Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prion-induced photoreceptor degeneration begins with misfolded prion protein accumulation in cones at two distinct sites: cilia and ribbon synapses

Prion-induced photoreceptor degeneration begins with misfolded prion protein accumulation in... Accumulation of misfolded host proteins is central to neuropathogenesis of numerous human brain diseases includ- ing prion and prion-like diseases. Neurons of retina are also affected by these diseases. Previously, our group and others found that prion-induced retinal damage to photoreceptor cells in mice and humans resembled pathology of human retinitis pigmentosa caused by mutations in retinal proteins. Here, using confocal, epifluorescent and electron microscopy we followed deposition of disease-associated prion protein (PrPSc) and its association with damage to critical retinal structures following intracerebral prion inoculation. The earliest time and place of retinal PrPSc deposi- tion was 67 days post-inoculation (dpi) on the inner segment (IS) of cone photoreceptors. At 104 and 118 dpi, PrPSc was associated with the base of cilia and swollen cone inner segments, suggesting ciliopathy as a pathogenic mecha- nism. By 118 dpi, PrPSc was deposited in both rods and cones which showed rootlet damage in the IS, and photore- ceptor cell death was indicated by thinning of the outer nuclear layer. In the outer plexiform layer (OPL) in uninfected mice, normal host PrP (PrPC) was mainly associated with cone bipolar cell processes, but in infected mice, at 118 dpi, PrPSc was detected on cone and rod bipolar cell dendrites extending into ribbon synapses. Loss of ribbon synapses in cone pedicles and rod spherules in the OPL was observed to precede destruction of most rods and cones over the next 2–3 weeks. However, bipolar cells and horizontal cells were less damaged, indicating high selectivity among neurons for injury by prions. PrPSc deposition in cone and rod inner segments and on the bipolar cell processes par- ticipating in ribbon synapses appear to be critical early events leading to damage and death of photoreceptors after prion infection. These mechanisms may also occur in human retinitis pigmentosa and prion-like diseases, such as AD. Keywords: Prion, Prion-like, Ribbon synapses, Retinitis pigmentosa, Alzheimer, Parkinson, Ciliopathy, Scrapie, Necrosis, Apoptosis Introduction Prion diseases are progressive neurodegenerative dis- eases which affect humans as well as numerous wild and domestic animal species. These diseases are characterized by vacuolar degeneration of the grey matter and gliosis primarily in grey matter of the CNS. In addition, abnor- *Correspondence: bchesebro@niaid.nih.gov mal aggregation of a host prion protein (PrPC) leads to Laboratory of Persistent Viral Diseases, Rocky Mountain Laboratories, deposition of a protease-resistant PrP isoform (PrPSc) National Institute of Allergy and Infectious Diseases, National Institutes in the nervous system and other organs, which results in of Health, 903 South Fourth Street, Hamilton, MT 59840, USA Full list of author information is available at the end of the article damage by unclear mechanisms [25]. Protein aggregation © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, 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 changes were made. 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 http://creat iveco mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Striebel et al. acta neuropathol commun (2021) 9:17 Page 2 of 26 in prion disease proceeds by a process referred to as Photoreceptors are also damaged in human retinitis seeded polymerization or PrP conversion, where initial pigmentosa, which is a major cause of human blindness small aggregates of PrPSc catalyze extension of protein resulting in a retinal pathology similar to prion diseases. aggregation to generate additional PrPSc [9]. In some forms of retinitis pigmentosa, microglia are A similar seeded polymerization mechanism has been known to become activated by the misfolding of mutant noted in several other more common neurodegenera- host proteins such as rhodopsin, and microglia have also tive diseases including Alzheimer’s disease, Parkinson’s been suspected to be important in the pathogenic pro- disease and tauopathies, but in these diseases, the aggre- cess [40, 46]. In prion diseases, microglia activated by the gated host proteins are amyloid beta (Aβ), α-synuclein deposition of aggregated prion protein have also been and tau respectively [4, 11, 26]. Because of the similarities suspected to be a possible mechanism of pathogenesis in the protein aggregation process in these diseases and [18]. However, in our recent studies using PLX5622 to prion diseases, they have been referred to as “prion-like” eliminate microglia in  vivo, removal of microglia led to diseases. Some people have suspected that prion diseases a decrease in survival time due to accelerated brain prion and prion-like diseases might be susceptible to simi- pathogenesis in mice [7]. Thus, microglia appeared to be lar therapeutic interventions [22, 51]. Thus, at this time, mainly helpful, not harmful, to the host during prion dis- there is high interest in the mechanisms of pathogenesis ease. Furthermore, in our studies of retinal prion disease of all these diseases. in mice, elimination of microglia gave similar results, Prion and “prion-like” diseases are known to cause reti- showing accelerated retinal degeneration when microglia nal damage in humans and other species, but each dis- were reduced or absent [54]. Müller glial cells are also ease affects the retina in unique ways [44]. In Alzheimer’s activated during prion disease but are likely a response disease patients, Aβ deposition has been associated with to damage rather than the cause of pathogenesis [28, degeneration of the retinal ganglion cell layer, photore- 29, 58]. u Th s, in vivo damage induced by prion infection ceptors and the retinal pigmented epithelium [3, 45]. Rat- might be due to direct effects of aggregated prion protein nayaka et  al. have also shown that Aβ plaques may be a on photoreceptor cells rather than indirect effects of acti - key factor in Age-related Macular Degeneration (AMD). vated microglia or astroglia. In AD, phosphorylated Tau deposits have also been In the present paper, we studied direct early events observed from the outer plexiform layer to the ganglion of prion protein deposition in mouse retina following cell layer [31, 32]. Likewise, abnormal α-synuclein aggre- intracerebral prion injection. For these studies, we used gates have been detected in retina of Parkinson’s disease confocal and epifluorescence microscopy with detec - patients, and these are implicated in the degeneration of tion of multiple targets on the same section to localize the nerve fiber layer (ganglion cell axons), ganglion cell PrP in its normal and disease-associated forms, while body layer and inner plexiform layer [57]. also observing multiple subcellular components of reti- In prion disease, results from our lab and others have nal rods, cones, bipolar cells and horizontal cells. Ultra- confirmed deposition of disease-associated PrP (PrPSc) structural studies were also done to confirm some of in human, bovine, primate, ovine, cervid, and rodent the observations. The results showed early deposition of retina by RT-QuIC, western blot and/or immunohis- PrPSc in two distinct, subcellular areas of retina as well tochemistry [5, 20, 21, 23, 24, 29, 41, 52, 54, 55, 58, 60]. as interesting novel events of photoreceptor damage in These studies represent examples from both natural dis - both these areas which preceded the death of both rods ease (human, elk, deer, sheep) and experimental disease and cones while sparing other nearby cell types. (primate, bovine, sheep, rodent). In these studies, dam- age appeared to affect primarily photoreceptor rods and Materials and methods cones, and damage to other retinal neuronal populations Ethics statement was not clear [50]. In our previous work, mouse photore- All mice were housed at the Rocky Mountain Laborato- ceptors were shown to die mainly by the process of apop- ries (RML) in an AAALAC-accredited facility in compli- tosis which coincided with PrPSc deposition [29, 54]. ance with guidelines provided by the Guide for the Care Previously, defective iron transport has been proposed and Use of Laboratory Animals (Institute for Laboratory as a mechanism of damage in prion-infected eye [47]. Animal Research Council). Experimentation followed Since photoreceptors rely heavily on iron-containing RML Animal Care and Use Committee approved proto- enzymes for the biochemical reactions of the phototrans- col 2016-042. duction pathways, disruption of iron transport by prion infection may also play a role in retinal damage. Such a Mice mechanism might explain the selective damage to photo- Retinas used in these experiments were obtained receptor cells by prions. from two strains of mice, C57BL/10SnJ and S triebel et al. acta neuropathol commun (2021) 9:17 Page 3 of 26 C57BL/6  J-TgGFP/RFP. 79A scrapie-induced retinal 95  °C. Staining for PrP was done using human anti- degeneration is very similar in these strains and was PrP monoclonal antibody D13 [33] which was obtained characterized in our previous publication [54]. An in- from tissue culture supernatants made in our labora- house breeding colony supplied C57BL/10SnJ mice,and tory from CHO cells expressing the D13 antibody con- C57BL/6 J–TgGFP/RFP were also bred in-house as pre- struct, which were kindly provided by Dr. R. Anthony viously described [54]. The GFP and RFP fluorescence Williamson, The Scripps Research Institute, La Jolla, was not important to the questions addressed here. All CA. D13 culture fluid was used at a dilution of 1:100 mice were group housed in transparent cages in a 12 h (diluted in PBS with 1% normal goat serum and 0.1% light (250-300lux) /12 h dark cycle and food and water Triton X-100) for 2 h at 37 °C. The secondary antibody were available ad libitium. was biotinylated goat anti-human IgG at 1:500 dilu- tion (Jackson ImmunoResearch, West Grove, PA.), and avidin-horseradish peroxidase was used with DAB as Scrapie inoculation model chromogen (DAB Map kit; Ventana Medical Systems, Scrapie inoculations were carried out as previously Tucson, AZ.). described [54]. Briefly, mice (4–6  weeks old) were For immunofluorescent staining, antigen retrieval injected intracerebrally (i.c.) in the left hemisphere with for all targets was performed using a Biocare Medi- 30  μl of a 1% (wt/vol) dilution of brain homogenate cal DC2002 Decloaking chamber with sodium citrate pools from C57BL mice terminally ill from 79A scra- buffer at pH 6.0(0.01  M) for 20  min at 120° C / 20 PSI pie. Brain homogenates contained 1.0 × 10 ID50 / 30 and cooled to 50 °C. For each of the following steps, 250– ul after they were diluted for inoculation in phosphate- 300  µl of solution was applied to each slide and covered buffered balanced saline (PBBS) pH 7.2, supplemented with a temporary plastic coverslip and incubated for a set with 2% fetal bovine serum (Hyclone, Logan, UT). amount of time. Tissues were blocked first with a normal The course of disease, details on the clinical symptoms donkey serum blocking solution (2% donkey serum, 1% and retinal degeneration were previously well docu- BSA, 0.1% Triton X-100, 0.05% Tween 20 in 0.01 M PBS) mented in the mouse strains used in this study in our pre- for 1  h at room temperature and then in 0.1  M Glycine vious publication [54]. Briefly, in the 79A mouse-adapted in 0.01 M PBS for 30 min at room temperature. Primary scrapie model, mice begin showing clinical signs con- antibodies (Table 1.) were diluted in donkey serum block- sistent with scrapie between 105 and 120dpi and reach ing solution and applied for 1  h at room temperature. clinical endpoint disease at approximately 160dpi. Thin - Alexafluor (ThermoFisher) secondary antibodies were ning of the retina begins around 118dpi and likely causes diluted to 1:250 in donkey serum solution and applied blindness by the disease endpoint, though this diagnosis for 1  h. In dual or triple stainings, primary antibodies is difficult to determine without conducting further tests. were applied simultaneously, as were secondary antibod- The 79A scrapie strain was previously compared to seven ies. After each antibody incubation, slides were washed other strains of mouse-adapted scrapie and shown to be 3 times in 1X PBS for 10 min. Coverslips were mounted the most retinal-tropic [15]. At pre-clinical and clinical with ProLong Gold with DAPI (Life Technologies) and time-points, mice were euthanized by isoflurane anesthe - examined and photographed using an Olympus BX51 sia overdose followed by perfusion with 10 ml of saline. microscope/Olympus CellSens software or using a confo- cal microscope as described below. Immunohistochemistry and Immunofluorescence For immunohistochemistry and immunofluorescence, Numbers of mice studied eyes were removed, placed in 10% neutral buffered for - The numbers of animals analyzed at each timepoint are malin for 3 to 5 days and then processed by dehydration presented in Table  2. Uninfected mice used as controls and embedded in paraffin as a single block. Next, 5 μm were of similar age to experimental animals, evidence of sections were cut using a standard Leica microtome, age-related retinal changes was not observed in the age placed on positively charged glass slides, and air-dried range of control animals used. overnight at room temperature. The following day slides were heated in an oven at 60  °C for 20  min. A Nomenclature and detection of PrP, PrPC and PrPSc Ventana automated Discovery XT stainer was used for Monoclonal antibody D13 was used in immunostain- deparaffinization, antigen retrieval and immunohisto - ing of tissue sections to detect PrP. In tissues of unin- chemical staining. fected mice, PrP detected was assumed to be the normal For immunohistochemical staining of PrP antigens PrP isoform, PrPC. In infected tissues, PrP detected were exposed by incubation in CC1 buffer (Ventana) in locations different from those seen uninfected mice containing Tris–Borate-EDTA, pH 8.0 for 100  min at was assumed to be disease-associated PrPSc, and PrP Striebel et al. acta neuropathol commun (2021) 9:17 Page 4 of 26 Table 1 Primary Antibodies used in immunofluorescent staining Antibody Specificity (antigen/cell type or structure) Dilution Host species Source D13 Prion protein 1:100 Human Ref, Matsunaga Cone Arrestin Arrestin 3/Cone photoreceptors 1:100 Rabbit Millipore, AB15282 Cone Opsin Red, Green opsins/Cone photoreceptors 1:100 Rabbit Chemicon, AB5404 Rhodopsin Rhodopsin/Rod photoreceptors 1:100 Rabbit Millipore, MABN15 GNAT1 G protein subunit alpha transducin 1 1:100 Rabbit Abcam, ab74059 /Rod photoreceptors GNAT2 G protein subunit alpha transducin 2 /Cone photoreceptors 1:100 Rabbit Thermofisher, PA5-22,340 GLUT1 Glucose transporter 1/Cell membranes of many cell types 1:100 Rabbit Abcam, ab115730 CtBP2 C-terminal binding protein 2/ribeye protein of ribbon synapses 1:100 Rabbit, mouse Invitrogen, PA-79086 Santa Cruz, sc-17759 PKCα Protein kinase C/Rod bipolar cells 1:100 Rabbit, mouse Invitrogen PA5-17,551, MA1-157 Calbindin Calbindin/Horizontal cells 1:100 Rabbit Abcam, ab108404 Rootletin Rootletin/photoreceptor rootlets 1:50 Mouse Millipore, ABN1714 Centrin3 Centrin3/cilia and basal bodies in photoreceptor inner segments 1:100 Rabbit Thermofisher, PA5-35,865 Secretagogin (SCGN) Cone bipolar cells 1:100 Rabbit Thermofisher, PA5-30,393 SCGN marks 8 of the 12 subtypes of mouse cone bipolar cells [13, 42] detected in similar locations to those found in uninfected 1.518. Image acquisition settings including laser power mice was assumed to be either or both isoforms. and gain were optimized for minimal background and cross-talk, and kept constant within an experiment for Quantification of bipolar and horizontal cells all timepoints and samples to enable direct comparisons. To quantify rod bipolar cells throughout the timecourse Stacks were collected with a lateral resolution of 43  nm of disease, two sections of retina from a mouse at each and z-spacing of 130  nm except for: quantification of timepoint were stained with DAPI, anti-PKCα primary anti-D13 and anti-CtBP2 signals, where stacks were col- antibody and secondary antibody Alexa Fluor 488 as lected with a lateral resolution of 71 nm and z-spacing of described above. The PKCα-positive rod bipolar cell 367  nm, with five representative fields of view acquired bodies were counted in four 20X fields per timepoint for each timepoint, and anti-Cone Opsin, anti-GNAT1, and averaged. Horizontal cell numbers were determined anti-Cone Arrestin, anti-Centrin3 which were collected by staining retinal sections with DAPI, anti-calbindin with a lateral resolution/z-spacing of 18  nm/250  nm, primary antibody and Alexa Fluor 488 secondary anti- 132  nm/250  nm, 65  nm/500  nm and 70  nm/500  nm body as described above. Calbindin-positive cell bodies respectively. were counted along two entire retinal sections from one mouse per timepoint. Cone bipolar cells were counted by staining retinal sections with anti-secretagogin antibody, Image processing and analysis which labels 8 of the 12 types of cone bipolar cells [13, Image stacks were exported from ZEN software and 42] and counting cell bodies on two retinal sections from deconvolved with Huygens Professional v. 20.04 (Scien- at least one mouse per timepoint (see figure legend for tific Volume Imaging, The Netherlands) using the CMLE n values). One-way ANOVA statistical analysis was per- algorithm, with SNR = 20 and a maximum of 40 itera- formed using GraphPad Prism software. tions. The deconvolved datasets were imported to Ima - ris x86_64 v.9.5.1 (Bitplane AG, Zürich, Switzerland) for segmentation, surface rendering, visualization, and quan- Confocal microscopy tification. The average number of CtBP2 ribbons based All samples were handled and chemically decontami- on anti-CtBP2 signal and amount of total integrated anti- nated according to established scrapie protocols in con- D13 signal was calculated per micron length of retina. sultation with RML Biosafety. Samples were imaged Data were imported into Microsoft Excel for compila- using a Zeiss laser scanning confocal (LSM 880) micro- tion, and statistical analysis was performed using Graph- scope driven by ZEN v.2.3 software (Carl Zeiss Micros- Pad Prism v 8.3.0 (La Jolla, CA). copy). A Plan Apochromat 63X/NA1.4 oil immersion lens was used, with immersion oil at a refractive index of S triebel et al. acta neuropathol commun (2021) 9:17 Page 5 of 26 Table 2 Number of retinas analyzed by immunofluorescence at timepoints during disease a b TimepointAntigen tested PrP (D13) Cone Cone Opsin SCGN CtBP2 PKCα Calbindin Rhodopsin GNAT1 GNAT2 Rootletin Centrin3 Arrestin Early (67,82,104) 9 3 3 3 3 2 1 3 1 1 3 3 Mid (118, 125, 129, 131) 8 3 2 2 2 3 2 3 1 nd 1 2 Late (144, 153, 159, 163, 165) 8 2 nd 2 3 2 1 2 3 nd nd 1 Uninfected 5 2 3 2 2 1 1 2 2 1 2 1 nd not done Timepoints are shown in days post inoculation (dpi) with 79A mouse adapted scrapie. In the 79A mouse-adapted scrapie model, mice begin showing clinical signs consistent with scrapie around 105-120dpi and reach clinical endpoint disease at approximately 160dpi. Thinning of the retina begins around 118dpi and likely causes blindness by the disease endpoint. Antigens detected with antibodies described in Table 1 Number of mice tested with each antibody at timepoint range shown Data not shown Striebel et al. acta neuropathol commun (2021) 9:17 Page 6 of 26 Electron microscopy sample preparation the optic tract and optic nerve [6, 16]. Progression of C57BL/10SnJ mice were perfused with 2% paraformalde- retinal infection was followed by immunohistochemistry hyde + 2% glutaraldehyde in 0.1 M Sorensen’s phosphate (IHC) or immunofluorescence (IF) with anti-PrP mono - buffer (Electron Microscopy Sciences, Pennsylvania). clonal antibody D13 to detect PrP. In uninfected PrPKO Eyes were enucleated and placed in fresh fixative for at mice, no PrP signal was detected by IHC or IF (Fig.  1a, least 30  min before further for processing. The ante - b). However, in uninfected mice expressing PrP, stain- rior portions of eyes were dissected and discarded. The ing of PrP was clearly detectable by IHC and IF in an remaining posterior eye cups were rinsed in phosphate irregular clumpy distribution in the outer plexiform layer buffer, followed by embedment in 2.5% low-melt agar (Precisionary, Massachusetts) made in PBS. 200  µm sections were cut with a VT1000S vibrating blade microtome (Leica Biosystems, Illinois). Sections were processed for transmission electron microscopy as fol- lows: postfixation with 0.05% osmium tetroxide + 0.08% potassium ferrocyanide in 0.1  M phosphate buffer for 1 h, rinsed with buffer, then dehydrated in a graded etha - nol series to 100%, infiltrated with LRWhite (Electron Microscopy Sciences, Pennsylvania) and polymerized overnight in homemade flat-embedding molds covered with aclar sheets at 50 °C in a vacuum oven. For electron microscopy studies the following num- bers of retinas from C57BL/10SnJ mice were taken at the given timepoints; uninfected (n = 22), 84 dpi (n = 2), 89 dpi (n = 4), 98 dpi (n = 2), 104 dpi (n = 3), 112 dpi (n = 2), 122 dpi (n = 4), 126 dpi (n = 2), 132 dpi (n = 3), 137 dpi (n = 3), 140 dpi (n = 1), 151 dpi (n = 3), 154 dpi (n = 1), 165 dpi (n = 3). These retinas were embedded in various resins (Durcupan, Araldite, HM20, and LRWhite) and examined. The LRWhite embedded retinas were selected for the imaging and comparisons shown in this paper. The numbers at each timepoint were; uninfected (n = 4), 104dpi (n = 1), 126dpi (n = 1), 132dpi (n = 1), 137dpi(n = 2), 151dpi(n = 2). Transmission electron microscopy Flat-embedded vibratome sections were excised and super-glued onto resin stubs such that ultramicrotomy sections would be in the desired orientation. 70 nm sec- tions were cut with a Ultracut UCT (Leica Biosystems, Illinois) ultramicrotome and picked up on Formvar coated 100 hex mesh copper grids (Electron Microscopy Sciences, Pennsylvania). Micrographs were acquired on a HT7800 (Hitachi, Oregon) operating at 80  kV with an XR-81B CMOS digital camera (AMT Imaging Systems, Fig. 1 PrPC expression in uninfected PrP knockout (PrPKO) and Massachusetts). wild-type ( WT ) retina using D13 antibody. a, b In PrPKO retina, normal host PrP (PrPC) was not detected by immunofluorescent stain (red) or immunohistochemical stain (brown). However, photoreceptors Results in OS of PrPKO retina showed red autofluorescence, a common but Detection of PrPSc in retina after intracerebral scrapie variable artifact of immunofluorescence studies of retina. c, d In WT injection. retina, PrPC was observed by both staining techniques in the OPL In the current experiments, prion infection of retina and IPL, and weaker levels were detected in the INL and IS. Scale bar was achieved by intracerebral inoculation of mice with in a = 20 µm. OS outer segment, IS inner segment, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, IPL inner 1.0 × 10 ID50 units of scrapie strain 79A. With this plexiform layer method, prions are known to spread to the retina via S triebel et al. acta neuropathol commun (2021) 9:17 Page 7 of 26 Fig. 2 Timecourse of PrP staining with D13 antibody in retina at various times after infection with strain 79A scrapie prions. a Uninfected mouse showing PrPC (magenta) mainly in OPL. b, c At 82 and 104 days post infection (dpi) the misfolded, disease-associated form of PrP (PrPSc) (magenta) can now be seen in the IS and OPL at progressively wider areas. d At 118 dpi, PrPSc is widespread in the IS and deposits are not restricted to discrete individual cells. Some small deposits are visible in the ONL and ONL is beginning to thin as photoreceptors die. e At 131 dpi, PrPSc is deposited in the IS and OPL, and ONL is much thinner. f PrPSc staining appears less at 162 dpi in IS and OPL, and ONL is dramatically thinner. Punctate PrPSc is present in IPL (arrowhead). Scale bar in a = 20 µm (OPL) and in a more diffuse pattern in the inner plexi - form layer (IPL) (Fig. 1c, d). This was similar to what has been described previously by others [10, 17, 19]. In addi- tion, a smooth faint signal was observed in the inner seg- ment (IS) of the photoreceptor layer (Fig.  1c, d) [54]. All of these sites were likely to represent normal cellular PrP (PrPC) as they were not seen in PrPKO mice (Fig. 1a, b). At 82  days post-infection (dpi), bright punctate PrP staining in patchy aggregates of varying sizes were seen in an irregular scattered distribution along the IS and OPL layers (Fig.  2b). This staining was likely to be dis - ease-associated PrPSc, as it was not seen in uninfected mice (Figs. 1c, d, 2a). At 104 and 118 dpi, the PrPSc stain- ing in the IS and OPL was more extensive, and at 118 dpi, PrPSc staining was often seen on the entire circum- ference of the IS and OPL (Fig.  2c, d). At 118 dpi, there was also a decrease in the width of the outer nuclear layer (ONL) indicating a loss of photoreceptor cell nuclei. At 131 dpi, the ONL was further reduced in size, and the PrPSc staining in the IS and OPL was decreased in inten- sity (Fig.  2e). At 162 dpi, all retinal layers were thinned (Fig.  2f ), and the ONL was now only 2–3 cells thick in most areas. Faint PrPSc staining was further decreased in the IS and OPL regions but was now slightly more detect- able in the inner plexiform layer (IPL). Interestingly, this type of neurodegeneration was not seen in uninfected PrPKO mice, which suggested that loss of PrP does not itself cause photoreceptor degeneration (unpublished results from our group). The time course of PrPSc deposition and retinal degen - eration seen in these experiments using indirect IF was similar to our previous data using detection of PrP by IHC [54]. However, the details of the PrPSc aggregate morphology was seen more clearly by IF. Since photore- ceptor rods and cones are the main cells normally pre- sent in the IS region and these cells also have processes ◂ Striebel et al. acta neuropathol commun (2021) 9:17 Page 8 of 26 Fig. 3 PrPSc accumulates first in cone photoreceptor inner segments. a At 67 dpi, rare small punctate PrPSc deposits (arrow) are present on cone photoreceptors marked with Cone Arrestin (green). Magenta in outer segment is autofluorescence of rhodopsin. b Cone opsin (white) marks cone outer segments with PrPSc (magenta) deposits associated mainly with cone inner segments (yellow arrow) at 104 dpi. The transition from faint to intense cone opsin staining marks the boundary between cone IS and cone OS (blue arrow). c Another cone-specific outer segment protein GNAT2 (green) shows obvious connection with PrPSc (magenta) staining cone inner segments at 104 dpi (arrows). d At 104 dpi, rod-specific marker, GNAT1 (green) stains rod inner and outer segments, but spares cones (arrowheads). PrPSc (magenta) accumulations are present in the dark GNAT1-free areas (arrows). e A separate experiment done with 22L scrapie strain shows same association of PrPSc with cone inner segments at 123 dpi, suggesting cone specificity is not strain-specific. Scale bar = 5 µm. a, b, d are confocal z-stacks, c, e are widefield images extending into the OPL, early PrPSc deposition was likely to be associated with one or both of these photoreceptor cell types. To study the cell types associated with PrPSc in this model, we used the IF method with dual staining for PrP and antibodies reactive with cone- or rod-specific proteins (Tables 1, 2). Deposition of PrPSc occurs first in cone photoreceptors At 67 dpi, PrPSc was detected in a few individual cells in the IS region, and this PrPSc was associated with detec- tion of cone arrestin in the same individual cells (Fig. 3a). Similarly, at 104 dpi, PrPSc staining was associated with detection of cone opsin (Fig.  3b) and GNAT2 (trans- ducin alpha-2), another cone-specific marker (Fig.  3c). In both of these examples, the PrPSc staining was in the inner portion of the IS, whereas both the cone opsin and GNAT2 proteins were mostly in the outer segment (OS), i.e. distal to the PrPSc but appearing to be in the same individual cells as the cone-specific marker proteins. In contrast, GNAT1 (transducin alpha-1), a rod-specific protein, did not co-associate with PrPSc at 104 dpi which was located in the dark spaces not stained by GNAT1, i.e. cones (Fig. 3d). Because scrapie strains have been shown to show cell-specific infectivity [8], we tested the 22L strain, which also targeted cone photoreceptors before rods (Fig.  3e). These observations indicated that PrPSc deposition appeared first in cone photoreceptor inner segments. Additional studies showed that rods were also infected starting around 118 dpi (see below). Retinas were next studied by confocal microscopy to look for cone damage by dual staining with anti-PrP and S triebel et al. acta neuropathol commun (2021) 9:17 Page 9 of 26 Fig. 4 Early PrPSc accumulation associated with cone photoreceptor damage. a Uninfected retina shows normal cone morphology and distribution of cone opsin (green) in outer segment. b At 104 dpi early PrPSc deposits are associated with cone inner segments (arrow) and some PrPSc appears to localize with cone opsin (white areas in × 2 magnified inset). c At 118 dpi, a swollen, dystrophic cone with PrPSc is seen (yellow box) and some PrPSc is also present in rods (blue arrow). d Serial confocal sections spaced 0.5 µm apart, magnify the swollen cone inner segment from c and suggests the PrPSc is inside the inner segment along with mistrafficked cone opsin. Scale bar in a and c = 5 µm. Scale bar in d series = 2 µm. a–c are confocal z-stacks anti-cone opsin. At 104 dpi, swelling in the inner segment (Fig.  4d). However, the detection of cone opsin outlin- portions of some PrPSc-positive cones was observed ing the outer edge of the cytoplasm was not seen in cells (Fig. 4b). This swelling was never seen in uninfected reti - lacking PrPSc, suggesting that this unusual distribution nas (Fig. 4a), and therefore appeared to be a sign of dam- of cone opsin may be a manifestation of prion-induced age due to prion infection. This swelling was typical of damage in this cone. necrotic cell death as described in cone degeneration in a rd10 mouse model [37]. At 118 dpi abundant PrPSc was Detection of PrPSc‑associated damage in the inner seen in cones and in the inner segment cells lacking cone segment opsin (Fig. 4c), indicating that prion infection had spread To understand how PrPSc deposition might be caus- to rods by this time. Six serial optical sections through ing damage to cone (and rod) photoreceptors we looked the cone outlined in Fig. 4c revealed that PrPSc (red) and for association of PrPSc with additional key inner seg- cone opsin (green) were mostly separated inside this cone ment structures. First, the relationship between PrPSc (See figure on next page.) Fig. 5 Early PrPSc accumulation and damage in the inner segment. a Cartoon of cone photoreceptor shows key structures related to PrPSc deposition. cc connecting cilium, bb basal body, r rootlet, m mitochondria, cp cone pedicle, rs ribbon synapse. b Anti-centrin3 antibody (green) marks the connecting cilium and basal body (small green dots) of all photoreceptors. PrPSc (magenta) accumulation at entrance to connecting cilium (arrow), magnified in inset. c Serial 0.5 µm confocal sections (c1–6) showing relative localization of PrPSc, cilium and basal body. d Cone arrestin (green) staining of uninfected cone photoreceptor. Arrow indicates likely position of connecting cilium. e 118 dpi shows retina stained for cone arrestin (green) and PrP (magenta). Asterisks mark cones missing outer segments, yellow arrows point to position of connecting cilia and associated PrPSc (magenta) deposit, and arrowheads show dystrophic outer segments. f, g Confocal analysis showing xy and yz planes of swollen cones from e confirm the presence of PrPSc at the location of the connecting cilium (arrow) between the IS and OS. h In an uninfected retina, anti-rootletin (green) antibody stains the rootlets of photoreceptors. i, j At later days post infection, PrPSc (magenta) is increased, rootlets are fewer in number and misshapen. Scale bars: b, e = 5 µm; c = 1 µm; d = 2 µm; f, g = 2 µm; h = 10 µm. b, d, e Maximum intensity projections (MIP) of confocal z-stacks h, i, j are widefield images Striebel et al. acta neuropathol commun (2021) 9:17 Page 10 of 26 deposition and the cilium connecting the inner and the detected near the entry point of the connecting cilium outer segments of the photoreceptor cells (Fig.  5a) was (Fig.  5b). In some cases, PrPSc was directly adjacent to examined by dual staining for PrP and Centrin3, a pro- the cilium, as confocal analysis showed that PrPSc, basal tein located within the cilia of rods and cones. Early in body and cilium could be found within the same section disease, at 104 dpi, deposition of PrPSc could often be (Fig.  5c). The deposition of PrPSc at this site may have S triebel et al. acta neuropathol commun (2021) 9:17 Page 11 of 26 Fig. 6 Timecourse of damage to cone photoreceptors. a Anti-Cone arrestin (green) stains cone photoreceptor, inner segments (yellow arrow) and pedicles (green arrow) in an uninfected mouse. Anti-PrP antibody D13, stains PrPC (magenta) at the base of cone pedicles in the OPL (purple arrow). Autofluorescence is also present in the OS. b At 67 dpi cone arrestin and PrP are distributed similarly to uninfected retina (magenta arrow). c Small deposits of PrPSc are present in the IS at 104 dpi and are associated with cones (arrows). d At 118 dpi, the number of cones is reduced (green), most cone pedicles have disappeared from the OPL and the remaining cones are dystrophic (yellow arrow), missing their outer segments. PrPSc is widespread in the IS and ONL. e Few cones (green) remain at 131 dpi, pedicles (arrow) are much smaller in size and associated with dense clumps of PrPSc (arrowhead). f Cones are not detectable at 162 dpi, and ONL is 2–3 nuclei thick, suggesting rod photoreceptors have mostly died. A macrophage or microglial cell (yellow arrow) containing PrPSc is present in the remnants of the OS. Scale bar in a = 20 µm, applicable to all images affected the cilium’s ability to transport phototransduc - To test this idea, we stained for cone arrestin, a cone- tion proteins between the IS and OS. Ciliary dysfunction specific phototransduction protein, known to be traf - is known to cause photoreceptor death in other forms of ficked between the IS and OS [2, 49]. In uninfected retinal degeneration [43]. retinas, cone arrestin was present in the OS, IS, ONL (Fig.  5d) and pedicles (Fig.  6). The cone outer segments Striebel et al. acta neuropathol commun (2021) 9:17 Page 12 of 26 of a 118 dpi retina appeared shrunken and malformed and smaller in size (Fig.  6d). PrPSc was abundant in the compared to an uninfected cone and in some cases, IS region on both cones and rods, and PrPSc staining outer segments were absent (Fig.  5d–g). Inner segments in the OPL was both punctate and diffuse and appeared of cones in the 118 dpi retina were often swollen, and to be independent of cone pedicles (Fig.  6d). At 131 and PrPSc deposits could be found at the constriction point 162 dpi, PrPSc staining in IS and OPL remained strong between the IS and OS, i.e. the location of the connect- and was similar to 118 dpi. However, at these times, most ing cilium (Fig.  5d–g). Together these data suggest that cone pedicles were gone, and the ONL showed thin- deposition of PrPSc in the inner segment may have ning due to a large loss in both cone and rod cell nuclei affected the ciliary transport of cone arrestin and/or (Fig. 6e, f ). These data suggested that PrPSc deposition in other proteins. We also investigated the association of the IS and OPL regions might have induced pathogenic PrPSc deposits with another prominent inner segment effects resulting in death of both cones and rods. structure, the rootlet, which functions to stabilize the connecting cilium [59] (Fig. 5a). For this, we used an anti- rootletin antibody together with D13 anti-PrP. As PrPSc Loss of cone pedicles and rod spherules visualized appeared to accumulate in the inner segment at 104 and by GLUT1 staining 118 dpi, rootlet morphology changed, and the density of Glucose transporter 1 (GLUT1) is known to be an excel- rootlets decreased (Fig. 5h–j). While PrPSc was not usu- lent marker for cell surface visualization in tissues. There - ally associated with rootletin, these data suggested that fore, we used anti-GLUT1 and anti-PrP dual staining to PrPSc accumulation may have damaged the inner seg- study the details of PrP localization relative to cone pedi- ment and its structures including the rootlets. cles and rod spherules in the OPL area following prion retinal infection. Prior to prion infection, cone pedicles were detected as triangular dark shaped spaces outlined Detection of PrPSc‑associated damage in the outer by GLUT1, and just vitread to many of these pedicles a plexiform layer patch of clustered coarse PrPC staining was observed The early accumulation of PrPSc and damage to cone (Supp Fig. 1a). In addition, in the same figure, numerous inner segments led us to examine cone photoreceptors rod cell endings, i.e. spherules, were also seen, and these more closely. We used dual staining with anti-PrP plus appeared to be smaller rounded dark shapes outlined by anti-cone arrestin to follow the progression of PrPSc GLUT1 mainly sclerad (toward the sclera) to cone pedi- deposition and changes in overall cone morphology over cles (Additional file  1: Fig.  1a). Following prion infec- the time course of prion infection from 67 to 162 dpi. tion at 118 dpi, cone pedicles were difficult to recognize In uninfected retina, small dense patches of PrP stain- or had disappeared and rod spherules were decreased ing were seen in the OPL just vitread (toward the vit- in number and size (Additional file  1: Fig.  1b). Later, at reous) to many of the cone pedicles expressing cone 153 dpi, most of the spherules and pedicles were miss- arrestin (Fig.  6a), and this was assumed to be PrPC as it ing, and all the photoreceptor layers (OS, IS, ONL, OPL) was seen in uninfected mice. Similar staining was also were degenerated (Additional file  1: Fig. 1c). These results seen at 67 dpi (Fig.  6b). At 104 dpi, PrP staining in the demonstrated progressive damage to both rods and cones OPL was seen basal to some of the cone pedicles simi- associated with the presence of PrPSc deposition sclerad lar to uninfected mice. More obvious changes were seen to and within the OPL. at 118 dpi, where cone pedicles were fewer in number (See figure on next page.) Fig. 7 In uninfected retina, PrPC expression is concentrated at the base of cone pedicles. a In the OPL, PrPC (yellow arrowheads) staining is present at the base of cone pedicles highlighted with anti-cone arrestin (arrows). b Electron micrograph ( TEM) of the OPL showing a cone pedicle (cp), with multiple round mitochondria (m) and three ribbon synapses (arrows). A yellow dotted line surrounds an estimated region of PrPC from a, which contains many dendritic processes from bipolar and horizontal cells. Some dendrites synapse at ribbons (yellow arrowheads) or make flat contact type synapses (magenta arrowhead) with cone pedicles. Ribbon synapses (arrows) are also visible in rod spherules (r). c Confocal z-stack of uninfected retina shows gross relationship of PrP to rod bipolar cells (PKCα) and horizontal cells (Calbindin). d–h A high magnification single 0.130 µm optical section taken from area in box in panel c, shows PrP is not associated with PKCα and that some calbindin-positive processes (arrow) lie within the PrP-positive area, but association is not clear. i Confocal z-stack image of retina stained for cone bipolar cell dendrites (SCGN), ribbons (CtBP2) and PrP, shows a dense patch of PrPC vitread to a cluster of short ribbons (box) in a cone pedicle. j–l A single 0.130 µm optical section taken from within the box in panel e shows PrPC in close association with SCGN-positive dendrites, and some colocalization is seen as white (arrows). Scale bar in a = 10 µm, b = 1 µm, c = 2 µm, d = 1 µm, i = 4 µm, j = 0.5 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 13 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 14 of 26 Comparative detection of PrP, ribbon synapses and other results suggested that PrPC is highly expressed very near structures in uninfected retina cone pedicles and thus may explain the early accumula- In the photoreceptor layer bordering the OPL, cone and tion of PrPSc in cones versus rods. rod photoreceptor cells are connected with processes from horizontal cells and rod or cone bipolar cells to Damage to ribbon synapses seen by staining form ribbon synapses, which have unique electrophysi- with anti‑CtBP2 ological and morphological features [36]. Initially we Using dual staining immunofluorescence combining compared the locations of several of these structures in anti-PrP with anti-CtBP2 antibodies specific for rib - uninfected mice to determine their normal morphology, bon structures, we studied the development of PrPSc- as observed by both confocal and electron microscopy. associated alterations on ribbon synapse components Dual staining of an uninfected mouse for cone arrestin over time after prion infection. In both uninfected and and PrP showed PrPC located in large dense clusters vit- 82 dpi retinas, ribbon synapses were seen sclerad to the read to cone pedicles (Fig.  7a). Electron microscopy of OPL as a continuous band of linear or horseshoe-shaped the OPL demonstrated rod spherules with single large structures, and PrP was mostly in bunched loose aggre- ribbon synapses and cone pedicles containing multiple gates vitread to the lower-most ribbons (Fig.  8a, b). At short ribbon synapses (Fig.  7b). Notably, comparisons of 104 dpi, ribbons appeared to be similar to 82 dpi, but EM and light microscopy images suggested that areas of PrP staining was slightly more prominent in the ribbon intertwined dendrites from bipolar and horizontal cells area than previously (Fig.  8c). At 118 dpi (Fig.  8d), rib- (Fig.  7b) corresponded to the clustered areas of PrPC bons were significantly decreased in number (Fig.  8l), and seen in Fig. 7a. PrP was deposited among the ribbons rather than below In order to understand the association of PrPC with them (Fig.  8d, j) compared to uninfected retina and ear- these structures, uninfected retina was stained simul- lier timepoints. This PrP was very likely PrPSc as staining taneously with antibodies against PrP and proteins was not seen in this location in uninfected retina. Inter- expressed by horizontal and bipolar cells (Fig.  7c–f ). estingly, at higher magnification, small accumulations of Confocal microscopy analysis of full z-stacks and individ- PrPSc were detected within the horseshoe-like curves of ual 130  nm optical sections was used to reveal the rela- many of the ribbons (Fig.  8j, k). Moreover, PrPSc accu- tive locations of the processes and PrP. Dendrites of rod mulations were never found directly on ribbons, which bipolar cells were marked with anti-PKCα and although is where synaptic vesicles are located. This suggested they were present very near the PrP-stained area beneath that PrPSc was on the postsynaptic side of the ribbon cone pedicles, they were not directly associated with synapses (Fig.  8j). At later time points, 131, 153 and 163 PrP (Fig.  7c, d). Likewise, Calbindin-positive horizontal dpi, ribbons were progressively destroyed, as were the cell processes were not clearly associated with the dense cone and rod cell nuclei in the ONL (Fig. 8e–g, l). Disap- clusters of PrP (Fig.  7c–h). Next, we triple-stained with pearance of ribbons correlated with significant increases anti-PrP, anti-CtBP2, a marker for ribbons and anti- of PrPSc at 118 dpi (Fig.  8l vs m). Thus, the appearance Secretagogin, a protein expressed in 8 of the 12 types of of PrP within the ribbon horseshoes correlated with the cone bipolar cells [13, 42]. In this experiment, dendrites onset of loss of ribbon synapses and suggested this PrPSc of cone bipolar cells expressing secretagogin (SCGN) deposition might be responsible for this damage. showed a close association with PrPC (Fig.  7i–l). These (See figure on next page.) Fig. 8 Timecourse of disappearance of ribbon synapses in OPL. a, h Uninfected retina shows many horseshoe-shaped ribbons stained with CtBP2 (green), sclerad to PrPC (magenta) densities. Most ribbons are not associated with PrPC see panel h. b, i At 82 dpi, the number of ribbons is similar to uninfected, but at cone pedicles PrP appears brighter and more punctate suggestive of new PrPSc accumulation. c 104 dpi retina appears similar to 82 dpi. d, j 118 dpi, ribbons in OPL appear disorganized and are significantly reduced in number, and PrPSc deposits are visible within horseshoe arcs, but not touching, most ribbons (j, yellow arrows). e, k At 131 dpi PrPSc accumulation has peaked and most ribbons have disappeared. Remaining horseshoe-shaped ribbons show association with PrPSc (k, yellow arrow). f, g At late dpi, ONL has thinned dramatically, few ribbons remain, PrPSc is widespread. l Graph shows # of ribbons/µm of OPL of scrapie infected retina at dpi vs uninfected. Each dot represents count from one field of view. m Quantitation of PrP (magenta) fluorescence integrated intensity/µm of OPL of scrapie infected retina at dpi vs uninfected. Each dot represents integrated intensity measurements from one field of view. Lines median value, ns not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 00.0001. Statistics were calculated using one-way ANOVA with multiple comparisons. Scale bars: a–g = 4 µm; h–k = 1 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 15 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 16 of 26 Association of PrPSc to rod and cone bipolar cells dendritic boutons disappeared along with the cell bod- near ribbon synapses ies. This was in marked contrast to the rods and cones The location of PrPSc, adjacent to but not touching rib - themselves which were mostly eliminated at this time bons, suggested these deposits were associated with the post-infection. postsynaptic portion of the ribbon synapses. Dendrites of The dendrites of cone bipolar cells synapse mainly with rod and cone bipolar cells and horizontal cells are known cone photoreceptor cells in the OPL and can be marked to make up the postsynaptic elements of ribbon synapses using anti-Secretagogin (SCGN) antibody [13]. To check in the OPL [36]. Here, investigation of the possible inter- for an association of PrPSc deposits with cone bipolar actions of PrPSc with rod bipolar cells was done by dual cell dendrites we triple-stained retina for PrP, SCGN and staining with anti-PrP and anti-PKCα, which is specific CtBP2 (Fig. 11a–d). As was also shown in Fig. 7, PrPC in for rod bipolar cells. uninfected mice was found to be in close association with In uninfected and 82 dpi prion-infected mice, cell SCGN-positive dendrites under clusters of short ribbons bodies of rod bipolar cells were seen just vitread to the (Fig.  11a). Likewise, single confocal sections, showed OPL and dendritic processes from these cells extended PrPC intermingled among SCGN-positive cone bipo- up through the OPL to the area of the cone pedicles and lar cell dendrites (Fig.  11b). In infected mice at 104 and rod spherules (Fig. 9a, b). PrP was detected mostly in the 118 dpi, staining with anti-PrP antibody D13 revealed an OPL sclerad to the rod bipolar cells (Fig. 9a, b). However, increase in PrPSc among SCGN-positive dendrites where starting at 104 dpi and continuing at 118 dpi, PrP depos- they make synaptic connections with cone pedicles its were seen in a new location, at the tips of the dendritic (Fig. 11a–f ), and this deposition was coincident with the processes of the rod bipolar cells that synapse with rib- previous finding of PrPSc on the tips on rod bipolar cell bons (Fig.  9c, d). This was also visible in more detail by dendrites beneath the horseshoe-shaped ribbon synapses confocal microscopy (Compare Fig.  9f vs g). Because present in rod photoreceptors (Fig. 9h–l). The timing and deposits in these locations were only seen after prion location of this increase in PrP staining suggested these infection, they were thought to be PrPSc. deposits were likely to be the disease-associated form of At higher magnification this PrPSc had the appear - PrP. Despite the destruction of most ribbon synapses in ance of beads on a string extending along the processes cones and the disappearance of cone bipolar cell den- (Fig. 9h–k), and PrPSc seemed to displace or overlap the drites, the number of SCGN-positive cone bipolar cells original PKCα staining of the processes. This was spe - did not decrease during disease (Fig.  11g–k). Thus, both cific for infected retinas at the 104–118 dpi time-points rods and cones were uniquely sensitive to prion-induced (Fig. 9c, d, g) and was not seen in the uninfected or 82 dpi damage compared to other retinal neurons. retinas (Fig. 9a, b). The precise location of PrPSc deposits, at the ribbon synapse on the tips of rod bipolar cell den- Identification of PrPSc near horizontal cell dendrites drites, was confirmed using confocal analysis of a section and boutons triple stained for PrP, PKCα and CtBP2 (Fig.  9l). At late Along with dendritic processes from rod or cone bipolar times in disease (131 dpi), the majority of the rod bipolar cells, dendrites from two horizontal cell neurons (HC) dendritic structures had degenerated (Fig.  9e), however, make up the triad of processes that invaginate the pho- rod bipolar cell bodies appeared to be decreased only by toreceptor at each ribbon synapse [36]. Therefore, HC about 50% starting at 118 dpi and they did not decrease were also studied during prion retinal infection using further after this time (Fig. 10a–e). Axons, dendrites and dual staining with anti-PrP and anti-calbindin, which (See figure on next page.) Fig. 9 PrPSc accumulation at tips of rod bipolar cell dendrites and disappearance of dendrites. Epifluorescent microscopy analysis. a In uninfected retina, PrP (magenta) (blue arrowhead) and PKCα-positive rod bipolar cells (green) are seen near each other in the OPL. Many boutons (blue arrow) are present at dendritic tips of rod bipolar cells. b At 82 dpi, OPL appears similar to uninfected. Asterisk marks a rod bipolar cell body in INL. c At 104 dpi, PrP staining appears at dendritic tips. d At 118 dpi, most PKCα-positive dendrites display PrP at tips of processes (boutons) (arrow). e By 131 dpi most dendrites and dendritic boutons have disappeared, bipolar cell bodies remain (arrow), and PrPSc is distributed in large aggregates. Confocal analysis. f Maximum intensity projection shows PKCα and PrP staining of uninfected retina. Arrows indicate PKCα-positive rod bipolar cells with obvious boutons at the ends of dendrites. g A 118 dpi retina shows PrPSc arranged in bead-like deposits on dendrites, often replacing boutons. Magnification of boxes on left and right are shown below in panels h–k. h Confocal image shows association of PrPSc with dendrites and dendritic tips in area of right box from g. i–k Confocal images show merged, PrP and PKC staining of left box area in g. l In a high magnification of confocal image, anti-CtBP2 antibody marks ribbons (arrows) and confirms PrPSc deposits (magenta) on tips of dendrites (green) invaginating, but not touching, ribbons. Scale bars: a–e = 5 µm; f, g = 3 µm; h–k = 2 µm; l = 0.5 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 17 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 18 of 26 is expressed on the cell body, dendrites and boutons of HC. In uninfected mice, HC bodies were vitread to the OPL, and processes and boutons were mostly in the OPL (Fig.  12a, b), and the clumped aggregates of PrP seen previously did not associate with either the HC or their processes. After scrapie infection at 104 dpi, HC bod- ies, processes, and boutons were unchanged (Fig.  12c, d). As expected, the PrP clumps seen previously in unin- fected mice were missing, and PrP was now mostly dis- seminated in small punctate or linear deposits in the OPL which were near, but slightly separated from the HC boutons (Fig.  12c, d). This difference in PrP morphology suggested that this material was PrPSc. At 118 dpi, the HC boutons were less distinct and PrPSc was still simi- lar to 104 dpi (Fig.  12e, f ). There was still no association between HC bodies and PrP. This relationship was con - firmed using confocal analysis of a triple stained section (Fig.  12i–l), where at high magnification, it was evident that calbindin-positive boutons did not contact PrPSc (Fig.  12j), whereas PrPSc was in contact with PKCα, the marker for rod bipolar cells (Fig.  12k). At 131 and 153 dpi the HC processes and the entire OPL region was atrophied, and damage appeared to be severe. However, PrPSc deposits were more consolidated and surprisingly, most HC bodies were intact (Fig. 12g, h, m). In summary, at 118 dpi there was evidence of deposi- tion of PrPSc in the OPL, at the ribbon synapses of rod photoreceptors. While PrPSc was not directly in contact with the presynaptic ribbons (CtBP2), it was associated with the postsynaptic elements of the synapses. PrPSc was found on the dendritic boutons of rod bipolar cells (PKCα), where they invaginated ribbons synapses and on cone bipolar cell dendrites (SCGN) adjacent to ribbon synapses. HC boutons (Calbindin) did not show accumu- lations of PrPSc. These data suggested that prion infec - tion and PrPSc deposition on rod and cone bipolar cell processes might have a toxic effect on ribbon synapses leading to damage and death of both rods and cones. Electron microscopy of outer plexiform layer confirms loss of synapses Fig. 10 Changes in rod bipolar cells during infection. a anti-PKCα staining (green) in an uninfected mouse shows normal distribution Analysis of the OPL by transmission electron microscopy and morphology of rod bipolar cells. Numerous dendrites (arrow) confirmed many of the findings determined by confocal reach upward to synapse with photoreceptors in ONL from each microscopy. TEM was advantageous in that it allowed a cell body (arrowhead). b At 104 dpi rod bipolar cells appear similar view of all structures simultaneously. Sections from unin- to uninfected mouse. c By 118 dpi the number of bipolar cells has fected mice showed many photoreceptor axon terminals dropped to about half vs uninfected and some cell bodies have few if any associated dendrites (arrowhead). d Late in disease, at 163 (rod spherules and cone pedicles) containing ribbon dpi, rod bipolar cell number is significantly reduced compared to synapses, most were connected to a triad of invaginat- uninfected retina, and few dendrites reach into OPL (arrow). e The ing dendritic processes from horizontal and bipolar cells. number of rod bipolar cells per 100 µm of retina over time. Each dot Rod spherules contained a single large, round mito- represents counts from one field of view, line is median, ns = not chondrion and a ribbon synapse, while cones contained significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p 0.0001 statistics by one-way ANOVA with multiple comparisons test. Scale bar in multiple small mitochondria and 3–5 ribbon synapses a = 20 µm (Fig.  13a). Early in disease (104 dpi), overall morphology S triebel et al. acta neuropathol commun (2021) 9:17 Page 19 of 26 was similar to uninfected mice (Fig.  13b), however in the IS and OS regions of photoreceptors as has been pro- some cone pedicles, swollen, dystrophic dendritic pro- posed in retinitis pigmentosa and ciliopathies [34, 40]. In cesses were noted around ribbon synapses, suggesting support of this mechanism, we found unusual distribu- that cone photoreceptors were beginning to degenerate tions of opsin and cone arrestin (Figs. 4, 5, 6), suggesting (Fig.  13b). At later timepoints (Fig.  13c–f ), typical triad interference with the normal trafficking of these proteins ribbon synapses were rarely noted and both rod spher- between the IS and OS. ules and cone pedicles, if present, were often dystrophic. Beginning at 104 dpi, abnormal PrPSc deposition was Frequent whorl structures, suggestive of autophago- also detected in the OPL region near and within the somes, were noted adjacent to, or in place of ribbon cone pedicle and rod spherules. Although this process synapses in both rods and cones. In short, these ultras- appeared to begin later than the process in the IS region, tructural findings supported the conclusions from fluo - it still coincided with the detection of photoreceptor cell rescent microscopy that at late timepoints all elements of pathology at 104–118 dpi and thus might play a role in ribbon synapses had degenerated. damage and death of rods and cones. In the OPL of unin- fected mice, PrP was normally detected on or near the Discussion dendrites of cone and rod bipolar cells located vitread to In the present study, we followed the timing and loca- rod spherules and cone pedicles, but after prion infec- tion of PrPSc deposition in retina following intracere- tion, PrP detection shifted to sites on the tips of bipolar bral prion injection in mice using immunofluorescence cell dendritic boutons, where they invaginate rods to and electron microscopy. The earliest PrPSc deposits form ribbon synapses. Likewise, PrP deposits increased in retina were found at 67 dpi and were located in the among the dendrites of cone bipolar cells where they IS region of cone photoreceptor cells. At 104 dpi PrPSc synapse with cone pedicles. This different PrP distribu - could be detected adjacent to the cilium, which con- tion suggested that this material was disease-associated nects the IS and OS portions of each cone and rod cell. PrPSc. During this same time, starting at 104–118 dpi These events were followed by cell swelling and altera - and continuing up through 153 dpi, there was progres- tion of organelles, such as rootlets, at 104 dpi, and pro- sive loss of ribbons in rods and cones associated with loss gressive loss of cone and rod cell nuclei in the ONL of over 90% of photoreceptor cone and rod nuclei in the starting at 118 dpi. Similar events with the exception ONL. The mechanism of damage induced by the pres - of swelling were seen in rods on a time course about ence of PrPSc on the tips of bipolar cell dendritic bou- 14 days behind cones. tons is not clear. However, the accumulation of protein The association of PrPSc with cilia may be an impor - aggregates at sites of synaptic transmission might con- tant clue to the pathogenic process of prion infection in found the synaptic transmission process or alter neuritic retina. This abnormal accumulation might interfere with connectivity [53, 61]. Similar speculations have been sug- or damage the ciliary protein transport system between gested for hippocampal synapses in prion-infected brain (See figure on next page.) Fig. 11 Association of PrPSc with cone bipolar cells. a Confocal z-stack of an uninfected retina shows secretagogin-positive cone bipolar cell bodies (magenta arrows) and their green dendritic processes associated with PrP (magenta) in dense clusters (yellow arrows) which include short ribbon synapses (yellow). b A single confocal section taken from the box in panel a shows PrP in close association with cone bipolar cell dendrites (green), beneath clusters of short ribbons (yellow) typically found in cone pedicles. Horseshoe-shaped ribbons in rod spherules (blue arrows) are not associated with PrP or SCGN. c In a confocal z-stack, at 118 dpi, cone bipolar cell bodies (magenta arrow) appear normal but PrP staining is increased in the OPL. d A single confocal section taken from the box in panel c shows increase in PrP among the SCGN-positive dendrites with some PrP overlapping SCGN, suggesting colocalization (white arrow). PrP is also present within the horseshoe-shaped ribbon synapses in rods (blue arrows). e, f Widefield images at 104 dpi show numerous examples of PrP clusters localizing with the ends of SCGN-positive dendrites (blue arrows) originating from cone bipolar cells (magenta arrows). g Number of cone bipolar cells per mm vs dpi. Each symbol represents the count from one field of view. Different symbols represent different animals. Line represents median. h–k anti-Secretagogin stained retina from uninfected and three different dpi show relative numbers of SCGN-positive cells. ns = not significant, statistics by one-way ANOVA with multiple comparisons test. Scale bars: a, c = 4 µm; b, d = 1 µm; e,f = 10 µm; h–k = 25 µm Striebel et al. acta neuropathol commun (2021) 9:17 Page 20 of 26 S triebel et al. acta neuropathol commun (2021) 9:17 Page 21 of 26 [12, 27, 48]. However, this has not previously been sug- cones [1, 35, 44]. Interestingly, cones make up only 3% of gested for retinal photoreceptor ribbon synapses, where photoreceptors in mouse retina, so the cone cell specific - the molecular components can be visualized in better ity in our study was surprising. This could be due to local detail. differences in PrP expression which might influence the Interestingly other nearby retinal neurons were dam- timing of PrP conversion to PrPSc. We have not been able aged less by prion infection. For example, cell bodies of to measure differences in normal PrP expression in the cone bipolar cells remained constant throughout disease, IS region of uninfected rods and cones. However, in the rod bipolar cell bodies decreased by 50%, and horizontal OPL region, uninfected mice have high PrPC expression cell bodies decreased less than 10%. However, dendrites in the dendritic processes of bipolar cells located directly in the outer plexiform layer of all these cells showed sig- beneath the cone pedicles. These are a mixture of rod and nificant damage. Ganglion cells and amacrine cells also cone bipolar cells which all appear to deposit PrPSc on did not appear to be damaged. These latter two cell types their dendrites after prion infection, but while there are were not specifically stained and counted, but cellularity hundreds of synaptic connections to each cone pedicle, in the regions where they typically reside appeared to be each rod spherule has no more than seven [38]. This dif - normal. One explanation for the extreme sensitivity of ference might increase the efficiency of prion travel to the cone and rod cells to prion infection might be the PrPSc IS regions of cones vs rods. In addition, prion travel from association and blockage of cilia which are not found in ganglion cells to cone cells can be direct, whereas inter- the other above-mentioned retinal neurons. Alterna- mediate cells, such as amacrine cells, are often involved tively, the level of PrPC expression in various cell types in connecting ganglion cells to rod bipolar cells [14]. The might also play a role in these differences. In contrast, more direct route to cones might increase the tempo of photoreceptor degeneration was not observed in our prion infection, favoring cones over rods. studies of prion-infected transgenic mice expressing GPI- The damage mechanism(s) seen in prion-infected ret - anchorless PrP, suggesting expression of anchored PrPC ina and brain are not well understood [25]. Our earlier on the cell membrane may be critical to PrPSc-induced studies indicated the presence of apoptotic cells in the damage to certain cell types [30]. ONL after prion infection [29, 54]. Since these observa- Another question raised by these results is why the tions were made later in the disease course, these cells cones preceded rods in the PrPSc deposition and dam- were likely to be mostly rods which are 32 × more abun- age process. In some rare retinal degenerations, known dant than cones in mouse retina. However, in our studies as Cone and Cone-rod dystrophies, cones also precede using cone opsin to detect cones, damaged PrPSc-posi- rods in damage and death, however it is more common tive cones in the IS region at 104 and 118 dpi were swol- for rods to begin the degenerative process, as in retinitis len suggesting a necrotic rather than apoptotic cell death pigmentosa [39, 56]. In other neurodegenerative diseases process (Figs. 4, 5, 13). Interestingly, this same pathology affecting retina, such as AD and PD, photoreceptors are preceding cone necrosis is similar to previous results of not typically affected first and specificity for cones or Murikami et  al. [37] studying the human retinitis pig- rods is not clear, though a recent study found accumula- mentosa rd10 mutation in a mouse model and in human tions of phosphorylated tau specifically in ageing primate retinitis pigmentosa patients. (See figure on next page.) Fig. 12 Timecourse and confocal analysis of PrP association with horizontal cells. a,b In uninfected mouse retina, anti-calbindin labels horizontal cell bodies (yellow arrow), dendrites (blue arrows) and boutons (arrowheads). PrPC magenta densities (long yellow arrow) are surrounded by horizontal cell boutons and dendrites. c, d At 104 dpi, calbindin-positive boutons (arrowheads) are near, but not touching, new PrPSc deposits (magenta). e, f At 118 dpi, PrPSc (magenta) is more widely distributed, and number of boutons is significantly reduced. PrPSc is near, but not in contact with horizontal cell components. g, h At later timepoints 131 and 153 dpi, ONL is thinned due to loss of photoreceptor nuclei. Few, if any, horizontal cell boutons are present. Large PrPSc deposits (arrow) are visible in the OPL, but not in horizontal cell bodies. i At very high magnification, a 118 dpi retina shows a PKCα-positive rod bipolar cell dendrite (green) near calbindin-positive (yellow) horizontal cell boutons (blue arrows). j A PrPSc deposit (magenta) is very close to, but not touching the horizontal cell boutons (blue arrowhead). k In a merge of three channels the white color indicates the association of a PrPSc deposit (magenta) with the dendritic tip of a rod bipolar cell (green). l Cartoon depicts preceding three panels, with addition of presumed location of rod spherules (rs) and ribbons (r). m Surprisingly, the number of horizontal cell bodies does not change significantly over time. Scale bars: a, c, e, g, h = 5 µm; b, d, f = 2 µm; i–l = 0.5 µm Striebel et al. acta neuropathol commun (2021) 9:17 Page 22 of 26 S triebel et al. acta neuropathol commun (2021) 9:17 Page 23 of 26 Fig. 13 Transmission electron microscopy shows timeline of changes in outer plexiform layer. a Uninfected retina shows numerous rod spherules (r) each containing a single ribbon synapse (arrow) and mitochondrion (m). Cone pedicles (cp) contain multiple ribbon synapses (yellow arrows) and mitochondria (m). Inset shows magnified cone ribbon synapse with ribbon (arrow), and a typical triad consisting of two invaginating horizontal cell processes (h) and one bipolar cell process (bp). Many bipolar and horizontal cell processes are present vitread to cone pedicles (asterisks). b At 104 dpi, most rod spherules appear normal with ribbon synapses present. A cone pedicle (cp) at center, has swollen dystrophic dendritic processes at ribbon synapses (red arrow). Bipolar and horizontal cell processes are present beneath the pedicle (asterisk). c At 137 dpi the OPL appears disorganized with abnormal rod spherules (red arrow), a floating ribbon synapse (blue arrowhead) without invaginating dendritic processes, autophagic whorl (blue arrow) and a dystrophic dendrite (red arrowhead) at a ribbon synapse in a cone pedicle. Microglial cell (mg). d, e, f Examples of autophagic-like whorls (red arrows) in rod spherules at 137 dpi. Scale bars a‑ c = 2 µm; d, e, f = 1 µm between the inner and outer segments of PR cells as well Conclusions as damage to ribbon synapses found in the synaptic end- The present experiments report two new areas of depo - feet of rods and cones near the OPL. Possible synergy sition of abnormal disease-associated PrPSc in prion- and specificity between these two mechanisms remains infected retina. These regions both involve photoreceptor to be worked out. These mechanisms might be active in cone and rod cells which then go on to die as a part of other human retinal degenerative diseases where pro- the disease process. The location of the PrPSc deposition tein misfolding occurs, such as retinitis pigmentosa and suggests that the damage mechanisms may involve inter- prion-like diseases, such as AD and PD. ruption of the ciliary transport pathways of molecules Striebel et al. acta neuropathol commun (2021) 9:17 Page 24 of 26 Received: 4 December 2020 Accepted: 9 January 2021 Supplementary information The online version contains supplementary material available at https ://doi. org/10.1186/s4047 8-021-01120 -x. References Additional file 1: Fig. 1 Cone pedicles and rod spherules disappear as 1. Aboelnour A, Van der Spuy J, Powner M, Jeffery G (2017) Primate retinal PrPSc accumulates in the OPL. a In an uninfected mouse, concentra- cones express phosphorylated tau associated with neuronal degen- tions of PrPC (magenta, yellow arrows) lie vitread to cone pedicles (dark eration yet survive in old age. Exp Eye Res 165:105–108. https ://doi. space, yellow arrowhead) outlined by glucose transporter 1 (GLUT1) a org/10.1016/j.exer.2017.09.013 membrane-associated protein. Rod spherules (red arrowhead) often have 2. Arshavsky VY (2003) Protein translocation in photoreceptor light adapta- tiny faint patches of PrPC (red arrow) adjacent to their membranes. b At tion: a common theme in vertebrate and invertebrate vision. Sci STKE 118 dpi, cone pedicles are less frequent and the distribution of PrP is more 2003:PE43. doi:https ://doi.org/10.1126/stke.2003.204.pe43 punctate and widespread, suggesting PrPSc deposits. c at 153 dpi OPL has 3. Ashok A, Singh N, Chaudhary S, Bellamkonda V, Kritikos AE, Wise AS, Rana thinned significantly, most rod spherules have disappeared, and PrPSc is N, McDonald D, Ayyagari R (2020) Retinal degeneration and Alzheimer’s now in dense punctate patches. ONL has thinned to 1-2 nuclei. Scale bar disease: an evolving link. Int J Mol Sci. https ://doi.org/10.3390/ijms2 11972 = 5 µm. a, b, c are images of optical slices with total z-depth =1 µm 4. Brundin P, Melki R, Kopito R (2010) Prion-like transmission of protein aggregates in neurodegenerative diseases. Nat Rev Mol Cell Biol Acknowledgments 11:301–307. https ://doi.org/10.1038/nrm28 73 The authors thank Katie Williams for technical assistance with mouse anes- 5. Buyukmihci N, Goehring-Harmon F, Marsh RF (1982) Photoreceptor thesia, observation and necropsy, Nancy Kurtz, Lori Lubke and Dan Long, degeneration preceding clinical scrapie encephalopathy in hamsters. J for technical histology assistance, Dave Dorward for electron microscopy Comp Neurol 205:49–54. https ://doi.org/10.1002/cne.90205 0105 technical assistance, Tylor Lewis from the Department of Ophthalmology at 6. Buyukmihci N, Rorvik M, Marsh RF (1980) Replication of the scrapie agent Duke University Medical Center for technical advice on vibratome sectioning in ocular neural tissues. Proc Natl Acad Sci USA 77:1169–1171 of retinal tissue for electron microscopy processing, and Jeffrey Severson for 7. Carroll JA, Race B, Williams K, Striebel J, Chesebro B (2018) Microglia animal husbandry, and Drs. James Carroll, Byron Caughey, Cathryn Haigh, are critical in host defense against prion disease. J Virol. https ://doi. Suzette Priola, and Clayton Winkler for advice on manuscript preparation. org/10.1128/JVI.00549 -18 8. Carroll JA, Striebel JF, Rangel A, Woods T, Phillips K, Peterson KE, Race B, Author contributions Chesebro B (2016) Prion strain differences in accumulation of PrPSc on JS designed study, gathered and analyzed pathology data from all sources, neurons and glia are associated with similar expression profiles of neu- and wrote manuscript. BR designed study, carried out animal experiments, roinflammatory genes: comparison of three prion strains. PLoS Pathog and edited manuscript. JL gathered and analyzed confocal datasets and 12:e1005551. https ://doi.org/10.1371/journ al.ppat.10055 51 edited manuscript. CS prepared samples for electron microscopy, gathered 9. Caughey B, Baron GS, Chesebro B, Jeffrey M (2009) Getting a grip on and analyzed electron microscopy data and edited manuscript. BC designed prions: oligomers, amyloids, and pathological membrane interactions. study, analyzed pathology data, and wrote manuscript. Annu Rev Biochem 78:177–204. https ://doi.org/10.1146/annur ev.bioch em.78.08290 7.14541 0 Funding 10. Chishti MA, Strome R, Carlson GA, Westaway D (1997) Syrian ham- Open Access funding provided by the National Institutes of Health (NIH).. ster prion protein (PrP(C)) is expressed in photoreceptor cells of the This research was supported by the Intramural Research Program (DIR) of adult retina. Neurosci Lett 234:11–14. https ://doi.org/10.1016/s0304 the National Institute of Allergy and Infectious Diseases (NIAID) of the United -3940(97)00669 -1 States, National Institutes of Health (NIH). 11. Clavaguera F, Duyckaerts C, Haik S (2020) Prion-like properties of Tau assemblies. Curr Opin Neurobiol 61:49–57. https ://doi.org/10.1016/j. Availability of data and material conb.2019.11.022 The data supporting the conclusions of this article are included within the 12. Cunningham C, Deacon R, Wells H, Boche D, Waters S, Diniz CP, Scott H, article. Original slides, tissues and photographs are retained. All reagents and Rawlins JN, Perry VH (2003) Synaptic changes characterize early behav- animals used in this study are available from scientific supply companies, ioural signs in the ME7 model of murine prion disease. Eur J Neurosci except the anti-PrP antibody D13, which depending on supply, may be avail- 17:2147–2155. https ://doi.org/10.1046/j.1460-9568.2003.02662 .x able upon request. 13. Euler T, Haverkamp S, Schubert T, Baden T (2014) Retinal bipolar cells: elementary building blocks of vision. Nat Rev Neurosci 15:507–519. https Ethics statement ://doi.org/10.1038/nrn37 83 All mice were housed at the Rocky Mountain Laboratories (RML) in an 14. Fain G, Sampath AP (2018) Rod and cone interactions in the retina. AAALAC-accredited facility in compliance with guidelines provided by the F1000Res 7. doi:https ://doi.org/10.12688 /f1000 resea rch.14412 .1 Guide for the Care and Use of Laboratory Animals (Institute for Laboratory 15. Foster JD, Fraser H, Bruce ME (1986) Retinopathy in mice with experimen- Animal Research Council). Experimentation followed RML Animal Care and tal scrapie. Neuropathol Appl Neurobiol 12:185–196 Use Committee approved protocol 2016–042. 16. Fraser H (1982) Neuronal spread of scrapie agent and targeting of lesions within the retino-tectal pathway. Nature 295:149–150. https ://doi. Consent for publication org/10.1038/29514 9a0 Not applicable. 17. Frigg R, Wenzel A, Samardzija M, Oesch B, Wariwoda H, Navarini AA, Seeliger MW, Tanimoto N, Reme C, Grimm C (2006) The prion protein Competing interest is neuroprotective against retinal degeneration in vivo. Exp Eye Res The authors declare that they have no conflict of interest. 83:1350–1358. https ://doi.org/10.1016/j.exer.2006.07.010 18. Gomez-Nicola D, Fransen NL, Suzzi S, Perry VH (2013) Regulation of Author details microglial proliferation during chronic neurodegeneration. J Neurosci Laboratory of Persistent Viral Diseases, Rocky Mountain Laboratories, National 33:2481–2493. https ://doi.org/10.1523/JNEUR OSCI.4440-12.2013 Institute of Allergy and Infectious Diseases, National Institutes of Health, 19. Gong J, Jellali A, Forster V, Mutterer J, Dubus E, Altrock WD, Sahel JA, 903 South Fourth Street, Hamilton, MT 59840, USA. Research Technologies Rendon A, Picaud S (2007) The toxicity of the PrP106-126 prion peptide Branch, Rocky Mountain Laboratories, National Institute of Allergy and Infec- on cultured photoreceptors correlates with the prion protein distribu- tious Diseases, National Institutes of Health, Hamilton, MT 59840, USA. tion in the mammalian and human retina. Am J Pathol 170:1314–1324. https ://doi.org/10.2353/ajpat h.2007.06034 0 S triebel et al. acta neuropathol commun (2021) 9:17 Page 25 of 26 20. Greenlee MH, Smith JD, Platt EM, Juarez JR, Timms LL, Greenlee JJ 37. Murakami Y, Ikeda Y, Nakatake S, Tachibana T, Fujiwara K, Yoshida N, (2015) Changes in retinal function and morphology are early clinical Notomi S, Nakao S, Hisatomi T, Miller JW, Vavvas DG, Sonoda KH, Ishibashi signs of disease in cattle with bovine spongiform encephalopathy. T (2015) Necrotic enlargement of cone photoreceptor cells and the PLoS ONE 10:e0119431. https ://doi.org/10.1371/journ al.pone.01194 31 release of high-mobility group box-1 in retinitis pigmentosa. Cell Death 21. Hamir AN, Miller JM, Cutlip RC, Kunkle RA, Jenny AL, Stack MJ, Chaplin Discov 1:15058. https ://doi.org/10.1038/cddis cover y.2015.58 MJ, Richt JA (2004) Transmission of sheep scrapie to elk (Cervus ela- 38. Narayan DS, Chidlow G, Wood JP, Casson RJ (2017) Glucose metabolism phus nelsoni) by intracerebral inoculation: final outcome of the experi- in mammalian photoreceptor inner and outer segments. Clin Exp Oph- ment. J Vet Diagn Invest 16:316–321. https ://doi.org/10.1177/10406 thalmol 45:730–741. https ://doi.org/10.1111/ceo.12952 38704 01600 410 39. Narayan DS, Wood JP, Chidlow G, Casson RJ (2016) A review of the 22. Hasegawa M, Nonaka T, Masuda-Suzukake M (2017) Prion-like mechanisms of cone degeneration in retinitis pigmentosa. Acta Ophthal- mechanisms and potential therapeutic targets in neurodegenerative mol 94:748–754. https ://doi.org/10.1111/aos.13141 disorders. Pharmacol Ther 172:22–33. https ://doi.org/10.1016/j.pharm 40. Newton F, Megaw R (2020) Mechanisms of photoreceptor death in retini- thera .2016.11.010 tis pigmentosa. Genes (Basel). https ://doi.org/10.3390/genes 11101 120 23. Head MW, Northcott V, Rennison K, Ritchie D, McCardle L, Bunn 41. Orru CD, Soldau K, Cordano C, Llibre-Guerra J, Green AJ, Sanchez H, TJ, McLennan NF, Ironside JW, Tullo AB, Bonshek RE (2003) Prion Groveman BR, Edland SD, Safar JG, Lin JH, Caughey B, Geschwind MD, protein accumulation in eyes of patients with sporadic and variant Sigurdson CJ (2018) Prion seeds distribute throughout the eyes of spo- Creutzfeldt–Jakob disease. Invest Ophthalmol Vis Sci 44:342–346 radic Creutzfeldt–Jakob disease patients. MBio. https ://doi.org/10.1128/ 24. Head MW, Peden AH, Yull HM, Ritchie DL, Bonshek RE, Tullo AB, Ironside mBio.02095 -18 JW (2005) Abnormal prion protein in the retina of the most commonly 42. Puthussery T, Gayet-Primo J, Taylor WR (2010) Localization of the calcium- occurring subtype of sporadic Creutzfeldt–Jakob disease. Br J Ophthal- binding protein secretagogin in cone bipolar cells of the mammalian mol 89:1131–1133. https ://doi.org/10.1136/bjo.2004.06349 5 retina. J Comp Neurol 518:513–525. https ://doi.org/10.1002/cne.22234 25. Hughes D, Halliday M (2017) What is our current understanding of 43. Rachel RA, Li T, Swaroop A (2012) Photoreceptor sensory cilia and cili- PrP(Sc)-associated neurotoxicity and its molecular underpinnings? opathies: focus on CEP290. RPGR and their interacting proteins Cilia 1:22. Pathogens. https ://doi.org/10.3390/patho gens6 04006 3https ://doi.org/10.1186/2046-2530-1-22 26. Jaunmuktane Z, Brandner S (2020) Invited review: the role of prion-like 44. Ramirez AI, de Hoz R, Salobrar-Garcia E, Salazar JJ, Rojas B, Ajoy D, Lopez- mechanisms in neurodegenerative diseases. Neuropathol Appl Neuro- Cuenca I, Rojas P, Trivino A, Ramirez JM (2017) The role of microglia in reti- biol 46:522–545. https ://doi.org/10.1111/nan.12592 nal neurodegeneration: Alzheimer’s disease, parkinson, and glaucoma. 27. Jeffrey M, Halliday WG, Bell J, Johnston AR, MacLeod NK, Ingham C, Front Aging Neurosci 9:214. https ://doi.org/10.3389/fnagi .2017.00214 Sayers AR, Brown DA, Fraser JR (2000) Synapse loss associated with 45. Ratnayaka JA, Serpell LC, Lotery AJ (2015) Dementia of the eye: the role of abnormal PrP precedes neuronal degeneration in the scrapie-infected amyloid beta in retinal degeneration. Eye (Lond) 29:1013–1026. https :// murine hippocampus. Neuropathol Appl Neurobiol 26:41–54. https :// doi.org/10.1038/eye.2015.100 doi.org/10.1046/j.1365-2990.2000.00216 .x 46. Silverman SM, Wong WT (2018) Microglia in the retina: roles in develop- 28. Kercher L, Favara C, Chan CC, Race R, Chesebro B (2004) Differences in ment, maturity, and disease. Annu Rev Vis Sci 4:45–77. https ://doi. scrapie-induced pathology of the retina and brain in transgenic mice org/10.1146/annur ev-visio n-09151 7-03442 5 that express hamster prion protein in neurons, astrocytes, or multiple 47. Singh N, Chaudhary S, Ashok A, Lindner E (2020) Prions and prion cell types. Am J Pathol 165:2055–2067. https ://doi.org/10.1016/S0002 diseases: insights from the eye. Exp Eye Res 199:108200. https ://doi. -9440(10)63256 -7org/10.1016/j.exer.2020.10820 0 29. Kercher L, Favara C, Striebel JF, LaCasse R, Chesebro B (2007) Prion pro- 48. Siskova Z, Page A, O’Connor V, Perry VH (2009) Degenerating synaptic tein expression differences in microglia and astroglia influence scrapie- boutons in prion disease: microglia activation without synaptic stripping. induced neurodegeneration in the retina and brain of transgenic mice. Am J Pathol 175:1610–1621. https ://doi.org/10.2353/ajpat h.2009.09037 2 J Virol 81:10340–10351. https ://doi.org/10.1128/JVI.00865 -07 49. Slepak VZ, Hurley JB (2008) Mechanism of light-induced translocation of 30. Klingeborn M, Race B, Meade-White KD, Rosenke R, Striebel JF, Chese- arrestin and transducin in photoreceptors: interaction-restricted diffusion. bro B (2011) Crucial role for prion protein membrane anchoring in the IUBMB Life 60:2–9. https ://doi.org/10.1002/iub.7 neuroinvasion and neural spread of prion infection. J Virol 85:1484– 50. Smith JD, Greenlee JJ, Hamir AN, West Greenlee MH (2008) Retinal cell 1494. https ://doi.org/10.1128/JVI.02167 -10 types are differentially affected in sheep with scrapie. J Comp Pathol 31. Liu B, Rasool S, Yang Z, Glabe CG, Schreiber SS, Ge J, Tan Z (2009) Amyloid- 138:12–22. https ://doi.org/10.1016/j.jcpa.2007.09.002 peptide vaccinations reduce {beta}-amyloid plaques but exacerbate 51. Soto C, Pritzkow S (2018) Protein misfolding, aggregation, and conforma- vascular deposition and inflammation in the retina of Alzheimer’s trans- tional strains in neurodegenerative diseases. Nat Neurosci 21:1332–1340. genic mice. Am J Pathol 175:2099–2110. https ://doi.org/10.2353/ajpat https ://doi.org/10.1038/s4159 3-018-0235-9 h.2009.09015 9 52. Spraker TR, Zink RR, Cummings BA, Wild MA, Miller MW, O’Rourke KI 32. Liu C, Zhang CW, Zhou Y, Wong WQ, Lee LC, Ong WY, Yoon SO, Hong W, (2002) Comparison of histological lesions and immunohistochemical Fu XY, Soong TW, Koo EH, Stanton LW, Lim KL, Xiao ZC, Dawe GS (2018) staining of proteinase-resistant prion protein in a naturally occurring APP upregulation contributes to retinal ganglion cell degeneration spongiform encephalopathy of free-ranging mule deer (Odocoileus via JNK3. Cell Death Differ 25:661–676. https ://doi.org/10.1038/s4141 hemionus) with those of chronic wasting disease of captive mule deer. 8-017-0005-3 Vet Pathol 39:110–119. https ://doi.org/10.1354/vp.39-1-110 33. Matsunaga Y, Peretz D, Williamson A, Burton D, Mehlhorn I, Groth D, 53. Steinert JR (2015) Prion protein as a mediator of synaptic transmis- Cohen FE, Prusiner SB, Baldwin MA (2001) Cryptic epitopes in N-termi- sion. Commun Integr Biol 8:e1063753. https ://doi.org/10.1080/19420 nally truncated prion protein are exposed in the full-length molecule: 889.2015.10637 53 dependence of conformation on pH. Proteins 44:110–118 54. Striebel JF, Race B, Williams K, Carroll JA, Klingeborn M, Chesebro B (2019) 34. Mendes HF, van der Spuy J, Chapple JP, Cheetham ME (2005) Mecha- Microglia are not required for prion-induced retinal photoreceptor nisms of cell death in rhodopsin retinitis pigmentosa: implications for degeneration. Acta Neuropathol Commun 7:48. https ://doi.org/10.1186/ therapy. Trends Mol Med 11:177–185. https ://doi.org/10.1016/j.molme s4047 8-019-0702-x d.2005.02.007 55. Takao M, Kimura H, Kitamoto T, Mihara B (2018) PrP(res) deposition 35. Mirzaei N, Shi H, Oviatt M, Doustar J, Rentsendorj A, Fuchs DT, Sheyn J, in the retina is a common finding of sporadic, familial and iatrogenic Black KL, Koronyo Y, Koronyo-Hamaoui M (2020) Alzheimer’s retin- Creutzfeldt–Jakob diseases (CJD). Acta Neuropathol Commun 6:78. https opathy: seeing disease in the eyes. Front Neurosci 14:921. https ://doi.://doi.org/10.1186/s4047 8-018-0582-5 org/10.3389/fnins .2020.00921 56. Thiadens AA, Phan TM, Zekveld-Vroon RC, Leroy BP, van Born LI, Hoyng 36. Moser T, Grabner CP, Schmitz F (2020) Sensory processing at ribbon CB, Klaver CC, Writing Committee for the Cone Disorders study Group C, synapses in the retina and the cochlea. Physiol Rev 100:103–144. https :// Roosing S, Pott JW, van Schooneveld MJ, van Moll-Ramirez N, van Gen- doi.org/10.1152/physr ev.00026 .2018 deren MM, Boon CJ, van Hollander AI, Bergen AA, De Baere E, Cremers FP, Lotery AJ (2012) Clinical course, genetic etiology, and visual outcome in Striebel et al. acta neuropathol commun (2021) 9:17 Page 26 of 26 cone and cone-rod dystrophy. Ophthalmology 119(819):826. https ://doi. of sensory cilia. Mol Cell Biol 25:4129–4137. https ://doi.org/10.1128/ org/10.1016/j.ophth a.2011.10.011 MCB.25.10.4129-4137.2005 57. Veys L, Vandenabeele M, Ortuno-Lizaran I, Baekelandt V, Cuenca N, Moons 60. Ye X (2009) Visual pathology in animal prion diseases. Histol Histopathol L, De Groef L (2019) Retinal alpha-synuclein deposits in Parkinson’s dis- 24:1563–1577. https ://doi.org/10.14670 /HH-24.1563 ease patients and animal models. Acta Neuropathol 137:379–395. https :// 61. Zamponi E, Pigino GF (2019) Protein misfolding, signaling abnormali- doi.org/10.1007/s0040 1-018-01956 -z ties and altered fast axonal transport: implications for Alzheimer and 58. West Greenlee MH, Lind M, Kokemuller R, Mammadova N, Kondru N, prion diseases. Front Cell Neurosci 13:350. https ://doi.org/10.3389/fncel Manne S, Smith J, Kanthasamy A, Greenlee J (2016) Temporal resolu- .2019.00350 tion of misfolded prion protein transport, accumulation, glial activation, and neuronal death in the retinas of mice inoculated with scrapie. Am J Publisher’s Note Pathol 186:2302–2309. https ://doi.org/10.1016/j.ajpat h.2016.05.018 Springer Nature remains neutral with regard to jurisdictional claims in pub- 59. Yang J, Gao J, Adamian M, Wen XH, Pawlyk B, Zhang L, Sanderson MJ, Zuo lished maps and institutional affiliations. J, Makino CL, Li T (2005) The ciliary rootlet maintains long-term stability Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Acta Neuropathologica Communications Springer Journals

Prion-induced photoreceptor degeneration begins with misfolded prion protein accumulation in cones at two distinct sites: cilia and ribbon synapses

Loading next page...
 
/lp/springer-journals/prion-induced-photoreceptor-degeneration-begins-with-misfolded-prion-jLl7gZSrNW
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2051-5960
DOI
10.1186/s40478-021-01120-x
Publisher site
See Article on Publisher Site

Abstract

Accumulation of misfolded host proteins is central to neuropathogenesis of numerous human brain diseases includ- ing prion and prion-like diseases. Neurons of retina are also affected by these diseases. Previously, our group and others found that prion-induced retinal damage to photoreceptor cells in mice and humans resembled pathology of human retinitis pigmentosa caused by mutations in retinal proteins. Here, using confocal, epifluorescent and electron microscopy we followed deposition of disease-associated prion protein (PrPSc) and its association with damage to critical retinal structures following intracerebral prion inoculation. The earliest time and place of retinal PrPSc deposi- tion was 67 days post-inoculation (dpi) on the inner segment (IS) of cone photoreceptors. At 104 and 118 dpi, PrPSc was associated with the base of cilia and swollen cone inner segments, suggesting ciliopathy as a pathogenic mecha- nism. By 118 dpi, PrPSc was deposited in both rods and cones which showed rootlet damage in the IS, and photore- ceptor cell death was indicated by thinning of the outer nuclear layer. In the outer plexiform layer (OPL) in uninfected mice, normal host PrP (PrPC) was mainly associated with cone bipolar cell processes, but in infected mice, at 118 dpi, PrPSc was detected on cone and rod bipolar cell dendrites extending into ribbon synapses. Loss of ribbon synapses in cone pedicles and rod spherules in the OPL was observed to precede destruction of most rods and cones over the next 2–3 weeks. However, bipolar cells and horizontal cells were less damaged, indicating high selectivity among neurons for injury by prions. PrPSc deposition in cone and rod inner segments and on the bipolar cell processes par- ticipating in ribbon synapses appear to be critical early events leading to damage and death of photoreceptors after prion infection. These mechanisms may also occur in human retinitis pigmentosa and prion-like diseases, such as AD. Keywords: Prion, Prion-like, Ribbon synapses, Retinitis pigmentosa, Alzheimer, Parkinson, Ciliopathy, Scrapie, Necrosis, Apoptosis Introduction Prion diseases are progressive neurodegenerative dis- eases which affect humans as well as numerous wild and domestic animal species. These diseases are characterized by vacuolar degeneration of the grey matter and gliosis primarily in grey matter of the CNS. In addition, abnor- *Correspondence: bchesebro@niaid.nih.gov mal aggregation of a host prion protein (PrPC) leads to Laboratory of Persistent Viral Diseases, Rocky Mountain Laboratories, deposition of a protease-resistant PrP isoform (PrPSc) National Institute of Allergy and Infectious Diseases, National Institutes in the nervous system and other organs, which results in of Health, 903 South Fourth Street, Hamilton, MT 59840, USA Full list of author information is available at the end of the article damage by unclear mechanisms [25]. Protein aggregation © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, 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 changes were made. 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 http://creat iveco mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Striebel et al. acta neuropathol commun (2021) 9:17 Page 2 of 26 in prion disease proceeds by a process referred to as Photoreceptors are also damaged in human retinitis seeded polymerization or PrP conversion, where initial pigmentosa, which is a major cause of human blindness small aggregates of PrPSc catalyze extension of protein resulting in a retinal pathology similar to prion diseases. aggregation to generate additional PrPSc [9]. In some forms of retinitis pigmentosa, microglia are A similar seeded polymerization mechanism has been known to become activated by the misfolding of mutant noted in several other more common neurodegenera- host proteins such as rhodopsin, and microglia have also tive diseases including Alzheimer’s disease, Parkinson’s been suspected to be important in the pathogenic pro- disease and tauopathies, but in these diseases, the aggre- cess [40, 46]. In prion diseases, microglia activated by the gated host proteins are amyloid beta (Aβ), α-synuclein deposition of aggregated prion protein have also been and tau respectively [4, 11, 26]. Because of the similarities suspected to be a possible mechanism of pathogenesis in the protein aggregation process in these diseases and [18]. However, in our recent studies using PLX5622 to prion diseases, they have been referred to as “prion-like” eliminate microglia in  vivo, removal of microglia led to diseases. Some people have suspected that prion diseases a decrease in survival time due to accelerated brain prion and prion-like diseases might be susceptible to simi- pathogenesis in mice [7]. Thus, microglia appeared to be lar therapeutic interventions [22, 51]. Thus, at this time, mainly helpful, not harmful, to the host during prion dis- there is high interest in the mechanisms of pathogenesis ease. Furthermore, in our studies of retinal prion disease of all these diseases. in mice, elimination of microglia gave similar results, Prion and “prion-like” diseases are known to cause reti- showing accelerated retinal degeneration when microglia nal damage in humans and other species, but each dis- were reduced or absent [54]. Müller glial cells are also ease affects the retina in unique ways [44]. In Alzheimer’s activated during prion disease but are likely a response disease patients, Aβ deposition has been associated with to damage rather than the cause of pathogenesis [28, degeneration of the retinal ganglion cell layer, photore- 29, 58]. u Th s, in vivo damage induced by prion infection ceptors and the retinal pigmented epithelium [3, 45]. Rat- might be due to direct effects of aggregated prion protein nayaka et  al. have also shown that Aβ plaques may be a on photoreceptor cells rather than indirect effects of acti - key factor in Age-related Macular Degeneration (AMD). vated microglia or astroglia. In AD, phosphorylated Tau deposits have also been In the present paper, we studied direct early events observed from the outer plexiform layer to the ganglion of prion protein deposition in mouse retina following cell layer [31, 32]. Likewise, abnormal α-synuclein aggre- intracerebral prion injection. For these studies, we used gates have been detected in retina of Parkinson’s disease confocal and epifluorescence microscopy with detec - patients, and these are implicated in the degeneration of tion of multiple targets on the same section to localize the nerve fiber layer (ganglion cell axons), ganglion cell PrP in its normal and disease-associated forms, while body layer and inner plexiform layer [57]. also observing multiple subcellular components of reti- In prion disease, results from our lab and others have nal rods, cones, bipolar cells and horizontal cells. Ultra- confirmed deposition of disease-associated PrP (PrPSc) structural studies were also done to confirm some of in human, bovine, primate, ovine, cervid, and rodent the observations. The results showed early deposition of retina by RT-QuIC, western blot and/or immunohis- PrPSc in two distinct, subcellular areas of retina as well tochemistry [5, 20, 21, 23, 24, 29, 41, 52, 54, 55, 58, 60]. as interesting novel events of photoreceptor damage in These studies represent examples from both natural dis - both these areas which preceded the death of both rods ease (human, elk, deer, sheep) and experimental disease and cones while sparing other nearby cell types. (primate, bovine, sheep, rodent). In these studies, dam- age appeared to affect primarily photoreceptor rods and Materials and methods cones, and damage to other retinal neuronal populations Ethics statement was not clear [50]. In our previous work, mouse photore- All mice were housed at the Rocky Mountain Laborato- ceptors were shown to die mainly by the process of apop- ries (RML) in an AAALAC-accredited facility in compli- tosis which coincided with PrPSc deposition [29, 54]. ance with guidelines provided by the Guide for the Care Previously, defective iron transport has been proposed and Use of Laboratory Animals (Institute for Laboratory as a mechanism of damage in prion-infected eye [47]. Animal Research Council). Experimentation followed Since photoreceptors rely heavily on iron-containing RML Animal Care and Use Committee approved proto- enzymes for the biochemical reactions of the phototrans- col 2016-042. duction pathways, disruption of iron transport by prion infection may also play a role in retinal damage. Such a Mice mechanism might explain the selective damage to photo- Retinas used in these experiments were obtained receptor cells by prions. from two strains of mice, C57BL/10SnJ and S triebel et al. acta neuropathol commun (2021) 9:17 Page 3 of 26 C57BL/6  J-TgGFP/RFP. 79A scrapie-induced retinal 95  °C. Staining for PrP was done using human anti- degeneration is very similar in these strains and was PrP monoclonal antibody D13 [33] which was obtained characterized in our previous publication [54]. An in- from tissue culture supernatants made in our labora- house breeding colony supplied C57BL/10SnJ mice,and tory from CHO cells expressing the D13 antibody con- C57BL/6 J–TgGFP/RFP were also bred in-house as pre- struct, which were kindly provided by Dr. R. Anthony viously described [54]. The GFP and RFP fluorescence Williamson, The Scripps Research Institute, La Jolla, was not important to the questions addressed here. All CA. D13 culture fluid was used at a dilution of 1:100 mice were group housed in transparent cages in a 12 h (diluted in PBS with 1% normal goat serum and 0.1% light (250-300lux) /12 h dark cycle and food and water Triton X-100) for 2 h at 37 °C. The secondary antibody were available ad libitium. was biotinylated goat anti-human IgG at 1:500 dilu- tion (Jackson ImmunoResearch, West Grove, PA.), and avidin-horseradish peroxidase was used with DAB as Scrapie inoculation model chromogen (DAB Map kit; Ventana Medical Systems, Scrapie inoculations were carried out as previously Tucson, AZ.). described [54]. Briefly, mice (4–6  weeks old) were For immunofluorescent staining, antigen retrieval injected intracerebrally (i.c.) in the left hemisphere with for all targets was performed using a Biocare Medi- 30  μl of a 1% (wt/vol) dilution of brain homogenate cal DC2002 Decloaking chamber with sodium citrate pools from C57BL mice terminally ill from 79A scra- buffer at pH 6.0(0.01  M) for 20  min at 120° C / 20 PSI pie. Brain homogenates contained 1.0 × 10 ID50 / 30 and cooled to 50 °C. For each of the following steps, 250– ul after they were diluted for inoculation in phosphate- 300  µl of solution was applied to each slide and covered buffered balanced saline (PBBS) pH 7.2, supplemented with a temporary plastic coverslip and incubated for a set with 2% fetal bovine serum (Hyclone, Logan, UT). amount of time. Tissues were blocked first with a normal The course of disease, details on the clinical symptoms donkey serum blocking solution (2% donkey serum, 1% and retinal degeneration were previously well docu- BSA, 0.1% Triton X-100, 0.05% Tween 20 in 0.01 M PBS) mented in the mouse strains used in this study in our pre- for 1  h at room temperature and then in 0.1  M Glycine vious publication [54]. Briefly, in the 79A mouse-adapted in 0.01 M PBS for 30 min at room temperature. Primary scrapie model, mice begin showing clinical signs con- antibodies (Table 1.) were diluted in donkey serum block- sistent with scrapie between 105 and 120dpi and reach ing solution and applied for 1  h at room temperature. clinical endpoint disease at approximately 160dpi. Thin - Alexafluor (ThermoFisher) secondary antibodies were ning of the retina begins around 118dpi and likely causes diluted to 1:250 in donkey serum solution and applied blindness by the disease endpoint, though this diagnosis for 1  h. In dual or triple stainings, primary antibodies is difficult to determine without conducting further tests. were applied simultaneously, as were secondary antibod- The 79A scrapie strain was previously compared to seven ies. After each antibody incubation, slides were washed other strains of mouse-adapted scrapie and shown to be 3 times in 1X PBS for 10 min. Coverslips were mounted the most retinal-tropic [15]. At pre-clinical and clinical with ProLong Gold with DAPI (Life Technologies) and time-points, mice were euthanized by isoflurane anesthe - examined and photographed using an Olympus BX51 sia overdose followed by perfusion with 10 ml of saline. microscope/Olympus CellSens software or using a confo- cal microscope as described below. Immunohistochemistry and Immunofluorescence For immunohistochemistry and immunofluorescence, Numbers of mice studied eyes were removed, placed in 10% neutral buffered for - The numbers of animals analyzed at each timepoint are malin for 3 to 5 days and then processed by dehydration presented in Table  2. Uninfected mice used as controls and embedded in paraffin as a single block. Next, 5 μm were of similar age to experimental animals, evidence of sections were cut using a standard Leica microtome, age-related retinal changes was not observed in the age placed on positively charged glass slides, and air-dried range of control animals used. overnight at room temperature. The following day slides were heated in an oven at 60  °C for 20  min. A Nomenclature and detection of PrP, PrPC and PrPSc Ventana automated Discovery XT stainer was used for Monoclonal antibody D13 was used in immunostain- deparaffinization, antigen retrieval and immunohisto - ing of tissue sections to detect PrP. In tissues of unin- chemical staining. fected mice, PrP detected was assumed to be the normal For immunohistochemical staining of PrP antigens PrP isoform, PrPC. In infected tissues, PrP detected were exposed by incubation in CC1 buffer (Ventana) in locations different from those seen uninfected mice containing Tris–Borate-EDTA, pH 8.0 for 100  min at was assumed to be disease-associated PrPSc, and PrP Striebel et al. acta neuropathol commun (2021) 9:17 Page 4 of 26 Table 1 Primary Antibodies used in immunofluorescent staining Antibody Specificity (antigen/cell type or structure) Dilution Host species Source D13 Prion protein 1:100 Human Ref, Matsunaga Cone Arrestin Arrestin 3/Cone photoreceptors 1:100 Rabbit Millipore, AB15282 Cone Opsin Red, Green opsins/Cone photoreceptors 1:100 Rabbit Chemicon, AB5404 Rhodopsin Rhodopsin/Rod photoreceptors 1:100 Rabbit Millipore, MABN15 GNAT1 G protein subunit alpha transducin 1 1:100 Rabbit Abcam, ab74059 /Rod photoreceptors GNAT2 G protein subunit alpha transducin 2 /Cone photoreceptors 1:100 Rabbit Thermofisher, PA5-22,340 GLUT1 Glucose transporter 1/Cell membranes of many cell types 1:100 Rabbit Abcam, ab115730 CtBP2 C-terminal binding protein 2/ribeye protein of ribbon synapses 1:100 Rabbit, mouse Invitrogen, PA-79086 Santa Cruz, sc-17759 PKCα Protein kinase C/Rod bipolar cells 1:100 Rabbit, mouse Invitrogen PA5-17,551, MA1-157 Calbindin Calbindin/Horizontal cells 1:100 Rabbit Abcam, ab108404 Rootletin Rootletin/photoreceptor rootlets 1:50 Mouse Millipore, ABN1714 Centrin3 Centrin3/cilia and basal bodies in photoreceptor inner segments 1:100 Rabbit Thermofisher, PA5-35,865 Secretagogin (SCGN) Cone bipolar cells 1:100 Rabbit Thermofisher, PA5-30,393 SCGN marks 8 of the 12 subtypes of mouse cone bipolar cells [13, 42] detected in similar locations to those found in uninfected 1.518. Image acquisition settings including laser power mice was assumed to be either or both isoforms. and gain were optimized for minimal background and cross-talk, and kept constant within an experiment for Quantification of bipolar and horizontal cells all timepoints and samples to enable direct comparisons. To quantify rod bipolar cells throughout the timecourse Stacks were collected with a lateral resolution of 43  nm of disease, two sections of retina from a mouse at each and z-spacing of 130  nm except for: quantification of timepoint were stained with DAPI, anti-PKCα primary anti-D13 and anti-CtBP2 signals, where stacks were col- antibody and secondary antibody Alexa Fluor 488 as lected with a lateral resolution of 71 nm and z-spacing of described above. The PKCα-positive rod bipolar cell 367  nm, with five representative fields of view acquired bodies were counted in four 20X fields per timepoint for each timepoint, and anti-Cone Opsin, anti-GNAT1, and averaged. Horizontal cell numbers were determined anti-Cone Arrestin, anti-Centrin3 which were collected by staining retinal sections with DAPI, anti-calbindin with a lateral resolution/z-spacing of 18  nm/250  nm, primary antibody and Alexa Fluor 488 secondary anti- 132  nm/250  nm, 65  nm/500  nm and 70  nm/500  nm body as described above. Calbindin-positive cell bodies respectively. were counted along two entire retinal sections from one mouse per timepoint. Cone bipolar cells were counted by staining retinal sections with anti-secretagogin antibody, Image processing and analysis which labels 8 of the 12 types of cone bipolar cells [13, Image stacks were exported from ZEN software and 42] and counting cell bodies on two retinal sections from deconvolved with Huygens Professional v. 20.04 (Scien- at least one mouse per timepoint (see figure legend for tific Volume Imaging, The Netherlands) using the CMLE n values). One-way ANOVA statistical analysis was per- algorithm, with SNR = 20 and a maximum of 40 itera- formed using GraphPad Prism software. tions. The deconvolved datasets were imported to Ima - ris x86_64 v.9.5.1 (Bitplane AG, Zürich, Switzerland) for segmentation, surface rendering, visualization, and quan- Confocal microscopy tification. The average number of CtBP2 ribbons based All samples were handled and chemically decontami- on anti-CtBP2 signal and amount of total integrated anti- nated according to established scrapie protocols in con- D13 signal was calculated per micron length of retina. sultation with RML Biosafety. Samples were imaged Data were imported into Microsoft Excel for compila- using a Zeiss laser scanning confocal (LSM 880) micro- tion, and statistical analysis was performed using Graph- scope driven by ZEN v.2.3 software (Carl Zeiss Micros- Pad Prism v 8.3.0 (La Jolla, CA). copy). A Plan Apochromat 63X/NA1.4 oil immersion lens was used, with immersion oil at a refractive index of S triebel et al. acta neuropathol commun (2021) 9:17 Page 5 of 26 Table 2 Number of retinas analyzed by immunofluorescence at timepoints during disease a b TimepointAntigen tested PrP (D13) Cone Cone Opsin SCGN CtBP2 PKCα Calbindin Rhodopsin GNAT1 GNAT2 Rootletin Centrin3 Arrestin Early (67,82,104) 9 3 3 3 3 2 1 3 1 1 3 3 Mid (118, 125, 129, 131) 8 3 2 2 2 3 2 3 1 nd 1 2 Late (144, 153, 159, 163, 165) 8 2 nd 2 3 2 1 2 3 nd nd 1 Uninfected 5 2 3 2 2 1 1 2 2 1 2 1 nd not done Timepoints are shown in days post inoculation (dpi) with 79A mouse adapted scrapie. In the 79A mouse-adapted scrapie model, mice begin showing clinical signs consistent with scrapie around 105-120dpi and reach clinical endpoint disease at approximately 160dpi. Thinning of the retina begins around 118dpi and likely causes blindness by the disease endpoint. Antigens detected with antibodies described in Table 1 Number of mice tested with each antibody at timepoint range shown Data not shown Striebel et al. acta neuropathol commun (2021) 9:17 Page 6 of 26 Electron microscopy sample preparation the optic tract and optic nerve [6, 16]. Progression of C57BL/10SnJ mice were perfused with 2% paraformalde- retinal infection was followed by immunohistochemistry hyde + 2% glutaraldehyde in 0.1 M Sorensen’s phosphate (IHC) or immunofluorescence (IF) with anti-PrP mono - buffer (Electron Microscopy Sciences, Pennsylvania). clonal antibody D13 to detect PrP. In uninfected PrPKO Eyes were enucleated and placed in fresh fixative for at mice, no PrP signal was detected by IHC or IF (Fig.  1a, least 30  min before further for processing. The ante - b). However, in uninfected mice expressing PrP, stain- rior portions of eyes were dissected and discarded. The ing of PrP was clearly detectable by IHC and IF in an remaining posterior eye cups were rinsed in phosphate irregular clumpy distribution in the outer plexiform layer buffer, followed by embedment in 2.5% low-melt agar (Precisionary, Massachusetts) made in PBS. 200  µm sections were cut with a VT1000S vibrating blade microtome (Leica Biosystems, Illinois). Sections were processed for transmission electron microscopy as fol- lows: postfixation with 0.05% osmium tetroxide + 0.08% potassium ferrocyanide in 0.1  M phosphate buffer for 1 h, rinsed with buffer, then dehydrated in a graded etha - nol series to 100%, infiltrated with LRWhite (Electron Microscopy Sciences, Pennsylvania) and polymerized overnight in homemade flat-embedding molds covered with aclar sheets at 50 °C in a vacuum oven. For electron microscopy studies the following num- bers of retinas from C57BL/10SnJ mice were taken at the given timepoints; uninfected (n = 22), 84 dpi (n = 2), 89 dpi (n = 4), 98 dpi (n = 2), 104 dpi (n = 3), 112 dpi (n = 2), 122 dpi (n = 4), 126 dpi (n = 2), 132 dpi (n = 3), 137 dpi (n = 3), 140 dpi (n = 1), 151 dpi (n = 3), 154 dpi (n = 1), 165 dpi (n = 3). These retinas were embedded in various resins (Durcupan, Araldite, HM20, and LRWhite) and examined. The LRWhite embedded retinas were selected for the imaging and comparisons shown in this paper. The numbers at each timepoint were; uninfected (n = 4), 104dpi (n = 1), 126dpi (n = 1), 132dpi (n = 1), 137dpi(n = 2), 151dpi(n = 2). Transmission electron microscopy Flat-embedded vibratome sections were excised and super-glued onto resin stubs such that ultramicrotomy sections would be in the desired orientation. 70 nm sec- tions were cut with a Ultracut UCT (Leica Biosystems, Illinois) ultramicrotome and picked up on Formvar coated 100 hex mesh copper grids (Electron Microscopy Sciences, Pennsylvania). Micrographs were acquired on a HT7800 (Hitachi, Oregon) operating at 80  kV with an XR-81B CMOS digital camera (AMT Imaging Systems, Fig. 1 PrPC expression in uninfected PrP knockout (PrPKO) and Massachusetts). wild-type ( WT ) retina using D13 antibody. a, b In PrPKO retina, normal host PrP (PrPC) was not detected by immunofluorescent stain (red) or immunohistochemical stain (brown). However, photoreceptors Results in OS of PrPKO retina showed red autofluorescence, a common but Detection of PrPSc in retina after intracerebral scrapie variable artifact of immunofluorescence studies of retina. c, d In WT injection. retina, PrPC was observed by both staining techniques in the OPL In the current experiments, prion infection of retina and IPL, and weaker levels were detected in the INL and IS. Scale bar was achieved by intracerebral inoculation of mice with in a = 20 µm. OS outer segment, IS inner segment, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, IPL inner 1.0 × 10 ID50 units of scrapie strain 79A. With this plexiform layer method, prions are known to spread to the retina via S triebel et al. acta neuropathol commun (2021) 9:17 Page 7 of 26 Fig. 2 Timecourse of PrP staining with D13 antibody in retina at various times after infection with strain 79A scrapie prions. a Uninfected mouse showing PrPC (magenta) mainly in OPL. b, c At 82 and 104 days post infection (dpi) the misfolded, disease-associated form of PrP (PrPSc) (magenta) can now be seen in the IS and OPL at progressively wider areas. d At 118 dpi, PrPSc is widespread in the IS and deposits are not restricted to discrete individual cells. Some small deposits are visible in the ONL and ONL is beginning to thin as photoreceptors die. e At 131 dpi, PrPSc is deposited in the IS and OPL, and ONL is much thinner. f PrPSc staining appears less at 162 dpi in IS and OPL, and ONL is dramatically thinner. Punctate PrPSc is present in IPL (arrowhead). Scale bar in a = 20 µm (OPL) and in a more diffuse pattern in the inner plexi - form layer (IPL) (Fig. 1c, d). This was similar to what has been described previously by others [10, 17, 19]. In addi- tion, a smooth faint signal was observed in the inner seg- ment (IS) of the photoreceptor layer (Fig.  1c, d) [54]. All of these sites were likely to represent normal cellular PrP (PrPC) as they were not seen in PrPKO mice (Fig. 1a, b). At 82  days post-infection (dpi), bright punctate PrP staining in patchy aggregates of varying sizes were seen in an irregular scattered distribution along the IS and OPL layers (Fig.  2b). This staining was likely to be dis - ease-associated PrPSc, as it was not seen in uninfected mice (Figs. 1c, d, 2a). At 104 and 118 dpi, the PrPSc stain- ing in the IS and OPL was more extensive, and at 118 dpi, PrPSc staining was often seen on the entire circum- ference of the IS and OPL (Fig.  2c, d). At 118 dpi, there was also a decrease in the width of the outer nuclear layer (ONL) indicating a loss of photoreceptor cell nuclei. At 131 dpi, the ONL was further reduced in size, and the PrPSc staining in the IS and OPL was decreased in inten- sity (Fig.  2e). At 162 dpi, all retinal layers were thinned (Fig.  2f ), and the ONL was now only 2–3 cells thick in most areas. Faint PrPSc staining was further decreased in the IS and OPL regions but was now slightly more detect- able in the inner plexiform layer (IPL). Interestingly, this type of neurodegeneration was not seen in uninfected PrPKO mice, which suggested that loss of PrP does not itself cause photoreceptor degeneration (unpublished results from our group). The time course of PrPSc deposition and retinal degen - eration seen in these experiments using indirect IF was similar to our previous data using detection of PrP by IHC [54]. However, the details of the PrPSc aggregate morphology was seen more clearly by IF. Since photore- ceptor rods and cones are the main cells normally pre- sent in the IS region and these cells also have processes ◂ Striebel et al. acta neuropathol commun (2021) 9:17 Page 8 of 26 Fig. 3 PrPSc accumulates first in cone photoreceptor inner segments. a At 67 dpi, rare small punctate PrPSc deposits (arrow) are present on cone photoreceptors marked with Cone Arrestin (green). Magenta in outer segment is autofluorescence of rhodopsin. b Cone opsin (white) marks cone outer segments with PrPSc (magenta) deposits associated mainly with cone inner segments (yellow arrow) at 104 dpi. The transition from faint to intense cone opsin staining marks the boundary between cone IS and cone OS (blue arrow). c Another cone-specific outer segment protein GNAT2 (green) shows obvious connection with PrPSc (magenta) staining cone inner segments at 104 dpi (arrows). d At 104 dpi, rod-specific marker, GNAT1 (green) stains rod inner and outer segments, but spares cones (arrowheads). PrPSc (magenta) accumulations are present in the dark GNAT1-free areas (arrows). e A separate experiment done with 22L scrapie strain shows same association of PrPSc with cone inner segments at 123 dpi, suggesting cone specificity is not strain-specific. Scale bar = 5 µm. a, b, d are confocal z-stacks, c, e are widefield images extending into the OPL, early PrPSc deposition was likely to be associated with one or both of these photoreceptor cell types. To study the cell types associated with PrPSc in this model, we used the IF method with dual staining for PrP and antibodies reactive with cone- or rod-specific proteins (Tables 1, 2). Deposition of PrPSc occurs first in cone photoreceptors At 67 dpi, PrPSc was detected in a few individual cells in the IS region, and this PrPSc was associated with detec- tion of cone arrestin in the same individual cells (Fig. 3a). Similarly, at 104 dpi, PrPSc staining was associated with detection of cone opsin (Fig.  3b) and GNAT2 (trans- ducin alpha-2), another cone-specific marker (Fig.  3c). In both of these examples, the PrPSc staining was in the inner portion of the IS, whereas both the cone opsin and GNAT2 proteins were mostly in the outer segment (OS), i.e. distal to the PrPSc but appearing to be in the same individual cells as the cone-specific marker proteins. In contrast, GNAT1 (transducin alpha-1), a rod-specific protein, did not co-associate with PrPSc at 104 dpi which was located in the dark spaces not stained by GNAT1, i.e. cones (Fig. 3d). Because scrapie strains have been shown to show cell-specific infectivity [8], we tested the 22L strain, which also targeted cone photoreceptors before rods (Fig.  3e). These observations indicated that PrPSc deposition appeared first in cone photoreceptor inner segments. Additional studies showed that rods were also infected starting around 118 dpi (see below). Retinas were next studied by confocal microscopy to look for cone damage by dual staining with anti-PrP and S triebel et al. acta neuropathol commun (2021) 9:17 Page 9 of 26 Fig. 4 Early PrPSc accumulation associated with cone photoreceptor damage. a Uninfected retina shows normal cone morphology and distribution of cone opsin (green) in outer segment. b At 104 dpi early PrPSc deposits are associated with cone inner segments (arrow) and some PrPSc appears to localize with cone opsin (white areas in × 2 magnified inset). c At 118 dpi, a swollen, dystrophic cone with PrPSc is seen (yellow box) and some PrPSc is also present in rods (blue arrow). d Serial confocal sections spaced 0.5 µm apart, magnify the swollen cone inner segment from c and suggests the PrPSc is inside the inner segment along with mistrafficked cone opsin. Scale bar in a and c = 5 µm. Scale bar in d series = 2 µm. a–c are confocal z-stacks anti-cone opsin. At 104 dpi, swelling in the inner segment (Fig.  4d). However, the detection of cone opsin outlin- portions of some PrPSc-positive cones was observed ing the outer edge of the cytoplasm was not seen in cells (Fig. 4b). This swelling was never seen in uninfected reti - lacking PrPSc, suggesting that this unusual distribution nas (Fig. 4a), and therefore appeared to be a sign of dam- of cone opsin may be a manifestation of prion-induced age due to prion infection. This swelling was typical of damage in this cone. necrotic cell death as described in cone degeneration in a rd10 mouse model [37]. At 118 dpi abundant PrPSc was Detection of PrPSc‑associated damage in the inner seen in cones and in the inner segment cells lacking cone segment opsin (Fig. 4c), indicating that prion infection had spread To understand how PrPSc deposition might be caus- to rods by this time. Six serial optical sections through ing damage to cone (and rod) photoreceptors we looked the cone outlined in Fig. 4c revealed that PrPSc (red) and for association of PrPSc with additional key inner seg- cone opsin (green) were mostly separated inside this cone ment structures. First, the relationship between PrPSc (See figure on next page.) Fig. 5 Early PrPSc accumulation and damage in the inner segment. a Cartoon of cone photoreceptor shows key structures related to PrPSc deposition. cc connecting cilium, bb basal body, r rootlet, m mitochondria, cp cone pedicle, rs ribbon synapse. b Anti-centrin3 antibody (green) marks the connecting cilium and basal body (small green dots) of all photoreceptors. PrPSc (magenta) accumulation at entrance to connecting cilium (arrow), magnified in inset. c Serial 0.5 µm confocal sections (c1–6) showing relative localization of PrPSc, cilium and basal body. d Cone arrestin (green) staining of uninfected cone photoreceptor. Arrow indicates likely position of connecting cilium. e 118 dpi shows retina stained for cone arrestin (green) and PrP (magenta). Asterisks mark cones missing outer segments, yellow arrows point to position of connecting cilia and associated PrPSc (magenta) deposit, and arrowheads show dystrophic outer segments. f, g Confocal analysis showing xy and yz planes of swollen cones from e confirm the presence of PrPSc at the location of the connecting cilium (arrow) between the IS and OS. h In an uninfected retina, anti-rootletin (green) antibody stains the rootlets of photoreceptors. i, j At later days post infection, PrPSc (magenta) is increased, rootlets are fewer in number and misshapen. Scale bars: b, e = 5 µm; c = 1 µm; d = 2 µm; f, g = 2 µm; h = 10 µm. b, d, e Maximum intensity projections (MIP) of confocal z-stacks h, i, j are widefield images Striebel et al. acta neuropathol commun (2021) 9:17 Page 10 of 26 deposition and the cilium connecting the inner and the detected near the entry point of the connecting cilium outer segments of the photoreceptor cells (Fig.  5a) was (Fig.  5b). In some cases, PrPSc was directly adjacent to examined by dual staining for PrP and Centrin3, a pro- the cilium, as confocal analysis showed that PrPSc, basal tein located within the cilia of rods and cones. Early in body and cilium could be found within the same section disease, at 104 dpi, deposition of PrPSc could often be (Fig.  5c). The deposition of PrPSc at this site may have S triebel et al. acta neuropathol commun (2021) 9:17 Page 11 of 26 Fig. 6 Timecourse of damage to cone photoreceptors. a Anti-Cone arrestin (green) stains cone photoreceptor, inner segments (yellow arrow) and pedicles (green arrow) in an uninfected mouse. Anti-PrP antibody D13, stains PrPC (magenta) at the base of cone pedicles in the OPL (purple arrow). Autofluorescence is also present in the OS. b At 67 dpi cone arrestin and PrP are distributed similarly to uninfected retina (magenta arrow). c Small deposits of PrPSc are present in the IS at 104 dpi and are associated with cones (arrows). d At 118 dpi, the number of cones is reduced (green), most cone pedicles have disappeared from the OPL and the remaining cones are dystrophic (yellow arrow), missing their outer segments. PrPSc is widespread in the IS and ONL. e Few cones (green) remain at 131 dpi, pedicles (arrow) are much smaller in size and associated with dense clumps of PrPSc (arrowhead). f Cones are not detectable at 162 dpi, and ONL is 2–3 nuclei thick, suggesting rod photoreceptors have mostly died. A macrophage or microglial cell (yellow arrow) containing PrPSc is present in the remnants of the OS. Scale bar in a = 20 µm, applicable to all images affected the cilium’s ability to transport phototransduc - To test this idea, we stained for cone arrestin, a cone- tion proteins between the IS and OS. Ciliary dysfunction specific phototransduction protein, known to be traf - is known to cause photoreceptor death in other forms of ficked between the IS and OS [2, 49]. In uninfected retinal degeneration [43]. retinas, cone arrestin was present in the OS, IS, ONL (Fig.  5d) and pedicles (Fig.  6). The cone outer segments Striebel et al. acta neuropathol commun (2021) 9:17 Page 12 of 26 of a 118 dpi retina appeared shrunken and malformed and smaller in size (Fig.  6d). PrPSc was abundant in the compared to an uninfected cone and in some cases, IS region on both cones and rods, and PrPSc staining outer segments were absent (Fig.  5d–g). Inner segments in the OPL was both punctate and diffuse and appeared of cones in the 118 dpi retina were often swollen, and to be independent of cone pedicles (Fig.  6d). At 131 and PrPSc deposits could be found at the constriction point 162 dpi, PrPSc staining in IS and OPL remained strong between the IS and OS, i.e. the location of the connect- and was similar to 118 dpi. However, at these times, most ing cilium (Fig.  5d–g). Together these data suggest that cone pedicles were gone, and the ONL showed thin- deposition of PrPSc in the inner segment may have ning due to a large loss in both cone and rod cell nuclei affected the ciliary transport of cone arrestin and/or (Fig. 6e, f ). These data suggested that PrPSc deposition in other proteins. We also investigated the association of the IS and OPL regions might have induced pathogenic PrPSc deposits with another prominent inner segment effects resulting in death of both cones and rods. structure, the rootlet, which functions to stabilize the connecting cilium [59] (Fig. 5a). For this, we used an anti- rootletin antibody together with D13 anti-PrP. As PrPSc Loss of cone pedicles and rod spherules visualized appeared to accumulate in the inner segment at 104 and by GLUT1 staining 118 dpi, rootlet morphology changed, and the density of Glucose transporter 1 (GLUT1) is known to be an excel- rootlets decreased (Fig. 5h–j). While PrPSc was not usu- lent marker for cell surface visualization in tissues. There - ally associated with rootletin, these data suggested that fore, we used anti-GLUT1 and anti-PrP dual staining to PrPSc accumulation may have damaged the inner seg- study the details of PrP localization relative to cone pedi- ment and its structures including the rootlets. cles and rod spherules in the OPL area following prion retinal infection. Prior to prion infection, cone pedicles were detected as triangular dark shaped spaces outlined Detection of PrPSc‑associated damage in the outer by GLUT1, and just vitread to many of these pedicles a plexiform layer patch of clustered coarse PrPC staining was observed The early accumulation of PrPSc and damage to cone (Supp Fig. 1a). In addition, in the same figure, numerous inner segments led us to examine cone photoreceptors rod cell endings, i.e. spherules, were also seen, and these more closely. We used dual staining with anti-PrP plus appeared to be smaller rounded dark shapes outlined by anti-cone arrestin to follow the progression of PrPSc GLUT1 mainly sclerad (toward the sclera) to cone pedi- deposition and changes in overall cone morphology over cles (Additional file  1: Fig.  1a). Following prion infec- the time course of prion infection from 67 to 162 dpi. tion at 118 dpi, cone pedicles were difficult to recognize In uninfected retina, small dense patches of PrP stain- or had disappeared and rod spherules were decreased ing were seen in the OPL just vitread (toward the vit- in number and size (Additional file  1: Fig.  1b). Later, at reous) to many of the cone pedicles expressing cone 153 dpi, most of the spherules and pedicles were miss- arrestin (Fig.  6a), and this was assumed to be PrPC as it ing, and all the photoreceptor layers (OS, IS, ONL, OPL) was seen in uninfected mice. Similar staining was also were degenerated (Additional file  1: Fig. 1c). These results seen at 67 dpi (Fig.  6b). At 104 dpi, PrP staining in the demonstrated progressive damage to both rods and cones OPL was seen basal to some of the cone pedicles simi- associated with the presence of PrPSc deposition sclerad lar to uninfected mice. More obvious changes were seen to and within the OPL. at 118 dpi, where cone pedicles were fewer in number (See figure on next page.) Fig. 7 In uninfected retina, PrPC expression is concentrated at the base of cone pedicles. a In the OPL, PrPC (yellow arrowheads) staining is present at the base of cone pedicles highlighted with anti-cone arrestin (arrows). b Electron micrograph ( TEM) of the OPL showing a cone pedicle (cp), with multiple round mitochondria (m) and three ribbon synapses (arrows). A yellow dotted line surrounds an estimated region of PrPC from a, which contains many dendritic processes from bipolar and horizontal cells. Some dendrites synapse at ribbons (yellow arrowheads) or make flat contact type synapses (magenta arrowhead) with cone pedicles. Ribbon synapses (arrows) are also visible in rod spherules (r). c Confocal z-stack of uninfected retina shows gross relationship of PrP to rod bipolar cells (PKCα) and horizontal cells (Calbindin). d–h A high magnification single 0.130 µm optical section taken from area in box in panel c, shows PrP is not associated with PKCα and that some calbindin-positive processes (arrow) lie within the PrP-positive area, but association is not clear. i Confocal z-stack image of retina stained for cone bipolar cell dendrites (SCGN), ribbons (CtBP2) and PrP, shows a dense patch of PrPC vitread to a cluster of short ribbons (box) in a cone pedicle. j–l A single 0.130 µm optical section taken from within the box in panel e shows PrPC in close association with SCGN-positive dendrites, and some colocalization is seen as white (arrows). Scale bar in a = 10 µm, b = 1 µm, c = 2 µm, d = 1 µm, i = 4 µm, j = 0.5 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 13 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 14 of 26 Comparative detection of PrP, ribbon synapses and other results suggested that PrPC is highly expressed very near structures in uninfected retina cone pedicles and thus may explain the early accumula- In the photoreceptor layer bordering the OPL, cone and tion of PrPSc in cones versus rods. rod photoreceptor cells are connected with processes from horizontal cells and rod or cone bipolar cells to Damage to ribbon synapses seen by staining form ribbon synapses, which have unique electrophysi- with anti‑CtBP2 ological and morphological features [36]. Initially we Using dual staining immunofluorescence combining compared the locations of several of these structures in anti-PrP with anti-CtBP2 antibodies specific for rib - uninfected mice to determine their normal morphology, bon structures, we studied the development of PrPSc- as observed by both confocal and electron microscopy. associated alterations on ribbon synapse components Dual staining of an uninfected mouse for cone arrestin over time after prion infection. In both uninfected and and PrP showed PrPC located in large dense clusters vit- 82 dpi retinas, ribbon synapses were seen sclerad to the read to cone pedicles (Fig.  7a). Electron microscopy of OPL as a continuous band of linear or horseshoe-shaped the OPL demonstrated rod spherules with single large structures, and PrP was mostly in bunched loose aggre- ribbon synapses and cone pedicles containing multiple gates vitread to the lower-most ribbons (Fig.  8a, b). At short ribbon synapses (Fig.  7b). Notably, comparisons of 104 dpi, ribbons appeared to be similar to 82 dpi, but EM and light microscopy images suggested that areas of PrP staining was slightly more prominent in the ribbon intertwined dendrites from bipolar and horizontal cells area than previously (Fig.  8c). At 118 dpi (Fig.  8d), rib- (Fig.  7b) corresponded to the clustered areas of PrPC bons were significantly decreased in number (Fig.  8l), and seen in Fig. 7a. PrP was deposited among the ribbons rather than below In order to understand the association of PrPC with them (Fig.  8d, j) compared to uninfected retina and ear- these structures, uninfected retina was stained simul- lier timepoints. This PrP was very likely PrPSc as staining taneously with antibodies against PrP and proteins was not seen in this location in uninfected retina. Inter- expressed by horizontal and bipolar cells (Fig.  7c–f ). estingly, at higher magnification, small accumulations of Confocal microscopy analysis of full z-stacks and individ- PrPSc were detected within the horseshoe-like curves of ual 130  nm optical sections was used to reveal the rela- many of the ribbons (Fig.  8j, k). Moreover, PrPSc accu- tive locations of the processes and PrP. Dendrites of rod mulations were never found directly on ribbons, which bipolar cells were marked with anti-PKCα and although is where synaptic vesicles are located. This suggested they were present very near the PrP-stained area beneath that PrPSc was on the postsynaptic side of the ribbon cone pedicles, they were not directly associated with synapses (Fig.  8j). At later time points, 131, 153 and 163 PrP (Fig.  7c, d). Likewise, Calbindin-positive horizontal dpi, ribbons were progressively destroyed, as were the cell processes were not clearly associated with the dense cone and rod cell nuclei in the ONL (Fig. 8e–g, l). Disap- clusters of PrP (Fig.  7c–h). Next, we triple-stained with pearance of ribbons correlated with significant increases anti-PrP, anti-CtBP2, a marker for ribbons and anti- of PrPSc at 118 dpi (Fig.  8l vs m). Thus, the appearance Secretagogin, a protein expressed in 8 of the 12 types of of PrP within the ribbon horseshoes correlated with the cone bipolar cells [13, 42]. In this experiment, dendrites onset of loss of ribbon synapses and suggested this PrPSc of cone bipolar cells expressing secretagogin (SCGN) deposition might be responsible for this damage. showed a close association with PrPC (Fig.  7i–l). These (See figure on next page.) Fig. 8 Timecourse of disappearance of ribbon synapses in OPL. a, h Uninfected retina shows many horseshoe-shaped ribbons stained with CtBP2 (green), sclerad to PrPC (magenta) densities. Most ribbons are not associated with PrPC see panel h. b, i At 82 dpi, the number of ribbons is similar to uninfected, but at cone pedicles PrP appears brighter and more punctate suggestive of new PrPSc accumulation. c 104 dpi retina appears similar to 82 dpi. d, j 118 dpi, ribbons in OPL appear disorganized and are significantly reduced in number, and PrPSc deposits are visible within horseshoe arcs, but not touching, most ribbons (j, yellow arrows). e, k At 131 dpi PrPSc accumulation has peaked and most ribbons have disappeared. Remaining horseshoe-shaped ribbons show association with PrPSc (k, yellow arrow). f, g At late dpi, ONL has thinned dramatically, few ribbons remain, PrPSc is widespread. l Graph shows # of ribbons/µm of OPL of scrapie infected retina at dpi vs uninfected. Each dot represents count from one field of view. m Quantitation of PrP (magenta) fluorescence integrated intensity/µm of OPL of scrapie infected retina at dpi vs uninfected. Each dot represents integrated intensity measurements from one field of view. Lines median value, ns not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 00.0001. Statistics were calculated using one-way ANOVA with multiple comparisons. Scale bars: a–g = 4 µm; h–k = 1 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 15 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 16 of 26 Association of PrPSc to rod and cone bipolar cells dendritic boutons disappeared along with the cell bod- near ribbon synapses ies. This was in marked contrast to the rods and cones The location of PrPSc, adjacent to but not touching rib - themselves which were mostly eliminated at this time bons, suggested these deposits were associated with the post-infection. postsynaptic portion of the ribbon synapses. Dendrites of The dendrites of cone bipolar cells synapse mainly with rod and cone bipolar cells and horizontal cells are known cone photoreceptor cells in the OPL and can be marked to make up the postsynaptic elements of ribbon synapses using anti-Secretagogin (SCGN) antibody [13]. To check in the OPL [36]. Here, investigation of the possible inter- for an association of PrPSc deposits with cone bipolar actions of PrPSc with rod bipolar cells was done by dual cell dendrites we triple-stained retina for PrP, SCGN and staining with anti-PrP and anti-PKCα, which is specific CtBP2 (Fig. 11a–d). As was also shown in Fig. 7, PrPC in for rod bipolar cells. uninfected mice was found to be in close association with In uninfected and 82 dpi prion-infected mice, cell SCGN-positive dendrites under clusters of short ribbons bodies of rod bipolar cells were seen just vitread to the (Fig.  11a). Likewise, single confocal sections, showed OPL and dendritic processes from these cells extended PrPC intermingled among SCGN-positive cone bipo- up through the OPL to the area of the cone pedicles and lar cell dendrites (Fig.  11b). In infected mice at 104 and rod spherules (Fig. 9a, b). PrP was detected mostly in the 118 dpi, staining with anti-PrP antibody D13 revealed an OPL sclerad to the rod bipolar cells (Fig. 9a, b). However, increase in PrPSc among SCGN-positive dendrites where starting at 104 dpi and continuing at 118 dpi, PrP depos- they make synaptic connections with cone pedicles its were seen in a new location, at the tips of the dendritic (Fig. 11a–f ), and this deposition was coincident with the processes of the rod bipolar cells that synapse with rib- previous finding of PrPSc on the tips on rod bipolar cell bons (Fig.  9c, d). This was also visible in more detail by dendrites beneath the horseshoe-shaped ribbon synapses confocal microscopy (Compare Fig.  9f vs g). Because present in rod photoreceptors (Fig. 9h–l). The timing and deposits in these locations were only seen after prion location of this increase in PrP staining suggested these infection, they were thought to be PrPSc. deposits were likely to be the disease-associated form of At higher magnification this PrPSc had the appear - PrP. Despite the destruction of most ribbon synapses in ance of beads on a string extending along the processes cones and the disappearance of cone bipolar cell den- (Fig. 9h–k), and PrPSc seemed to displace or overlap the drites, the number of SCGN-positive cone bipolar cells original PKCα staining of the processes. This was spe - did not decrease during disease (Fig.  11g–k). Thus, both cific for infected retinas at the 104–118 dpi time-points rods and cones were uniquely sensitive to prion-induced (Fig. 9c, d, g) and was not seen in the uninfected or 82 dpi damage compared to other retinal neurons. retinas (Fig. 9a, b). The precise location of PrPSc deposits, at the ribbon synapse on the tips of rod bipolar cell den- Identification of PrPSc near horizontal cell dendrites drites, was confirmed using confocal analysis of a section and boutons triple stained for PrP, PKCα and CtBP2 (Fig.  9l). At late Along with dendritic processes from rod or cone bipolar times in disease (131 dpi), the majority of the rod bipolar cells, dendrites from two horizontal cell neurons (HC) dendritic structures had degenerated (Fig.  9e), however, make up the triad of processes that invaginate the pho- rod bipolar cell bodies appeared to be decreased only by toreceptor at each ribbon synapse [36]. Therefore, HC about 50% starting at 118 dpi and they did not decrease were also studied during prion retinal infection using further after this time (Fig. 10a–e). Axons, dendrites and dual staining with anti-PrP and anti-calbindin, which (See figure on next page.) Fig. 9 PrPSc accumulation at tips of rod bipolar cell dendrites and disappearance of dendrites. Epifluorescent microscopy analysis. a In uninfected retina, PrP (magenta) (blue arrowhead) and PKCα-positive rod bipolar cells (green) are seen near each other in the OPL. Many boutons (blue arrow) are present at dendritic tips of rod bipolar cells. b At 82 dpi, OPL appears similar to uninfected. Asterisk marks a rod bipolar cell body in INL. c At 104 dpi, PrP staining appears at dendritic tips. d At 118 dpi, most PKCα-positive dendrites display PrP at tips of processes (boutons) (arrow). e By 131 dpi most dendrites and dendritic boutons have disappeared, bipolar cell bodies remain (arrow), and PrPSc is distributed in large aggregates. Confocal analysis. f Maximum intensity projection shows PKCα and PrP staining of uninfected retina. Arrows indicate PKCα-positive rod bipolar cells with obvious boutons at the ends of dendrites. g A 118 dpi retina shows PrPSc arranged in bead-like deposits on dendrites, often replacing boutons. Magnification of boxes on left and right are shown below in panels h–k. h Confocal image shows association of PrPSc with dendrites and dendritic tips in area of right box from g. i–k Confocal images show merged, PrP and PKC staining of left box area in g. l In a high magnification of confocal image, anti-CtBP2 antibody marks ribbons (arrows) and confirms PrPSc deposits (magenta) on tips of dendrites (green) invaginating, but not touching, ribbons. Scale bars: a–e = 5 µm; f, g = 3 µm; h–k = 2 µm; l = 0.5 µm S triebel et al. acta neuropathol commun (2021) 9:17 Page 17 of 26 Striebel et al. acta neuropathol commun (2021) 9:17 Page 18 of 26 is expressed on the cell body, dendrites and boutons of HC. In uninfected mice, HC bodies were vitread to the OPL, and processes and boutons were mostly in the OPL (Fig.  12a, b), and the clumped aggregates of PrP seen previously did not associate with either the HC or their processes. After scrapie infection at 104 dpi, HC bod- ies, processes, and boutons were unchanged (Fig.  12c, d). As expected, the PrP clumps seen previously in unin- fected mice were missing, and PrP was now mostly dis- seminated in small punctate or linear deposits in the OPL which were near, but slightly separated from the HC boutons (Fig.  12c, d). This difference in PrP morphology suggested that this material was PrPSc. At 118 dpi, the HC boutons were less distinct and PrPSc was still simi- lar to 104 dpi (Fig.  12e, f ). There was still no association between HC bodies and PrP. This relationship was con - firmed using confocal analysis of a triple stained section (Fig.  12i–l), where at high magnification, it was evident that calbindin-positive boutons did not contact PrPSc (Fig.  12j), whereas PrPSc was in contact with PKCα, the marker for rod bipolar cells (Fig.  12k). At 131 and 153 dpi the HC processes and the entire OPL region was atrophied, and damage appeared to be severe. However, PrPSc deposits were more consolidated and surprisingly, most HC bodies were intact (Fig. 12g, h, m). In summary, at 118 dpi there was evidence of deposi- tion of PrPSc in the OPL, at the ribbon synapses of rod photoreceptors. While PrPSc was not directly in contact with the presynaptic ribbons (CtBP2), it was associated with the postsynaptic elements of the synapses. PrPSc was found on the dendritic boutons of rod bipolar cells (PKCα), where they invaginated ribbons synapses and on cone bipolar cell dendrites (SCGN) adjacent to ribbon synapses. HC boutons (Calbindin) did not show accumu- lations of PrPSc. These data suggested that prion infec - tion and PrPSc deposition on rod and cone bipolar cell processes might have a toxic effect on ribbon synapses leading to damage and death of both rods and cones. Electron microscopy of outer plexiform layer confirms loss of synapses Fig. 10 Changes in rod bipolar cells during infection. a anti-PKCα staining (green) in an uninfected mouse shows normal distribution Analysis of the OPL by transmission electron microscopy and morphology of rod bipolar cells. Numerous dendrites (arrow) confirmed many of the findings determined by confocal reach upward to synapse with photoreceptors in ONL from each microscopy. TEM was advantageous in that it allowed a cell body (arrowhead). b At 104 dpi rod bipolar cells appear similar view of all structures simultaneously. Sections from unin- to uninfected mouse. c By 118 dpi the number of bipolar cells has fected mice showed many photoreceptor axon terminals dropped to about half vs uninfected and some cell bodies have few if any associated dendrites (arrowhead). d Late in disease, at 163 (rod spherules and cone pedicles) containing ribbon dpi, rod bipolar cell number is significantly reduced compared to synapses, most were connected to a triad of invaginat- uninfected retina, and few dendrites reach into OPL (arrow). e The ing dendritic processes from horizontal and bipolar cells. number of rod bipolar cells per 100 µm of retina over time. Each dot Rod spherules contained a single large, round mito- represents counts from one field of view, line is median, ns = not chondrion and a ribbon synapse, while cones contained significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p 0.0001 statistics by one-way ANOVA with multiple comparisons test. Scale bar in multiple small mitochondria and 3–5 ribbon synapses a = 20 µm (Fig.  13a). Early in disease (104 dpi), overall morphology S triebel et al. acta neuropathol commun (2021) 9:17 Page 19 of 26 was similar to uninfected mice (Fig.  13b), however in the IS and OS regions of photoreceptors as has been pro- some cone pedicles, swollen, dystrophic dendritic pro- posed in retinitis pigmentosa and ciliopathies [34, 40]. In cesses were noted around ribbon synapses, suggesting support of this mechanism, we found unusual distribu- that cone photoreceptors were beginning to degenerate tions of opsin and cone arrestin (Figs. 4, 5, 6), suggesting (Fig.  13b). At later timepoints (Fig.  13c–f ), typical triad interference with the normal trafficking of these proteins ribbon synapses were rarely noted and both rod spher- between the IS and OS. ules and cone pedicles, if present, were often dystrophic. Beginning at 104 dpi, abnormal PrPSc deposition was Frequent whorl structures, suggestive of autophago- also detected in the OPL region near and within the somes, were noted adjacent to, or in place of ribbon cone pedicle and rod spherules. Although this process synapses in both rods and cones. In short, these ultras- appeared to begin later than the process in the IS region, tructural findings supported the conclusions from fluo - it still coincided with the detection of photoreceptor cell rescent microscopy that at late timepoints all elements of pathology at 104–118 dpi and thus might play a role in ribbon synapses had degenerated. damage and death of rods and cones. In the OPL of unin- fected mice, PrP was normally detected on or near the Discussion dendrites of cone and rod bipolar cells located vitread to In the present study, we followed the timing and loca- rod spherules and cone pedicles, but after prion infec- tion of PrPSc deposition in retina following intracere- tion, PrP detection shifted to sites on the tips of bipolar bral prion injection in mice using immunofluorescence cell dendritic boutons, where they invaginate rods to and electron microscopy. The earliest PrPSc deposits form ribbon synapses. Likewise, PrP deposits increased in retina were found at 67 dpi and were located in the among the dendrites of cone bipolar cells where they IS region of cone photoreceptor cells. At 104 dpi PrPSc synapse with cone pedicles. This different PrP distribu - could be detected adjacent to the cilium, which con- tion suggested that this material was disease-associated nects the IS and OS portions of each cone and rod cell. PrPSc. During this same time, starting at 104–118 dpi These events were followed by cell swelling and altera - and continuing up through 153 dpi, there was progres- tion of organelles, such as rootlets, at 104 dpi, and pro- sive loss of ribbons in rods and cones associated with loss gressive loss of cone and rod cell nuclei in the ONL of over 90% of photoreceptor cone and rod nuclei in the starting at 118 dpi. Similar events with the exception ONL. The mechanism of damage induced by the pres - of swelling were seen in rods on a time course about ence of PrPSc on the tips of bipolar cell dendritic bou- 14 days behind cones. tons is not clear. However, the accumulation of protein The association of PrPSc with cilia may be an impor - aggregates at sites of synaptic transmission might con- tant clue to the pathogenic process of prion infection in found the synaptic transmission process or alter neuritic retina. This abnormal accumulation might interfere with connectivity [53, 61]. Similar speculations have been sug- or damage the ciliary protein transport system between gested for hippocampal synapses in prion-infected brain (See figure on next page.) Fig. 11 Association of PrPSc with cone bipolar cells. a Confocal z-stack of an uninfected retina shows secretagogin-positive cone bipolar cell bodies (magenta arrows) and their green dendritic processes associated with PrP (magenta) in dense clusters (yellow arrows) which include short ribbon synapses (yellow). b A single confocal section taken from the box in panel a shows PrP in close association with cone bipolar cell dendrites (green), beneath clusters of short ribbons (yellow) typically found in cone pedicles. Horseshoe-shaped ribbons in rod spherules (blue arrows) are not associated with PrP or SCGN. c In a confocal z-stack, at 118 dpi, cone bipolar cell bodies (magenta arrow) appear normal but PrP staining is increased in the OPL. d A single confocal section taken from the box in panel c shows increase in PrP among the SCGN-positive dendrites with some PrP overlapping SCGN, suggesting colocalization (white arrow). PrP is also present within the horseshoe-shaped ribbon synapses in rods (blue arrows). e, f Widefield images at 104 dpi show numerous examples of PrP clusters localizing with the ends of SCGN-positive dendrites (blue arrows) originating from cone bipolar cells (magenta arrows). g Number of cone bipolar cells per mm vs dpi. Each symbol represents the count from one field of view. Different symbols represent different animals. Line represents median. h–k anti-Secretagogin stained retina from uninfected and three different dpi show relative numbers of SCGN-positive cells. ns = not significant, statistics by one-way ANOVA with multiple comparisons test. Scale bars: a, c = 4 µm; b, d = 1 µm; e,f = 10 µm; h–k = 25 µm Striebel et al. acta neuropathol commun (2021) 9:17 Page 20 of 26 S triebel et al. acta neuropathol commun (2021) 9:17 Page 21 of 26 [12, 27, 48]. However, this has not previously been sug- cones [1, 35, 44]. Interestingly, cones make up only 3% of gested for retinal photoreceptor ribbon synapses, where photoreceptors in mouse retina, so the cone cell specific - the molecular components can be visualized in better ity in our study was surprising. This could be due to local detail. differences in PrP expression which might influence the Interestingly other nearby retinal neurons were dam- timing of PrP conversion to PrPSc. We have not been able aged less by prion infection. For example, cell bodies of to measure differences in normal PrP expression in the cone bipolar cells remained constant throughout disease, IS region of uninfected rods and cones. However, in the rod bipolar cell bodies decreased by 50%, and horizontal OPL region, uninfected mice have high PrPC expression cell bodies decreased less than 10%. However, dendrites in the dendritic processes of bipolar cells located directly in the outer plexiform layer of all these cells showed sig- beneath the cone pedicles. These are a mixture of rod and nificant damage. Ganglion cells and amacrine cells also cone bipolar cells which all appear to deposit PrPSc on did not appear to be damaged. These latter two cell types their dendrites after prion infection, but while there are were not specifically stained and counted, but cellularity hundreds of synaptic connections to each cone pedicle, in the regions where they typically reside appeared to be each rod spherule has no more than seven [38]. This dif - normal. One explanation for the extreme sensitivity of ference might increase the efficiency of prion travel to the cone and rod cells to prion infection might be the PrPSc IS regions of cones vs rods. In addition, prion travel from association and blockage of cilia which are not found in ganglion cells to cone cells can be direct, whereas inter- the other above-mentioned retinal neurons. Alterna- mediate cells, such as amacrine cells, are often involved tively, the level of PrPC expression in various cell types in connecting ganglion cells to rod bipolar cells [14]. The might also play a role in these differences. In contrast, more direct route to cones might increase the tempo of photoreceptor degeneration was not observed in our prion infection, favoring cones over rods. studies of prion-infected transgenic mice expressing GPI- The damage mechanism(s) seen in prion-infected ret - anchorless PrP, suggesting expression of anchored PrPC ina and brain are not well understood [25]. Our earlier on the cell membrane may be critical to PrPSc-induced studies indicated the presence of apoptotic cells in the damage to certain cell types [30]. ONL after prion infection [29, 54]. Since these observa- Another question raised by these results is why the tions were made later in the disease course, these cells cones preceded rods in the PrPSc deposition and dam- were likely to be mostly rods which are 32 × more abun- age process. In some rare retinal degenerations, known dant than cones in mouse retina. However, in our studies as Cone and Cone-rod dystrophies, cones also precede using cone opsin to detect cones, damaged PrPSc-posi- rods in damage and death, however it is more common tive cones in the IS region at 104 and 118 dpi were swol- for rods to begin the degenerative process, as in retinitis len suggesting a necrotic rather than apoptotic cell death pigmentosa [39, 56]. In other neurodegenerative diseases process (Figs. 4, 5, 13). Interestingly, this same pathology affecting retina, such as AD and PD, photoreceptors are preceding cone necrosis is similar to previous results of not typically affected first and specificity for cones or Murikami et  al. [37] studying the human retinitis pig- rods is not clear, though a recent study found accumula- mentosa rd10 mutation in a mouse model and in human tions of phosphorylated tau specifically in ageing primate retinitis pigmentosa patients. (See figure on next page.) Fig. 12 Timecourse and confocal analysis of PrP association with horizontal cells. a,b In uninfected mouse retina, anti-calbindin labels horizontal cell bodies (yellow arrow), dendrites (blue arrows) and boutons (arrowheads). PrPC magenta densities (long yellow arrow) are surrounded by horizontal cell boutons and dendrites. c, d At 104 dpi, calbindin-positive boutons (arrowheads) are near, but not touching, new PrPSc deposits (magenta). e, f At 118 dpi, PrPSc (magenta) is more widely distributed, and number of boutons is significantly reduced. PrPSc is near, but not in contact with horizontal cell components. g, h At later timepoints 131 and 153 dpi, ONL is thinned due to loss of photoreceptor nuclei. Few, if any, horizontal cell boutons are present. Large PrPSc deposits (arrow) are visible in the OPL, but not in horizontal cell bodies. i At very high magnification, a 118 dpi retina shows a PKCα-positive rod bipolar cell dendrite (green) near calbindin-positive (yellow) horizontal cell boutons (blue arrows). j A PrPSc deposit (magenta) is very close to, but not touching the horizontal cell boutons (blue arrowhead). k In a merge of three channels the white color indicates the association of a PrPSc deposit (magenta) with the dendritic tip of a rod bipolar cell (green). l Cartoon depicts preceding three panels, with addition of presumed location of rod spherules (rs) and ribbons (r). m Surprisingly, the number of horizontal cell bodies does not change significantly over time. Scale bars: a, c, e, g, h = 5 µm; b, d, f = 2 µm; i–l = 0.5 µm Striebel et al. acta neuropathol commun (2021) 9:17 Page 22 of 26 S triebel et al. acta neuropathol commun (2021) 9:17 Page 23 of 26 Fig. 13 Transmission electron microscopy shows timeline of changes in outer plexiform layer. a Uninfected retina shows numerous rod spherules (r) each containing a single ribbon synapse (arrow) and mitochondrion (m). Cone pedicles (cp) contain multiple ribbon synapses (yellow arrows) and mitochondria (m). Inset shows magnified cone ribbon synapse with ribbon (arrow), and a typical triad consisting of two invaginating horizontal cell processes (h) and one bipolar cell process (bp). Many bipolar and horizontal cell processes are present vitread to cone pedicles (asterisks). b At 104 dpi, most rod spherules appear normal with ribbon synapses present. A cone pedicle (cp) at center, has swollen dystrophic dendritic processes at ribbon synapses (red arrow). Bipolar and horizontal cell processes are present beneath the pedicle (asterisk). c At 137 dpi the OPL appears disorganized with abnormal rod spherules (red arrow), a floating ribbon synapse (blue arrowhead) without invaginating dendritic processes, autophagic whorl (blue arrow) and a dystrophic dendrite (red arrowhead) at a ribbon synapse in a cone pedicle. Microglial cell (mg). d, e, f Examples of autophagic-like whorls (red arrows) in rod spherules at 137 dpi. Scale bars a‑ c = 2 µm; d, e, f = 1 µm between the inner and outer segments of PR cells as well Conclusions as damage to ribbon synapses found in the synaptic end- The present experiments report two new areas of depo - feet of rods and cones near the OPL. Possible synergy sition of abnormal disease-associated PrPSc in prion- and specificity between these two mechanisms remains infected retina. These regions both involve photoreceptor to be worked out. These mechanisms might be active in cone and rod cells which then go on to die as a part of other human retinal degenerative diseases where pro- the disease process. The location of the PrPSc deposition tein misfolding occurs, such as retinitis pigmentosa and suggests that the damage mechanisms may involve inter- prion-like diseases, such as AD and PD. ruption of the ciliary transport pathways of molecules Striebel et al. acta neuropathol commun (2021) 9:17 Page 24 of 26 Received: 4 December 2020 Accepted: 9 January 2021 Supplementary information The online version contains supplementary material available at https ://doi. org/10.1186/s4047 8-021-01120 -x. References Additional file 1: Fig. 1 Cone pedicles and rod spherules disappear as 1. Aboelnour A, Van der Spuy J, Powner M, Jeffery G (2017) Primate retinal PrPSc accumulates in the OPL. a In an uninfected mouse, concentra- cones express phosphorylated tau associated with neuronal degen- tions of PrPC (magenta, yellow arrows) lie vitread to cone pedicles (dark eration yet survive in old age. Exp Eye Res 165:105–108. https ://doi. space, yellow arrowhead) outlined by glucose transporter 1 (GLUT1) a org/10.1016/j.exer.2017.09.013 membrane-associated protein. Rod spherules (red arrowhead) often have 2. Arshavsky VY (2003) Protein translocation in photoreceptor light adapta- tiny faint patches of PrPC (red arrow) adjacent to their membranes. b At tion: a common theme in vertebrate and invertebrate vision. Sci STKE 118 dpi, cone pedicles are less frequent and the distribution of PrP is more 2003:PE43. doi:https ://doi.org/10.1126/stke.2003.204.pe43 punctate and widespread, suggesting PrPSc deposits. c at 153 dpi OPL has 3. Ashok A, Singh N, Chaudhary S, Bellamkonda V, Kritikos AE, Wise AS, Rana thinned significantly, most rod spherules have disappeared, and PrPSc is N, McDonald D, Ayyagari R (2020) Retinal degeneration and Alzheimer’s now in dense punctate patches. ONL has thinned to 1-2 nuclei. Scale bar disease: an evolving link. Int J Mol Sci. https ://doi.org/10.3390/ijms2 11972 = 5 µm. a, b, c are images of optical slices with total z-depth =1 µm 4. Brundin P, Melki R, Kopito R (2010) Prion-like transmission of protein aggregates in neurodegenerative diseases. Nat Rev Mol Cell Biol Acknowledgments 11:301–307. https ://doi.org/10.1038/nrm28 73 The authors thank Katie Williams for technical assistance with mouse anes- 5. Buyukmihci N, Goehring-Harmon F, Marsh RF (1982) Photoreceptor thesia, observation and necropsy, Nancy Kurtz, Lori Lubke and Dan Long, degeneration preceding clinical scrapie encephalopathy in hamsters. J for technical histology assistance, Dave Dorward for electron microscopy Comp Neurol 205:49–54. https ://doi.org/10.1002/cne.90205 0105 technical assistance, Tylor Lewis from the Department of Ophthalmology at 6. Buyukmihci N, Rorvik M, Marsh RF (1980) Replication of the scrapie agent Duke University Medical Center for technical advice on vibratome sectioning in ocular neural tissues. Proc Natl Acad Sci USA 77:1169–1171 of retinal tissue for electron microscopy processing, and Jeffrey Severson for 7. Carroll JA, Race B, Williams K, Striebel J, Chesebro B (2018) Microglia animal husbandry, and Drs. James Carroll, Byron Caughey, Cathryn Haigh, are critical in host defense against prion disease. J Virol. https ://doi. Suzette Priola, and Clayton Winkler for advice on manuscript preparation. org/10.1128/JVI.00549 -18 8. Carroll JA, Striebel JF, Rangel A, Woods T, Phillips K, Peterson KE, Race B, Author contributions Chesebro B (2016) Prion strain differences in accumulation of PrPSc on JS designed study, gathered and analyzed pathology data from all sources, neurons and glia are associated with similar expression profiles of neu- and wrote manuscript. BR designed study, carried out animal experiments, roinflammatory genes: comparison of three prion strains. PLoS Pathog and edited manuscript. JL gathered and analyzed confocal datasets and 12:e1005551. https ://doi.org/10.1371/journ al.ppat.10055 51 edited manuscript. CS prepared samples for electron microscopy, gathered 9. Caughey B, Baron GS, Chesebro B, Jeffrey M (2009) Getting a grip on and analyzed electron microscopy data and edited manuscript. BC designed prions: oligomers, amyloids, and pathological membrane interactions. study, analyzed pathology data, and wrote manuscript. Annu Rev Biochem 78:177–204. https ://doi.org/10.1146/annur ev.bioch em.78.08290 7.14541 0 Funding 10. Chishti MA, Strome R, Carlson GA, Westaway D (1997) Syrian ham- Open Access funding provided by the National Institutes of Health (NIH).. ster prion protein (PrP(C)) is expressed in photoreceptor cells of the This research was supported by the Intramural Research Program (DIR) of adult retina. Neurosci Lett 234:11–14. https ://doi.org/10.1016/s0304 the National Institute of Allergy and Infectious Diseases (NIAID) of the United -3940(97)00669 -1 States, National Institutes of Health (NIH). 11. Clavaguera F, Duyckaerts C, Haik S (2020) Prion-like properties of Tau assemblies. Curr Opin Neurobiol 61:49–57. https ://doi.org/10.1016/j. Availability of data and material conb.2019.11.022 The data supporting the conclusions of this article are included within the 12. Cunningham C, Deacon R, Wells H, Boche D, Waters S, Diniz CP, Scott H, article. Original slides, tissues and photographs are retained. All reagents and Rawlins JN, Perry VH (2003) Synaptic changes characterize early behav- animals used in this study are available from scientific supply companies, ioural signs in the ME7 model of murine prion disease. Eur J Neurosci except the anti-PrP antibody D13, which depending on supply, may be avail- 17:2147–2155. https ://doi.org/10.1046/j.1460-9568.2003.02662 .x able upon request. 13. Euler T, Haverkamp S, Schubert T, Baden T (2014) Retinal bipolar cells: elementary building blocks of vision. Nat Rev Neurosci 15:507–519. https Ethics statement ://doi.org/10.1038/nrn37 83 All mice were housed at the Rocky Mountain Laboratories (RML) in an 14. Fain G, Sampath AP (2018) Rod and cone interactions in the retina. AAALAC-accredited facility in compliance with guidelines provided by the F1000Res 7. doi:https ://doi.org/10.12688 /f1000 resea rch.14412 .1 Guide for the Care and Use of Laboratory Animals (Institute for Laboratory 15. Foster JD, Fraser H, Bruce ME (1986) Retinopathy in mice with experimen- Animal Research Council). Experimentation followed RML Animal Care and tal scrapie. Neuropathol Appl Neurobiol 12:185–196 Use Committee approved protocol 2016–042. 16. Fraser H (1982) Neuronal spread of scrapie agent and targeting of lesions within the retino-tectal pathway. Nature 295:149–150. https ://doi. Consent for publication org/10.1038/29514 9a0 Not applicable. 17. Frigg R, Wenzel A, Samardzija M, Oesch B, Wariwoda H, Navarini AA, Seeliger MW, Tanimoto N, Reme C, Grimm C (2006) The prion protein Competing interest is neuroprotective against retinal degeneration in vivo. Exp Eye Res The authors declare that they have no conflict of interest. 83:1350–1358. https ://doi.org/10.1016/j.exer.2006.07.010 18. Gomez-Nicola D, Fransen NL, Suzzi S, Perry VH (2013) Regulation of Author details microglial proliferation during chronic neurodegeneration. J Neurosci Laboratory of Persistent Viral Diseases, Rocky Mountain Laboratories, National 33:2481–2493. https ://doi.org/10.1523/JNEUR OSCI.4440-12.2013 Institute of Allergy and Infectious Diseases, National Institutes of Health, 19. Gong J, Jellali A, Forster V, Mutterer J, Dubus E, Altrock WD, Sahel JA, 903 South Fourth Street, Hamilton, MT 59840, USA. Research Technologies Rendon A, Picaud S (2007) The toxicity of the PrP106-126 prion peptide Branch, Rocky Mountain Laboratories, National Institute of Allergy and Infec- on cultured photoreceptors correlates with the prion protein distribu- tious Diseases, National Institutes of Health, Hamilton, MT 59840, USA. tion in the mammalian and human retina. Am J Pathol 170:1314–1324. https ://doi.org/10.2353/ajpat h.2007.06034 0 S triebel et al. acta neuropathol commun (2021) 9:17 Page 25 of 26 20. Greenlee MH, Smith JD, Platt EM, Juarez JR, Timms LL, Greenlee JJ 37. Murakami Y, Ikeda Y, Nakatake S, Tachibana T, Fujiwara K, Yoshida N, (2015) Changes in retinal function and morphology are early clinical Notomi S, Nakao S, Hisatomi T, Miller JW, Vavvas DG, Sonoda KH, Ishibashi signs of disease in cattle with bovine spongiform encephalopathy. T (2015) Necrotic enlargement of cone photoreceptor cells and the PLoS ONE 10:e0119431. https ://doi.org/10.1371/journ al.pone.01194 31 release of high-mobility group box-1 in retinitis pigmentosa. Cell Death 21. Hamir AN, Miller JM, Cutlip RC, Kunkle RA, Jenny AL, Stack MJ, Chaplin Discov 1:15058. https ://doi.org/10.1038/cddis cover y.2015.58 MJ, Richt JA (2004) Transmission of sheep scrapie to elk (Cervus ela- 38. Narayan DS, Chidlow G, Wood JP, Casson RJ (2017) Glucose metabolism phus nelsoni) by intracerebral inoculation: final outcome of the experi- in mammalian photoreceptor inner and outer segments. Clin Exp Oph- ment. J Vet Diagn Invest 16:316–321. https ://doi.org/10.1177/10406 thalmol 45:730–741. https ://doi.org/10.1111/ceo.12952 38704 01600 410 39. Narayan DS, Wood JP, Chidlow G, Casson RJ (2016) A review of the 22. Hasegawa M, Nonaka T, Masuda-Suzukake M (2017) Prion-like mechanisms of cone degeneration in retinitis pigmentosa. Acta Ophthal- mechanisms and potential therapeutic targets in neurodegenerative mol 94:748–754. https ://doi.org/10.1111/aos.13141 disorders. Pharmacol Ther 172:22–33. https ://doi.org/10.1016/j.pharm 40. Newton F, Megaw R (2020) Mechanisms of photoreceptor death in retini- thera .2016.11.010 tis pigmentosa. Genes (Basel). https ://doi.org/10.3390/genes 11101 120 23. Head MW, Northcott V, Rennison K, Ritchie D, McCardle L, Bunn 41. Orru CD, Soldau K, Cordano C, Llibre-Guerra J, Green AJ, Sanchez H, TJ, McLennan NF, Ironside JW, Tullo AB, Bonshek RE (2003) Prion Groveman BR, Edland SD, Safar JG, Lin JH, Caughey B, Geschwind MD, protein accumulation in eyes of patients with sporadic and variant Sigurdson CJ (2018) Prion seeds distribute throughout the eyes of spo- Creutzfeldt–Jakob disease. Invest Ophthalmol Vis Sci 44:342–346 radic Creutzfeldt–Jakob disease patients. MBio. https ://doi.org/10.1128/ 24. Head MW, Peden AH, Yull HM, Ritchie DL, Bonshek RE, Tullo AB, Ironside mBio.02095 -18 JW (2005) Abnormal prion protein in the retina of the most commonly 42. Puthussery T, Gayet-Primo J, Taylor WR (2010) Localization of the calcium- occurring subtype of sporadic Creutzfeldt–Jakob disease. Br J Ophthal- binding protein secretagogin in cone bipolar cells of the mammalian mol 89:1131–1133. https ://doi.org/10.1136/bjo.2004.06349 5 retina. J Comp Neurol 518:513–525. https ://doi.org/10.1002/cne.22234 25. Hughes D, Halliday M (2017) What is our current understanding of 43. Rachel RA, Li T, Swaroop A (2012) Photoreceptor sensory cilia and cili- PrP(Sc)-associated neurotoxicity and its molecular underpinnings? opathies: focus on CEP290. RPGR and their interacting proteins Cilia 1:22. Pathogens. https ://doi.org/10.3390/patho gens6 04006 3https ://doi.org/10.1186/2046-2530-1-22 26. Jaunmuktane Z, Brandner S (2020) Invited review: the role of prion-like 44. Ramirez AI, de Hoz R, Salobrar-Garcia E, Salazar JJ, Rojas B, Ajoy D, Lopez- mechanisms in neurodegenerative diseases. Neuropathol Appl Neuro- Cuenca I, Rojas P, Trivino A, Ramirez JM (2017) The role of microglia in reti- biol 46:522–545. https ://doi.org/10.1111/nan.12592 nal neurodegeneration: Alzheimer’s disease, parkinson, and glaucoma. 27. Jeffrey M, Halliday WG, Bell J, Johnston AR, MacLeod NK, Ingham C, Front Aging Neurosci 9:214. https ://doi.org/10.3389/fnagi .2017.00214 Sayers AR, Brown DA, Fraser JR (2000) Synapse loss associated with 45. Ratnayaka JA, Serpell LC, Lotery AJ (2015) Dementia of the eye: the role of abnormal PrP precedes neuronal degeneration in the scrapie-infected amyloid beta in retinal degeneration. Eye (Lond) 29:1013–1026. https :// murine hippocampus. Neuropathol Appl Neurobiol 26:41–54. https :// doi.org/10.1038/eye.2015.100 doi.org/10.1046/j.1365-2990.2000.00216 .x 46. Silverman SM, Wong WT (2018) Microglia in the retina: roles in develop- 28. Kercher L, Favara C, Chan CC, Race R, Chesebro B (2004) Differences in ment, maturity, and disease. Annu Rev Vis Sci 4:45–77. https ://doi. scrapie-induced pathology of the retina and brain in transgenic mice org/10.1146/annur ev-visio n-09151 7-03442 5 that express hamster prion protein in neurons, astrocytes, or multiple 47. Singh N, Chaudhary S, Ashok A, Lindner E (2020) Prions and prion cell types. Am J Pathol 165:2055–2067. https ://doi.org/10.1016/S0002 diseases: insights from the eye. Exp Eye Res 199:108200. https ://doi. -9440(10)63256 -7org/10.1016/j.exer.2020.10820 0 29. Kercher L, Favara C, Striebel JF, LaCasse R, Chesebro B (2007) Prion pro- 48. Siskova Z, Page A, O’Connor V, Perry VH (2009) Degenerating synaptic tein expression differences in microglia and astroglia influence scrapie- boutons in prion disease: microglia activation without synaptic stripping. induced neurodegeneration in the retina and brain of transgenic mice. Am J Pathol 175:1610–1621. https ://doi.org/10.2353/ajpat h.2009.09037 2 J Virol 81:10340–10351. https ://doi.org/10.1128/JVI.00865 -07 49. Slepak VZ, Hurley JB (2008) Mechanism of light-induced translocation of 30. Klingeborn M, Race B, Meade-White KD, Rosenke R, Striebel JF, Chese- arrestin and transducin in photoreceptors: interaction-restricted diffusion. bro B (2011) Crucial role for prion protein membrane anchoring in the IUBMB Life 60:2–9. https ://doi.org/10.1002/iub.7 neuroinvasion and neural spread of prion infection. J Virol 85:1484– 50. Smith JD, Greenlee JJ, Hamir AN, West Greenlee MH (2008) Retinal cell 1494. https ://doi.org/10.1128/JVI.02167 -10 types are differentially affected in sheep with scrapie. J Comp Pathol 31. Liu B, Rasool S, Yang Z, Glabe CG, Schreiber SS, Ge J, Tan Z (2009) Amyloid- 138:12–22. https ://doi.org/10.1016/j.jcpa.2007.09.002 peptide vaccinations reduce {beta}-amyloid plaques but exacerbate 51. Soto C, Pritzkow S (2018) Protein misfolding, aggregation, and conforma- vascular deposition and inflammation in the retina of Alzheimer’s trans- tional strains in neurodegenerative diseases. Nat Neurosci 21:1332–1340. genic mice. Am J Pathol 175:2099–2110. https ://doi.org/10.2353/ajpat https ://doi.org/10.1038/s4159 3-018-0235-9 h.2009.09015 9 52. Spraker TR, Zink RR, Cummings BA, Wild MA, Miller MW, O’Rourke KI 32. Liu C, Zhang CW, Zhou Y, Wong WQ, Lee LC, Ong WY, Yoon SO, Hong W, (2002) Comparison of histological lesions and immunohistochemical Fu XY, Soong TW, Koo EH, Stanton LW, Lim KL, Xiao ZC, Dawe GS (2018) staining of proteinase-resistant prion protein in a naturally occurring APP upregulation contributes to retinal ganglion cell degeneration spongiform encephalopathy of free-ranging mule deer (Odocoileus via JNK3. Cell Death Differ 25:661–676. https ://doi.org/10.1038/s4141 hemionus) with those of chronic wasting disease of captive mule deer. 8-017-0005-3 Vet Pathol 39:110–119. https ://doi.org/10.1354/vp.39-1-110 33. Matsunaga Y, Peretz D, Williamson A, Burton D, Mehlhorn I, Groth D, 53. Steinert JR (2015) Prion protein as a mediator of synaptic transmis- Cohen FE, Prusiner SB, Baldwin MA (2001) Cryptic epitopes in N-termi- sion. Commun Integr Biol 8:e1063753. https ://doi.org/10.1080/19420 nally truncated prion protein are exposed in the full-length molecule: 889.2015.10637 53 dependence of conformation on pH. Proteins 44:110–118 54. Striebel JF, Race B, Williams K, Carroll JA, Klingeborn M, Chesebro B (2019) 34. Mendes HF, van der Spuy J, Chapple JP, Cheetham ME (2005) Mecha- Microglia are not required for prion-induced retinal photoreceptor nisms of cell death in rhodopsin retinitis pigmentosa: implications for degeneration. Acta Neuropathol Commun 7:48. https ://doi.org/10.1186/ therapy. Trends Mol Med 11:177–185. https ://doi.org/10.1016/j.molme s4047 8-019-0702-x d.2005.02.007 55. Takao M, Kimura H, Kitamoto T, Mihara B (2018) PrP(res) deposition 35. Mirzaei N, Shi H, Oviatt M, Doustar J, Rentsendorj A, Fuchs DT, Sheyn J, in the retina is a common finding of sporadic, familial and iatrogenic Black KL, Koronyo Y, Koronyo-Hamaoui M (2020) Alzheimer’s retin- Creutzfeldt–Jakob diseases (CJD). Acta Neuropathol Commun 6:78. https opathy: seeing disease in the eyes. Front Neurosci 14:921. https ://doi.://doi.org/10.1186/s4047 8-018-0582-5 org/10.3389/fnins .2020.00921 56. Thiadens AA, Phan TM, Zekveld-Vroon RC, Leroy BP, van Born LI, Hoyng 36. Moser T, Grabner CP, Schmitz F (2020) Sensory processing at ribbon CB, Klaver CC, Writing Committee for the Cone Disorders study Group C, synapses in the retina and the cochlea. Physiol Rev 100:103–144. https :// Roosing S, Pott JW, van Schooneveld MJ, van Moll-Ramirez N, van Gen- doi.org/10.1152/physr ev.00026 .2018 deren MM, Boon CJ, van Hollander AI, Bergen AA, De Baere E, Cremers FP, Lotery AJ (2012) Clinical course, genetic etiology, and visual outcome in Striebel et al. acta neuropathol commun (2021) 9:17 Page 26 of 26 cone and cone-rod dystrophy. Ophthalmology 119(819):826. https ://doi. of sensory cilia. Mol Cell Biol 25:4129–4137. https ://doi.org/10.1128/ org/10.1016/j.ophth a.2011.10.011 MCB.25.10.4129-4137.2005 57. Veys L, Vandenabeele M, Ortuno-Lizaran I, Baekelandt V, Cuenca N, Moons 60. Ye X (2009) Visual pathology in animal prion diseases. Histol Histopathol L, De Groef L (2019) Retinal alpha-synuclein deposits in Parkinson’s dis- 24:1563–1577. https ://doi.org/10.14670 /HH-24.1563 ease patients and animal models. Acta Neuropathol 137:379–395. https :// 61. Zamponi E, Pigino GF (2019) Protein misfolding, signaling abnormali- doi.org/10.1007/s0040 1-018-01956 -z ties and altered fast axonal transport: implications for Alzheimer and 58. West Greenlee MH, Lind M, Kokemuller R, Mammadova N, Kondru N, prion diseases. Front Cell Neurosci 13:350. https ://doi.org/10.3389/fncel Manne S, Smith J, Kanthasamy A, Greenlee J (2016) Temporal resolu- .2019.00350 tion of misfolded prion protein transport, accumulation, glial activation, and neuronal death in the retinas of mice inoculated with scrapie. Am J Publisher’s Note Pathol 186:2302–2309. https ://doi.org/10.1016/j.ajpat h.2016.05.018 Springer Nature remains neutral with regard to jurisdictional claims in pub- 59. Yang J, Gao J, Adamian M, Wen XH, Pawlyk B, Zhang L, Sanderson MJ, Zuo lished maps and institutional affiliations. J, Makino CL, Li T (2005) The ciliary rootlet maintains long-term stability Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

Acta Neuropathologica CommunicationsSpringer Journals

Published: Jan 29, 2021

There are no references for this article.