Protective role of Galectin‐3 in prion disease through regulation of microglial PrP^Sc uptake

Abstract

Microglia constitute a major innate defense mechanism against prion infection; however, the molecular pathways regulating microglial responses during disease progression remain incompletely understood. Galectin‐3 (Gal3), a β‐galactoside‐binding lectin associated with reactive microglia in multiple neurodegenerative disorders, has been implicated in phagocytosis, inflammatory signaling, and lysosomal homeostasis. Here, we investigated the role of Gal3 in prion disease pathogenesis using prion‐infected mice. Gal3 expression was undetectable in healthy brain but became upregulated beginning at late preclinical stages, increasing with disease progression. Gal3 localized predominantly to a subpopulation of reactive IBA1‐positive microglia, particularly within the thalamus, and inversely correlated with expression of the homeostatic microglial markers P2Y12 and TMEM119, consistent with acquisition of a reactive phenotype. Microglia engaged in neuronal envelopment displayed elevated Gal3 expression during terminal disease. Constitutive deletion of Gal3 significantly accelerated clinical disease progression without altering total PrP^Sc accumulation, reactive gliosis, neuronal envelopment, or overall microglial and astrocytic activation. However, Gal3 deficiency markedly reduced microglial uptake of PrP^Sc, resulting in a lower intracellular‐to‐extracellular PrP^Sc ratio, supporting a role for Gal3 in phagocytic sequestration of prions. In contrast, Gal3 deficiency did not impair lysosomal activity, lysosomal membrane integrity, or expression of genes involved in lysosomal repair pathways. Likewise, selective inhibition of autophagy in myeloid cells exerted only minor effects on disease progression. Collectively, these findings identify Gal3 as a sensitive marker of reactive microglia that contributes to microglial uptake of PrP^Sc and exerts a protective role during prion disease progression.

Galectin‐3 (Gal3) is induced in reactive microglia during prion disease and promotes microglial uptake of PrP^Sc. Gal3 deficiency reduces prion sequestration and accelerates disease progression, revealing a protective role for Gal3.

1. INTRODUCTION

Prion diseases, also known as transmissible spongiform encephalopathies (TSEs), comprise a group of fatal transmissible neurodegenerative disorders affecting both humans and animals [1]. Despite extensive research efforts, no effective therapies are currently available. Prion diseases are caused by prions, or PrP^Sc, the misfolded and aggregated form of the cellular prion protein (PrP^C), a host‐encoded sialoglycoprotein [2]. Disease progression is driven by the self‐propagating conversion of PrP^C into β‐sheet‐rich PrP^Sc conformers, leading to the accumulation and spread of prions throughout the central nervous system (CNS) [3]. Progressive neurodegeneration in prion‐infected individuals and animals is closely associated with the accumulation of PrP^Sc and its neurotoxic effects on neurons [4, 5, 6, 7, 8, 9, 10].

Microglia are the principal innate immune cells of the CNS, where they continuously monitor the brain microenvironment and mediate the clearance of damaged cells and protein aggregates. Accumulating evidence supports the concept that microglia constitute a major host‐defense mechanism against prion infection [11, 12, 13]. Indeed, depletion of microglia either before prion inoculation or during the early stages of disease accelerates disease progression and shortens survival time [14, 15, 16, 17]. However, other studies suggest that microglia may exert deleterious effects during later disease stages. Partial inhibition of microglial proliferation and reactivity during the preclinical phase delayed the onset of neurological symptoms and prolonged survival [18]. Similarly, pharmacological suppression of microglial activation at disease onset reduced reactive gliosis and extended survival in humanized mice infected with human prions [19]. Together, these findings support the emerging view that microglia may transition from protective to pathogenic phenotypes as neurodegeneration progresses, a concept increasingly recognized across multiple neurodegenerative disorders, including Alzheimer's disease (AD).

Galectin‐3 (Gal3) is a multifunctional β‐galactoside‐binding lectin implicated in microglial activation, innate immune signaling, and tissue remodeling within the CNS [20, 21, 22]. Gal3 preferentially recognizes β‐galactoside‐containing glycans and exhibits high affinity for branched complex N‐glycans lacking terminal sialylation [23, 24, 25, 26]. Under physiological conditions, Gal3 expression in the CNS is low; however, injury, infection, and protein misfolding trigger robust upregulation of Gal3, particularly in activated microglia [20, 21, 22, 27]. Gal3 functions both intracellularly and extracellularly, acting as an opsonin. By binding to TLR4 and integrins, Gal3 amplifies pro‐inflammatory signaling cascades and promotes the transition of microglia to an amoeboid, phagocytic phenotype [28]. Through its glycan‐binding properties, Gal3 is believed to enhance the uptake of extracellular matrix components, apoptotic cells, myelin debris, and misfolded protein aggregates, contributing to tissue remodeling during chronic neuroinflammation.

Intracellularly, Gal3 has been implicated in the maintenance of lysosomal integrity. Upon lysosomal membrane damage, Gal3 rapidly accumulates on ruptured lysosomes, where it recruits components of the endosomal sorting complex required for transport machinery and autophagy pathways to coordinate membrane repair and lysophagy [29, 30]. These activities suggest that Gal3 may play an important role in regulating cellular responses to proteotoxic stress and impaired protein degradation.

In neurodegenerative disorders, including AD, Parkinson's disease, Huntington's disease, multiple sclerosis, and prion diseases, Gal3 is consistently upregulated in reactive microglia [20, 27, 31, 32, 33, 34]. Single‐cell transcriptomic studies have identified Gal3 as a characteristic marker of an advanced stage of disease‐associated microglia (DAM) [21]. However, the functional consequences of Gal3 upregulation appear to be context‐dependent. While Gal3 can facilitate debris clearance and tissue repair, sustained Gal3 expression may perpetuate inflammatory signaling, promote synaptic remodeling, and contribute to chronic glial activation [22, 31, 33]. Elevated Gal3 levels in cerebrospinal fluid and brain tissue correlate with disease progression and neuroinflammatory burden, highlighting its potential utility as both a biomarker and therapeutic target [27, 35].

Consistent with these observations, genetic deletion or pharmacological inhibition of Gal3 attenuated disease phenotypes in several experimental models of neurodegeneration. In the 5XFAD mouse model of AD, Gal3 expression was markedly elevated in plaque‐associated microglia, whereas Gal3 deletion reduced amyloid‐β burden and improved cognitive performance [31]. Gal3 depletion also suppressed tau propagation, neuroinflammation, and associated functional impairments [33]. In models of retinal degeneration, inhibition of Gal3 reduced microglial activation and delayed neurodegeneration [36]. Similarly, in Huntington's disease models, Gal3 deletion diminished inflammatory responses, reduced huntingtin aggregation, improved motor function, and prolonged survival [27]. Collectively, these findings suggest that Gal3 contributes to neurodegenerative pathology and may represent a promising therapeutic target for slowing disease progression.

In prion disease, Gal3 expression is strongly upregulated in affected brain regions characterized by prion deposition and neurodegeneration [34, 37, 38, 39, 40, 41]. Elevated Gal3 expression localizes predominantly to myeloid cells, including reactive microglia [39, 41]. An earlier study reported a substantial prolongation of survival in Gal3‐deficient mice infected with the 139A prion strain, despite no detectable changes in PrP^Sc accumulation or glial activation [42]. This finding was particularly intriguing given that Gal3 is expressed by only a small subpopulation of microglia in prion‐infected brains [34, 43]. The observation that genetic targeting of a pathway expressed by a limited microglial subpopulation could markedly alter disease trajectory raised the possibility that Gal3 represents a highly selective disease‐modifying or therapeutic target and provided a strong rationale for further investigation. However, to our knowledge, these initial findings have not been followed by subsequent mechanistic studies. We therefore independently re‐examined the role of Gal3 in prion disease using the SSLOW prion strain. Compared with other mouse‐adapted prion strains, SSLOW has the shortest incubation time and elicits especially pronounced and widespread neuroinflammation, providing a robust experimental system in which microglial activation and PrP^Sc handling can be examined [40, 44, 45]. Importantly, SSLOW infection is associated with a high prevalence of neuronal envelopment by microglia and a prominent Gal3‐expressing microglial population. In addition to assessing the effects of Gal3 deficiency on disease progression, we investigated potential mechanisms through which Gal3 might influence prion pathogenesis, including microglial PrP^Sc uptake, neuronal envelopment, lysosomal function, and microglial autophagy.

2. METHODS

2.1. Animals

Ten percentage (w/v) SSLOW brain homogenates (BH) for inoculations were prepared in PBS, pH 7.4, using glass/Teflon homogenizers attached to a cordless 12 V compact drill. Immediately before inoculation, each inoculum was further dispersed by 30 s of indirect sonication at approximately 200 watts in a microplate horn of a sonicator (Qsonica, Newtown, CT) and diluted in PBS, pH 7.4. Brain‐derived materials were inoculated as 1% via i.c. (intracerebral) or i.p. (intraperitoneal) route, as noted. Male and female Gal3 KO mice (#006338, Jackson Laboratory, Bar Harbor, ME) and C57BL/6J (Veterinary Resources, University of Maryland, Baltimore, MD, USA), 6–7 weeks old, received 200 μL of inoculum i.p. or 20 μL of inoculum i.c. under 3% isoflurane anesthesia and were regularly monitored for clinical signs, which included clasping hind legs, difficulty walking, abnormal gait, nesting problems, and weight loss. Naïve age‐matched C57BL/6J and Gal3 KO mice were used as non‐infected controls. The mice were considered terminal and euthanized when they were unable to rear and/or lost 20% of their weight. Becn1 cKO (Becn1^flox/floxLyz2^Cre/Cre) and control (Lyz2^Cre/Cre) female mice [46] were inoculated with 1% SSLOW ic and assessed as above. P2Y12 KO and CD11b KO mice were described previously [40, 47].

The study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The animal protocol was approved by the Institutional Animal Care and Use Committee of the University of Maryland, Baltimore (assurance number A32000‐01; permit number 0215002).

2.2. Sex as a biological variable

Majority of the experiments were performed on male and female mice, and similar findings are reported for both sexes. Becn1 cKO (Becn1^flox/flox; Lyz2^Cre/Cre) and control (Lyz2^Cre/Cre) mice were only females.

2.3. Antibodies

Primary antibodies used for immunofluorescence, immunohistochemistry, and immunoblotting were as follows: mouse monoclonal anti‐prion protein, clone SAF‐84 (189775, Cayman Chemicals, Ann Arbor, MI); rabbit monoclonal anti‐prion protein, clone 3D17 (ZRB1268, Millipore Sigma, Rockville, MD); rabbit polyclonal anti‐IBA1 (013‐27691, Fujifilm Wako Chemicals USA, Richmond, VA); goat polyclonal anti‐IBA1 (NB100‐1028, Bio‐techne Minneapolis, MN); rat monoclonal anti‐Gal3, clone M3/38 (sc‐23938, Santa Cruz Biotechnology Inc.); rabbit anti‐P2Y12 (#55043A, Anaspec, Fremont, CA); TMEM119 (#90840, Cell Signaling Technology, Danvers, MA); rabbit anti‐Vim (#5741, Cell Signaling Technology); mouse monoclonal anti‐NeuN, clone A60 (MAB377, Millipore Sigma); rat monoclonal anti‐LAMP1, clone 1D4B (121601, BioLegend, San Diego, CA); rabbit polyclonal anti‐CD11b (ab128797, Abcam, Waltham, MA); rabbit polyclonal anti‐interferon‐induced transmembrane protein 3 (11714‐1‐AP, Proteintech, Rosemont, IL); rabbit monoclonal anti‐cathepsin D (69854, Cell Signaling Technology); goat polyclonal anti‐glycoprotein non‐metastatic melanoma protein B (anti‐GPNMB) (AF2320, Bio‐techne); and mouse monoclonal anti‐β‐actin, clone AC‐15 (A5441, Millipore Sigma). The secondary antibodies for immunofluorescence were Alexa Fluor 488, 546, and 647 labeled (Thermo Fisher Scientific, Waltham, MA).

2.4. Immunofluorescence and diaminobenzidine staining of mouse brains

Formalin‐fixed brains (sagittal or coronal 3 mm slices) were treated for 1 h in 96% formic acid before being embedded in paraffin using standard procedures; 4 μm sections produced with Leica RM2235 microtome (Leica Biosystems, Deer Park, IL) were mounted on Superfrost Plus Microscope slides (22‐037‐246, Thermo Fisher Scientific) and processed for immunohistochemistry according to standard protocols. To expose epitopes, slides were subjected to 20 min of hydrated autoclaving at 121°C in citrate buffer, pH 6.0, antigen retriever (C9999, Sigma‐Aldrich, St. Louis, MO). For the detection of disease‐associated PrP, an additional 3 min treatment in concentrated formic acid was applied.

For immunofluorescence, an Autofluorescence Eliminator Reagent (2160, Sigma‐Aldrich) and Signal Enhancer (11932, Cell Signaling Technology) were used on slides according to the original protocols to reduce background fluorescence. Images were collected using an inverted microscope Nikon Eclipse TE2000‐U (Nikon Instruments Inc.) equipped with an illumination system X‐cite 120 (EXFO Photonics Solutions Inc.) and a cooled 12‐bit CoolSnap HQ CCD camera (Photometrics), or a Leica Mica widefield microscope (Leica Microsystems, Boston, MA). Images were processed using ImageJ software (1.54p, NIH). Analyze Particles function of ImageJ was used for segmentation and defining regions of interest (ROIs) for the calculation of individual cell values. Cathepsin D and LAMP1 colocalization was performed with JACOP plugin of ImageJ. Full‐brain tile scan images were acquired and stitched using the Leica MICA widefield microscope with the 20× dry objective.

For detection with diaminobenzidine (DAB), horseradish peroxidase‐labeled secondary antibodies and DAB Quanto chromogen and substrate (TA125QHDX, Thermo Fisher Scientific) were used. The images were acquired with a Canon G7 digital camera equipped with a microscope adaptor (Bower, NY).

2.5. Confocal microscopy and three‐dimensional image reconstruction

Confocal images were acquired using a Leica TCS SP8 microscope using the ×40/1.30 or ×63/1.40 oil immersion objective lenses with laser lines 405, 488, 552, and 638 as needed. Image resolution was 1024 × 1024 or 2048 × 2048 pixels at a 400 Hz scan speed. For three‐dimensional (3D) reconstruction, Z‐stack thickness ranged from 6.2 to 10.45 μm taken at the system‐optimized number of steps. Images were processed using the LAS X, version 1.4.5.27713, and ImageJ software. For Figure 6, 3D reconstructions of the confocal images were prepared using the surface rendering option in Imaris software (Oxford Instruments plc, High Wycombe, UK).

FIGURE 6.

Analyses of PrP^Sc uptake by microglia. (A) Representative immunofluorescence images from SSLOW‐infected terminally ill C57Bl/6J mice and age‐matched non‐infected control mice stained for IBA1 and PrP^Sc (3D17 antibody). Scale bar 50 μm. (B) Representative confocal microscopy field of view used for quantification of microglia‐associated PrP^Sc, with enlarged images showing PrP^Sc (3D17, green) colocalized with microglia (IBA1, red) in SSLOW‐infected terminally ill Galectin‐3 knockout (Gal3 KO) and wild type (WT) mice. (C–E) Quantification of the total number of PrP^Sc puncta (C), microglia‐associated PrP^Sc puncta (D), and the ratio of PrP^Sc puncta associated with microglia per puncta not associated with microglia in SSLOW‐infected terminally ill Gal3 KO and WT mice. Quantification was performed using integrated fluorescence densities of 3D17‐positive puncta. N = 5 animals per group, with three fields of view per animal and 60–80 microglia per field of view. **p < 0.01, ns, not significant by unpaired Student's t‐test. Colors represent individual brains; dots represent individual fields of view; average values for each brain are shown as circles; means ± standard deviation are marked by black lines. DAPI, 4′,6‐diamidino‐2‐phenylindole.

2.6. Quantification of envelopment

To estimate the percentage of neurons enveloped by microglia and the percentage of microglial cells involved in envelopment, brains were co‐immunostained for NeuN and IBA1, and immunofluorescence images from cortices of each brain were collected. Using ImageJ software, NeuN and IBA1 channel images were subjected to a threshold, converted to binary, and then individual neurons and microglia cells were identified as ROIs using the Analyze Particles function of ImageJ. The area of IBA1 signal in ROIs defined for individual neurons was measured, and the neurons were counted as undergoing envelopment if their NeuN signal overlapped by IBA1. The total number of identified NeuN+ cells was recorded as a neuronal count. The area of NeuN signal in ROIs defined for individual microglia cells was measured, and the microglia were counted as involved in envelopment if they had IBA1 signal overlapped by NeuN. The total number of identified microglia cells was recorded as a microglia count.

To calculate the percent of envelopment for Gal3⁻ and Gal3⁺ cells, the brains were co‐immunostained for NeuN, IBA1, and Gal3, and images were taken with a Leica Mica microscope using a 20× objective. ROIs for microglia cells were created with the Particle Analysis function of ImageJ and measured on binary Gal3 and NeuN channels. The cells were considered Gal3‐positive if the Gal3 signal covered >15% of the IBA1 area. The microglia cell was considered enveloping a neuron if the NeuN signal area overlapped the IBA1 area by >5%.

2.7. Quantification of PrP^Sc in microglia

Brains were co‐immunostained with anti‐PrP (3D17 or SAF‐84) and IBA1 antibodies. The images from IBA1 and anti‐PrP channels were thresholded, converted to binary, and the Analyze Particles function of ImageJ was used to identify individual microglia cells and PrP^Sc puncta on binary images from corresponding channels. The PrP^Sc puncta's ROIs were measured on binary IBA1 images to classify the identified puncta as associated or not with microglia cells.

PrP^Sc content in Gal3⁻ and Gal3⁺ microglia was calculated in cortices of brains co‐immunostained for PrP^Sc (3D17), IBA1, and Gal3, and imaged with Leica Mica microscope using 20× objective. For each cortex, the microglia cells were analyzed in three representative fields of view. ROIs for microglia cells were created with the Particle Analysis function of ImageJ and measured on binary Gal3 and PrP^Sc channels. The cells were considered Gal3‐positive if the Gal3 signal covered >15% of the microglia cell area. If PrP^Sc signal in cell was >5 pixels, the cell was counted as containing PrP^Sc. The combined number of cells analyzed in three fields of view was plotted as well.

2.8. Western blot

For Western blots, 10% (w/v) BH were prepared using radioimmunoprecipitation assay (RIPA) Lysis Buffer (20‐188, Millipore Sigma). To analyze brain‐derived PrP^Sc, BH aliquots were diluted with RIPA buffer to achieve a 5% BH final concentration and treated with 20 μg/mL proteinase K (P8107S, New England BioLabs, Ipswich, MA) in the presence of 50 mM Tris, pH 7.5, and 2% Sarcosyl, for 30 min at 37°C. To analyze other proteins, BH was diluted with RIPA buffer to 1% and proteinase digestion was omitted. The resulting samples were supplemented with 4× sodium dodecylsulfate (SDS) loading buffer and heated for 10 min in a boiling water bath before loading onto NuPAGE 12% Bis‐Tris gels (NP0341BOX, Thermo Fisher Scientific). Wet transfer onto polyvinylidene difluoride (PVDF) membranes and probing of Western blots was done according to standard procedures. The signals were visualized by Immobilon Forte Western HRP Substrate (WBLUF0100, Millipore Sigma) or SuperSignal West pico PLUS Chemiluminescent Substrate (34577, Thermo Fisher Scientific) using Invitrogen iBright 1500 imager, and quantified with iBright Analysis software (Thermo Fisher Scientific). Intensity data were presented as normalized by actin, except for PrP^Sc Western blots treated with protease K.

2.9. RT‐qPCR

Total RNA was isolated from 10% BH in RIPA buffer. One hundred microliters of BH aliquots were further homogenized within RNase‐free 1.5‐mL tubes in 200 μL of Trizol (15‐596‐026, Thermo Fisher Scientific), using RNase‐free disposable pestles (12‐141‐364, Thermo Fisher Scientific). After homogenization, an additional 600 μL of Trizol was added to each homogenate, and the samples were centrifuged at 11,400 × g for 5 min at 4°C. The supernatant was collected, incubated for 5 min at room temperature, then supplemented with 160 μL of cold chloroform and vigorously shaken for 30 s by hand. After an additional 5‐min incubation at room temperature, the samples were centrifuged at 11,400 × g for 15 min at 4°C. The top layer was transferred to new RNase‐free tubes and mixed with an equal amount of 70% ethanol. Subsequent steps were performed using an Aurum Total RNA Mini Kit (7326820, Bio‐Rad, Hercules, CA) following the manufacturer instructions. Isolated total RNA was subjected to DNase I digestion. RNA purity and concentrations were estimated using a NanoDrop One Spectrophotometer (Thermo Fisher Scientific). Complementary DNA (cDNA) synthesis was performed using iScript cDNA Synthesis Kit (1708890, Bio‐Rad) as described elsewhere. The cDNA was amplified with CFX96 Touch Real‐Time PCR Detection System (Bio‐Rad) using SsoAdvanced Universal SYBR Green Supermix and the primers listed in Table 1. The polymerase chain reaction (PCR) protocol consisted of 95°C incubation for 2 min followed by 40 amplification cycles at 95°C for 5 s and 60°C for 30 s. The data were analyzed using CFX96 Touch Real‐Time PCR Detection System Software (Bio‐Rad).

TABLE 1.

RT‐qPCR primers.

Gene	Forward	Reverse
Alix	5′‐GGACTCCATCCAATGACCTGTAC‐3′	5′‐CGGCTTACACAGAAGTGCGATG‐3′
Ccl2	5′‐GCTACAAGAGGATCACCAGCAG‐3′	5′‐GTCTGGACCCATTCCTTCTTGG‐3′
CD68	5′‐ACTGGTGTAGCCTAGCTGGT‐3′	5′‐CCTTGGGCTATAAGCGGTCC‐3′
Gapdh	5′‐CATCACTGCCACCCAGAAGACTG‐3′	5′‐ATGCCAGTGAGCTTCCCGTTCAG‐3′
GPNMB	5′‐GGCTACTTCAGAGCCACCATCA‐3′	5′‐CTTTGCAGGTCACAGTGAAGTCC‐3′
Tlr4	5′‐AGCTTCTCCAATTTTTCAGAACTTC‐3′	5′‐TGAGAGGTGGTGTAAGCCATGC‐3′
Trim16	5′‐CAACACTGTGCTGTTCCTGGAG‐3′	5′‐ATGACCTTGCGGATGCCTGAGA‐3′
Tgs101	5′‐GGTTATCCTGGCTGTCCTTACC‐3′	5′‐GCACGGATAGTGTCCTCACTGA‐3′

2.10. Co‐immunoprecipitation

The immunoprecipitation experiment using Protein A/G Magnetic Beads (88803, Thermo Fisher Scientific) was conducted essentially as recommended by the manufacturer. Fifty microliters of magnetic beads slurry was used in each reaction. The beads were pre‐incubated with 5 μL of rat monoclonal anti‐Gal3 antibody, clone M3/38 (sc‐23938, Santa Cruz Biotechnology Inc.) for 2 h at 4°C on a rotator. Ten percent BH was prepared from a freshly collected brain from a terminally ill mouse, using 1× PBS, pH 7.4, supplemented with 150 mM sodium chloride, 1 mM ethylenediamine tetraacetic acid (EDTA), 1% Triton X‐100, and protease inhibitors cocktail (P8340, Millipore Sigma), and further dispersion by 30 s indirect sonication of 120 μL aliquots at approximately 200 W in a microplate horn of a sonicator (Qsonica, Newtown, CT). After short spin centrifugation, the supernatant was combined, and 300 μL of 10% BH was added to the tubes with prepared beads for 2 h of incubation at room temperature on a rotator. The beads were washed 5 times with the same buffer, and the Gal3 complexes were recovered by boiling in SDS sample buffer for Western blots. Successful immunoprecipitation of Gal3 was confirmed by rabbit monoclonal anti‐Gal3 antibody, clone E7B6R (89572, Cell Signaling Technology), while PrP^Sc detection in the same samples was performed with 3D17.

2.11. Statistics

Statistical analyses and plotting of the data were performed using GraphPad Prism software, versions 8.4.2‐10.4.1 for Windows (GraphPad Software, San Diego, CA) or Excel 2016—version 2302. Statistical comparison of two groups with normal distribution of data points and equal variances was done with an unpaired two‐tailed t‐test. Alternatively, the non‐parametric two‐tailed Mann–Whitney test was used. In case of unequal variances, the comparison of two groups with a normal distribution of data points was performed with an unpaired two‐tailed t‐test with Welch's correction. For statistical comparison of more than two groups, one‐way analysis of variance (ANOVAs) were followed by multiple comparison tests as mentioned in figure legends, and the resulting p‐values were reported. For statistical analysis of data presented in Superplots, means ± standard deviation (SD) were calculated using biological replicas, where n = the number of brains or animals analyzed.

3. RESULTS

3.1. Upregulation of Gal3 expression in a subpopulation of myeloid cells during disease progression

To establish the temporal profile of Gal3 upregulation, Gal3 expression was analyzed in C57BL/6J mice challenged i.p. with the SSLOW prion strain. Mild upregulation was detected as early as 92 dpi; however, robust increases were observed at 122 dpi, coinciding with clinical onset [44] (Figure 1A). Progressive elevation of Gal3 levels after 122 dpi further supported its association with disease progression. Immunohistochemical analysis revealed a regional correlation between PrP^Sc accumulation, microgliosis, and Gal3 immunoreactivity (Figure 1B). Although Gal3⁺ cells were widely distributed across all prion‐affected brain regions, Gal3 expression was especially prominent in the thalamus, the brain region most severely affected in prion‐infected mice and sporadic Creutzfeldt‐Jakob disease (sCJD) [34, 45, 48, 49, 50] (Figures 1B and S1A).

FIGURE 1.

Upregulation of Galectin‐3 (Gal3) expression in prion‐infected mice. (A) Upper panel: Analysis of Gal3 expression in whole‐brain homogenates of C57Bl/6J mice infected with SSLOW via intraperitoneal route at different stages of disease (dpi, days post‐inoculation; terminal state, 157–166 dpi) by Western blot. Lower panel: Representative Western blots of Gal3 and actin. N = 4–5 animals per time point. *p < 0.05, **p < 0.01, ns, not significant; comparison to non‐infected brains by Brown‐Forsythe and Welch's analysis of variance with Dunnett's T3 multiple comparison test. (B) Immunohistochemistry of C57Bl/6J mice infected with SSLOW via intracerebral route, showing microglia reactivity (IBA1) and Gal3⁺ cells in areas with PrP^Sc deposition (SAF‐84 antibody). Bottom row shows high‐magnification views of boxed areas. Scale bar 200 μm. (C) Western blot analysis of Gal3 expression in whole‐brain homogenates from SSLOW‐infected P2Y12 knockout (KO) mice at clinical onset, CD11b KO mice at terminal stage, and corresponding wild type (WT) control groups. N = 5–7 animals per group, *p < 0.05 by unpaired Welch's t‐test (P2Y12 KO) and Student's t‐test (CD11b KO). (D) Immunofluorescence staining of SSLOW‐infected C57Bl/6J mice illustrating Gal3 expression in a subpopulation of IBA1⁺ cells in the thalamus (Th) and cortex (Ctx), and non‐infected cortex lacking detectable Gal3 signal. Scale bar 100 μm. (E) Confocal images illustrating Gal3‐positive and Gal3‐negative IBA1⁺ cells in SSLOW‐infected and non‐infected C57Bl/6J mice, correspondingly.

Gal3 expression was highly responsive to genetic perturbations of microglia. Specifically, deletion of P2Y12, a marker of homeostatic microglia, or CD11b, a receptor involved in cell adhesion and phagocytosis, further increased Gal3 expression in prion‐infected mice relative to corresponding prion‐infected controls (Figure 1C). Collectively, these data identify Gal3 as a sensitive marker of both prion infection and altered microglial phenotype in infected animals.

Co‐immunostaining for IBA1 and Gal3 demonstrated that, in prion‐infected brains, Gal3 was expressed by a subpopulation of IBA1⁺ myeloid cells (Figures 1D,E and S1B). In addition, Gal3 signal colocalized with a subset of Vim⁺ cells (Figure S2A), suggesting expression in reactive astrocytes under chronic inflammatory conditions. Gal3⁺/Vim⁺ cells were particularly abundant in the striatum (Figure S2A). Given recent single‐cell RNA sequencing (scRNA‐seq) studies identifying Vim‐expressing microglial populations in prion‐infected mice [34, 43], we examined the cellular distribution of Gal3 using triple immunostaining for IBA1, Vim, and Gal3 (Figure S2B,C). We observed similar patterns in the thalamus, which exhibits the most advanced neuropathology, and in the striatum. In both regions, Vim did not colocalize with IBA1⁺ cells, arguing against substantial expression of Vim in IBA1⁺ microglia on a protein level (Figure S2D). In the thalamus, Gal3 predominantly colocalized with IBA1⁺ cells, as expected, whereas considerably less Gal3 signal was detected in Vim⁺ cells (Figure S2B,D). In contrast, in the striatum, Gal3 was found predominantly in Vim⁺ cells and showed little colocalization with IBA1⁺ cells (Figure S2C,D). Together, these findings demonstrate that Gal3 is expressed not only by reactive microglia but also by Vim⁺ reactive astrocytes. Consistent with previous observations that astrocytic responses to prion infection are region dependent [49, 50, 51], astrocyte‐associated Gal3 expression was particularly prominent in the striatum (Figure S2C).

3.2. Gal3 is upregulated in resident microglia transitioning to a reactive state

Confocal imaging of prion‐infected brains revealed predominantly intracellular localization of Gal3 (Figures 1E and 2A,C,D). To define the identity of Gal3⁺ cells, co‐immunostaining for Gal3 together with TMEM119 or P2Y12, both markers of resident microglia, was performed. Under homeostatic conditions, TMEM119 and P2Y12 are highly expressed but progressively decline as microglia transition into chronically reactive states [43, 47]. In contrast, Gal3 was undetectable under homeostatic conditions (Figures 1E and S1A) and increased progressively beginning at late preclinical stages (Figure 1A). Consistent with these opposing expression dynamics, most Gal3⁺ cells displayed only trace levels of TMEM119 and P2Y12, whereas TMEM119⁺ or P2Y12⁺ cells that had not yet acquired a reactive phenotype exhibited little or no Gal3 signal (Figure 2A–D). Nevertheless, IBA1⁺ cells with mild levels of both Gal3 and TMEM119/P2Y12 were also observed (Figure 2A,C,D), suggesting transitional states during conversion to reactive microglia.

FIGURE 2.

Galectin‐3 (Gal3) is expressed by resident microglia. (A) Confocal images of cortex from SSLOW‐infected C57Bl/6J mice immunostained for Gal3, TMEM119 and IBA1. Scale bar 50 μm. (B) Quantification of TMEM119 mean fluorescence intensity in Gal3⁻/IBA1⁺ and Gal⁺/IBA1⁺ cells in the cortex and thalamus of SSLOW‐infected C57Bl/6J mice. N = 3 animals per group, *p < 0.05, **p < 0.01 by unpaired Student's t‐test. colors represent individual brains; dots represent individual fields of view, average values for each brain are shown as circles; means ± standard deviation are marked by black lines. (C, D) Confocal images of thalamus from SSLOW‐infected C57Bl/6J mice stained for Gal3, P2Y12 and IBA1. Scale bar 50 μm. In (A) and (D), arrows indicate IBA1⁺ cell co‐expressing Gal3 and TMEM119, or Gal3 and P2Y12 positivity, respectively.

3.3. Microglia that envelop neurons express Gal3

We previously demonstrated that, in prion disease, reactive microglia establish close body‐to‐body interactions with neurons, partially wrapping around neuronal somata in a process termed neuronal envelopment [40]. These interactions are dynamic and reversible, persist from minutes to hours [52], and occur across all brain regions affected by prions [40]. Neuronal envelopment is observed across multiple prion strains (SSLOW, 22L, RML, and ME7) as well as in human sCJD, suggesting a conserved mechanism [40]. Notably, the vast majority of envelopment events do not culminate in neuronal phagocytosis. Although the prevalence of neuronal envelopment in the cortex increases significantly with disease progression, individual envelopment events are reversible, and cortical neuronal density does not decline [40]. Importantly, the prevalence of neuronal envelopment correlates with the severity of neuroinflammation, and genetic manipulations that increase neuronal envelopment accelerate disease progression [47]. Because Gal3 functions as an opsonin and has been implicated in interactions between reactive microglia and neurons [53, 54, 55, 56], we next investigated whether microglia involved in neuronal envelopment express Gal3.

Microglia engaged in neuronal envelopment exhibited a broad range of Gal3 expression, ranging from undetectable to strongly positive (Figures 3A–C and S3). At clinical onset, no differences in Gal3 expression were detected between enveloping and non‐enveloping microglia (Figure 3D), potentially reflecting relatively low overall Gal3 levels at this stage. However, at terminal disease stages, microglia engaged in body‐to‐body interactions with neurons displayed significantly higher Gal3 levels than non‐enveloping microglia (Figure 3D). Consistent with previous scRNA‐seq analyses [34, 43], immunostaining revealed that only a subset of IBA1⁺ microglia expressed measurable levels of Gal3 (Figure 3E). Nevertheless, the proportion of microglia engaged in neuronal envelopment tended to be higher within the Gal3⁺ than the Gal3⁻ microglial population, although this difference did not reach statistical significance (Figure 3E). Together, these observations suggest that Gal3 is not required for the initial docking of microglia to neurons, but may be induced following microglia–neuron contact. Alternatively, Gal3‐dependent interactions represent one of several mechanisms that mediate microglial docking to neurons. These observations prompted us to examine whether loss of Gal3 alters disease progression.

FIGURE 3.

Analysis of neuronal envelopment by microglia. (A) Confocal maximum‐intensity projection images of cortex (Ctx) and thalamus (Th) from SSLOW‐infected mice immunostained for IBA1, NeuN and Galectin‐3 (Gal3), demonstrating Gal3 expression in microglia enveloping neuronal somata. Scale bar 5 μm. (B) Representative gallery of individual IBA1⁺ cells showing Gal3⁺ and Gal3⁻ cells not engaged in neuronal envelopment, as well as Gal3⁺ and Gal3⁻ cells engaged in envelopment. (C, D) Representative field of view used for quantification (C), and quantification of Gal3 mean fluorescence intensity in non‐enveloping and enveloping IBA1⁺ microglia at clinical onset and terminal stages (D). Scale bar 50 μm. N = 3 brains per group, three fields of view per animal, *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant by ordinary one‐way analysis of variance with Tukey's multiple comparison tests. (E) Quantification of the numbers of Gal3⁻ and Gal3⁺ microglia and the percentage of microglia engaged in neuronal envelopment within the Gal3⁻ and Gal3⁺ subpopulations. N = 3 brains per group, three fields of view per animal, p by paired t‐test. In (D) and (E), colors represent individual brains; dots represent individual fields of view, average values for each brain are shown as circles; means ± standard deviation are marked by black lines.

3.4. Gal3 deficiency accelerates prion disease progression

The impact of Gal3 deficiency was investigated using constitutive Gal3 knockout (Gal3 KO) mice challenged i.c. with the mouse‐adapted SSLOW strain. Clinical onset, assessed using a composite neurological score evaluating posture, mobility, clasping, and gait abnormalities, occurred at similar times in SSLOW‐infected Gal3 KO and wild type (WT) mice. However, following onset, disease progression was significantly accelerated in Gal3 KO animals (survival time mean ± SD 119.9 ± 6.2) relative to WT controls (125.4 ± 3.1) (Figure 4A). Sex‐stratified analysis showed accelerated disease progression in Gal3 KO mice of both sexes (females: Gal3 KO 116.3 ± 5.8, WT 123.0 ± 2.5; males: Gal3 KO 122.0 ± 5.6, WT 126.6 ± 2.80) (Figure S4).

FIGURE 4.

Galectin‐3 (Gal3) deficiency accelerates disease progression in SSLOW‐infected mice. (A) Survival curves of Gal3 knockout (Gal3 KO) and wild type (WT) mice inoculated with SSLOW via intracerebral route. Statistical analysis was performed using the Mantel‐Cox log‐rank test. (B) Representative Western blots of whole‐brain homogenates from SSLOW‐infected terminally ill Gal3 KO and WT control mice. (C) Quantification of Gal3, PrP^Sc, CD11b, IFITM3, and Vim in brains of SSLOW‐infected terminally ill Gal3 KO and WT mice by Western blots. Protein levels were normalized to actin. N = 6 animals per group, **p < 0.01, ns, non‐significant, by unpaired Student's t‐test, except for Gal3, where the non‐parametric Mann–Whitney test was used. Data for non‐infected brains are shown for reference (N = 2). (D) RT‐qPCR analysis of gene expression in SSLOW‐infected terminally ill Gal3 KO and WT mice and non‐infected age‐matched control groups. N = 3 animals per group, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, non‐significant, by ordinary one‐way analysis of variance with Tukey's multiple comparison tests.

No differences in total PrP^Sc accumulation were observed between Gal3 KO and WT mice at either early clinical or terminal stages (Figures 4B,C and S5A,B). Likewise, expression of interferon‐induced transmembrane protein 3 (IFITM3) and CD11b, markers upregulated in reactive microglia [57], did not differ between genotypes by Western blot analysis (Figure 4B,C). Similarly, quantitative reverse transcription polymerase chain reaction (RT‐qPCR) analysis demonstrated significant upregulation of neuroinflammation‐associated genes (Tlr4, Ccl2, CD68, and GPNMB) in prion‐infected relative to non‐infected mice, but no significant differences between Gal3 KO and WT animals (Figure 4D). Astrocyte reactivity, assessed by Vim expression, was also unchanged (Figure 4B,C). Analysis of CD11b and Vim expression during early clinical stage similarly revealed no genotype‐dependent differences, despite significant increases relative to non‐infected mice (Figure S5B).

3.5. Gal3 deficiency does not alter neuronal envelopment by microglia

To investigate mechanisms underlying accelerated disease progression in Gal3 KO mice, we next examined neuronal envelopment. Neuronal and microglial densities did not differ between SSLOW‐infected Gal3 KO and WT mice (Figure 5B,D). Confocal imaging confirmed extensive neuronal envelopment in both genotypes (Figure 5A). Quantification demonstrated that the percentage of neurons exhibiting body‐to‐body contacts with microglia, as well as the percentage of microglia engaged in envelopment, was comparable between Gal3 KO and WT groups (Figure 5C,E).

FIGURE 5.

Analysis of neuronal envelopment by microglia. (A) Confocal three‐dimensional reconstruction images of SSLOW‐infected terminally ill Galectin‐3 knockout (Gal3 KO) and wild type (WT) mice showing examples of neuronal envelopment, in which microglia (IBA1, green) partially wrap around neuronal somata (NeuN, red). (B–E) Quantification of neuronal density (B), percentage of neurons undergoing envelopment (C), microglial density (D), and percentage of enveloping microglia (E) in the cortex of SSLOW‐infected terminally ill Gal3 KO and WT mice. N = 5 animals per group, ns, not significant by unpaired Student's t‐test. Colors represent individual brains; dots represent individual fields of view; average values for each brain are shown as circles; means ± standard deviation are marked by black lines. Scale bars 25 μm (top) and 10 μm (bottom). DAPI, 4′,6‐diamidino‐2‐phenylindole.

3.6. Gal3 deficiency reduces microglial uptake of PrP^Sc

Because Gal3 is a component of the phagocytic machinery responsible for uptake of extracellular targets, including pathogens, cellular debris, and protein aggregates [55, 58], we next investigated whether Gal3 deficiency affects PrP^Sc uptake. Although microglia do not replicate prions in vivo, they acquire PrP^Sc positivity through phagocytic uptake [11, 12, 40]. We previously demonstrated that microglia internalize PrP^Sc beginning at early subclinical disease stages and accumulate substantial amounts of PrP^Sc within phagolysosomes by terminal disease [40].

Intracellular PrP^Sc puncta associated with microglia were readily detected in both SSLOW‐infected Gal3 KO and WT mice (Figure 6A,B). Quantification revealed no differences in combined extracellular and intracellular microglia‐associated PrP^Sc puncta between genotypes (Figure 6B,C), consistent with Western blot analyses showing similar total PrP^Sc levels (Figure 4B,C). However, the proportion of microglia‐associated PrP^Sc puncta was markedly reduced in Gal3 KO mice relative to WT controls (Figure 6D). Consequently, the ratio of intracellular‐to‐extracellular PrP^Sc was significantly lower in Gal3‐deficient mice, indicating that Gal3 contributes to microglial phagocytic uptake of PrP^Sc (Figure 6E).

We next examined the relationship between Gal3 expression and PrPSc accumulation at the level of individual microglia. PrP^Sc was detected in both Gal3⁺ and Gal3⁻ microglia within the same fields of view in SSLOW‐infected WT animals (Figure 7A,C). Three‐dimensional reconstruction of confocal microscopy images of individual IBA1⁺ cells revealed considerable heterogeneity in the relationship between Gal3 expression and PrP^Sc accumulation. This included microglia with abundant PrP^Sc deposition but minimal Gal3 signal, microglia with strong Gal3 and PrP^Sc signals showing substantial spatial overlap, and microglia with very strong Gal3 expression but relatively sparse PrP^Sc puncta (Figure 7C). Because Gal3 can be secreted into the extracellular space, Gal3⁻ microglia may nevertheless internalize PrP^Sc through a mechanism involving extracellular Gal3‐mediated opsonization [54]. Thus, Gal3‐dependent PrP^Sc uptake may not be restricted to microglia that exhibit detectable intracellular Gal3 upregulation.

FIGURE 7.

Analysis of PrP^Sc associated with Gal3⁺ and Gal3⁻ microglia. (A) Representative immunofluorescence images of the cerebral cortex from SSLOW‐infected terminally ill C57Bl/6J mice stained for IBA1, Gal3 and PrP^Sc. High‐magnification view of the boxed areas is shown on the right. (B) Quantification of the numbers of Gal3⁻ and Gal3⁺ microglia within the same fields of view and the percentage of PrP^Sc+ microglia within the Gal3⁻ and Gal3⁺ subpopulations. N = 4 animals with three fields of view analyzed per animal. **p < 0.01 by paired t‐test. Colors represent individual brains; dots represent individual fields of view; average values for each brain are shown as circles; means ± standard deviation are marked by black lines. (C) Representative gallery of individual PrP^Sc‐positive IBA1⁺ cells generated via three‐dimensional reconstruction of confocal microscopy images from cortexes of SSLOW‐infected terminally ill C57Bl/6J mice stained for IBA1, Gal3 and PrP^Sc. (D) Western blot analysis of Gal3 and PrP^Sc co‐immunoprecipitated in the presence or absence of anti‐Gal3 antibodies. Serial dilutions of SSLOW‐infected brain homogenate are shown as input references.

Because PrP^Sc is relatively resistant to proteolytic degradation, it progressively accumulates within microglia following its initial uptake at early disease stages, resulting in a subset of cells containing very large PrP^Sc deposits, yet limited Gal3 staining. Quantification revealed a trend toward a higher proportion of PrP^Sc‐positive cells among Gal3⁺ microglia compared with Gal3⁻ microglia (Figure 7B), supporting the role of Gal3 in PrP^Sc uptake.

To determine whether Gal3 binds PrP^Sc, we performed an immunoprecipitation assay using anti‐Gal3 antibodies conjugated to magnetic beads to pull down Gal3 from pre‐cleared SSLOW‐infected BH. PrP^Sc was co‐immunoprecipitated with Gal3, whereas beads lacking anti‐Gal3 antibodies failed to precipitate PrP^Sc (Figure 7D), indicating that Gal3 can bind PrP^Sc in SSLOW‐infected brain tissue. (Figure 7D).

A recent study identified a subpopulation of microglia characterized by high expression of GPNMB and transcriptional signatures associated with phagocytic activity [34]. This population was particularly enriched in the thalamus and spatially overlapped with Lgals3‐expressing regions. To determine whether GPNMB and Gal3 mark the same microglial subpopulation, prion‐infected C57BL/6J brains were co‐immunostained for GPNMB, Gal3, and IBA1 (Figure 6A). Quantification within the thalamus, which exhibited the highest proportions of GPNMB⁺ and Gal3⁺ cells, revealed partial overlap between these populations: approximately 20% of IBA1⁺ cells were Gal3⁺/ GPNMB⁻, ~10% of IBA1⁺ cells were Gal3⁻/ GPNMB⁺, whereas 24% of cells were Gal3⁺/GPNMB⁺ (Figure S6B).

3.7. Gal3 deficiency does not impair lysosomal functions

Accumulation of PrP^Sc within phagolysosomal compartments depends not only on uptake efficiency but also on degradation rates mediated by lysosomal activity. Protein aggregates, including α‐synuclein, have previously been shown to disrupt lysosomal membranes and impair lysosomal function [59]. Gal3 has been implicated in sensing and promoting repair of damaged lysosomal membranes [29, 30, 60]. We therefore investigated whether Gal3 deficiency affects lysosomal function during prion disease.

Lysosomal activity was assessed by analyzing expression of LAMP1 and cathepsin D, both previously shown to be upregulated in microglia during prion disease [40, 61, 62]. Western blot analysis revealed no differences in expression of either protein between SSLOW‐infected Gal3 KO and WT mice at either terminal or early clinical stages (Figures 8A,B and S3A,C). In addition, the ratio of pro‐cathepsin D (inactive precursor) to mature cathepsin D (active enzyme), a measure of lysosomal processing efficiency and autophagy‐lysosome pathway activity [63], was unchanged between genotypes at both disease stages (Figure 8B).

FIGURE 8.

Analyses of lysosomal function in prion‐infected Galectin‐3 knockout (Gal3 KO) mice. (A) Representative Western blot of whole‐brain homogenates from SSLOW‐infected terminally ill Gal3 KO and wild type (WT) control mice. Arrows indicate pro‐cathepsin D and mature cathepsin D heavy chain. (B) Quantification of mature cathepsin D, the ratio of pro‐cathepsin D to mature cathepsin D, and LAMP1 expression by Western blots. Protein levels were normalized to actin. N = 6 animals per group, ns, non‐significant, by unpaired Student's t‐test. Data from non‐infected brains are shown as a reference (N = 2). (C) Representative confocal microscopy images showing colocalization of cathepsin D (CathD, red) with lysosomes (LAMP1, gray) within microglia (IBA1, green) in SSLOW‐infected terminally ill Gal3 KO and WT mice, and age‐matched non‐infected WT mice. Scale bar 20 μm. (D–F) Quantification of the mean fluorescence intensities of microglia‐associated LAMP1 (D), CathD (E), and fraction of CathD colocalized with LAMP1 (F). N = 3 animals per group. ns, not significant by unpaired Student's t‐test. Colors represent individual brains; dots represent individual fields of view; average values for each brain are shown as circles; means ± standard deviation are marked by black lines. (G) RT‐qPCR analysis of gene expression in SSLOW‐infected terminally ill Gal3 KO and WT mice and non‐infected age‐matched groups. N = 3 animals per group, *p < 0.05, **p < 0.01, ns, non‐significant, by ordinary one‐way analysis of variance with Tukey's multiple comparison tests.

Immunofluorescence quantification of LAMP1 and cathepsin D within microglia confirmed that intracellular levels of these proteins were comparable between SSLOW‐infected Gal3 KO and WT mice (Figure 8C–E). Because lysosomal membrane rupture results in leakage of lysosomal contents and reduced colocalization of cathepsin D with LAMP1 [60], we next analyzed their colocalization within IBA1⁺ cells. Again, no genotype‐dependent differences were observed (Figure 8F). Finally, expression of genes involved in lysosomal function, sensing, and repair of damaged lysosomes (Alix, Tsg101, and Trim16) [29] was unchanged between genotypes (Figure 8G). Together, these findings indicate that Gal3 deficiency does not measurably impair lysosomal activity or lysosomal membrane repair in prion‐infected microglia. However, these results do not inform whether microglial lysosomes are damaged in prion diseases and whether their damage contributes to disease progression.

3.8. Deficiency in microglial autophagy does not alter disease progression

Autophagy protects cells by removing damaged leaking lysosomes [64, 65, 66]. Therefore, next we investigated whether impairment of microglial autophagy influences prion disease progression. Beclin 1 acts as a central regulator of autophagy initiation, promoting autophagosome formation around damaged organelles. To inhibit autophagy selectively in myeloid cells, we used homozygous Becn1^flox/floxLyz2^Cre/Cre mice (Becn1 cKO), in which Beclin 1 is deleted in myeloid lineages [46].

SSLOW‐infected Becn1 cKO mice exhibited a modest reduction in incubation time to terminal disease relative to control Lyz2^Cre/Cre mice; however, this difference did not reach statistical significance (Figure 9A). Both groups accumulated comparable levels of PrP^Sc as assessed by Western blotting (Figure 9C,D). Confocal imaging demonstrated the presence of PrP^Sc aggregates within microglia in Becn1 cKO mice, similar to control and C57BL/6J mice (Figure 9B,E). Quantification by immunofluorescence revealed a trend toward reduced microglia‐associated PrP^Sc in Becn1 cKO mice, although the difference was not statistically significant (Figure 9E).

FIGURE 9.

Analyses of disease progression in Becn1 cKO mice with impaired microglial autophagy. (A) Survival curves of Becn1 cKO and control mice inoculated with SSLOW via intracerebral route. Statistical analysis was performed using the Mantel‐Cox log‐rank test. (B) Representative immunofluorescent images showing PrP^Sc puncta (SAF84, red) associated with microglia (IBA1, green) in SSLOW‐infected terminally ill Becn1 cKO mice. Scale bar 10 μm. (C) Representative Western blots of whole‐brain homogenates from SSLOW‐infected terminally ill Becn1 cKO and control mice. In the LAMP1 blot, the arrow marks the high‐molecular‐weight isoform attributed to microglia (Mg LAMP1). In the cathepsin D blot, arrows point at pro‐cathepsin D and mature cathepsin D heavy chain. (D) Quantification of PrP^Sc, CD11b, Vim, total LAMP1, and the ratio of microglial LAMP1 to total LAMP1 in brains of SSLOW‐infected Becn1 cKO and control mice by Western blot analysis. Protein levels were normalized to actin. N = 7 animals per group, **p < 0.01, ns, non‐significant, by unpaired Student's t‐test. (E) Quantification of integrated fluorescence densities of microglia‐associated PrP^Sc in SSLOW‐infected terminally ill Becn1 KO and control mice. N = 3 animals per group. ns, not significant by unpaired Student's t‐test. Colors represent individual brains; dots represent individual fields of view; average values for each brain are shown as circles; means ± standard deviation are marked by black lines. (F) Quantification of mature cathepsin D, the ratio of pro‐cathepsin D to mature cathepsin D, and Gal3 by Western blot analysis. Protein levels were normalized to actin. N = 7 animals per group, ns, non‐significant, by unpaired Student's t‐test.

No differences were detected between Becn1 cKO and control mice in expression of CD11b, Vim, total LAMP1, Gal3, cathepsin D, or the ratio of pro‐cathepsin D to mature cathepsin D (Figure 9C,D,F). Notably, microglial LAMP1, which is more heavily glycosylated than neuronal LAMP1 and can therefore be distinguished by Western blotting [67], was increased relative to total LAMP1 in Becn1 cKO mice, suggesting lysosomal accumulation within microglia (Figure 9D), consistent with impaired lysophagy. Collectively, these findings indicate that inhibition of autophagy in microglia exerts only minor effects and does not substantially alter prion disease progression.

4. DISCUSSION

The current study demonstrates that Gal3 upregulation in prion disease becomes detectable at late preclinical stages and gradually increases with disease progression. Although Gal3 expression was observed in a subpopulation of reactive microglia throughout all brain regions exhibiting neuroinflammation (Figure S1A), it was most prominent in the thalamus, consistent with previous transcriptomic studies [34, 43]. The inverse correlation between cellular Gal3 levels and the expression of the homeostatic microglial markers P2Y12 and TMEM119 within individual cells further suggests that Gal3 is associated with an advanced reactive microglial phenotype.

A recent study proposed that the GPNMB⁺/Gal3⁺ microglial subpopulation is responsible for phagocytosis of apoptotic neurons and cellular debris [34]. In the present work, both GPNMB and Gal3 were strongly upregulated in the thalamus, where substantial overlap between GPNMB⁺ and Gal3⁺ microglial populations was observed using co‐immunostaining (Figure S6). Because the GPNMB⁺ phenotype is thought to be induced by apoptotic or dead neurons [34], and Gal3 has also been implicated in microglial phagocytosis of neurons [53, 54], we hypothesized that Gal3 deficiency might disrupt the extensive body‐to‐body interactions between reactive microglia and neurons observed during prion disease. Contrary to this expectation, Gal3 deficiency had no effect on neuronal envelopment by microglia (Figure 6), despite significantly reducing microglial uptake of PrP^Sc (Figure 5). These findings provide functional evidence that neuronal envelopment and phagocytosis are mechanistically distinct processes. Specifically, neuronal envelopment does not necessarily represent an intermediate stage preceding neuronal engulfment or phagocytosis, but rather constitutes a specialized microglia–neuron interaction that can occur independently of Gal3‐mediated phagocytic machinery.

In the current study, loss of Gal3 reduced PrP^Sc uptake by microglia and accelerated disease progression. Gal3 KO animals displayed a significantly lower ratio of intracellular microglia‐associated to extracellular PrP^Sc (Figure 5), supporting a role for Gal3 in PrP^Sc internalization. Immunoprecipitation of Gal3 using anti‐Gal3 antibody co‐immunoprecipitated PrP^Sc, confirming Gal3 binding to PrP^Sc. Gal3 exhibits high affinity for uncapped β‐galactose residues on branched glycans lacking terminal sialylation and can function as an opsonin that promotes phagocytic uptake [23, 24, 25, 26]. Both PrP^C and PrP^Sc carry up to two branched N‐linked glycans per molecule [68, 69, 70, 71], with each glycan containing up to five negatively charged sialic acid residues [69, 70, 71]. In the mammalian brain, however, only ~40% of N‐linked glycan branches are sialylated, leaving a substantial proportion of uncapped β‐galactose residues exposed [72]. Previously, we demonstrated that among the large diversity of PrP^C sialoglycoforms present in the brain, prion strains recruit PrP^C sialoglycoforms selectively [73, 74, 75], and that selectivity is dictated by N‐glycan sialylation status [76]. Mouse‐adapted strains, as well as PrP^Sc associated with distinct CJD subtypes, display a broad range of selectivity profiles. SSLOW is among the most selective strains, preferentially recruiting less sialylated PrP^C isoforms enriched in uncapped β‐galactose residues [44]. Given that Gal3 mediates phagocytosis through recognition of uncapped β‐galactose‐containing glycans, the present findings provide experimental evidence that Gal3 contributes to the phagocytic machinery responsible for PrP^Sc uptake by microglia.

In addition to marked strain‐specific differences [76, 77], PrP^Sc sialylation also varies across brain regions within the same strain [78]. Although these regional differences are modest, the relative ranking of PrP^Sc sialylation levels across brain regions is conserved across multiple strains [78]. Regardless of strain, thalamus‐derived PrP^Sc consistently exhibits lower sialylation than PrP^Sc isolated from the cortex or hippocampus [78]. The observations in the current and previous studies that the thalamus displays the most pronounced neuroinflammatory response, including the strongest Gal3 upregulation [34, 50], together with the finding that thalamic PrP^Sc is the least sialylated [78], support the hypothesis that microglia respond to exposed galactose residues on PrP^Sc glycans and that Gal3 contributes to PrP^Sc uptake.

Consistent with this interpretation, the proportion of PrP^Sc‐positive microglia tended to be higher among Gal3⁺ than Gal3⁻ microglia, although PrP^Sc was also detected in Gal3⁻ microglia. According to the current mechanism of Gal3‐mediated phagocytosis [54, 58], microglia do not necessarily need to express Gal3 to internalize a Gal3‐opsonized substrate. Secreted Gal3 can bind glycan‐containing targets in the extracellular space and function as an opsonin, thereby promoting their uptake by neighboring phagocytes, including microglia that do not themselves express Gal3. Thus, the contribution of Gal3 to PrP^Sc uptake may extend beyond the Gal3‐expressing microglial subpopulation, potentially explaining the presence of intracellular PrP^Sc in Gal3⁻ microglia.

A subpopulation of microglia exhibited very significant upregulation of Gal3 (Figure 7C), yet limited PrP^Sc immunoreactivity. Such a pattern suggests that a subpopulation of cells specialized in Gal3 production and that its role extends beyond PrP^Sc opsonization.

While the difference in disease incubation time between Gal3 KO and WT mice is modest, it is biologically informative, particularly because it is accompanied by a defined cellular phenotype. The reduction in microglial uptake of PrP^Sc provides a mechanistic explanation for how Gal3 deficiency may accelerate disease progression without altering the overall abundance of PrP^Sc in the brain. Microglial uptake of PrP^Sc is not necessarily expected to reduce total brain PrP^Sc levels; rather, microglia may function as a cellular sink, sequestering extracellular prions and thereby limiting their availability to interact with susceptible neuronal cells. Thus, the absence of a detectable difference in total PrP^Sc accumulation does not diminish the physiological significance of altered cellular uptake. On the contrary, the accelerated disease progression in Gal3‐deficient mice, together with reduced microglial association with PrP^Sc, supports the conclusion that microglial sequestration of PrP^Sc represents a biologically meaningful host‐defense mechanism.

Our findings contrast with those reported by Mok et al. [42], who observed the opposite effect of Gal3 deficiency on prion disease progression. In the present study, Gal3 deficiency accelerated, rather than delayed, disease progression and was associated with reduced microglial uptake of PrP^Sc. These findings argue against therapeutic inhibition of the Gal3 pathway in prion disease and instead support a protective role for Gal3‐mediated microglial sequestration of PrP^Sc. The basis for the divergent outcomes between the two studies remains unclear, particularly because the mechanism underlying the prolonged survival of Gal3‐deficient mice reported by Mok et al. was not investigated.

Upregulation of Gal3 is not specific to prion disease, as it has been previously shown to be a marker of generalized microglial activation. Gal3 is upregulated in response to a broad range of CNS insults, including stroke, traumatic brain injuries, and neurotropic viral infections, and has been reported in both acute and chronic states of microglial activation [79, 80, 81, 82, 83]. Notably, neurotropic viruses such as Herpes Simplex Virus 1 can induce acute and profound infiltration of monocytes, which also upregulate Gal3, further demonstrating that Gal3 expression is not restricted to resident microglia [79]. Thus, we would expect Gal3 upregulation to occur in response to other infectious insults, although the magnitude and functional consequences may depend on the nature and stage of the underlying pathology.

Maintenance of lysosomal integrity is essential for phagocytic cells to internalize and process extracellular material. Previous studies in cultured cells demonstrated that Gal3 participates in sensing lysosomal membrane rupture and coordinating lysosomal repair and autophagic degradation [29]. We reasoned that accumulation of PrP^Sc within phagolysosomal compartments might compromise lysosomal integrity and thereby affect lysosomal function. However, we did not observe evidence of more severe lysosomal damage or impaired repair in prion‐infected Gal3 KO mice relative to wild‐type controls. Expression levels of LAMP1 and cathepsin D, markers of lysosomal activity, were not altered by Gal3 deficiency. Moreover, colocalization of cathepsin D with LAMP1, a parameter indicative of lysosomal membrane rupture, was unchanged in the absence of Gal3. Expression of genes involved in lysosomal damage sensing and repair (Alix, TSG101, and Trim16) was also comparable between prion‐infected Gal3 KO and wild‐type animals. These findings suggest that, in vivo, Gal3 is not essential for maintaining lysosomal integrity during prion infection. One possible explanation is that the role of Gal3 in lysosomal repair, previously described in cultured cells [29], may differ under in vivo conditions, where additional compensatory mechanisms are likely to exist. Alternatively, lysosomal membrane repair pathways in vivo may be substantially more redundant than those operating in cultured systems.

Autophagy protects cells from toxicity by eliminating damaged and leaking lysosomes. To determine whether lysosomal damage influences disease progression, we examined prion disease in Becn1^flox/floxLyz2^Cre/Cre mice, in which autophagy is impaired through conditional deletion of Becn1 in microglia. An increased ratio of microglial LAMP1 to total LAMP1 signal illustrated that inhibition of microglial lysophagy was achieved in Becn1 cKO mice. Nevertheless, only a modest, statistically non‐significant acceleration of disease progression was observed, suggesting that inhibition of autophagy in microglia has limited impact on disease progression. PrP^Sc accumulation within microglia begins during early preclinical stages [40]. In the absence of efficient degradation of PrP^Sc, microglial proliferation may increase the overall capacity of the brain to sequester PrP^Sc. However, despite this increase in sequestration capacity, PrP^Sc replication appears to become uncontrolled by the early clinical stage [40]. Thus, the ability of microglia to sequester PrP^Sc during subclinical stages likely influences disease trajectory and the timing of terminal disease. By contrast, autophagic clearance of damaged lysosomes may become relevant only at later stages, when individual microglial cells approach their storage capacity. At this point, preservation of phagolysosomal integrity may no longer substantially alter disease progression.

In summary, the present study demonstrates that Gal3 contributes to the uptake of PrP^Sc by microglia and that disruption of this pathway accelerates disease progression. Because substantial amounts of PrP^Sc still accumulated within microglia in Gal3 KO animals, additional Gal3‐independent mechanisms of PrP^Sc phagocytosis likely exist, providing functional redundancy for this critical microglial activity. Functional redundancy would therefore be expected to attenuate the magnitude of the survival effect without negating the role of Gal3 in PrP^Sc uptake. Although Gal3 is upregulated only in a subset of microglia, its expression serves as a sensitive indicator of neuroinflammation across both disease stages and brain regions. Furthermore, Gal3 upregulation during disease progression is sensitive to alterations in microglial phenotype induced by deletion of microglia‐associated genes.

AUTHOR CONTRIBUTIONS

NM and IVB conceived and designed the study; NM, TS, OB, OM, AS, and NPP, performed experiments; KM managed mouse colony, performed animal procedures and scored the disease signs; NM, TS, and OB analyzed the data; MML, provided animal model; IVB and NM wrote the manuscript. All authors read, edited and approved the final manuscript.

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to declare.

Supporting information

Data S1. Supporting Information.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available upon reasonable request.