Stromal CD21/35 Facilitates Lymphoid Colonization and Pathogenesis

This information is current as Mark D. Zabel, Mathias Heikenwalder, Marco Prinz, Isabelle of September 27, 2021. Arrighi, Petra Schwarz, Jan Kranich, Adriana von Teichman, Karen M. Haas, Nicolas Zeller, Thomas F. Tedder, John H. Weis and Adriano Aguzzi J Immunol 2007; 179:6144-6152; ;

doi: 10.4049/jimmunol.179.9.6144 Downloaded from http://www.jimmunol.org/content/179/9/6144

References This article cites 74 articles, 25 of which you can access for free at: http://www.jimmunol.org/content/179/9/6144.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 27, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Stromal Complement Receptor CD21/35 Facilitates Lymphoid Prion Colonization and Pathogenesis1

Mark D. Zabel,2* Mathias Heikenwalder,* Marco Prinz,† Isabelle Arrighi,* Petra Schwarz,* Jan Kranich,* Adriana von Teichman,* Karen M. Haas,‡ Nicolas Zeller,* Thomas F. Tedder,‡ John H. Weis,§ and Adriano Aguzzi2*

We have studied the role of CD21/35, which bind derivatives of complement factors C3 and C4, in extraneural prion replication and neuroinvasion. Upon administration of small prion inocula, CD21/35؊/؊ mice experienced lower attack rates and delayed disease over both wild-type (WT) mice and mice with combined C3 and C4 deficiencies. Early after inoculation, CD21/35؊/؊ were devoid of infectivity. Reciprocal adoptive bone marrow transfers between WT and CD21/35؊/؊ mice revealed that protection from prion infection resulted from ablation of stromal, but not hemopoietic, CD21/35. Further adoptive transfer experiments between WT mice and mice devoid of both the cellular prion PrPC and CD21/35 showed that splenic retention Downloaded from of inoculum depended on stromal CD21/35 expression. Because both PrPC and CD21/35 are highly expressed on follicular den- dritic cells, CD21/35 appears to be involved in targeting to follicular dendritic cells and expediting neuroinvasion following peripheral exposure to prions. The Journal of Immunology, 2007, 179: 6144–6152.

ransmissible spongiform encephalopathies (TSEs)3 are fa- Extracerebral prion accumulation and replication precedes neu- tal neurodegenerative diseases for which no early diagnosis roinvasion in many cases of TSEs. The immune system plays an http://www.jimmunol.org/ or treatment other than palliation exists. Accumulation of important role in PrPSc neuroinvasion from peripheral sites in mu- TSc PrP , a proteinase K (PK) resistant form of the normal cellular prion rine models of (25–27). Prion accumulation and replication protein, PrPC, is common to most instances of TSEs. Interspecies occur in lymphoid follicles or inflammatory foci containing follic- transmission was seen in TSEs both in the field and in the laboratory, ular dendritic cells (FDCs) that express PrPC (28–32). FDCs are and transmission from bovine spongiform encephalopathy-infected cells of stromal origin that trap immune complexes on their elab- cattle to humans has almost certainly occurred (1–5). The infectious orate projections and present them to B cells, with which they agent, termed prion, has been detected in nervous, lymphoid, and closely associate (33–37). They may retain Ag on their cell sur-

muscle tissues (6–12), as well as , urine, and saliva (13–20). faces for prolonged periods, maximizing presentation to B cells. by guest on September 27, 2021 in cervids and scrapie in appear to be FDCs accumulate PrP-immunoreactive material, and depletion unique among TSEs because of their high transmission efficiency and of FDCs suppresses lymphoid prion titers. These observations sug- prevalence (21–24). gest that FDCs replicate prions (30), although this has not been formally proven. Splenic PrPSc accumulation requires B cells that, despite replicating little or no PrPSc themselves, are required for neuroinvasion (38). Because most B cells express little or no PrPC *Institute for Neuropathology, University Hospital of Zu¨rich, Zu¨rich, Switzerland; (39), this requirement presumably relates to B cells supplying †Institute of Neuropathology, Georg-August University, Go¨ttingen, Germany; ‡De- partment of Immunology, Duke University Medical Center, Durham, NC 27710; and FDCs with lymphotoxins necessary for their maturation and main- §Department of Pathology, University of Utah, Salt Lake City, UT 84132 tenance (40–45). The mechanism by which FDCs trap and prop- Received for publication June 15, 2007. Accepted for publication August 16, 2007. agate prions is poorly understood, and it is unknown whether mol- C The costs of publication of this article were defrayed in part by the payment of page ecules other than PrP are involved. Both B cells and FDCs charges. This article must therefore be hereby marked advertisement in accordance express and interact with components of the with 18 U.S.C. Section 1734 solely to indicate this fact. shown to be important for prion neuroinvasion and pathogenesis. 1 M.D.Z. was supported by the Human Frontiers in Science Foundation Long-Term Fellowship and the Volkswagen Foundation. M.H. was supported by grants of the The serum complement protein C1q binding immune complexes Bonizzi-Theler Foundation, the Schweizer MS Foundation, and the Prof. Max-Cloe¨tta triggers the classical complement pathway. Mice deficient in C1q Foundation. A.A. was supported by the Volkswagen Foundation, the Swiss National exhibit dramatically delayed PrPSc accumulation and onset of ter- Foundation, and the Ernst-Jung Foundation. minal disease when inoculated with scrapie prions (46, 47). C1q 2 Address correspondence and reprint requests to Dr. Mark D. Zabel at the current C address: Department of Microbiology, Immunology and Pathology, College of Vet- binds PrP in a conformation- and density-dependent manner and erinary Medicine and Biomedical Sciences, Colorado State University, 1619 Campus PrPSc activates the classical complement pathway (48, 49). Cleav- Delivery, Fort Collins, CO 80523-1619; E-mail address: [email protected] or age products of the soluble complement C3 and C4 co- Dr. Adriano Aguzzi, Institute of Neuropathology, Department of Pathology, Univer- sita¨tspital Zu¨rich, Schmelzbergstrasse 12, CH-8091 Zu¨rich, Switzerland; E-mail valently attach to microbial surfaces and immune complexes, which address: [email protected] are then presented as Ags on the surface of B cells and FDCs via the 3 Abbreviations used in this paper: TSE, transmissible spongiform encephalopathy; complement receptors CD21/35 (50–52). Pharmacological or genetic PK, proteinase K; FDC, follicular ; BM, bone marrow; dpi, days posti- noculation; SCEPA, scrapie cell assay in endpoint format; WT, wild type; IHC, im- ablation of C3 and full-length membrane-bound receptors CD21/35 munohistochemistry; MBA, mouse bioassay; i.c., intracerebral; NaPTA, sodium delays prion pathogenesis (46, 47). These data suggest that comple- phosphotungstic acid; IF, immunofluorescence stain; PrP, prion protein; PrPC, cellular ment may mediate FDC trapping of prions. prion protein; PrPSc, scrapie-associated, misfolded prion protein. In this study, we show that complete elimination of the comple- Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 ment proteins that trap Ag on FDCs significantly delays splenic prion www.jimmunol.org The Journal of Immunology 6145

accumulation and terminal prion disease in mice inoculated i.p. with (DakoCytomation), and diaminobenzidine (Sigma-Aldrich). PrP was visual- the Rocky Mountain laboratory (RML) strain of prions. Ablation of ized with anti-PrP SAF-84 mAb (A03208, 1/200; SPI Bio). FDCs were complement receptors CD21/35 affected prion trapping and disease stained with anti-FDC-M1 mAb (clone 4C11, 1/50; BD Biosciences) and B cells and FDCs, with rat anti-CD21/35 mAb (7G6, 1/100; BD Pharm- more profoundly than ablating their ligand sources, C3 and C4, sug- ingen) on frozen acetone-fixed sections of spleens and visualized by incu- gesting a role for CD21/35 in peripheral prion pathogenesis indepen- bation with goat Ab against rat IgG (Milan) and alkaline phosphatase- dent of their endogenous ligands. To assess the relative importance of conjugated donkey Abs against goat IgG with fast red. Pairs of consecutive CD21/35 on hemopoietic and stromal cell types, we performed re- sections were used to quantify CD21/35 and FDC-M1 expression on in- dividual splenic follicles from three distinct areas in spleens from two mice ciprocal reconstitution experiments by bone marrow (BM) transplan- per group. Quantification of CD21/35 and FDC-M1 IHC signal intensities tation. CD21/35 expression exclusively on FDCs in white pulp folli- was performed using the CMYK color model (58) in Adobe Photoshop. cles resulted in prion titers, PrPSc retention, and disease kinetics and Two-color immunofluorescence was performed with rabbit anti-mouse PrP severity similar to those of wild-type (WT) mice. CD21/35 expression mAb XN (1/1000) and FITC-conjugated rat anti-mouse CD21/35 mAb 7G6 on hemopoietic cells also significantly affected these parameters, but (1/100) on frozen acetone-fixed sections. Alexa 546-conjugated goat IgG against rabbit IgG (Molecular Probes) was used to visualize PrP. For far less than FDC expression of CD21/35. Therefore, complement- controls, preimmune sera were used or primary Abs were omitted. Histoblot mediated Ag trapping on FDCs is an important mechanism for lym- analysis was performed as previously described (59). Briefly, frozen spleen phoid prion accumulation. sections were transferred to nitrocellulose membranes soaked in lysis buffer using 60 s of constant pressure. Membranes were dried overnight, soaked in Materials and Methods TBST for 1 h, then incubated in digestion buffer containing 0–100 ␮g/ml proteinase K for 4 h. Membranes were rinsed three times for 10 min each in Mice TBST, blocked for1hin5%milk, incubated overnight with anti-PrP Ab 6H4 C57BL/6 ϫ 129sv (B6 129 SF2/J), C3Ϫ/Ϫ and C4Ϫ/Ϫ (C3tm1Crr/J and (1/2000; Prionics), washed again with TBST, and incubated 1 h with alkaline Downloaded from C4tm1Crr, C57BL/6 strain) mice were purchased from The Jackson Labo- phosphatase-goat anti-mouse IgG1 (DO486, 1/1000; DakoCytomation) and ratory. C3Ϫ/Ϫ and C4Ϫ/Ϫ mice were crossed to produce C3/C4Ϫ/Ϫ mice. washed a final time. PrP signals were developed with 5-bromo-4-chloro-3- TgA20, which overexpress mouse PrP, CD21/35Ϫ/Ϫ and Prnpo/o mice have indolyl phosphate/nitro-blue tetrazolium chloride for 1 h. Ϫ/Ϫ o/o been previously described (53–55). CD21/35 and Prnp mice were BM reconstitution and FACS analysis crossed to produce Prnpo/oCD21/35Ϫ/Ϫ mice. Breeding and experiments were performed in compliance with the animal experimentation guidelines BM was taken from tibiae and femurs of respective donor groups and Ն107

of the Kanton of Zu¨rich. cells were injected into tail veins of lethally irradiated (1100 radians for 10 http://www.jimmunol.org/ min) recipient mice as previously described (60). Six to 8 weeks after Prion inoculations reconstitution, 2 ϫ 106 peripheral blood cells from mice of each group Transgenic and control (C57BL/6 ϫ 129sv) animals were infected i.p. with were stained at 4°C with 1/100 dilutions of FITC-labeled rat anti-mouse 100 ␮l of brain homogenate diluted in 320 mM sucrose, containing 3 or 6 log CD21/35 Ab 7G6, PE-labeled rat anti-mouse B220 mAb (BD Pharmingen) or Cy5-labeled mouse anti-mouse PrP mAb (POM-1) in FACS buffer LD50 U of the Rocky Mountain laboratory (RML) scrapie strain passage 5 (RML 5.0), the titer of which was previously assessed by intracerebral (i.c.) (0.1% BSA, 10 mM EDTA in PBS). RBC were lysed using FACSLyse solution and the remaining cells analyzed using a FACSCalibur flow cy- inoculation into TgA20 mice and found to be 8.9 log LD50/g of brain tissue. Mice were monitored every other day, and scrapie were diagnosed according tometer and CellQuest software (BD Biosciences). Live were to clinical criteria including , kyphosis, tail rigidity, and hind leg paresis. gated based on forward scatter and side scatter properties and analyzed for CD21/35, B220, and/or PrP expression using FlowJo software.

Mice were sacrificed at the onset of terminal disease. by guest on September 27, 2021 Sc Infectivity mouse bioassay (MBA) Sodium phosphotungstic acid (NaPTA) precipitation of PrP Assays were performed on 1% spleen homogenates. Spleen tissues were Ten percent spleen homogenates were prepared in PBS as described above. ϫ ␮ homogenized in sterile 320 mM sucrose (1/10) with a RiboLyser (Hybaid), Gross cellular debris was removed by centrifugation at 80 g and 500 l centrifuged for 5 min at 500 ϫ g. Cleared supernatants were diluted 1/10 of supernatant mixed 1:1 with 4% sarkosyl in PBS. Samples were incu- in sterile 5% BSA in PBS and 30 ␮l was injected i.c. into each of four bated for 15 min at 37°C with constant agitation, then incubated with 50 TgA20 mice per homogenate (55). Titers were determined using the rela- U/ml benzonase and 12.75 mM MgCl2 for 30 min at 37°C with constant ϭ Ϫ agitation. Prewarmed NaPTA stock solution (pH 7.4) was added to a final tionship: y 11.45 0.088x, where y is log LD50 and x is the incubation time in days to terminal disease (56). Histological and immunohistochem- concentration of 0.3% and the sample was incubated at 37°C for 30 min with ical analyses using H&E, glial fibrillary acidic protein, and SAF84 staining constant agitation and centrifuged at 37°C for 30 min at maximum speed in an ␮ revealed spongiosis, gliosis, and PrP deposition in all scrapie symptomatic Eppendorf microcentrifuge. The pellet was resuspended in 30 l of 0.1% sar- ␮ mice and none of these histopathological features in asymptomatic mice kosyl in PBS and digested with 20 g/ml PK for 30 min at 37°C. sacrificed 180–200 days postinoculation (dpi). Immunoblot analysis Scrapie cell assay in endpoint format (SCEPA) Tissue homogenates were NaPTA precipitated or adjusted to 5 mg/ml pro- ␮ ␮ Replicate aliquots of highly prion-susceptible neuroblastoma cells (sub- tein and 50 g treated or not with 20 g/ml PK for 30 min at 37°C. clone N2aPK1; Ref. 57) were placed into 12 wells of a 96-well plate and Samples were heated at 95°C for 5 min in SDS-PAGE loading buffer and exposed to prion samples for 3 days, split 1:3 three times every 2 days, and pipetted into wells of 12% or 4–12% gradient Novex SDS polyacrylamide 1:10 three times every 3 days. After reaching confluence, 2.5 ϫ 104 cells gels (Invitrogen Life Technologies) and electrophoresed. Proteins were from each well were filtered onto the membrane of an ELISPOT plate, transferred to nitrocellulose membranes (Schleicher-Schuell) by wet blot- treated with PK, denatured, and individual infected (PrPSc -positive) cells ting, blocked with TBST containing 5% Topblock (Juro) incubated with were detected immunochemically using Ab ICSM-18 to PrP. Wells were anti-mouse PrP mAb POM-1 (1/2000) overnight at 4°C. Membranes were scored positive if the spot number exceeded mean background values plus washed and incubated with HRP-conjugated rabbit anti-mouse IgG1 for 1 h five times the SD. From the proportion of negative to total wells, the num- at room temperature. Bands were detected by chemiluminescence (Pierce) ber of “infectious tissue culture units” per aliquot was calculated using the and visualized by the VersaDoc imaging system (Bio-Rad). Bands were Poisson equation. The potency of the SCEPA is based on the finding that quantified using QuantityOne software (Bio-Rad). the proportion of infected cells, and with it the signal-to-background ratio, Statistical analyses increases on average ϳ25% per day during culturing (57). One-way ANOVA and Student’s t test were performed where appropriate Histology and immunohistochemistry (IHC) using the software packages GraphPad Prism and Microsoft Excel. A total of 2 ␮m of paraffin or 5–10 ␮m of frozen sections from brain and spleen were stained with H&E. Infected, formalin-fixed brain samples were Results treated with concentrated formic acid for 60 min to inactivate prion infec- Absence of CD21/35 or its ligands delays prion disease tivity. Astrocytes were stained with rabbit anti-glial fibrillary acidic protein mAb (DakoCytomation; 1/300) and visualized with biotinylated swine Mice deficient for both C3 and C4 were obtained by serial crosses Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ serum against rabbit IgG (DakoCytomation; 1/250), avidin-peroxidase of C3 / and C4 / mice. The resulting C3/4 / offspring were 6146 CD12/35 EXPEDITES PRION NEUROINVASION

Table I. Prion disease progression in complement-deficient mice press a surface-localized, biologically active hypomorphic isoform of CD21/35 (CD21/35hypo) at 40% of the level of full-length

6 Log LD50 3 Log LD50 CD21/35 in WT mice (62). We reported previously that CD21/ 35hypo mice exhibited minimally delayed incubation times after Host Genotypea Incidenceb dpic Ϯ SD Incidence dpi Ϯ SD i.p. prion inoculation (46 days) with limiting doses of RML 5 C3Ϫ/Ϫ 8/8 193 Ϯ 10 5/5 276 Ϯ 19d prions (3 log LD50) and no delay at saturating doses (46). C3/4Ϫ/Ϫ 6/6 249 Ϯ 38d,e 6/6 290 Ϯ 1d Ϫ Ϫ Because the above studies used mice that express functional CD21/35 / 10/10 255 Ϯ 26d,e 15/20 350 Ϯ 55e–g WT 8/8 206 Ϯ 8 15/15 238 Ϯ 8 CD21/35 molecules, they may underestimate the contributions of CD21/35 to prion pathogenesis. Therefore, we now used mice a Mice were inoculated i.p. with RML 5 prions. b Incidence ϭ Number of terminally sick mice/number of animals inoculated. completely deficient in CD21/35 expression (54). When inoculated Ϫ Ϫ c dpi, Days postinoculation to terminal disease. i.p. with limiting doses of prions, 25% of CD21/35 / mice did d Value of p Ͻ 0.05 compared to WT. ϭ e Value of p Ͻ 0.05 compared to C3Ϫ/Ϫ. not contract disease (Table I, n 20). Those mice that developed f Value of p Ͻ 0.01 compared to WT. scrapie reached the terminal stage of disease (350 Ϯ 55 days, n ϭ Ϫ Ϫ g Value of p Ͻ 0.05 compared to C3/4 / . 15) 112 days later than congenic WT controls (238 Ϯ 8, n ϭ 8). CD21/35Ϫ/Ϫ contracted disease 74 days later than C3Ϫ/Ϫ Ϯ ϭ Ϫ/Ϫ Ϯ inoculated with saturating (6 log LD ) or limiting (3 log LD ) (276 19, n 5) and 60 days later than C3/4 (290 1, 50 50 ϭ doses of RML 5 prions i.p. Upon high-dose prion challenge (Table n 6) mice, all of which died from both groups. When inoc- Ϫ/Ϫ Ϫ/Ϫ ulated with saturating doses of prions, CD21/35 mice con- I), C3 mice contracted terminal prion disease with an incuba- Downloaded from Ϯ ϭ tion time similar to WT controls (193 Ϯ 10 and 206 Ϯ 8 days, tracted disease (255 26, n 10) 49 days later than WT Ϫ/Ϫ Ϯ ϭ respectively, n ϭ 8/group). In contrast, C3/4 mice contracted (206 8, n 8). Ϫ/Ϫ prion disease 43 days later than WT controls (249 Ϯ 38 days, n ϭ We examined terminally sick WT and CD21/35 mice for 6). Upon limiting-dose inoculation, C3Ϫ/Ϫ (276 Ϯ 19, n ϭ 5) and characteristic signs of prion neuropathology. Little or no vacuola- C3/4Ϫ/Ϫ (290 Ϯ 1, n ϭ 6) mice contracted disease 38 and 52 days tion, astrogliosis or PrPSc deposition was evident in sections of Ϫ Ϫ later, respectively, than WT controls (238 Ϯ 8, n ϭ 15). Taken from asymptomatic CD21/35 / sacrificed 242 days http://www.jimmunol.org/ together, these data indicate an involvement of both C3 and C4 after inoculation (Fig. 1, A–C). In contrast, microvacuolation and in extraneural prion pathogenesis and neuroinvasion. astrogliosis were evident in sections from terminally sick WT (Fig. Ϫ Ϫ We next analyzed the contribution of the CD21/35 receptors. In 1, D and E) and CD21/35 / mice (Fig. 1, G and H). Although the a previous study, we had used mice that had been originally re- pattern of PrPSc deposition was similar, brains of WT mice accu- ported to lack CD21/35 expression (61). However, these mice ex- mulated more PrPSc (Fig. 1F) than CD21/35Ϫ/Ϫ brains (Fig. 1I). by guest on September 27, 2021

FIGURE 1. Delayed neuropathology and PrPSc deposition in brains of CD21/35Ϫ/Ϫ mice. IHC of hippocampal sections from asymptomatic CD21/ 35Ϫ/Ϫ mice 242 dpi shows little or no vacuolation, astrogliosis, or PrPSc accumulation (A–C). Sections from terminally sick WT (244 dpi, D–F) and CD21/ 35Ϫ/Ϫ (475 dpi, G–I) mice reveal similar micro- vacuolation (D and G) and astrogliosis (E and H), but greater PrPSc accumulation in WT (F) than CD21/35Ϫ/Ϫ (I) brain. Scale bars, 100 ␮m. J, West- ern blot (WB) analysis of PrPSc content from 20 ␮g of brain homogenate from asymptomatic CD21/ 35Ϫ/Ϫ mice and terminally sick WT and CD21/ 35Ϫ/Ϫ mice. K, Quantification of band intensities from the WB in (J) confirms that brains from ter- minally sick WT mice contain significantly more PrPSc than brains from asymptomatic or terminally .(p Ͻ 0.05 ,ء) sick CD21/35Ϫ/Ϫ mice The Journal of Immunology 6147 Downloaded from

FIGURE 3. PrPSc accumulation correlates with CD21/35 expression at 47 dpi. A, Consecutive sections display splenic follicle (H&E) that contain PrPSc deposits (histoblots) and express both CD21/35 (green) and PrPC (red) in WT mice (IF). B,NoPrPSc deposition was detected in splenic http://www.jimmunol.org/ follicles expressing PrPC in CD21/35Ϫ/Ϫ mice. Higher magnifications of the boxed areas are shown below each figure. Scale bars, 200 ␮m.

15, 30, and 45 days earlier with 3 log LD50 of RML 5.0 prions. Both assays detected prion infectivity in WT spleens at all three time points, but no infectivity at the detection limits for the MBA Ϫ/Ϫ (1.5 log LD50) and SCEPA (2.4 log LD50) in CD21/35 spleens.

These data indicate severe impairment of prion accumulation in by guest on September 27, 2021 CD21/35Ϫ/Ϫ mice soon after infection. FIGURE 2. Transmission bioassays reveal no infectious prion titers in spleens from CD21/35Ϫ/Ϫ mice at 15, 30, and 45 dpi with 3 log LD of 50 U PrPSc colocalized with CD21/35 early after prion infection RML 5.0 administered i.p. A, Mouse infectivity bioassays (MBA). Each symbol represents results from four TgA20 indicator mice inoculated i.c. Because accumulation of prion infectivity correlated with with 30 ␮l of spleen homogenate from WT (circles) or CD21/35Ϫ/Ϫ (tri- CD21/35 expression in spleens of mice inoculated i.p. with prions, angles) mice. SDs were within 0.75 log units. Data points below the dotted we investigated whether early accumulation of PK-resistant PrPC line represent attack rates Ͻ100%. B, Scrapie cell end point assays (PrPSc) correlated with CD21/35 expression in splenic follicles. (SCEPA) confirm MBA. Highly scrapie-susceptible PK1 cells were incu- We analyzed WT and CD21/35Ϫ/Ϫ spleens for PrPSc accumulation ␮ Ϫ/Ϫ bated with 30 l of spleen homogenate from two WT (circles) or CD21 by histoblot (59). We detected PrPSc in WT (Fig. 3A), but not mice (triangles) at the indicated times after i.p. inoculation (dpi). The area between the dashed horizontal lines indicates the dynamic range of the SCEPA. Data points outside these lines are off-scale.

We quantified the amount of PrPSc in 20 ␮g of brain homoge- nate from each group by Western blot (Fig. 1J). We detected all three PrP glycoforms, with the expected resistance and shift in m.w. of PrPSc after PK digestion. Brain homogenates from terminally sick WT mice contained 15-fold more PrPSc than ho- mogenates from asymptomatic CD21/35Ϫ/Ϫ mice sacrificed at similar time points, and 4-fold more PrPSc than brains from ter- minally sick CD21/35Ϫ/Ϫ mice (Fig. 1K, p Ͻ 0.05).

Absence of CD21/35 impaired early splenic prion accumulation The reduced attack rate or delay in terminal disease of CD21/ 35Ϫ/Ϫ mice after i.p. prion challenge could stem from inefficient prion accumulation and replication in the periphery early after in- fection. We therefore examined prion loads in spleens of WT and FIGURE 4. FACS analysis of PBLs from WT, CD21/35Ϫ/Ϫ, and BM Ϫ/Ϫ CD21/35 mice shortly after peripheral prion inoculation. We chimeric mice. PBLs were stained with Abs recognizing the marker used the MBA (Fig. 2A) and scrapie cell endpoint assay (SCEPA; B220 and CD21/35, confirming presence or absence of CD21/35 on B Fig. 2B) to determine prion titers of spleens from mice inoculated cells. Numbers indicate percentage of double-positive cells. 6148 CD12/35 EXPEDITES PRION NEUROINVASION

Table II. Prion disease progression in mice differentially expressing CD21/35

6 Log LD50 3 Log LD50

Donor BMa Host Genotypeb CD21 Expressionc Incidence dpi Ϯ SD Incidence dpi Ϯ SD

CD21/35Ϫ/Ϫ WT FDCs 4/4 235 Ϯ 35 8/8 254 Ϯ 22d WT CD21/35Ϫ/Ϫ B cells 8/8 267 Ϯ 18d 8/8 334 Ϯ 42e,f CD21/35Ϫ/Ϫ CD21/35Ϫ/Ϫ none 2/2 256, 280 2/5 312, 367e,f WT WT B cells and FDCs 2/2 189, 195 7/7 206 Ϯ 12

a BM was isolated from mice of the indicated genotype. b Hemopoietic systems of lethally irradiated mice of the indicated genotypes were reconstituted with donor BM. c CD21/35 expression was restricted to the indicated cell types for each reconstitution group. d Value of p Ͻ 0.05 compared to WT3WT. e Value of p Ͻ 0.01 compared to WT3WT. f Value of p Ͻ 0.01 compared to CD21/35Ϫ/Ϫ3WT.

CD21/35Ϫ/Ϫ (Fig. 3B), spleens at 47 dpi. Accumulation of PrPSc FDCs express significant amounts of CD21/35 C correlated with CD21/35 and PrP expression in WT spleens (Fig. We attribute the impaired progression of prion disease in 3A, immunofluorescence stain (IF)). WT3CD21/35Ϫ/Ϫ mice to their lack of CD21/35 expression on Downloaded from Loss of CD21/35 expression on FDCs delayed prion disease progression To define the cellular compartment in which CD21/35 is important for prion replication, we prepared reciprocal BM chimeric mice with CD21/35 expression restricted to hemopoietic (WT3CD21/ Ϫ/Ϫ Ϫ/Ϫ 35 mice) or stromal (CD21/35 3WT mice) compartments http://www.jimmunol.org/ (Fig. 4). For control, WT mice were reconstituted with WT BM (WT3WT mice) and CD21/35Ϫ/Ϫ mice were reconstituted with CD21/35Ϫ/Ϫ BM (CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice). We con- firmed highly efficient BM engraftment by FACS analysis of PBLs from unmanipulated (WT and CD21/35Ϫ/Ϫ) and reconstituted mice (Fig. 4). BM from WT mice reconstituted 97% of CD21/35- expressing B cells in irradiated WT (WT3WT, 63%) and CD21/ 35Ϫ/Ϫ (WT3CD21/35Ϫ/Ϫ, 63%) mice compared with WT mice (65%). CD2135Ϫ/Ϫ BM reconstitution eliminated 94–97% of by guest on September 27, 2021 CD21/35-expressing B cells in irradiated WT (CD21/35Ϫ/Ϫ3 WT, 5%) mice compared with CD21/35Ϫ/Ϫ (1%) or CD21/ 35Ϫ/Ϫ3 CD21/35Ϫ/Ϫ mice (3%), respectively. After high- or low-dose prion challenge, control chimeric WT3WT and CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice displayed differ- ences in incubation times similar to those of unmanipulated WT and CD21/35Ϫ/Ϫ mice (Table II). At high-dose prion challenge, CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice contracted terminal disease 61–91 days later (256 and 280, n ϭ 2) than WT3WT mice (189 and 195, n ϭ 2). We observed incomplete attack rates (two of five) and a 106- to 161-day delay of terminal disease at low-dose chal- lenge of CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice (312 and 367, n ϭ 2) compared with WT3WT mice (206 Ϯ 12, n ϭ 7). At high-dose challenge, WT3CD21/35Ϫ/Ϫ mice contracted disease 72–78 days later (267 Ϯ 18, n ϭ 8) than WT3WT mice. At low-dose chal- lenge, WT3CD21/35Ϫ/Ϫ mice contracted disease 128 days later (334 Ϯ 42, n ϭ 8) than WT3WT mice and 80 days later than CD21/35Ϫ/Ϫ3WT mice (254 Ϯ 22, n ϭ 8). These data confirm the importance of CD21/35 expression on FDCs for facilitating FIGURE 5. FDCs express significant amounts of CD21/35. Immuno- terminal prion disease. Although the delay observed for these mice histochemical staining of consecutive splenic sections from WT, CD21/ (334 days), which lack CD21/35 expression on FDCs, is very sim- 35Ϫ/Ϫ, and CD21/35Ϫ/Ϫ3WT chimeric mice with Abs recognizing ilar to that observed for control CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice CD21/35 and the FDC marker FDC-M1. B cells and FDCs express Ϫ/Ϫ (312 and 367 days), we observed no decreased incidence (eight of CD21/35 in WT (A and B), but not CD21/35 (C and D), spleens. Al- Ϫ/Ϫ3 eight) as we did for the controls (two of five). In addition, CD21/ though splenic B cells from CD21/35 WT chimeric mice do not ex- Ϫ/Ϫ3 press CD21/35, FDCs from these mice express significant amounts of 35 WT mice, which lack CD21/35 expression on B cells, ␮ 3 CD21/35 (E and F) relative to WT spleens. Scale bar, 50 m. G, Quan- contracted disease 48 days later (254 days) than WT WT mice tification of CD21/35 and FDC-M1 signal intensities in splenic folli- (206 days) at low-dose prion challenge. Delayed disease progres- cles. The ratio of CD21/35:FDC-M1 staining reveals significant Ϫ/Ϫ sion in CD21/35 3WT mice was surprising, and may point to CD21/35 expression on B cells and FDCs in WT mice (3.1 Ϯ 1), none a previously unrecognized function for B cell expression of in CD21/35Ϫ/Ϫ mice (0.3 Ϯ 0.3), and CD21/35 expression exclusively CD21/35 in prion pathogenesis. on FDCs in CD21/35Ϫ/Ϫ3 WT mice (1.1 Ϯ 0.4). The Journal of Immunology 6149

(Fig. 5, A and B), while CD21/35Ϫ/Ϫ follicles expressed B220 (data not shown), FDC-M1 but no CD21/35 (Fig. 5, C and D). These data confirm the absence of CD21/35 on CD21/35Ϫ/Ϫ B cells, the BM progenitors of which we used to reconstitute the hemopoietic compartment in lethally irradiated WT mice (CD21/ 35Ϫ/Ϫ3WT). B cells did not express CD21/35, while FDCs ex- pressed substantial amounts of CD21/35 in CD21/35Ϫ/Ϫ3WT spleens (Fig. 5, E and F). Quantification of the ratio of CD21/35: FDC-M1 signal intensities from IHC in CD21/35Ϫ/Ϫ3WT splenic follicles (Fig. 5G) confirmed that FDC-M1 staining corre- lated very closely with CD21/35 (1.1 Ϯ 0.4, n ϭ 27 follicles) in splenic follicles, convincingly demonstrating significant CD21/35 expression by FDCs. The CD21/35:FDC-M1 ratio increased to 3.1 Ϯ 1 in WT follicles (n ϭ 27), most likely due to B cell ex- pression of CD21/35. The CD21/35:FDC-M1 ratio in CD21/35Ϫ/Ϫ follicles is 0.3 Ϯ 0.3 (n ϭ 29), confirming lack of CD21/35 expression. FIGURE 6. PrPSc accumulation correlates with CD21/35 expression on FDCs at 47 dpi. Consecutive sections display splenic follicles (H&E) that Loss of CD21/35 expression on FDCs impairs early splenic Downloaded from contain PrPSc deposits (histoblots) and express both CD21/35 (green) and PrPSc accumulation and replication PrPC (red) in WT3WT (A) and CD21/35Ϫ/Ϫ3WT (D) mice (IF), but no PrPSc deposition in splenic follicles lacking CD21/35 in CD21/35Ϫ/Ϫ3 We next investigated whether disease progression correlates with Sc CD21/35Ϫ/Ϫ (B) and WT3 CD21/35Ϫ/Ϫ (C) mice. Higher magnifica- PrP deposition and CD21/35 expression in reciprocal BM chi- tions of the boxed areas are shown to the right of each figure. Scale bars, meric mice (Fig. 6). Histoblot analyses revealed significant PrPSc 200 ␮m. deposition in WT3WT and CD21/35Ϫ/Ϫ3WT mice at 47 dpi, C whose splenic follicles coexpressed PrP and CD21/35 (Fig. 6, A http://www.jimmunol.org/ and D). No PrPSc was detected in CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ or FDCs. However, because CD21/35-expressing B cells closely as- WT3CD21/35Ϫ/Ϫ splenic follicles, which lack CD21/35 expres- sociate with FDCs, CD21/35 expression by FDCs has been diffi- sion entirely or on FDCs, respectively (Fig. 6, B and C). Thus, cult to ascertain. We analyzed spleens from CD21/35Ϫ/Ϫ3WT among the BM chimeras, we detected PrPSc only in spleens that mice to assess CD21/35 expression exclusively by FDCs. IHC express CD21/35 on FDCs, which colocally expressed CD21/35 confirmed that WT splenic follicles coexpressed the B cell marker and PrPC. We detected no PrPSc on FDCs expressing PrPC but not B220 (data not shown), CD21/35 and the FDC marker FDC-M1 CD21/35. by guest on September 27, 2021

FIGURE 7. Efficient PrPSc accumulation corre- lates with CD21/35 expression on FDCs. Western blot analysis of spleens from the indicated reconsti- tution groups 20 dpi (A and B)or47dpi(C and D) with 6 log LD50 RML 5.0 administered i.p. (A and B) PrPSc detection required NaPTA precipitation from 5 mg of spleen from BM chimeric mice (WT3CD21/35Ϫ/Ϫ, Prnpo/o3CD21/35Ϫ/Ϫ, CD21/ 35Ϫ/Ϫ3Prnpo/o, CD21/35Ϫ/Ϫ3WT, and CD21/ 35Ϫ/Ϫ3 CD21/35Ϫ/Ϫ) unless they expressed CD21/35 on both B cells and FDCs (WT3WT, WT3Prnpo/o, and Prnpo/o3WT mice). C, NaPTA- precipitated PrPSc was not detected 47 dpi in spleens from Prnpo/oCD21/35Ϫ/Ϫ BM chimeric mice (PϪ/Ϫ CϪ/Ϫ3PϪ/ϪCϪ/Ϫ and CϪ/Ϫ3 PϪ/ϪCϪ/Ϫ) unless they were reconstituted with CD21-expressing BM (PϪ/Ϫ3PϪ/ϪCϪ/Ϫ and WTϪ/Ϫ3PϪ/ϪCϪ/Ϫ mice), indicating CD21/35 capture of PrPSc. D, PrPSc de- tection required NaPTA precipitation from spleens of Prnpo/o or CD21/35Ϫ/Ϫ mice reconstituted with BM lacking both PrPC and CD21/35 (PϪ/Ϫ CϪ/Ϫ3PϪ/Ϫ and PϪ/ϪCϪ/Ϫ3CϪ/Ϫ mice, respec- tively), whose FDCs lack either PrPC or CD21/35. Expression of both CD21/35 and PrPC on FDCs (PϪ/ϪCϪ/Ϫ3WT and WT3WT mice) restored ef- ficient capture and replication of PrPSc, which was detected without NaPTA precipitation. 6150 CD12/35 EXPEDITES PRION NEUROINVASION

CD21/35 improves PrPSc retention these mice, a phenomenon known as skipping (63) occurred To examine retention of PrPSc by CD21/35 in more detail, we that caused inappropriate splicing of 8 and 11, of inoculated high doses of prions i.p. into reciprocal BM chimeric targeted exons 9 and 10, and creation of an in-frame transcript that hypo mice with segregated CD21/35 and PrP expression on B cells or produced CD21/35 proteins (CD21/35 ) that are 14 kDa FDCs. Twenty days after inoculation, most samples required smaller, yet retain domains required for CD19 interaction and li- hypo NaPTA precipitation of PrPSc from 5 mg of spleen for detection gand binding (62). CD21/35 mice exhibit modestly delayed (Fig. 7, A and B). Eliminating CD21/35 expression on either B incubation times after low-dose i.p. prion inoculation (46). Mice cells or FDCs significantly diminished PrPSc retention used in the current study possess a bona fide homozygous ablation (WT3CD21/35Ϫ/Ϫ, Prnpo/o3CD21/35Ϫ/Ϫ,WT3CD21/35Ϫ/Ϫ, of the CD21/35 and signal sequence that completely CD21/35Ϫ/Ϫ3Prnpo/o, and CD21/35Ϫ/Ϫ3CD21/35Ϫ/Ϫ mice). eliminates CD21/35 protein expression (54). After low-dose i.p. Ϫ/Ϫ PrPSc was detected in 50 ␮g of spleen homogenate without NaPTA prion inoculation, we now found that CD21/35 mice exhibited precipitation only in samples from mice expressing CD21/35 on dramatically prolonged incubation and incomplete attack rates Ϫ/Ϫ both B cells and FDCs (WT3WT, WT3Prnpo/o and Prnpo/o3 when compared with congenic controls. CD21/35 mice also WT mice). presented with significantly delayed prion neuropathology. The Splenic PrPSc detected at 20 dpi is most likely from the initial role of CD21/35 in prion pathogenesis may therefore be more prion inoculum, the capture of which may be aided by CD21/35. prominent than previously appreciated. Ϫ Ϫ To investigate whether early prion replication might contribute to CD21/35hypo and C3 / mice show similar delays in peripheral the detected PrPSc, we inoculated high doses of prions i.p. into BM prion pathogenesis even though FDCs failed to bind measurable chimeric mice lacking PrPC expression, but retaining CD21/35 ex- C3 in CD21/35hypo mice (61, 64). C4 could compensate for lack of Downloaded from Ϫ Ϫ pression. We crossed CD21/35Ϫ/Ϫ and Prnpo/o mice, and per- C3 binding on FDCs of CD21/35hypo mice. Indeed, C3/4 / and formed reciprocal BM reconstitutions with Prnpo/oCD21/35Ϫ/Ϫ, CD21/35Ϫ/Ϫ mice exhibited similar delays in prion pathogenesis Prnpo/o, CD21/35Ϫ/Ϫ, or WT mice. We then inoculated these BM when inoculated with high doses of prions. However, when inoc- Ϫ/Ϫ chimeras i.p. with 6 log LD50 of RML prions and analyzed their ulated with a lower prion dose, CD21/35 mice exhibited de- spleens for PrPSc. Because mice lacking PrPC cannot replicate pri- creased incidence and longer delay in disease progression than ons, any PrPSc detected in their spleen must originate from the orig- C3/4Ϫ/Ϫ mice. Thus, CD21/35 appear capable of enhancing prion http://www.jimmunol.org/ inal inocula. At 47 dpi, we detected no PrPSc in spleens from mice disease independent of known endogenous CD21/35 ligands, most Ϫ Ϫ lacking both PrPC and CD21/35 (Prnpo/oCD21/35 / 3Prnpo/o likely by enhancing prion retention. Ϫ Ϫ CD21/35 / ). Spleens lacking PrPC expression retained PrPSc We found substantial prion titers at very early time points in Ϫ Ϫ if CD21/35 was present (Prnpo/o3Prnpo/oCD21/35 / and Prnpo/o spleens from WT mice inoculated i.p. with prions, but no detect- Ϫ Ϫ CD21/35 / 3Prnpo/o mice), especially on FDCs (Fig. 7, C and able prion titers in spleens of mice lacking CD21/35. PrPSc cor- D). We detected PrPSc in spleens with B cells expressing only related with CD21/35 and PrPC expression in infectious splenic Ϫ Ϫ CD21/35 (Prnpo/o3Prnpo/oCD21/35 / mice), but not in spleens follicles, presumably on FDCs, which have been convincingly with B cells expressing only PrPC (CD21/35Ϫ/Ϫ3Prnpo/oCD21/ shown to mediate peripheral prion pathogenesis (28–32). We hy- by guest on September 27, 2021 Ϫ/Ϫ Sc 35 mice). This suggests not only that B cells retain PrP , but pothesized that FDCs require CD21/35 expression to maximize also that they do so more efficiently when they express CD21/35 prion accumulation and replication. However, FDC expression of C Sc rather than PrP . We detected significantly more PrP from CD21/35 has been difficult to evaluate because of their intimate C spleens with FDCs expressing CD21/35, PrP , or both (Fig. 7D). contact with germinal center B cells, which have confirmed Spleens incapable of prion replication that expressed CD21/35 on CD21/35 expression. We confirmed substantial CD21/35 expres- o/o Ϫ/Ϫ3 o/o Sc FDCs (Prnp CD21/35 Prnp mice) retained PrP as ef- sion on FDCs by IHC of splenic follicles from BM chimeric mice ficiently as replication-competent spleens that lacked CD21/35 lacking CD21/35 expression on B cells. These data corroborate o/o Ϫ/Ϫ3 Ϫ/Ϫ (Prnp CD21/35 CD21/35 mice). In the latter sample, FACS data from a recent report documenting CD21/35 expression however, we cannot exclude the possibility that prion replication, on purified FDCs from lethally irradiated WT mice (65). C in addition to retention, by PrP -expressing FDCs contributed to On B cells, CD21/35 function as Ag presentation molecules and Sc the PrP content that we detected. Samples from spleens express- as part of an important B cell coreceptor complex with CD19 and C o/o Ϫ/Ϫ3 ing both CD21/35 and PrP on FDCs (Prnp CD21/35 WT CD81 that helps activate B cells. On FDCs, they are believed to act and WT3WT mice) did not require NaPTA precipitation for de- Sc solely as Ag-presentation molecules. To determine its importance tection, strongly implicating FDCs as the primary source of PrP on each cell type during a prion infection, we restricted CD21/35 accumulation and replication. expression to either B cells or FDCs by reconstituting the hemo- poietic system of lethally irradiated CD21/35Ϫ/Ϫ or WT mice with Discussion BM from WT or CD21/35Ϫ/Ϫ mice, respectively. FDC expression Sc We investigated the cellular and molecular mechanisms underly- of CD21/35 was required for PrP accumulation in splenic folli- ing the involvement of FDCs, B cells, the CD21/35 receptors and cles early in infection, and resulted in the most expedient terminal their ligands in prion accumulation, replication, and disease pro- prion disease that we observed. From these and previous data, we gression. Simultaneous elimination of C3 and C4 extended the expected a more prominent role for CD21/35 on FDCs, which is delay to terminal prion disease observed when only C3 was elim- thought to use CD21/35 mainly to trap and display Ag because inated, suggesting a role for both CD21/35 ligand sources in prion they lack signaling components of the B cell coreceptor. More propagation upon i.p. inoculation. surprisingly, eliminating CD21/35 expression on B cells resulted Although mice deficient in both CD21/35 ligand sources exhibit in a significant delay in disease progression when compared with a more significant delay than C3Ϫ/Ϫ mice (46), they still suc- WT mice. Moreover, B cell-restricted CD21/35 expression abro- cumbed to disease significantly earlier than CD21/35Ϫ/Ϫ mice, a gated the decreased incidence of terminal disease that we observed significant number of which did not contract disease at all. These at low-dose inoculation of prions for mice completely lacking data contrast with data previously reported using mice which had CD21/35. These data point to a previously undiscovered function been erroneously reported to lack CD21/35 expression (61). In in prion pathogenesis for this receptor on B cells, the primary role The Journal of Immunology 6151 of which was thought to be to supply lymphotoxins to FDCs be- may bind prions directly to mediate accumulation, replication, and cause PrPC expression occurs on only a small subset of B cells (39) neuroinvasion. Experiments designed to test this idea may uncover and is dispensable to promote disease (43). Circulating B cells may the molecular basis for extraneural prion infections. also function as APCs, actively trapping and transporting PrPSc from peripheral sites to lymphoid follicles, a function proposed for Acknowledgments other circulating immune cells such as dendritic cells (66, 67) and We thank Denis Marino for help with the BM reconstitutions and Ed macrophages (45, 68–70). Hoover and Anne Avery for critical reading of the manuscript. Infectivity assays revealed no detectable prion titers in CD21/ 35Ϫ/Ϫ mice even at 15 dpi, suggesting that CD21/35 facilitates Disclosures prion retention on cells that express it. Using additional BM chi- The authors have no financial conflict of interest. meric mice with segregated CD21/35 and PrPC expression on B cells or FDCs, we investigated PrPSc retention in more detail. At References 20 dpi, we discovered that the most efficient PrPSc trapping re- 1. Baker, H. F., and R. M. Ridley. 1996. What went wrong in Bse–from prion quired CD21/35 expression on both B cells and FDCs. PrPSc de- disease to public disaster. Brain Res. Bull. 40: 237–244. 2. Will, R. G., J. W. Ironside, M. Zeidler, S. N. Cousens, K. Estibeiro, tection in spleens lacking CD21/35 on either cell type required A. Alperovitch, S. Poser, M. Pocchiari, A. Hofman, and P. G. Smith. 1996. A new concentrating PrPSc from 100-fold more splenic tissue. We further variant of Creutzfeldt-Jakob disease in the U.K. Lancet 347: 921–925. asked whether the PrPSc that we detected at this very early time 3. Bruce, M. E., R. G. Will, J. W. Ironside, I. McConnell, D. Drummond, A. Suttie, L. McCardle, A. Chree, J. Hope, C. Birkett, et al. 1997. Transmissions to mice Sc point was from residual inocula or nascent PrP . We found severe indicate that “new variant” CJD is caused by the BSE agent. Nature 389: impairment of prion retention in the absence of CD21/35 in 498–501. Downloaded from 4. Will, R. G. 1999. The transmission of prions to humans. Acta Paediatr. Suppl. 88: spleens incapable of prion replication at 47 dpi. Retention im- 28–32. proved dramatically when CD21/35 expression was restored, es- 5. Scott, M. R., R. Will, J. Ironside, H. O. Nguyen, P. Tremblay, S. J. DeArmond, pecially on FDCs. Surprisingly, we even detected inocula in rep- and S. B. Prusiner. 1999. Compelling transgenetic evidence for transmission of bovine spongiform encephalopathy prions to humans. Proc. Natl. Acad. Sci. USA lication-deficient spleens expressing CD21/35 on B cells, but not 96: 15137–15142. in spleens expressing PrPC on B cells but lacking CD21/35. We 6. Gajdusek, D. C., C. J. Gibbs, and M. Alpers. 1966. Experimental transmission of Sc a -like syndrome to chimpanzees. Nature 209: 794–796. conclude that CD21/35 retains PrP more efficiently than does http://www.jimmunol.org/ 7. Gajdusek, D. C., C. J. Gibbs, Jr., and M. Alpers. 1967. Transmission and passage C Sc PrP . Although CD21/35 improved PrP retention on both B cells of experimental “kuru” to chimpanzees. Science 155: 212–214. and FDCs, retention was much more efficient by CD21/35-ex- 8. Mould, D. L., A. M. Dawson, and J. C. Rennie. 1970. Very early replication of scrapie in lymphocytic tissue. Nature 228: 779–780. pressing FDCs. These data confirmed our histoblot experiments, 9. Gibbs, C. J., Jr., A. Joy, R. Heffner, M. Franko, M. Miyazaki, D. M. Asher, Sc where FDC expression of CD21/35 was required for PrP accu- J. E. Parisi, P. W. Brown, and D. C. Gajdusek. 1985. Clinical and pathological mulation. CD21/35 expression solely on B cells proved inadequate features and laboratory confirmation of Creutzfeldt-Jakob disease in a recipient of Sc pituitary-derived human growth hormone. N. Engl. J. Med. 313: 734–738. for accumulating detectable amounts of PrP . Clearly, CD21/35 10. Bosque, P. J., C. Ryou, G. Telling, D. Peretz, G. Legname, S. J. DeArmond, and affected prion accumulation such that, in their absence, replication S. B. Prusiner. 2002. Prions in skeletal muscle. Proc. Natl. Acad. Sci. USA 99: and neuroinvasion are impaired to the extent that 25% of infected 3812–3817. 11. Glatzel, M., E. Abela, M. Maissen, and A. Aguzzi. 2003. Extraneural pathologic by guest on September 27, 2021 mice never progress to terminal disease. We conclude that the lack prion protein in sporadic Creutzfeldt-Jakob disease. N. Engl. J. Med. 349: of CD21/35 sufficiently slows disease kinetics to allow a signifi- 1812–1820. 12. Angers, R. C., S. R. Browning, T. S. Seward, C. J. Sigurdson, M. W. Miller, cant number of mice to survive prion infection. E. A. Hoover, and G. C. Telling. 2006. Prions in skeletal muscles of with CD21/35 may be a receptor used by FDCs for trapping prions in chronic wasting disease. Science 311: 1117. lymphoid follicles. Increased retention of PrPSc by CD21/35 on 13. Clarke, M. C., and D. A. Haig. 1967. Presence of the transmissible agent of Sc scrapie in the serum of affected mice and rats. Vet. Rec. 80: 504. FDCs could induce a persistent state of PrP presentation to ad- 14. Hunter, N., J. Foster, A. Chong, S. McCutcheon, D. Parnham, S. Eaton, jacent B cells sufficient to cause an atypical germinal center re- C. MacKenzie, and F. Houston. 2002. Transmission of prion diseases by blood sponse previously reported (71). This aberrant germinal center en- transfusion. J. Gen. Virol. 83: 2897–2905. 15. Prinz, M., M. Heikenwalder, P. Schwarz, K. Takeda, S. Akira, and A. Aguzzi. vironment may prove to be the optimal prion bioreactor. 2003. Prion pathogenesis in the absence of Toll-like receptor signalling. EMBO Hypertrophic FDC dendrites bearing CD21/35 loaded with PrPSc Rep. 4: 195–199. C 16. Llewelyn, C. A., P. E. Hewitt, R. S. Knight, K. Amar, S. Cousens, J. Mackenzie, could present aggregates to PrP expressed on dendrites of the and R. G. Will. 2004. Possible transmission of variant Creutzfeldt-Jakob disease same cell or on neighboring cells. by . Lancet 363: 417–421. Although opsonization of prions by known CD21/35 ligands 17. Peden, A. H., M. W. Head, D. L. Ritchie, J. E. Bell, and J. W. Ironside. 2004. Preclinical vCJD after blood transfusion in a PRNP codon 129 heterozygous may mediate trapping, we conclude that it is not required. Whether patient. Lancet 364: 527–529. (an)other unidentified protein(s) can or must mediate CD21/35- 18. Seeger, H., M. Heikenwalder, N. Zeller, J. Kranich, P. Schwarz, A. Gaspert, enhanced prion disease has yet to be determined. The soluble com- B. Seifert, G. Miele, and A. Aguzzi. 2005. Coincident scrapie infection and ne- phritis lead to urinary prion excretion. Science 310: 324–326. plement protein C1q is one possible candidate. C1q deficiency 19. Wroe, S. J., S. Pal, D. Siddique, H. Hyare, R. Macfarlane, S. Joiner, dramatically impaired peripheral prion pathogenesis similar to J. M. Linehan, S. Brandner, J. D. Wadsworth, P. Hewitt, and J. Collinge. 2006. Clinical presentation and pre-mortem diagnosis of variant Creutzfeldt-Jakob dis- CD21/35 (46, 47) and in vitro studies have shown that C1q directly ease associated with blood transfusion: a case report. Lancet 368: 2061–2067. C binds specific conformers of PrP that are postulated to mimic 20. Mathiason, C. K., J. G. Powers, S. J. Dahmes, D. A. Osborn, K. V. Miller, PrPSc (48). More recently, C1q binding has also been shown to R. J. Warren, G. L. Mason, S. A. Hays, J. Hayes-Klug, D. M. Seelig, et al. 2006. C Infectious prions in the saliva and blood of deer with chronic wasting disease. depend on PrP density (72). C1q also depleted complement ac- Science 314: 133–136. tivation through the classical pathway and fixed C4 in a hemolytic 21. Miller, M. W., M. A. Wild, and E. S. Williams. 1998. Epidemiology of chronic assay. Given these data and ours, C1q appears more likely to me- wasting disease in captive Rocky Mountain elk. J. Wildl. Dis. 34: 532–538. Sc 22. Miller, M. W., and E. S. Williams. 2003. Prion disease: horizontal prion trans- diate a possible interaction between PrP and CD21/35 than C4. mission in mule deer. Nature 425: 35–36. C1q has been reported to be a ligand for human CD35 (73), al- 23. Woolhouse, M. E., L. Matthews, P. Coen, S. M. Stringer, J. D. Foster, and N. Hunter. 1999. Population dynamics of scrapie in a sheep flock. Philos. Trans. though this has yet to be confirmed in mice. Another possibility is R Soc. Lond. B Biol. Sci. 354: 751–756. mannan-binding lectin, a homolog of C1q that can also activate 24. Ryder, S., G. Dexter, S. Bellworthy, and S. Tongue. 2004. Demonstration of complement through the classical pathway. Mannan-binding lectin lateral transmission of scrapie between sheep kept under natural conditions using lymphoid tissue biopsy. Res. Vet. Sci. 76: 211–217. has been shown to bind human CD35 (74, 75), although direct 25. Kimberlin, R. H., and C. A. Walker. 1989. Pathogenesis of scrapie in mice after binding to PrPSc was not detected (72). Alternatively, CD21/35 intragastric infection. Virus Res. 12: 213–220. 6152 CD12/35 EXPEDITES PRION NEUROINVASION

26. Lasmezas, C., J. P. Deslys, O. Robain, R. Demairnay, K. T. Adjou, F. Lamoury, 51. Kalli, K. R., and D. T. Fearon. 1994. Binding of and C4b by the CR1-like J. Ironside, J. J. Hauw, and D. Dormont. 1996. BSE transmission to macaques. site in murine CR1. J. Immunol. 152: 2899–2903. Nature 381: 743–744. 52. Molina, H., T. Kinoshita, C. B. Webster, and V. M. Holers. 1994. Analysis of 27. Klein, M. A., R. Frigg, E. Flechsig, A. J. Raeber, U. Kalinke, H. Bluethmann, C3b/C3d binding sites and factor I cofactor regions within mouse complement F. Bootz, M. Suter, R. M. Zinkernagel, and A. Aguzzi. 1997. A crucial role for receptors 1 and 2. J. Immunol. 153: 789–795. B cells in neuroinvasive scrapie. Nature 390: 687–690. 53. Bu¨eler, H. R., M. Fischer, Y. Lang, H. Bluethmann, H. P. Lipp, S. J. DeArmond, 28. Glaysher, B. R., and N. A. Mabbott. 2007. Role of the GALT in scrapie agent S. B. Prusiner, M. Aguet, and C. Weissmann. 1992. Normal development and neuroinvasion from the intestine. J. Immunol. 178: 3757–3766. behaviour of mice lacking the neuronal cell-surface PrP protein. Nature 356: 29. Heikenwalder, M., N. Zeller, H. Seeger, M. Prinz, P. C. Klohn, P. Schwarz, 577–582. N. H. Ruddle, C. Weissmann, and A. Aguzzi. 2005. Chronic lymphocytic in- 54. Haas, K. M., M. Hasegawa, D. A. Steeber, J. C. Poe, M. D. Zabel, C. B. Bock, flammation specifies the organ tropism of prions. Science 307: 1107–1110. D. R. Karp, D. E. Briles, J. H. Weis, and T. F. Tedder. 2002. Complement 30. Brown, K. L., K. Stewart, D. L. Ritchie, N. A. Mabbott, A. Williams, H. Fraser, receptors CD21/35 link innate and protective immunity during Streptococcus W. I. Morrison, and M. E. Bruce. 1999. Scrapie replication in lymphoid tissues pneumoniae infection by regulating IgG3 antibody responses. Immunity 17: depends on prion protein-expressing follicular dendritic cells. Nat. Med.5: 713–723. 1308–1312. 55. Fischer, M., T. Ru¨licke, A. Raeber, A. Sailer, M. Moser, B. Oesch, S. Brandner, 31. Kitamoto, T., T. Muramoto, S. Mohri, K. Dohura, and J. Tateishi. 1991. Abnor- A. Aguzzi, and C. Weissmann. 1996. Prion protein (PrP) with amino-proximal mal isoform of prion protein accumulates in follicular dendritic cells in mice with deletions restoring susceptibility of PrP knockout mice to scrapie. EMBO J. 15: Creutzfeldt-Jakob disease. J. Virol. 65: 6292–6295. 1255–1264. 32. Montrasio, F., R. Frigg, M. Glatzel, M. A. Klein, F. Mackay, A. Aguzzi, and 56. Prusiner, S. B., S. P. Cochran, D. F. Groth, D. E. Downey, K. A. Bowman, and C. Weissmann. 2000. Impaired prion replication in spleens of mice lacking func- H. M. Martinez. 1982. Measurement of the scrapie agent using an incubation time tional follicular dendritic cells. Science 288: 1257–1259. interval assay. Ann. Neurol. 11: 353–358. 33. Kinet Denoel, C., E. Heinen, D. Radoux, and L. J. Simar. 1982. Follicular den- 57. Klohn, P. C., L. Stoltze, E. Flechsig, M. Enari, and C. Weissmann. 2003. A dritic cells in lymph nodes after x-irradiation. Int. J. Radiat. Biol. Relat. Stud. quantitative, highly sensitive cell-based infectivity assay for mouse scrapie pri- Phys. Chem. Med. 42: 121–130. ons. Proc. Natl. Acad. Sci. USA 100: 11666–11671. 34. Humphrey, J. H., D. Grennan, and V. Sundaram. 1984. The origin of follicular 58. Pham, N. A., A. Morrison, J. Schwock, S. Aviel-Ronen, V. Iakovlev, M. S. Tsao, dendritic cells in the mouse and the mechanism of trapping of immune complexes J. Ho, and D. W. Hedley. 2007. Quantitative image analysis of immunohisto- Downloaded from on them. Eur. J. Immunol. 14: 859–864. chemical stains using a CMYK color model. Diagn. Pathol. 2: 8–17. 35. Imazeki, N., A. Senoo, and Y. Fuse. 1992. Is the follicular dendritic cell a pri- 59. Taraboulos, A., K. Jendroska, D. Serban, S. L. Yang, S. J. DeArmond, and marily stationary cell? Immunology 76: 508–510. S. B. Prusiner. 1992. Regional mapping of prion proteins in brain. Proc. Natl. 36. Kapasi, Z. F., D. Qin, W. G. Kerr, M. H. Kosco-Vilbois, L. D. Shultz, J. G. Tew, Acad. Sci. USA 89: 7620–7624. and A. K. Szakal. 1998. Follicular dendritic cell (FDC) precursors in primary 60. Prinz, M., M. Heikenwalder, T. Junt, P. Schwarz, M. Glatzel, F. L. Heppner, lymphoid tissues. J. Immunol. 160: 1078–1084. Y. X. Fu, M. Lipp, and A. Aguzzi. 2003. Positioning of follicular dendritic cells 37. Park, C. S., and Y. S. Choi. 2005. How do follicular dendritic cells interact within the spleen controls prion neuroinvasion. Nature 425: 957–962.

intimately with B cells in the germinal centre? Immunology 114: 2–10. 61. Ahearn, J. M., M. B. Fischer, D. Croix, S. Goerg, M. Ma, J. Xia, X. Zhou, http://www.jimmunol.org/ 38. Montrasio, F., A. Cozzio, E. Flechsig, D. Rossi, M. A. Klein, T. Rulicke, R. G. Howard, T. L. Rothstein, and M. C. Carroll. 1996. Disruption of the Cr2 A. J. Raeber, C. A. Vosshenrich, J. Proft, A. Aguzzi, and C. Weissmann. 2001. results in a reduction in B-1a cells and in an impaired B cell response to B -restricted expression of prion protein does not enable prion rep- T-dependent . Immunity 4: 251–262. lication in prion protein knockout mice. Proc. Natl. Acad. Sci. USA 98: 62. Hasegawa, M., M. Fujimoto, J. C. Poe, D. A. Steeber, and T. F. Tedder. 2001. 4034–4037. CD19 can regulate B lymphocyte independent of complement 39. Liu, T., R. Li, B. S. Wong, D. Liu, T. Pan, R. B. Petersen, P. Gambetti, and activation. J. Immunol. 167: 3190–3200. M. S. Sy. 2001. Normal cellular prior protein is preferentially expressed on sub- 63. Hasty, P., and A. Bradley. 2000. targeting vectors for mammalian cells. In populations of murine hemopoietic cells. J. Immunol. 166: 3733–3742. Gene Targeting: A Practical Approach, Vol. 1. A. Joyner, ed. Oxford University 40. Le Hir, M., H. Bluethmann, M. H. Kosco-Vilbois, M. Muller, F. di Padova, Press, Oxford, p. 1. M. Moore, B. Ryffel, and H. P. Eugster. 1996. Differentiation of follicular den- 64. Guinamard, R., M. Okigaki, J. Schlessinger, and J. V. Ravetch. 2000. Absence of dritic cells and full antibody responses require tumor necrosis factor receptor-1 marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral

signaling. J. Exp. Med. 183: 2367–2372. response. Nat. Immunol. 1: 31–36. by guest on September 27, 2021 41. Matsumoto, M., Y. X. Fu, H. Molina, G. Huang, J. Kim, D. A. Thomas, 65. Sukumar, S., A. K. Szakal, and J. G. Tew. 2006. Isolation of functionally active M. H. Nahm, and D. D. Chaplin. 1997. Distinct roles of lymphotoxin ␣ and the murine follicular dendritic cells. J. Immunol. Methods 313: 81–95. type I tumor necrosis factor (TNF) receptor in the establishment of follicular 66. Aucouturier, P., F. Geissmann, D. Damotte, G. P. Saborio, H. C. Meeker, dendritic cells from non-bone marrow-derived cells. J. Exp. Med. 186: R. Kascsak, R. I. Carp, and T. Wisniewski. 2001. Infected splenic dendritic cells 1997–2004. are sufficient for prion transmission to the CNS in mouse scrapie. J. Clin. Invest. 42. Fu, Y. X., G. Huang, Y. Wang, and D. D. Chaplin. 1998. B lymphocytes induce 108: 703–708. the formation of follicular dendritic cell clusters in a lymphotoxin ␣-dependent 67. Huang, F. P., C. F. Farquhar, N. A. Mabbott, M. E. Bruce, and G. G. MacPherson. fashion. J. Exp. Med. 187: 1009–1018. 2002. Migrating intestinal dendritic cells transport PrP(Sc) from the gut. J. Gen. 43. Klein, M. A., R. Frigg, A. J. Raeber, E. Flechsig, I. Hegyi, R. M. Zinkernagel, Virol. 83: 267–271. C. Weissmann, and A. Aguzzi. 1998. PrP expression in B lymphocytes is not 68. Dealler, S. F., and R. W. Lacey. 1994. Suspected vertical transmission of BSE required for prion neuroinvasion. Nat. Med. 4: 1429–1433. (letter). Vet. Rec. 134: 151. 44. Mabbott, N. A., A. Williams, C. F. Farquhar, M. Pasparakis, G. Kollias, and 69. Beringue, V., M. Demoy, C. I. Lasmezas, B. Gouritin, C. Weingarten, M. E. Bruce. 2000. Tumor necrosis factor ␣-deficient, but not interleukin-6- J. P. Deslys, J. P. Andreux, P. Couvreur, and D. Dormont. 2000. Role of spleen deficient, mice resist peripheral infection with scrapie. J. Virol. 74: 3338–3344. macrophages in the clearance of scrapie agent early in pathogenesis. J. Pathol. 45. Prinz, M., F. Montrasio, M. A. Klein, P. Schwarz, J. Priller, B. Odermatt, 190: 495–502. K. Pfeffer, and A. Aguzzi. 2002. Lymph nodal prion replication and neuroinva- 70. Aguzzi, A., F. L. Heppner, M. Heikenwalder, M. Prinz, K. Mertz, H. Seeger, and sion in mice devoid of follicular dendritic cells. Proc. Natl. Acad. Sci. USA 99: M. Glatzel. 2003. Immune system and peripheral nerves in propagation of prions 919–924. to CNS. Br. Med. Bull. 66: 141–159. 46. Klein, M. A., P. S. Kaeser, P. Schwarz, H. Weyd, I. Xenarios, R. M. Zinkernagel, 71. McGovern, G., K. L. Brown, M. E. Bruce, and M. Jeffrey. 2004. Murine scrapie M. C. Carroll, J. S. Verbeek, M. Botto, M. J. Walport, et al. 2001. Complement infection causes an abnormal germinal centre reaction in the spleen. J. Comp. facilitates early prion pathogenesis. Nat. Med. 7: 488–492. Pathol. 130: 181–194. 47. Mabbott, N. A., M. E. Bruce, M. Botto, M. J. Walport, and M. B. Pepys. 2001. 72. Mitchell, D. A., L. Kirby, S. M. Paulin, C. L. Villiers, and R. B. Sim. 2007. Prion Temporary depletion of complement component C3 or genetic deficiency of C1q protein activates and fixes complement directly via the classical pathway: impli- significantly delays onset of scrapie. Nat. Med. 7: 485–487. cations for the mechanism of scrapie agent propagation in lymphoid tissue. Mol. 48. Blanquet-Grossard, F., N. M. Thielens, C. Vendrely, M. Jamin, and G. J. Arlaud. Immunol. 44: 2997–3004. 2005. Complement protein C1q recognizes a conformationally modified form of 73. Klickstein, L. B., S. F. Barbashov, T. Liu, R. M. Jack, and A. Nicholson-Weller. the prion protein. Biochemistry 44: 4349–4356. 1997. Complement receptor type 1 (CR1, CD35) is a receptor for C1q. Immunity 49. Sim, R. B., U. Kishore, C. L. Villiers, P. N. Marche, and D. A. Mitchell. 2007. 7: 345–355. C1q binding and complement activation by prions and amyloids. Immunobiology 74. Ikeda, K., T. Sannoh, N. Kawasaki, T. Kawasaki, and I. Yamashina. 1987. Serum 212: 355–362. lectin with known structure activates complement through the classical pathway. 50. Kinoshita, T., T. Fujita, and R. Tsunoda. 1991. Expression of complement re- J. Biol. Chem. 262: 7451–7454. ceptors CR1 and CR2 on murine follicular dendritic cells and B lymphocytes. In 75. Ghiran, I., S. F. Barbashov, L. B. Klickstein, S. W. Tas, J. C. Jensenius, and Dendritic Cells in Lymphoid Tissues. Y. Imai, J. G. Tew, and E. C. M. Hoefsmit, A. Nicholson-Weller. 2000. /CD35 is a receptor for man- eds. Elsevier, Amsterdam, p. 271. nan-binding lectin. J. Exp. Med. 192: 1797–1808.