation times among themselves at first passage, although this could be attributed to differences in the titer of each PMSA product inoculated. We could not derive any correlation between the presence or absence of a given PK-resistant fragment, or the relative intensity of the corresponding electrophoretic bands, and incubation times. With respect to the second passage, there were reductions in incubation time for every strain, as expected. Again, however, no pattern could be established. Strain Hep01 changed from 363 ± 7 to 86 ± 5 days, whereas Hep03 changed from 105 ± 2 to 95 ± 4 days, suggesting that the transmission barrier was easily surmounted in the second passage regardless of its original magnitude. All of this suggests that the biological properties of a given spontaneous rec-PrPSc conformer must be subtly encoded in its structure in such a way that the biochemical methods used in this study to assess and compare differences and similarities between the different conformers do not have sufficient resolution to allow for definitive conclusions. The high variability in PK-resistant fragment patterns and their lack of correlation with biological features may involve cofactor-PrPSc interactions. Recent cryo-EM structures of hamster and murine prions reveal poorly resolved densities near polybasic stretches (such as the K101-K110 cluster) [6, 8] that are replaced by phosphotungstic acid (PTA) in purified strains [9], suggesting these densities represent endogenous polyanionic cofactors—likely GAGs, nucleic acids, or lipids. Previous studies demonstrate that distinct polyanionic molecules exhibit varying PrPC-binding affinities and effects on prion misfolding in vitro [52, 53]. Since these cofactors are PK-resistant, they can sterically protect nearby cleavage sites from protease digestion, as demonstrated experimentally for dextran sulfate-associated recombinant prions [32]. Therefore, the observed electrophoretic differences might reflect varying cofactor binding patterns and protective effects rather than intrinsic structural variations among PMSA products. This mechanism could explain why biochemically distinct preparations converge upon in vivo transmission, highlighting the need for complementary characterization methods beyond electrophoretic analysis. Page 18 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 The quantitative morphological analysis of fibers visualized by TEM and negative staining presented here provides objective validation for the morphological diversity observed qualitatively through biochemical and biological characterization. The identification of five distinct morphotypes demonstrates that sulfated glycan cofactors can generate not only biochemically distinct conformers but also morphologically distinguishable fibril populations. The fact that different cofactors can produce similar morphotypes (e.g., Morphotype 1 across multiple conformers) while the same cofactor can generate distinct patterns (e.g., Hep01 vs. Hep02) further suggests that the relationship between cofactor chemistry and final fibril morphology is complex and likely influenced by stochastic nucleation events during PMSA. This morphological diversity may reflect subtle differences in PIRIBS core organization that are not readily apparent in biochemical analyses but could contribute to the distinct biological properties observed in vivo. Without doubt, resolution of the structures at an atomistic level by means of cryo-EM (ongoing studies) will provide answers. Previous studies using shaking or sonication-based methods with single cofactors have generated apparently stable conformations that eventually converged into a single strain [24, 54], contrasting with our PMSA results using dextran sulfate [29] or the three polyanionic cofactors described herein. Among the main differences with respect to the previous methods is the inclusion of glass beads that is critical to achieve spontaneous rec-PrP misfolding in our system together with the cofactor, which could raise doubts on their role in generating and keeping stable the distinct conformers characterized here. Glass beads could somehow overcome the effect of specific cofactors on defining and maintaining their biological properties, which in other cases have been demonstrated to converge to a predominant conformation. Nonetheless, the fact that the electrophoretic mobility pattern and PK-resistant fragments of our recombinant prions remain stable throughout their propagation in PMSA using zirconia silica beads (also observed in Eraña et al., 2023 [29]), suggests that the main drivers of the conformational differences in this system are the distinct cofactors, rather than the use of a specific type of beads. GAGs, and more specifically heparin, have long been known to facilitate spontaneous generation of neurodegenerative disease-relevant amyloids in vivo, particularly α-synuclein (α-syn) [55, 56] and tau [57, 58]. Besides acting as a scaffold for the assembly of the amyloid, negatively charged GAGs counter the stacked, electrostatically repulsive positive charges of similar basic residues in the PIRIBS. A recent cryo-EM-based study of α-syn fibrils in complex with heparin showed that heparin chains extend along the α-synuclein protofilament axis, with sulfonic acid groups forming salt bridges with two positively charged grooves formed by Lys43 and Lys45, and Lys58, Lys60, and Lys97, respectively (termed conformer P1 by the authors). However, heparin can also interact with Lys80, in which case the fold of the α-synuclein monomer changes and a different polymeric conformer, termed P4, is generated, with a surface-exposed stack of lys80 residues whose repelling positive charges are neutralized by heparin (termed P4). Two additional conformers, P2 and P3, can also emerge. P2 is characterized by lateral association of two P1-like protofilaments [56]. The study did not reveal what causes emergence of one or other amyloid strain. Another contemporary study solved the atomistic structures of α-synuclein fibers extracted from the brains of patients that had suffered from Parkinson´s disease and multiple system atrophy (MSA). It showed that filaments with a “Lewy fold” feature a non-peptidic density embedded in the characteristic Lys32, Lys34, Tyr39, Lys43, and Lys45 groove of the filament, whereas two strains with “MSA fold” also contained non-peptidic densities at the interface of two protofilaments, surrounded by Lys43 and Lys45 of each of the two α-synuclein units [59]. Both studies highlight the scaffolding capacity of heparin, likely sustaining polymerization of α-synuclein to generate more than one conformer. Taking all this into consideration, it is tempting to speculate that perhaps heparin (and the other two GAGs used in our experiments) might bind to the Lys-rich patch K101PSKPKTNIK110 within the unfolded N-terminal domain of PrPC, providing a scaffold for several such patches and their immediate unfolded Nand C-terminal residues to nucleate into a short PIRIBS (Fig. 6). Eventually, the entire ~ 90–121 tail stretch would be converted to the PIRIBS conformation. It has been previously shown that heparin binds the flexible N-terminal tail of recombinant PrPC through a pH-dependent interaction that is strongest at acidic pH values and sharply decreases at pH values above 7.5. At pH 6.5, the interaction was followed by the formation of PrP oligomeric complexes [60]. It is known that the central portion of PrPC, rich in amyloidogenic residues, has a high tendency to fibrillize into an amyloid conformation. In fact, a PrP106-126 peptide and its amyloid fibrils were used for a long time as a surrogate for PrPSc in dozens of studies in the 1990´s and 2000´s [61]. Paradoxically, when the whole PrP sequence, 23–230, is unfolded under denaturing conditions, fibrillization proceeds through the C-terminus and amyloids with a ~ 170–230 PIRIBS core are formed [20, 21, 62, 63]. This is likely a consequence of the effect of the Cys176Cys214 disulfide bond as a staple that fixes the 176–214 loop, restricting its motion and positioning it as a nucleation core that outcompetes the N-terminal ~ 90–120 region. However, in the presence of a fully folded globular Page 19 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 ~ 121–231 domain as in our experiments, such competition is not possible. Furthermore, the high local concentration of PrPC molecules adsorbed to the surface of glass beads [29] would favor nucleation around the GAG linear polymer (Fig. 6). Continuing with the speculation, once the N-terminal PIRIBS has been formed, the attached globular domains will need to unfold and refold into a C-terminal PIRIBS lobe. Using solution NMR, Vieira et al.. have shown that heparin binds to both the unfolded N-terminus at pH 7 and also to a specific site of the globular domain of mouse PrP at pH 5 [64]. While PMSA is carried out at pH = 7.4 in phosphate buffer, the glass beads used have been acidwashed by the manufacturer, and their surface might provide acidic microenvironments. Of note, when acidwashed beads were replaced by regular glass beads not subjected to acid washing, yields of spontaneous PrPSc generation decreased [29]. Binding of heparin (and other sulfated glycans) to the folded domain of PrPC might facilitate its unfolding or perhaps chaperone the process so as to prevent futile interactions of partially unfolding intermediates competing with ordered stacking, leading to completion of a full ∼90–231 PIRIBS core (Fig. 6). Our results demonstrate that sulfated glycans not only facilitate spontaneous conversion of PrPC into PrPSc but also enable the emergence of different, possibly closely related, but distinct PrPSc strains (see a summary of the differential properties of each preparation evaluated throughout this work in Supplementary Fig. 9). This is of clear relevance to spontaneous emergence of PrPSc in vivo: sulfated glycans (GAGs) such as heparan sulfate (of which heparin is a mimic), and others, are present in cell membranes and are very likely to interact with PrPC and with nascent spontaneous PrPSc. Besides their putative role in facilitating spontaneous emergence of PrPSc, already discussed, GAGs have been shown to play a relevant role in the process of interaction of PrPSc with cells, which is key to their toxicity. Experiments in cellula have shown that heparan sulfate is necessary for PrPSc formation in ScN2a cells and acts as a cellular receptor for purified infectious prions [65–67]. Furthermore, heparan sulfate is found in prion plaques in brain sections from patients affected by prion disease and prion-infected mice [68, 69]. It is therefore not surprising that shortening heparan sulfate chains through transgenic exostosin-1 haploinsufficiency prolongs survival and reduces parenchymal plaques in a model of prion disease [34]. On the other hand, paradoxically, some GAG mimetics have been proposed as anti-prion therapy [70–73]. This apparent contradiction is likely explained by the aforementioned dual action of GAGs and their mimetics: once PrPSc exists in the body, externally administered GAGs might compete with cellular GAGs for binding sites, preventing their interaction with cells and reducing their pathogenicity. Additionally, previous cell culture studies have shown that sulfated glycans can inhibit PrPSc accumulation by altering PrPC subcellular localization through clumping and internalization mechanisms, highlighting the complex and context-dependent nature of GAG-prion interactions [67, 70]. Fig. 6 Hypothetical modulation by GAGs of spontaneous emergence of distinct strains of rec-PrPScthrough PMSA. A: PrPC molecules adsorb to glass beads [29] generating a high local concentration (the ~ 23–90 segments that are disordered and PK-sensitive in both PrPC and PrPSc are omitted for simplicity). Negatively charged GAGs interact with the positively charged K101PSKPKTNIK110 patch within their unfolded 90–120 N-terminal tails, affording a nucleation scaffold. B: An N-terminal PIRIBS block is formed. C: Additional GAG chains adsorbed to the folded ~ 121–231 domains of PrPC molecules help align them and contribute to their unfolding, facilitating their refolding into a C-terminal PIRIBS domain D. An alternative binding pose of the GAG E, either because of an alternative binding interaction or because it is a different GAG with a different structure, nucleates and scaffolds a distinctly different conformation of the C-terminal PIRIBS domain F Page 20 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 In summary, this study demonstrates that sulfated glycans such as heparin, chondroitin sulfate, and pentosan polysulfate are potent cofactors capable of driving the spontaneous generation of multiple, structurally and biologically distinct recombinant prion strains. Using PMSA, we reproducibly obtained nine infectious rec-PrPSc conformers with distinct protease resistance profiles, fibril morphologies, incubation times, and neuroanatomical tropisms, confirming their classification as bona fide prion strains. Notably, some strains with similar properties arose from different cofactors, while multiple strains could also emerge from a single cofactor, highlighting both the permissiveness and selectivity of glycan-driven prion strain diversification. These findings offer significant insights into the molecular determinants of prion strain diversity and support the hypothesis that spontaneous emergence of distinct prion strains in vivo may be influenced by the presence and nature of endogenous polyanionic cofactors. Supplementary Information The online version contains supplementary material available at h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 / s 4 0 4 7 8 - 0 2 5 - 0 2 1 7 5 - w. Supplementary Material 1. Supplementary Material 2. Acknowledgements The authors would like to thank the following for their assistance and support: the IKERBasque Foundation; the vivarium, IT service and maintenance departments of CIC bioGUNE for their outstanding assistance; the Electron Microscopy Platform from CIC bioGUNE for excellent service and assistance; María de la Sierra Espinar and the rest of the IRTA-CReSA BSL3 facility personnel for their excellent technical support; and the staff of the biocontainment units at Neiker and CEBEGA for their excellent care and maintenance of the animals. The authors would also like to acknowledge the work from past laboratory members of the Prion Research Lab from CIC bioGUNE, that despite not directly involved in the manuscript have contributed along the years to the development of all the methods and techniques currently used in the laboratory.The present address of Nuria L. Lorenzo is: Galician Advanced Therapies Manufacturing Center, Galaria, Santiago de Compostela, Spain. And the present address of Sonia Veiga is: Food and Health Omics Lab, Faculty of Sciences, Ourense campus, University of Vigo, Spain. Author contributions Conception and design of the study was done by J.C., J.R.R. and H.E.; N.L.L., H.E., J.M.C., C.M.D-D, L.C. and S.V. performed the biochemical experiments and the analysis of biological samples, all participating in data acquisition, analysis and interpretation; N.L.L., L.P., S.V. and S.B. performed the Mass Spectrometry experiments including material preparation, data acquisition, analysis and interpretation; N.L.L., H.E., E.V., S.G., M.G. performed the in vivo studies and anatomopathological analyses; N.L.L., H.E., J.R.R. and J.C. drafted the initial versions of the manuscript and all authors reviewed and approved the submitted version. Funding The present work was partially funded by the following grants awarded by “Agencia Estatal de Investigación, Ministerio de Ciencia e Innovación” (Spanish Government): PID2024-160022OB-I00, PID2021-122201OB-C21 (granted to J.C.), PID2021-122201OB-C22 (granted to E.V.), and PID2020-117465GB-I00 (granted to J.R.R.), funded by MCIN/AEI/10.13039/501100011033 and co-financed by the European Regional Development Fund (ERDF). Data availability All data generated or analysed during this study are included in this published article and its supplementary information files. Declarations Ethics approval and consent to participate Mouse experiments were performed at multiple research facilities: CIC bioGUNE, Centro de Biomedicina Experimental (CEBEGA) of the University of Santiago de Compostela, Neiker—Basque Institute for Agricultural Research and Development, and IRTA-CReSA Animal Health Research Center. These procedures were approved by the respective institutional ethics committees with the following project codes: CIC bioGUNE (P-CBG-CBBA-0314 and 15,005/16/006), CEBEGA (15012/2023/002), Neiker (NEIKER-OEBA-2021-003), and IRTA-CReSA (5767, 11926 and 1124M2R). All experimental procedures performed until 2013 in Spain complied with the “Real Decreto 1201/2005 de 10 de Octubre” and “Real Decreto 214/1997 de 30 de Julio” on animal protection for experimental purposes, and those carried out from 2013 onwards complied with “Real Decreto 53/2013 de 1 de febrero” on protection of animals used for experimentation and other scientific purposes, which is based on European Directive 2010/63/EU on Laboratory Animal Protection. Competing interests Authors H.E. and J.M.C. are employed by the commercial company ATLAS Molecular Pharma SL. This does not alter our adherence to the Journal’s policies on sharing data and materials and did not influence in any way the work reported in this manuscript, given that the company had no role in study design, funding, and data analysis. The rest of the authors declare no competing interests. Author details 1CIMUS Biomedical Research Institute, University of Santiago de Compostela-IDIS, Santiago de Compostela, Spain 2Prion Research Lab, CIC BioGUNE, Basque Research and Technology Alliance (BRTA), Derio, Spain 3Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Carlos III National Health Institute, Madrid, Spain 4ATLAS Molecular Pharma S. L, Derio, Spain 5Unitat mixta d’Investigació IRTA-UAB en Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Campus de la Universitat Autònoma de Barcelona (UAB), Bellaterra, Catalonia, Spain 6IRTA. Programa de Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Universitat Autònoma de Barcelona (UAB), Campus, Bellaterra, Catalonia, Spain 7Proteomics Unit, Instituto de Investigaciones Sanitarias de Santiago (IDIS), Santiago de Compostela, Spain 8Animal Health Department, NEIKER-Basque Institute for Agricultural Research and Development , Basque Research and Technology Alliance (BRTA), Derio, Spain 9IKERBASQUE, Basque Foundation for Science, Bilbao, Spain Received: 12 June 2025 / Accepted: 23 October 2025 References 1. Prusiner SB, Prions (1998) Proc Natl Acad Sci 95:13363–13383 2. Aguzzi A, Calella AM (2009) Prions: protein aggregation and infectious diseases. Physiol Rev 89:1105–1152 3. Prusiner SB (1982) Novel proteinaceous infectious particles cause scrapie. Science 216:136–144 4. Mabbott NA (2017) Chapter ten immunology of prion protein and prions. Prog Mol Biol Transl Sci 150:203–240 5. Prusiner SB, DeArmond SJ (1991) Molecular biology and pathology of scrapie and the prion diseases of humans. Brain Pathol 1:297–310 6. Kraus A, Hoyt F, Schwartz CL, Hansen B, Artikis E, Hughson AG et al (2021) High-resolution structure and strain comparison of infectious mammalian prions. Mol Cell 81:4540–4551e6 Page 21 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 7. Hoyt F, Alam P, Artikis E, Schwartz CL, Hughson AG, Race B et al (2022) CryoEM of prion strains from the same genotype of host identifies conformational determinants. PLOS Pathog 18:e1010947 8. Hoyt F, Standke HG, Artikis E, Schwartz CL, Hansen B, Li K et al (2022) Cryo-EM structure of anchorless RML prion reveals variations in shared motifs between distinct strains. Nat Commun 13:4005 9. Manka SW, Zhang W, Wenborn A, Betts J, Joiner S, Saibil HR et al (2022) 2.7 Å cryo-EM structure of ex vivo RML prion fibrils. Nat Commun 13:4004 10. Manka SW, Wenborn A, Betts J, Joiner S, Saibil HR, Collinge J et al (2023) A structural basis for prion strain diversity. Nat Chem Biol 19:607–613 11. Prusiner SB, Scott MR, DeArmond SJ, Cohen FE (1998) Prion Protein Biology Cell 93:337–348 12. Donne DG, Viles JH, Groth D, Mehlhorn I, James TL, Cohen FE et al (1997) Structure of the Recombinant full-length hamster prion protein PrP(29–231): the N terminus is highly flexible. Proc Natl Acad Sci 94:13452–13457 13. Hornemann S, Korth C, Oesch B, Riek R, Wider G, Wüthrich K et al (1997) Recombinant full-length murine prion protein, mPrP(23–231): purification and spectroscopic characterization. FEBS Lett 413:277–281 14. Prion (2024) Conference Abstracts. Prion. 2024;18:1–47 15. O’Brien EP, Okamoto Y, Straub JE, Brooks BR, Thirumalai D (2009) Thermodynamic perspective on the Dock – Lock growth mechanism of amyloid fibrils. J Phys Chem B 113:14421–14430 16. Guijarro JI, Sunde M, Jones JA, Campbell ID, Dobson CM (1998) Amyloid fibril formation by an SH3 domain. Proc Natl Acad Sci 95:4224–4228 17. GRIFFITH JS (1967) Nature of the scrapie agent: Self-replication and scrapie. Nature 215:1043–1044 18. Come JH, Fraser PE, Lansbury PT (1993) A kinetic model for amyloid formation in the prion diseases: importance of seeding. Proc Natl Acad Sci 90:5959–5963 19. Legname G, Baskakov IV, Nguyen H-OB, Riesner D, Cohen FE, DeArmond SJ et al (2004) Synth Mammalian Prions Sci 305:673–676 20. Wang L-Q, Zhao K, Yuan H-Y, Wang Q, Guan Z, Tao J et al (2020) Cryo-EM structure of an amyloid fibril formed by full-length human prion protein. Nat Struct Mol Biol 27:598–602 21. Wang L-Q, Zhao K, Yuan H-Y, Li X-N, Dang H-B, Ma Y et al (2021) Genetic prion disease–related mutation E196K displays a novel amyloid fibril structure revealed by cryo-EM. Sci Adv 7:eabg9676 22. Deleault NR, Harris BT, Rees JR, Supattapone S (2007) Formation of native prions from minimal components in vitro. Proc Natl Acad Sci 104:9741–9746 23. Wang F, Wang X, Yuan C-G, Ma J (2010) Generating a prion with bacterially expressed Recombinant prion protein. Science 327:1132–1135 24. Deleault NR, Walsh DJ, Piro JR, Wang F, Wang X, Ma J et al (2012) Cofactor molecules maintain infectious conformation and restrict strain properties in purified prions. Proc Natl Acad Sci [Internet]. ;109:E1938–46. Available from: http:// www.pna s.org/c gi/d oi/ h t t p s : / / d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 1 2 0 6 9 9 9 1 0 9 25. Supattapone S (2014) Synthesis of High Titer Infectious Prions with Cofactor Molecules*. J Biol Chem [Internet]. ;289:19850–4. Available from: h t t p : / / p u b m e d . g o v / 2 4 8 6 0 0 9 7 26. Miller MB, Wang DW, Wang F, Noble GP, Ma J, Woods VL et al (2013) Cofactor Molecules Induce Structural Transformation during Infectious Prion Formation. Structure [Internet]. ;21:2061–8. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . s t r . 2 0 1 3 . 0 8 . 0 2 5 27. Eraña H, Charco JM, Bari MAD, Díaz-Domínguez CM, López-Moreno R, Vidal E et al (2019) Development of a new largely scalable in vitro prion propagation method for the production of infectious recombinant prions for high resolution structural studies. PLoS Pathog [Internet]. ;15:e1008117. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p p a t . 1 0 0 8 1 1 7 28. Fernández-Borges N, Bari MAD, Eraña H, Sánchez-Martín M, Pirisinu L, Parra B et al (2017) Cofactors influence the biological properties of infectious recombinant prions. Acta Neuropathol [Internet]. ;135:179–99. Available from: http:// link.sp ringer. com/ h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 0 0 4 0 1 - 0 1 7 - 1 7 8 2 - y 29. Eraña H, Díaz-Domínguez CM, Charco JM, Vidal E, González-Miranda E, PérezCastro MA et al (2023) Understanding the key features of the spontaneous formation of Bona Fide prions through a novel methodology that enables their swift and consistent generation. Acta Neuropathol Commun 11:145 30. Eraña H, Sampedro-Torres-Quevedo C, Charco JM, Díaz-Domínguez CM, Peccati F, San-Juan-Ansoleaga M et al (2024) A protein misfolding shaking Amplification-based method for the spontaneous generation of hundreds of Bona Fide prions. Nat Commun 15:2112 31. Fernández-Borges N, Eraña H, Elezgarai SR, Harrathi C, Venegas V, Castilla J, Prions (2017) Methods and Protocols. Methods Mol Biol [Internet]. ;1658:205–16. Available from: http://pubmed.gov/28861792 32. Pérez-Castro MÁ, Eraña H, Vidal E, Charco JM, Lorenzo NL, Gonçalves-Anjo N et al (2025) Cofactors facilitate Bona Fide prion misfolding in vitro but are not necessary for the infectivity of Recombinant murine prions. PLOS Pathog 21:e1012890 33. Vidal E, Sánchez-Martín MA, Eraña H, Lázaro SP, Pérez-Castro MA, Otero A et al (2022) Bona Fide atypical scrapie faithfully reproduced for the first time in a rodent model. Acta Neuropathol Commun 10:179 34. Aguilar-Calvo P, Sevillano AM, Bapat J, Soldau K, Sandoval DR, Altmeppen HC et al (2019) Shortening heparan sulfate chains prolongs survival and reduces parenchymal plaques in prion disease caused by mobile, ADAM10-cleaved prions. Acta Neuropathol [Internet]. ;139:527–46. Available from: h t t p : / / l i n k . s p r i n g e r . c o m / h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 0 0 4 0 1 - 0 1 9 - 0 2 0 8 5 - x 35. Vázquez-Fernández E, Alonso J, Pastrana MA, Ramos A, Stitz L, Vidal E et al (2012) Structural Organization of Mammalian Prions as Probed by Limited Proteolysis. Baskakov IV, editor. PLoS ONE [Internet]. ;7:e50111. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p o n e . 0 0 5 0 1 1 1 36. Silva CJ, Vázquez-Fernández E, Onisko B, Requena JR (2015) Proteinase K and the structure of prpsc: the good, the bad and the ugly. Virus Res 207:120–126 37. Sevillano AM, Fernández-Borges N, Younas N, Wang F, Elezgarai SR, Bravo S et al (2018) Recombinant PrPSc shares structural features with brain-derived prpsc: insights from limited proteolysis. PLoS Pathog 14:e1006797 38. Sajnani G, Pastrana MA, Dynin I, Onisko B, Requena JR (2008) Scrapie prion protein structural constraints obtained by limited proteolysis and mass spectrometry. J Mol Biol 382:88–98 39. Zanusso G, Farinazzo A, Prelli F, Fiorini M, Gelati M, Ferrari S et al (2004) Identification of distinct N-terminal truncated forms of prion protein in different Creutzfeldt-Jakob disease Subtypes*. J Biol Chem 279:38936–38942 40. Kocisko DA, Lansbury PT, Caughey B (1996) Partial unfolding and refolding of Scrapie-Associated prion protein: evidence for a critical 16-kDa C-Terminal domain †. Biochemistry 35:13434–13442 41. Zou W-Q, Capellari S, Parchi P, Sy M-S, Gambetti P, Chen SG (2003) Identification of novel proteinase K-resistant C-terminal fragments of PrP in Creutzfeldt-Jakob Disease*. J Biol Chem 278:40429–40436 42. Weissmann C, Li J, Mahal SP, Browning S (2011) Prions on the move. EMBO Rep [Internet]. ;12:1109–17. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 0 3 8 / e m b o r . 2 0 1 1 . 1 9 2 43. Rezaei H, Martin D, Herzog L, Reine F, Moreno AM, Moudjou M et al (2024) Species barrier as molecular basis for adaptation of synthetic prions with N-terminally truncated PrP. FEBS J 291:5051–5076 44. Klimova N, Makarava N, Baskakov IV (2015) The diversity and relationship of prion protein self-replicating states. Virus Res [Internet]. ;207:113–9. Available from: h t t p : / / l i n k i n g h u b . e l s e v i e r . c o m / r e t r i e v e / p i i / S 0 1 6 8 1 7 0 2 1 4 0 0 4 0 6 7 45. Makarava N, Kovacs GG, Bocharova O, Savtchenko R, Alexeeva I, Budka H et al (2010) Recombinant prion protein induces a new transmissible prion disease in wild-type animals. Acta Neuropathol 119:177–187 46. Makarava N, Kovacs GG, Savtchenko R, Alexeeva I, Budka H, Rohwer RG et al (2011) Genesis of Mammalian Prions: From Non-infectious Amyloid Fibrils to a Transmissible Prion Disease. Supattapone S, editor. PLoS Pathog [Internet]. 7:e1002419. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p p a t . 1 0 0 2 4 1 9 47. Deleault NR, Piro JR, Walsh DJ, Wang F, Ma J, Geoghegan JC et al (2012) Isolation of phosphatidylethanolamine as a solitary cofactor for prion formation in the absence of nucleic acids. Proc Natl Acad Sci. 109:8546–51. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 1 2 0 4 4 9 8 1 0 9 48. Timmes AG, Moore RA, Fischer ER, Priola SA (2013) Recombinant Prion Protein Refolded with Lipid and RNA Has the Biochemical Hallmarks of a Prion but Lacks In Vivo Infectivity. Chiesa R, editor. PLoS ONE [Internet]. ;8:e71081. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p o n e . 0 0 7 1 0 8 1 49. Wang F, Wang X, Orrú CD, Groveman BR, Surewicz K, Abskharon R et al (2017) Self-propagating, protease-resistant, Recombinant prion protein conformers with or without in vivo pathogenicity. PLoS Pathog 13:e1006491 50. Notari S, Capellari S, Langeveld J, Giese A, Strammiello R, Gambetti P et al (2007) A refined method for molecular typing reveals that co-occurrence of PrPSc types in Creutzfeldt–Jakob disease is not the rule. Lab Investig 87:1103–1112 51. Polymenidou M, Stoeck K, Glatzel M, Vey M, Bellon A, Aguzzi A (2005) Coexistence of multiple PrPSc types in individuals with Creutzfeldt-Jakob disease. Lancet Neurol 4:805–814 52. Wong C, Xiong L, Horiuchi M, Raymond L, Wehrly K, Chesebro B et al (2001) Sulfated glycans and elevated temperature stimulate PrPSc-dependent cell‐free formation of protease‐resistant prion protein. EMBO J [Internet]. ;20:377–86. Available from: h t t p : / / e m b o j . e m b o p r e s s . o r g / c g i / d o i / h t t p s : / / d o i . o r g / 1 0 . 1 0 9 3 / e m b o j / 2 0 . 3 . 3 7 7 Page 22 of 22Lorenzo et al. Acta Neuropathologica Communications (2025) 13:249 53. Caughey B, Brown K, Raymond GJ, Katzenstein GE, Thresher W (1994) Binding of the protease-sensitive form of PrP (prion protein) to sulfated glycosaminoglycan and congo red [corrected]. J Virol 68:2135–2141 54. Walsh DJ, Schwind AM, Noble GP, Supattapone S (2023) Conformational diversity in purified prions produced in vitro. PLOS Pathog 19:e1011083 55. Cohlberg JA, Li J, Uversky VN, Fink AL (2002) Heparin and other glycosaminoglycans stimulate the formation of amyloid fibrils from α-Synuclein in vitro †. Biochemistry 41:1502–1511 56. Tao Y, Sun Y, Lv S, Xia W, Zhao K, Xu Q et al (2022) Heparin induces α-synuclein to form new fibril polymorphs with attenuated neuropathology. Nat Commun 13:4226 57. Townsend D, Fullwood NJ, Yates EA, Middleton DA (2020) Aggregation kinetics and filament structure of a Tau fragment are influenced by the sulfation pattern of the cofactor heparin. Biochemistry 59:4003–4014 58. Scheres SH, Zhang W, Falcon B, Goedert M (2020) Cryo-EM structures of Tau filaments. Curr Opin Struct Biol 64:17–25 59. Yang Y, Shi Y, Schweighauser M, Zhang X, Kotecha A, Murzin AG et al (2022) Structures of α-synuclein filaments from human brains with lewy pathology. Nature 610:791–795 60. González-Iglesias R, Pajares MA, Ocal C, Espinosa JC, Oesch B, Gasset M (2002) Prion protein interaction with glycosaminoglycan occurs with the formation of oligomeric complexes stabilized by Cu(II) bridges. J Mol Biol 319:527–540 61. Forloni G, Chiesa R, Bugiani O, Salmona M, Tagliavini F (2019) Review: PrP 106126-25 years after. Neuropathol Appl Neurobiol 45:430–440 62. Tycko R, Savtchenko R, Ostapchenko VG, Makarava N, Baskakov IV (2010) The α-Helical C-Terminal domain of Full-Length Recombinant PrP converts to an In-Register parallel β-Sheet structure in PrP fibrils: evidence from solid state nuclear magnetic resonance. Biochemistry 49:9488–9497 63. Müller H, Brener O, Andreoletti O, Piechatzek T, Willbold D, Legname G et al (2014) Progress towards structural understanding of infectious sheep PrPamyloid. Prion [Internet]. ;8:344–58. Available from: h t t p : / / p u b m e d . g o v / 2 5 4 8 2 5 9 6 64. Vieira TCRG, Reynaldo DP, Gomes MPB, Almeida MS, Cordeiro Y, Silva JL (2011) Heparin binding by murine Recombinant prion protein leads to transient aggregation and formation of RNA-Resistant species. J Am Chem Soc 133:334–344 65. Caughey B, Raymond GJ (1993) Sulfated polyanion inhibition of scrapieassociated PrP accumulation in cultured cells. J Virol [Internet]. ;67:643–50. Available from: h t t p : / / j v i . a s m . o r g / c o n t e n t / 6 7 / 2 / 6 4 3 . s h o r t 66. Horonchik L, Tzaban S, Ben-Zaken O, Yedidia Y, Rouvinski A, Papy-Garcia D et al (2005) Heparan sulfate is a cellular receptor for purified infectious Prions. J Biol Chem 280:17062–17067 67. Ben-Zaken O, Tzaban S, Tal Y, Horonchik L, Esko JD, Vlodavsky I et al (2003) Cellular Heparan sulfate participates in the metabolism of Prions. J Biol Chem 278:40041–40049 68. McBride PA, Wilson MI, Eikelenboom P, Tunstall A, Bruce ME (1998) Heparan sulfate proteoglycan is associated with amyloid plaques and neuroanatomically targeted PrP pathology throughout the incubation period of ScrapieInfected mice. Exp Neurol 149:447–454 69. Snow AD, Wight TN, Nochlin D, Koike Y, Kimata K, DeArmond SJ et al (1990) Immunolocalization of Heparan sulfate proteoglycans to the prion protein amyloid plaques of Gerstmann-Straussler syndrome, Creutzfeldt-Jakob disease and scrapie. Lab Investig J Tech Methods Pathol 63:601–611 70. Adjou KT, Simoneau S, Salès N, Lamoury F, Dormont D, Papy-Garcia D et al (2003) A novel generation of heparan sulfate mimetics for the treatment of prion diseases. J Gen Virol [Internet]. ;84:2595–603. Available from: h t t p : / / v i r . s g m j o u r n a l s . o r g / c g i / d o i / h t t p s : / / d o i . o r g / 1 0 . 1 0 9 9 / v i r . 0 . 1 9 0 7 3 - 0 71. Doh-ura K, Ishikawa K, Murakami-Kubo I, Sasaki K, Mohri S, Race R et al (2004) Treatment of Transmissible Spongiform Encephalopathy by Intraventricular Drug Infusion in Animal Models. J Virol [Internet]. ;78:4999–5006. Available from: http:// jvi.asm .org/cg i/do i/ h t t p s : / / d o i . o r g / 1 0 . 1 1 2 8 / J V I . 7 8 . 1 0 . 4 9 9 9 - 5 0 0 6 . 2 0 0 4 72. Farquhar CF, Dickinson AG (1986) Prolongation of scrapie incubation period by an injection of dextran sulphate 500 within the month before or after infection. J Gen Virol. 67:463–473 73. Ladogana A, Casaccia P, Ingrosso L, Cibati M, Salvatore M, Xi Y et al (1992) Sulphate polyanions prolong the incubation period of scrapie-infected hamsters. J Gen Virol. 73:661–5. Available from: h t t p s : / / d o i . o r g / 1 0 . 1 0 9 9 / 0 0 2 2 - 1 3 1 7 - 7 3 - 3 - 6 6 1 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.