Complement protein C3 exacerbates prion disease in a mouse model of chronic wasting disease

Depletion of C3 ameliorates prion-associated chronic wasting disease.

Abstract

Accumulating evidence shows a critical role of the complement system in facilitating attachment of prions to both B cells and follicular dendritic cells and assisting in prion replication. Complement activation intensifies disease in prion-infected animals, and elimination of complement components inhibits prion accumulation, replication and pathogenesis. Chronic wasting disease (CWD) is a highly infectious prion disease of captive and free-ranging cervid populations that utilizes the complement system for efficient peripheral prion replication and most likely efficient horizontal transmission. Here we show that complete genetic or transient pharmacological depletion of C3 prolongs incubation times and significantly delays splenic accumulation in a CWD transgenic mouse model. Using a semi-quantitative prion amplification scoring system we show that C3 impacts disease progression in the early stages of disease by slowing the rate of prion accumulation and/or replication. The delayed kinetics in prion replication correlate with delayed disease kinetics in mice deficient in C3. Taken together, these data support a critical role of C3 in peripheral CWD prion pathogenesis.

Introduction

Chronic wasting disease (CWD) is a highly contagious prion disease of unknown origin affecting both wild and farmed raised deer, moose and elk (1, 2). First described as a disease entity by Williams et al. in the 1970s, CWD has become the most contagious and puzzling prion disease to date. Similar to other prion diseases, CWD is characterized by the accumulation PrP^RES, a proteinase K (PK) resistant form of the cellular prion protein, PrP^C (3). According to the Prion Hypothesis, PrP^RES is the non-genomic pathogen that causes prion diseases. CWD seems to be unique among prion diseases in its prevalence in both wild (≤ 45%) (4) and captive (≤ 90%) (5) animal populations.

Prions have been found in lymphoid and nervous tissue, muscle, blood, feces, urine and saliva (6–11). Much research has focused on lymphoid tissues, as peripheral prion accumulation and replication have been documented there. Prion accumulation and replication occur on follicular dendritic cells (FDCs) within lymphoid follicles of secondary lymphoid organs (SLOs) (12, 13). FDCs differentiate from perivascular precursors and display antigenic immune complexes on their dendritic processes to B cells to promote survival, immunoglobulin affinity maturation and activation to plasma cells (14).

FDCs and B cells are both important for optimal peripheral prion pathogenesis (15–18). Accumulating evidence shows a critical role of the complement system in facilitating attachment of prions to both B cells and FDCs and assisting in prion replication (19, 20). The complement system plays a vital role in immune-mediated defense against pathogens. Multiple pathways activate the complement system, all converging at C3 activation (21). C3 is the most abundant complement protein, present in the blood at mean physiological concentrations of 1.2mg/ml (22). C3 convertases asymmetrically cleave C3, revealing a thioester bond on the large fragment, C3b, that reacts with carbohydrates on microbial surfaces. Covalently bound C3b molecules opsonize microbial pathogens and mark them for phagocytosis by innate immune cells or lysis by the membrane attack complex. Murine complement receptors CR2 (CD21)/CR1 (CD35) expressed on B cells and FDCs trap pathogens coated with C3 cleavage products and mediate appropriate immune responses. Opsonization is critical for eliminating invading pathogens. However, many pathogens manipulate complement regulatory components to avoid being eliminated or promote their attachment to or infection of the host (23, 24). Complement activation exacerbates disease in prion-infected animals, and elimination of CD21/35 inhibits prion accumulation, replication and pathogenesis (25, 26).

In this study, we show that complete genetic or transient pharmacological depletion of C3 prolongs incubation times and significantly delays splenic accumulation in a CWD transgenic mouse model. Using a semi-quantitative prion amplification scoring system we show that C3 impacts disease progression in the early stages of disease by slowing the rate of prion accumulation and/or replication. The dilatory kinetics in prion replication correlate with dilatory disease kinetics in mice deficient in C3.

Methods

Mice

C3^−/− mice were purchased from Charles River (Wilmington, MA, USA), Prnp^o/o and Tg5037 mice were made as previously described (27, 28). Prnp^o/o and C3^−/− mice were bred to produce Prnp^o/oC3^−/− mice, which were crossed with Tg5037 mice to produce Tg5037;Prnp^o/oC3^−/− (Tg5037;C3^−/−) mice. All mice were bred and maintained at Lab Animal Resources, accredited by the Association for Assessment and Accreditation of Lab Animal Care International, in accordance with protocols approved by the Institutional Animal Care and Use Committee at Colorado State University.

Preparation and inoculation of cobra venom factor

Cobra venom factor (CVF) is a C3b homologue that forms a C3 convertase with Factor Bb, but is resistant to inactivation by complement regulatory proteins CR1 and Factor I, resulting in rapid and near complete temporary depletion of C3. We intra-peritoneally injected 100 μl of 900 μg/ml of CVF (Sigma) in PBS or vehicle as a control at least 24h before and 5, 10 and 15 days post CWD infection (DPI). Serum C3 concentrations were assayed by ELISA 0, 8 and 16 DPI.

Enzyme-linked immunosorbent assay

Double antibody sandwich ELISAs were performed as outlined by the manufacturer (Immunology Consultant Laboratory, Inc.). Briefly, sera collected from mice were diluted 1/50 000 and 100 μl of this serum transferred into an anti-C3 ELISA microtiter plate and incubated for 20min. Following incubation the contents were aspirated and wells were washed four times with diluted wash solution. Next, 100 μl of enzyme antibody conjugate was added and the samples incubated in the dark for 20min. The solution was aspirated and wells washed four times with PBS. A TMB substrate (100 μl) was added into each well and allowed to incubate in the dark for 20min. After incubation 100 μl of stop solution was added to each well. Fully developed plates were read at 405nm by an Opsys MR plate reader (Dynex Technologies, Chantilly, VA, USA).

Preparation of inoculum

Ten percent cervid brain homogenates infected with CWD prions were prepared in PMCA buffer (4-mM EDTA, 150-mM NaCl in PBS). These 10% homogenates were diluted 1:10 in 320mM sucrose supplemented with 100 units/ml penicillin and 100 µg/ml streptomycin (Gibco) in PBS immediately prior to inoculation.

Inoculations, clinical scoring and dissections

Mice were inoculated intra-peritoneally with 100 µl of CWD- infected brain homogenate using a 28G insulin syringe. Mice were observed throughout the study for clinical signs of prion disease, including tail rigidity, impaired extensor reflex, akinesia, tremors, ataxia and weight loss. Mice displaying any four of these signs or paralysis were scored terminally ill and sacrificed.

Inoculated mice were sacrificed at distinct time points by CO₂ inhalation. Spleens and brains collected from these mice were divided sagittally. One brain hemisphere and half a spleen were fixed in 4% paraformaldehyde in PBS for 5 days before histology. The other brain hemisphere and half spleen were homogenized and used for PK digestion, western blot analysis and PMCA.

Histology and immunocytochemistry

Slides were prepared and stained as previously described (29).

Protein misfolding cyclic amplification, PK digestion and western blotting

Protein misfolding cyclic amplification (PMCA) on spleen samples was performed and semiquantified as previously described (30). Briefly, 25 µl of 10% spleen homogenates prepared in PMCA buffer (150-mM NaCL, 4-mM EDTA, 1% Triton X-100 in PBS) were incubated with 25 µl of normal brain homogenate (NBH) from uninfected Tg5037 mice. We sonicated samples at 70–85% maximum power for 40 s in a microplate horn sonicator (Qsonica, Framingham, MA, USA), followed by a 30-min incubation at 37°C. We repeated this cycle for 24h, constituting one round of PMCA. We re-diluted 25 µl of sample into 25 µl of fresh NBH for each subsequent PMCA round for five total rounds. PK digestion and western blotting were performed as previously described (29).

We weighted sample scores according to the PMCA round at which they first appeared positive by western blotting and set a detection threshold distinguishing positive from negative PMCA samples based on the 99.9% confidence interval for designating NBH samples as negative, calculated from the mean PMCA score of 73 NBH control samples using the Student’s t-table. We estimated cervid PrP^RES (PrP^CWD) loads in spleens by first generating a standard curve by plotting serial dilutions of a known concentration of PrP^CWD titrated into uninfected spleen homogenates versus their PMCA scores generated after prion amplification. We then quantified PrP^CWD loads (y) by substituting PMCA scores (x) into the non-linear equation y = −0.6626x ^0.3771 + 17.87x ^0.1845 (r ² = 0.88).

Results

Genetic or pharmacological depletion of C3 results in significant delays in CWD pathogenesis

We tested whether complement protein C3 is important early in or throughout peripheral CWD prion infection, or both. To create mice deficient in C3 and susceptible to CWD prions, mice deficient in C3 (C3^−/−) and mouse PrP^C (Prnp^o/o) were crossed to produce C3^−/−Prnp^o/o mice. We then crossed Tg5037 mice that express high levels of elk, but no mouse, PrP^C, with C3^−/−Prnp^o/o mice and screened offspring for Tg5037;C3^−/− mice. To temporarily deplete C3, Tg5037 mice were inoculated with either 100 μl of 900 μg/ml of the C3 convertase-activating CVF (Sigma) or PBS at least 24h before and 5, 10 and 15 DPI with 100 µg brain homogenate from a cervid brain terminally infected with CWD prions (Fig. 1A). At all time points tested there was a significant decrease in C3 concentration in the serum of CVF-treated Tg5037 mice compared with PBS-treated Tg5037 mice. Although serum C3 concentrations were significantly different at 16 DPI, C3 concentrations of CVF-treated mice rose above normal physiological levels (~1mg/ml) indicating a probable antibody response against CVF that limited its effectiveness after the third administration. We estimate that C3 levels remained at or below 50% of PBS-treated mice for 7–10 DPI.

Fig. 1.

Mice genetically and pharmacologically deficient in C3 show delays in CWD prion infection. (A) Tg5037 mice were inoculated with either CVF (n = 10) or PBS (n = 32) at least 24h prior to i.p. inoculation of 100 µl of CWD-infected brain homogenate. After infection with CWD prions, additional CVF inoculations at 5, 10 and 15 days post infection were administered to maintain transient C3 depletion. At 0, 8 and 16 DPI we observed a significant depletion of C3 concentration in mice treated with CVF. (B) CVF-treated Tg5037 and Tg5037;C3^−/− mice (n=22) show a median survival time of 441 and 513 DPI, respectively, compared with PBS-treated and untreated Tg5037 mice, which show a median survival time of 301 DPI. (C) Western blot analysis of PrP^RES content in brains from PBS, Tg5037;C3^−/−, and CVF mice. All samples were digested with 50 µg/ml PK except lane 1. (D) IHC of brain sections from PBS (D and E), CVF (F and G), and C3^−/− (H and I) mice. (J) Densitometric analysis of PrP^RES content in brains from PBS and CVF-treated and Tg5037;C3^−/− mice.

Tg5037;C3^−/− and CVF-treated Tg5037 mice showed a median survival time of 513 and 441 DPI, respectively. These median survival times were significantly longer than Tg5037 control mice, which showed a median survival time of 301 DPI (Fig. 1B, p<0.001). Terminally ill Tg5037;C3^−/− mice, CVF-treated Tg5037 mice and control PBS-treated Tg5037 mice exhibited classic CWD neuropathology, including vacuolization, PrP^CWD deposition and severe astrogloisis (Fig. 1D–I). Western blot analyses revealed PrP^RES in all brains from terminally ill Tg5037;C3^−/−, CVF and PBS-treated Tg5037 mice (Fig. 1C and J).

Absence of C3 delays prion propagation in the spleen

We next used PMCA to evaluate the prion loads in the spleens of CWD-infected Tg5037, Tg5037;C3^−/− and CVF mice. This semi-quantitative prion amplification assay takes advantages of a prion’s ability to self propagate, using seeded protein fibrilization. We used this very sensitive diagnostic technique to amplify diminutive amounts PrP^CWD to detectable levels by stimulating its conversion from PrP^C (31). The starting PrP^CWD concentration in the sample strongly correlates with the PMCA round at which it is first detected by western blotting. We used this relationship to generate semi-quantitative PMCA scores and a standard curve to estimate PrP^CWD loads in spleens from mice at distinct intervals after CWD prion infection. Tg5037 mice at 15 DPI showed a significant difference in the amount of splenic PrP^CWD [Fig. 2, 63.49±8.160 relative PMCA units (rpu)] compared with Tg5037;C3^−/− mice (11.36±9.15 rpu). Using our standard curve for this assay generated as previously described (30), we show that the splenic PrP^CWD load in both Tg5037 mice is ~5500 pg/g of spleen tissue (Fig. 2). The PMCA score for Tg5037;C3^−/− mice falls out of the dynamic range of our assay, so we can only estimate the load to be <100 pg/g. Tg5037 mice at 30 DPI showed no significant difference in prion load (22.92±6.540 rpu, <100 pg/g) compared with Tg5037;C3^−/− mice (20.00±9.720 rpu, <100 pg/g). PrP^CWD increased significantly at 70 DPI (38.930±6.9 rpu, 350 pg/g) and 140 DPI (42.8±6.33 rpu, 500 pg/g) in Tg5037 mice. We detected significantly less PrP^CWD in the spleens of Tg5037;C3^−/− mice at 70 DPI (0±0 rpu, <100 pg/g) and 140 DPI (22.50±10.17 rpu, <100 pg/g).

Fig. 2.

C3-deficient mice showed delayed prion accumulation in the spleen at early time points. Tg5037 mice (n = 83) accumulated significantly more CWD prions in the spleen at 15, 70 and 140 DPI. compared with Tg5037;C3^−/− mice (15, 70 and 140 DPI. P < 0.05; n = 41). There was no significant difference in prion accumulation between PBS-treated (n = 24) Tg5037 mice and CVF-treated Tg5037 mice (n = 18) at 45, 70 and 151 DPI.

In addition to genetically depleting C3, we temporarily depleted C3 with CVF and checked for PrP^CWD accumulation in the spleen of Tg5037 mice. We administered one dose of CVF intra-peritoneally one and five DPI. Although not significantly different, CVF-treated mice at 45 (0±0 rpu, <100 pg/g), 70 (23.33±9.55 rpu, <100 pg/g) and 151 DPI (20.00±11.95 rpu, <100 pg/g) exhibited consistently lower mean prion loads compared with Tg5037 mice.

Discussion

We investigated the role of complement protein C3 in CWD prion accumulation, replication and disease progression. CWD development in mice with genetic or pharmacological depletion of C3 showed significant delays in disease development. In Tg5037;C3^−/− mice the median survival time was longer than that observed in CVF-treated mice. This difference may be attributed to both transient depletion and incomplete knockdown of C3 in CVF-treated mice. The delays in disease development in CWD-infected Tg5037;C3^−/− mice were more drastic than previous studies of C3-deficient mice infected with scrapie (19), even though Tg5037;C3^−/− mice expressed 5-fold less PrP^C. Complement components C1q and C3 have recently been shown to display similar strain preferences in vitro and in vivo (32). These results point to a vital role in prion pathogenesis for C3, the importance of which may differ by prion strain. Interestingly, in CWD and scrapie prion infections, depleting CD21/35 impacts disease progression significantly more than depleting its endogenous ligands, C3 and C4. Furthermore, CD21/35 was greatly enriched in NaPTA- precipitated PrP^RES preparations from spleens of terminally sick mice (19). These data strongly suggest a role of C21/35 in peripheral prion pathogenesis independent of its endogenous ligands. Recently, prion transmission barriers were shown to be more readily breached in the lymphoreticular system than in the nervous system (33). These cross-species infections resulted in distinct lymphotropic and neurotropic strains with differential host ranges. We hypothesized that this phenomenon may be due to the peripheral expression of CD21/35 by FDCs and B lymphocytes, which may play an important role in propagation, replication and strain selection in the spleen. In light of our current findings, we extend our recent hypothesis to include C3. As with CD21/35, lack of C3 may inhibit the efficiency of selection, propagation and replication of neurotropic prion strains. If true, this would have serious implications for cross-species transmission, subclinical infections and possible therapeutic approaches aimed at both complement proteins and their respective receptors.

To study the kinetics of splenic CWD prion accumulation, we amplified PrP^CWD from spleens of CWD prion infected Tg5037, Tg503;C3^−/− and CVF-treated mice at various time points throughout infection. At 15, 70 and 140 DPI, Tg5037;C3^−/− mice showed significantly less PrP^CWD than Tg5037 mice, whereas, no significant difference in splenic PrP^CWD could be detected in CVF-treated mice at any of the time points tested. We hypothesize that the incomplete transient knockdown of C3 in CVF-treated mice may contribute to higher concentrations PrP^CWD compared with Tg5037;C3^−/− mice. Although PrP^CWD was not significantly different between Tg5037 and CVF-treated mice at the time points tested, CVF-treated mice showed a biologically significant delay in disease development compared with Tg5037 mice. This observation indicates that, similar to Tg5037;C3^−/− mice, CVF-treated mice most likely showed a significant decrease in PrP^CWD accumulation before 45 DPI. These data point to a more pronounced role for C3 in the earlier stages of disease. This impairment in splenic prion replication and/or accumulation strongly corresponds with a delay in terminal disease.

While C3 exerts most of its effect early in prion disease, complete, sustained C3 depletion resulted in less severe prion pathogenesis and longer delays to terminal CWD than incomplete, transient depletion. This result is most likely due to insufficient depletion of C3 with CVF, resulting in not only residual endogenous ligands for CD21/35 early during infection, but also normal levels of C3 throughout most of the infection. Consequently, this transient, incomplete depletion of C3 may lead to higher accumulation and/or replication rates in the spleen later in infection. C3 may help to recapture newly synthesized prions from FDCs in SLOs, as well as prions emanating from the CNS in centrifugal movement back to SLOs. This may increase the rate of prionogenesis, resulting in higher prion titers and shorter incubation times in mice replete with C3 at later stages of disease. This result could also be attributed to C3 expression in the central nervous system (CNS) (34, 35). Unlike Tg5037;C3^−/− mice, CVF-treated mice likely express C3 in the CNS after transient depletion with CVF, which may exacerbate disease in the later stages of CWD.

Taken together, these data support a critical role of C3 in peripheral CWD pathogenesis. C3 opsonization of PrP^CWD may facilitate trapping by B lymphocytes and FDCs optimizing intranodal trafficking or peripheral prion replication. Recently, we have shown that B lymphocytes within lymph nodes trap large amounts of prions hours after infection, indicating a important role for these lymphocytes in intranodal trafficking. We have also shown that depleting C3 in these mice affects prion capture by APCs. We suspect that, similar to DCs and monocytes, C3 depletion may also affect prion uptake by CD21/35 positive follicular B cells. This may in turn delay transport of prions to FDCs, consequently leading to delays in prion accumulation, replication and perpheral prion pathogenesis. We are currently investigating the potential role of C3 in this process.

Funding

National Institute of Neurological Diseases and Stroke grant R01 NS056379 funded this work.