Accelerated Prion Replication in, but Prolonged Survival Times of, Prion-Infected CXCR3^−/− Mice^▿

Animals and scrapie infections.

Generation of CXCR3^−/− mice has been described previously (28). CXCR3^−/− mice and C57BL/6 wild-type controls, which were outbred from CXCR^+/− crosses, were kept in the local animal facilities. Five- to 6-week-old mice were intracerebrally (i.c.) inoculated with 20 μl of a 10⁻³- or 10⁻⁴-diluted 10% brain homogenate prepared from a terminally diseased C57BL/6 wild-type mouse infected with the scrapie strain 139A (courtesy of R. H. Kimberlin, Edinburgh, United Kingdom) as previously described (61). Mock infections were performed using similarly diluted brain homogenates obtained from uninfected, healthy wild-type mice. Infected animals were monitored thrice weekly for clinical signs until the symptomatic stage was reached (9) and daily once disease onset was diagnosed. Mice were sacrificed at 125 days postinfection (dpi) or at the terminal stage of disease, at which animals would naturally succumb to the disease within the next 48 h. Analysis and comparisons at the terminal stage were performed for mice, which were sacrificed on the same day or less than 5 days apart. The animal experiments and care protocols were approved by the institutional review committee Landesamt für Gesundheit und Soziales (Berlin, Germany). Survival times in all groups were statistically analyzed using the unpaired t test and the log rank test.

To determine prion titers an end point titration was carried out in which serial dilutions of a brain homogenate prepared from a terminally ill wild-type mouse were inoculated i.c. into tga20 mice (n = 4 per dilution) (22). The 50% lethal doses (LD₅₀s) were calculated according to the Spearman-Kaerber method and plotted against survival times. The resulting curve could be described by the equation y = −0.361x³ + 5.057x² − 28.67x + 123.83, where y is the survival time (in days postinfection) and x the log LD₅₀ (R² = 0.9979). Next, tga20 mice (n = 4 per sample) were i.c. infected with 20 μl of 10⁻²-diluted 10% brain homogenates prepared from CXCR3^−/− or wild-type control mice sacrificed at 125 dpi or at the terminal stage of the disease. Prion titers (LD₅₀/ml of 10% brain homogenate) in these samples were then calculated using the survival time/LD₅₀ relationship described above (52).

Histology and immunohistochemistry.

Immunohistochemistry was performed according to previously published procedures (31, 61). Serial sagittal paraffin sections (6 μm) for a minimum of three samples/group and time point were examined histologically by hematoxylin and eosin staining for spongiform changes (data not shown). For detection of activated astrocytes, sections were stained for glial fibrillary acidic protein (GFAP; 1:1,000; Dako, Glostrup, Denmark). Serial sagittal cryo sections (8 μm) were examined for microglia activation with an anti-Mac-1 (CD11b) monoclonal antibody (1:200; kindly provided by B. Engelhardt, Munster, Germany). The total microglia population was stained with an antibody against ionized calcium binding adapter molecule 1 (Iba-1; 1:100; Dako, Glostrup, Denmark). Stainings were visualized using the ABC method and diaminobenzidine or 3-amino-9-ethylcarbazol as chromogens. Differences between CXCR3^−/− and wild-type mice were evaluated by independent scoring of tissue samples by three investigators without previous knowledge of the group to which the mice belonged.

Paraffin-embeded tissue (PET) blot analysis.

Paraffin sections of formalin-fixed brain tissues were transferred onto a nitrocellulose membrane and stained as previously described (62). The immunodetection of PrP^Sc was performed with the anti-PrP antibody 6H4 (1:10,000; Prionics, Zürich, Switzerland) followed by incubation with an alkaline phosphatase-linked anti-mouse immunoglobulin antiserum (1:2,000; Dako, Glostrup, Denmark). The final staining was performed with nitro blue tetrazolium-5-bromo-4-chloro-3-indolyl phosphate.

Western blot analysis.

For Western blot analysis of proteinase K (PK)-resistant PrP^Sc, 10% brain homogenates were made in homogenization puffer (phosphate-buffered saline, 0.5% Triton X-100, and 0.05% sodium dodecyl sulfate) and subjected to PK digestion at a final concentration of 100 μg/ml for 1 h at 37°C (1). After sodium dodecyl sulfate-polyacrylamide gel electrophoresis and protein transfer, PrP^Sc bands were detected using the 6H4 antibody (1:5,000; Prionics, Schlieren-Zurich, Switzerland). For Western blot analysis of differences in GFAP expression, anti-GFAP antibody (1:5,000; Dako, Glostrup, Denmark) was used. To control for amounts of loaded protein, membranes were stripped by incubation in 0.2 M glycine-HCl (pH 2.5), 0.05% Tween 20 for 30 min at room temperature and reprobed with anti-β-actin antibody (1:5,000; Sigma, Hamburg, Germany). Blots were quantified and analyzed using Quantity One software following the manufacturer's instructions (Bio-Rad, Munich, Germany). Immunoprecipitations with the PrP^Sc-specific antibody 15B3 (kindly provided by A. Raeber, Prionics) prior to Western blot analysis were performed as previously described (47).

Quantitative real-time PCR.

Brains from scrapie-infected CXCR3^−/− and wild-type mice as well as from mock-infected and uninfected CXCR3^−/− and wild-type mice were removed 125 dpi and at the terminal stage of disease, frozen in liquid nitrogen, and stored at −80°C. Total RNA from whole brains was prepared with Trizol reagent (Invitrogen, Karlsruhe, Germany) according to the manufacturer's protocol. After digestion with DNase I for 45 min at 37°C, total RNA was purified with the RNeasy protect mini kit (Qiagen, Hilden, Germany) and 1.25 μg was reverse transcribed using the First-Strand cDNA synthesis kit (Amersham, Freiburg, Germany). Gene expression levels from each group of mice were subsequently determined by real-time PCR by employing a GeneAmp 5700 sequence detection system (Perkin-Elmer, Boston, MA) and the Sybr green PCR kit (Qiagen, Hilden Germany), following data analysis using the ΔΔC_t method. To compensate for variations in amounts of input RNA and efficiencies of reverse transcription, an endogenous housekeeping gene (glyceraldehyde-3-phosphate dehydrogenase) was quantified and results were normalized to these values. Quantification of signal was achieved by setting thresholds within the logarithmic phase of PCR and determined by the cycle number at which this threshold was reached (C_t). The C_t obtained for glyceraldehyde-3-phosphate dehydrogenase was subtracted from the C_t of the target gene to yield ΔC_t. The increase of target gene expression level was calculated according to the formula 2^{(ΔC_t1 − ΔC_t2)}, where ΔC_t1 is the normalized test sample value and ΔC_t2 the normalized mock-infected control sample value.

Primers.

The following primers were used: for GFAP (primer A, GCG GGA GTC GGC CAG TTA CC; primer B, GAC CTC ACC ATC CC GCA TCT); 2′,5′-oligoadenylate synthetase (OAS) (primer A, GGT CTC TGA GCT TCA AGC TGA G; primer B, TAC TGT GGA GGC AAT GGC TTC AA); Spi-2 (primer A, ATC TGC CCT GCT GTC CTC TG; primer B, GCG CTG GCA TTT CCT GTG TA); CXCL10 (primer A, GCA ACT GCA TCC ATA TCG ATG AC; primer B, TGT GCG TGG CTT CAC TCC A); CCL2 (primer A, CCC CAC TCA CCT GCT GCT AC; primer B, ACG GGT CAA CTT CAC ATT CA); CCL3 (primer A, ACT GCC CTT GCT GTT CTT CT; primer B, CTG CCG GTT TCT CTT AGT CA); CCL5 (primer A, TGC CCA CGT CAA GGA GTA TT; primer B, CAG GAC CGG AGT GGG AGT A).

Accelerated accumulation of PrP^Sc and infectivity but prolonged survival times in the absence of CXCR3.

To assess the role of CXCR3 in a prion disease of the CNS, intracerebral scrapie infection of mice deficient for CXCR3 was compared to that in similarly infected wild-type mice. CXCR3^−/− mice developed typical scrapie symptoms, including a characteristic reduction in mobility with progressive ataxia, progressive proprioceptive deficits, poor coat condition, and weight loss (61). The duration of this clinically overt symptomatic stage was similar to that in wild-type controls, but the onset of this stage was delayed (data not shown). CXCR3^−/− mice survived prion infections on average 20 to 30 days longer than control mice (P < 0.001) (Fig. 1).

The characteristic accumulation of PK-resistant PrP^Sc in scrapie infections of infected CXCR3^−/− and control mice was analyzed by Western and PET blotting. For both methods, CXCR3^−/− mice showed strongly increased PrP^Sc accumulation at 125 dpi, which appeared, however, to be equivalent at the terminal stage (Fig. 2 and 3). To assess PrP^Sc deposition and distribution, we examined scrapie-infected brains by PET blot analysis. Deposition started in the medulla oblongata and thalamus, followed by the hippocampus and cortex. Finally, typically diffuse PrP^Sc accumulations appeared in all brain areas, but staining was most intense in the cortex, hippocampus, and thalamus (Fig. 3). Overall, principal distribution and spread of PrP^Sc deposition was similar for both groups. To address the question of whether the more-rapid accumulation of PrP^Sc in receptor-deficient mice correlates with accumulation of prion infectivity, dilutions of 10% brain homogenates obtained from scrapie-infected CXCR3^−/− and control mice were inoculated into tga20 indicator mice. tga20 mice receiving brain homogenates from control animals obtained at 125 dpi showed a mean survival time of 117 days. In contrast, tga20 mice exposed to brain homogenates from CXCR3^−/− mice from the same stage (125 dpi) survived on average only 89 days (Table 1). This difference of 28 days corresponds to an at least 20 times higher titer of prion infectivity in the CXCR3^−/− mice at 125 dpi and is in agreement with the increased PrP^Sc accumulation at this time point (Fig. 2 and 3; Table 1). tga20 mice infected with brain homogenate from terminally ill knockout and control mice showed no significant difference in survival times (Table 1). tga20 mice inoculated with brain material from healthy mock-infected mice never developed disease.

Massive astrocytosis but reduced microglia activation in CXCR3^−/− mice.

To evaluate differences in the astrocytosis between CXCR3-deficient and wild-type mice, GFAP expression was determined by immunohistochemistry and Western blot analysis. CXCR3^−/− mice showed two- to fourfold-increased GFAP expression in immunohistochemistry at the terminal stage of disease (Fig. 4A to F and I). At the asymptomatic stage (125 dpi), GFAP expression was found to be similar for CXCR3^−/− and wild-type animals in the cerebellum, hippocampus, and cortex. However, in close correlation with brain areas of early PrP^Sc deposition, activation of astrocytes in CXCR3^−/− mice was more pronounced in the medulla oblongata and thalamus (data not shown). Elevated GFAP expression in CXCR3^−/− mice was confirmed by Western blot analysis, which showed, in agreement with the immunohistochemistry results, clear differences at the early time point and a sixfold augmentation at the terminal stage of the disease compared to the wild-type controls (Fig. 4G and H).

At the asymptomatic stage Mac-1-positive activated microglia were found in the midbrain, hippocampus, and striatum of infected control mice. CXCR3-deficient mice showed comparable Mac-1 staining patterns in the midbrain and striatum, whereas in the hippocampus numbers of Mac-1-expressing cells were clearly reduced in comparison to wild-type mice (data not shown). At the terminal stage controls showed up to threefold-increased Mac-1-positive cells in the whole brain, particularly in the hippocampus, striatum, and cortex, with strongest staining in the cerebellum and thalamus (Fig. 5).

Another characteristic alteration in scrapie infection is the spongiform vacuolization of brain tissue. To confirm microvacuolization, brain sections of infected animals were subjected to hematoxylin and eosin staining. Interestingly, although there were obvious differences in survival time, PrP^Sc accumulation, astrocytosis, and microgliosis, no differences in the amount or distribution of vacuolization were found between CXCR3^−/− and wild-type mice at any time point (data not shown).

For further analysis of disease-associated gliosis and its consequences, quantitative real-time PCR was performed to determine (inflammatory) marker gene mRNA levels in brains from prion-infected CXCR3^−/− and wild-type mice as well as in mock-infected controls (Fig. 6).

As expected, the strongly enhanced astrogliosis in CXCR3^−/− mice was also reflected in 30- to 140-fold-elevated GFAP transcript levels. However, expression levels of all other markers analyzed were found to be reduced in brain tissue of CXCR3^−/− mice at the asymptomatic stage (CXCL10, CCL5, and OAS) or even at both time points studied (CCL2, CCL3, and serine protease inhibitor 2 [Spi-2]), indicating a weaker inflammatory response in these animals.

Expression levels of these marker genes as well as prion protein and GFAP expression levels were similar for uninfected wild-type and CXCR3^−/− mice (data not shown), demonstrating that ablation of CXCR3 per se does not cause a gliosis or a general imbalance of brain homeostasis, which is in agreement with a previous study with CXCR3^−/− mice (55).