Absence of CD14 Delays Progression of Prion Diseases Accompanied by Increased Microglial Activation

Antibodies.

Anti-mouse prion protein (PrP) monoclonal antibodies (MAbs) 31C6 and 132 were prepared as described previously (18). Anti-Iba1 rabbit polyclonal antibodies were purchased from Wako (product no. 019-20001). Anti-glial fibrillary acidic protein (anti-GFAP) rabbit polyclonal antibodies were from Dako (product no. Z033401). Unconjugated and Alexa Fluor 488-conjugated anti-mouse CD11b rat MAb (clone M1/70), anti-mouse CD14 rat MAb (clone Sal 14-2), and anti-mouse F4/80 rat MAb (clone BM8) were from BioLegend. Anti-NeuN mouse MAb (clone A60) was from Millipore. Anti-mouse CD45 rat MAb (clone 13/2.3) was from Funakoshi. Anti-GFAP mouse MAb (clone GF5), anti-IL-10 rat MAb (clone JES5-2A5), anti-mouse CD68 rat MAb (clone FA-11), anti-TGF-β rabbit polyclonal antibodies (product no. 66043), and anti-IL-1β rabbit polyclonal antibodies (product no. 9722) were from Abcam. ECL horseradish peroxidase-labeled anti-mouse IgG was from Amersham Biosciences (product no. NA9310V). All Alexa Fluor-labeled antibodies were from Life Technologies.

Mice and prion inoculation.

C57BL/6J mice were purchased from Japan Clea Inc. CD14 knockout (CD14^−/−) mice of congenic strain B6.129S-Cd14^<tm1Frm>/J were purchased from Jackson Laboratories and were further maintained by inbreeding. All animal procedures were approved by the Institutional Animal Care and Use Committee of the Graduate School of Veterinary Medicine, Hokkaido University. Six-week-old female WT and CD14^−/− mice were inoculated intracerebrally with 20 μl of 2.5% brain homogenates of the Chandler- or Obihiro-infected C57BL/6J mice. The inoculation was carried out whenever the 6-week-old female CD14^−/− mice were available. The exact same aliquots of 2.5% brain homogenates were used for each inoculation. At the clinical stage, mice were observed every day, and the clinical endpoint of disease was defined as recumbency with severe emaciation within two consecutive days.

Quantitative RT-PCR.

Total RNA was extracted from the thalamus cut at a thickness of the coronal section with 5 mm around the level of bregma −1.82 mm (as indicated in The Mouse Brain in Stereotaxic Coordinates, 2nd ed. [19]) using TRIzol reagent (Life Technologies). First-strand cDNA was synthesized from 2 μg of the total RNA using a First-Strand cDNA synthesis kit (Amersham Biosciences) according to the manufacturer's instructions. The amplification reaction mixtures contained template cDNA, 1× predesigned TaqMan Gene Expression Assays for mouse Cd14 (Mm00438094_g1) and for mouse ACTB (catalog no. 4352933E), and 1× TaqMan Fast Universal PCR Master Mix. TaqMan assays were carried out using an ABI 7900HT Fast Real-Time PCR system (Applied Biosystems). The amplification profiles were analyzed using a threshold cycle relative quantification method and were normalized with the expression of mouse ACTB gene as described previously (20).

Bioassay.

Brain homogenates from 2 mice at each time point were prepared in sterile phosphate-buffered saline (PBS) to 1% (wt/vol). Twenty microliters of brain homogenates were inoculated intracerebrally into 6-week-old Tga20 mice. To obtain an infectivity-incubation time standard curve, 10-fold serial diluted brain homogenates used as the inocula of the Chandler and Obihiro strains were also injected into Tga 20 mice. The 50% lethal doses (LD₅₀) of the original 10% homogenates were estimated to be 10^9.3 and 10^9.5 LD₅₀/g brain tissue for the Chandler and Obihiro strains, respectively. In the Chandler infection, the standard curve for the incubation periods (χ) was fitted by the approximations of concentrations of original brain homogenates (y), determined as follows: y = e^{23.026−0.425χ} for the periods of up to 68 days, and y = e^{16.118−0.332χ} for the periods after 69 days. In the Obihiro infection, the approximation was y = e^{20.723−0.263χ} for the periods of up to107 days and y = e^{7.092−0.129χ} for the periods after 108 days.

PrP-res detection by immunoblotting.

For protease-resistant prion protein (PrP-res) detection, at 60, 90, and 120 dpi and at the terminal stage of the disease, brains were harvested and were homogenized in sterile PBS to prepare 10% (wt/vol) brain homogenates. The homogenates were treated with proteinase K (PK) and were subjected to SDS-PAGE and immunoblotting as described previously (20).

PrP^Sc-specific immunofluorescence staining.

The PrP^Sc-specific staining of frozen sections was performed by a method described previously (21) with some modifications. Briefly, the brains were embedded in OCT compound (Sakura Finetek, Japan) and were cut at a thickness of 10 μm. The samples were fixed with 4% paraformaldehyde (PFA) for 10 min at room temperature (rt). The slides were treated with 0.1% Triton X-100 for 10 min at rt and then with 5 M guanidine thiocyanate for 10 min at rt. After washing with PBS, the sections were blocked with 5% fetal bovine serum (FBS) in PBS at 37°C for 30 min. The sections were then incubated with MAb132 at 3 μg/ml at 37°C for 45 min. For the secondary antibody reaction, the sections were incubated with Alexa Fluor 488-labeled anti-mouse IgG (Fab)′₂ fragment of goat IgG at rt for 1 h. The samples were counterstained with 4′6-diamidino-2-phenylindole (DAPI), mounted with ProLong Gold antifade reagent (Invitrogen), and then observed with a laser scanning confocal microscope (LSM) (Carl Zeiss International).

Histopathological analysis.

The histopathological analysis was carried out as described previously (22). Briefly, brains were harvested at 90 and 120 dpi and at the terminal stage of the disease and fixed with 10% phosphate-buffered formalin. The brains were embedded in paraffin and cut at the level of bregma −1.82 mm. The samples were stained with hematoxylin and eosin (H&E).

Immunohistochemistry.

For the antigen retrieval of Iba1 detection in paraffin-embedded tissue, the slides were treated with microwaves at 600 W for 10 min in 0.01 M citrate buffer solution, pH 6.0. After washing with PBS, the slides were blocked with normal goat serum (Nichirei) and incubated overnight with primary antibody diluted at 1:1,000 in normal goat serum at 4°C. The slides were treated with 0.3% hydrogen peroxide in methanol to block endogenous peroxidase activity at rt for 10 min. Biotin-labeled anti-rabbit IgG goat polyclonal antibodies (Nichirei) were incubated for 10 min at rt. Peroxidase-labeled streptavidin (Nichirei) was reacted for 5 min at rt. The immunoreactivity was visualized with Impact DAB (Vector). For quantitative analysis of Iba1-positive microglia, cells with more than 10 μm² of Iba1 immunoreactivity were counted by Image J (U.S. National Institutes of Health, Bethesda, MD, USA; http://rsb.info.nih.gov/ij/).

For the immunohistochemistry of the microglial markers and cytokines in frozen tissue, brains were embedded in the OCT compound (Sakura Finetek, Japan) and were cut at a thickness of 10 μm. Sections were fixed with 4% PFA for 10 min at rt. For the antigen retrieval, samples were treated with 0.1% Tween 20 in PBS for 15 min at rt. The slides were blocked with 2% bovine serum albumin for 15 min at rt and incubated overnight with primary antibody at 4°C. Then, slides were incubated with secondary antibodies and DAPI for 1 h at rt. Finally, slides were mounted with ProLong Gold antifade reagent (Invitrogen) and observed with LSM700 (Zeiss). For the detection of CD14, CD45, and CD68, the antibodies were diluted at 1:800. For CD11b, TGF-β, and IL-1β, the antibodies were diluted at 1:200. For IL-10, the antibody was diluted at 1:100. For GFAP and F4/80, the antibodies were diluted at 1:1,000. For secondary antibodies, all Alexa Fluor-labeled antibodies were diluted at 1:1,000.

Expression of CD14 is increased in microglia and accompanied by progression of prion diseases.

Previous DNA microarray analyses have suggested that gene expression of CD14 was upregulated in prion-infected mouse brains in the early stage of infection (10, 11) (M. H. and C.-H. Song, unpublished). To confirm the gene expression of CD14, we performed a quantitative RT-PCR (Fig. 1A). The CD14 gene expression in the thalamus of the Chandler-infected mice was 2.5 ± 0.8, 3.6 ± 0.8, and 3.6 ± 0.6 higher than that in the thalamus of mock-infected mice, at 75, 90, and 120 dpi, respectively, confirming the upregulation of CD14 expression. We also performed immunofluorescence staining of frozen sections of the Chandler- and Obihiro-infected mice brains to confirm the protein expression of CD14 (Fig. 1B), and CD14-positive cells were most frequently detected in the cerebral peduncle at the observed time points (Fig. 1B). The CD14-positive cells were scattered in corpus callosum, internal capsules, and cerebral peduncles at 60 dpi, spreading more widely in the cerebral peduncles and periventricle areas at 90 dpi, and the spreading had expanded to the thalamus at 120 dpi (Fig. 1B and C). A few CD14-positive cells were also detected in the internal capsules and cerebral peduncles of mock-infected mouse brains, suggesting that these cells may be residents in these areas (Fig. 1C). To confirm the CD14 expression in the microglia, double staining of CD14 with CD11b for microglia and of GFAP for astrocytes was performed (Fig. 1D). Most of the CD14 immunoreactivity was detected in the CD11b-positive microglia, but it was not detected in the GFAP-positive astrocytes.

CD14 influences the progression of prion diseases.

To examine the effect of the lack of CD14 on the neuropathogenesis of prion diseases, we inoculated brain homogenates of the Chandler or Obihiro strain-infected mice into CD14^−/− mice (Fig. 2). The CD14^−/− mice infected either with the Chandler strain or with the Obihiro strain survived longer than the WT mice. The mean survival time of the Chandler strain-infected CD14^−/− mice (161.7 ± 3.7 days, n = 11) was significantly longer than that of the WT mice (153.8 ± 3.7 days, n = 12, P < 0.01, Student's t test) (Fig. 2A). The mean survival time of the Obihiro strain-infected CD14^−/− mice (172.3 ± 4.8 days, n = 10) was also significantly longer than that of WT mice (157.9 ± 7.8 days, n = 11, P < 0.01, Student's t test) (Fig. 2B). These results suggest that the lack of CD14 decelerates the progression of the disease.

To determine if knockout of CD14 influences the accumulation of PrP^Sc, we analyzed the PrP-res accumulation in the brains by immunoblotting (Fig. 3A and B). The intensity of the PrP^Sc signals in the brains of CD14^−/− mice infected with either the Chandler or the Obihiro strain was slightly reduced at 90 and 120 dpi and became comparable to that seen in the WT mice at the terminal stage.

We have reported that MAb132, recognizing amino acids 119 to 127 of mouse PrP, in combination with pretreatment of cells with 5 M guanidine thiocyanate, is useful for the PrP^Sc-specific immunofluorescence staining of prion-infected cell cultures (21). It is well established that PrP^Sc includes protease-sensitive and protease-resistant PrP^Sc (23, 24). The protease-sensitive PrP^Sc that cannot be detected by immunoblotting using proteinase K-treated samples could be detected by this PrP^Sc-specific immunofluorescence staining because of the omission of the protease treatment process. In the present study, we applied this method to frozen brain sections for the detection of PrP^Sc (Fig. 3C and D). Signals of PrP could be detected from the brains of prion-infected mice, but signals from the brains of uninfected mice remained at background levels under the same conditions (data not shown), demonstrating that this method can be applied to detect PrP^Sc from frozen brain sections. At 60 dpi, bright punctate staining of PrP^Sc was detected frequently in the thalamus and occasionally detected in the cerebral cortex of the WT mice infected with the Chandler strain. The PrP^Sc was also detected in parts of the thalamus of CD14^−/− mice, but the occurrence was less frequent than in the same areas of WT mice (Fig. 3C and D). At 90 dpi, PrP^Sc was distributed more widely; PrP^Sc could be detected in the hippocampus as well as in the thalamus and cerebral cortex of WT and CD14^−/− mice infected with the Chandler strain (Fig. 3D). However, the PrP^Sc staining in some brain areas of CD14^−/− mice seemed to be still weaker than in the WT mice (Fig. 3C and D). There were no marked differences in the PrP^Sc distribution between the WT and CD14^−/− mice at 120 dpi. Delay of the PrP^Sc accumulation was also observed in the brains of the Obihiro-infected CD14^−/− mice. At 60 dpi, PrP^Sc was detected in the hippocampus and thalamus of WT mice but not in these areas of the CD14^−/− mice. At 90 dpi, PrP^Sc was spread into the hypothalamus and amygdala of WT mice, but the presence of PrP^Sc in the CD14^−/− mice was restricted to the hippocampus and thalamus.

We also analyzed the prion infectivity in the brains of WT and CD14^−/− mice using Tga20 mice that overexpress mouse PrP and thus are highly susceptible to mouse-adapted prions (25). Prion infectivity in the brains of both of the Chandler-infected CD14^−/− mice at 60 dpi was significantly lower than that in both of the WT mice (Fig. 4). Prion infectivity in the brains of the Obihiro-infected mice also appeared to be lower than those from WT mice; the infectivity in the brain of one CD14^−/− mouse (no. 2) was significantly lower than that of one WT mouse (no.1). However, in the later stages, there were no significant differences in the prion infectivity of WT and CD14^−/− mice in the Chandler or Obihiro infection (Fig. 4).

We also performed a histopathological analysis of the brains of WT and CD14^−/− mice infected with the Chandler strain (Fig. 5). At 90 dpi, slight vacuolar degeneration of neuropil and neurons was only occasionally observed in the thalamus of Chandler-infected WT and CD14^−/− mice. Vacuolar degeneration was widely observed throughout the brains after 120 dpi; however, there was no apparent difference in the severity and distribution of vacuolar degeneration of the WT and CD14^−/− mice infected with the Chandler strain.

Microglial activation.

Depletion of CD14 in a mouse model of Alzheimer's disease (AD) was reported to result in a reduction in the number of microglia (17). To assess if depletion of CD14 altered microglial activation in prion diseases, we analyzed the expression of an activated microglial marker, Iba1, by immunohistochemistry (Fig. 6). At 60 dpi, some microglia of CD14^−/− mice had more protrusions than those of WT mice. Morphological differences in Iba1-positive microglia became more prominent at 90 dpi; microglia in CD14^−/− mice had a larger cytoplasm and more branched protrusions than those of WT mice. At 120 dpi, Iba1-positive microglia in WT mice had a smaller cytoplasm and fewer protrusions; however, microglia in CD14^−/− mice still had relatively larger cytoplasm with many protrusions. In the terminal stage, there were no obvious differences in the morphology of microglia of WT and CD14^−/− mice.

The quantitative analysis of Iba1-positive microglia showed that in the hippocampus and thalamus of the Chandler-infected WT and CD14^−/− mice the numbers of microglia increased from 60 to 120 dpi and then showed no further change or slightly decreased (Fig. 6B). Similar changes were observed in the Obihiro-infected WT and CD14^−/− mice. In the Chandler-infected mice, there was little difference in the numbers of Iba1-positive microglia between the WT and CD14^−/− mice by 90 dpi. At 120 dpi, Iba1-positive microglia increased more in the thalamus of CD14^−/− mice than in WT mice. In the terminal stage, numbers of the Iba1-positive microglia decreased both in WT and CD14^−/− mice, although those in the thalamus of CD14^−/− mice were still slightly larger than those of WT mice. In the Obihiro-infected mice, there were no differences in the numbers of Iba1-positive microglia of the WT and CD14^−/− mice at 60 dpi. At 90 and 120 dpi, numbers of Iba1-positive microglia in the thalamus of CD14^−/− mice appeared to be larger than in the thalamus of WT mice. In the terminal stage, there were no marked differences in the numbers of Iba1-positive microglia of WT and CD14^−/− mice.

To further characterize the differences in microglial activation of prion-infected WT and CD14^−/− mice, we performed immunofluorescence staining for other microglial markers (Fig. 7). Expression of CD11b, a commonly used microglial marker, was elevated in the brains of WT and CD14^−/− mice from 90 to 120 dpi, and the expression of CD11b in CD14^−/− mice was more intense than that in WT mice at each time point. A marker for macrophages and monocytes, F4/80, is also frequently used as a microglial marker. Similar to CD11b, F4/80 immunoreactivity was detected more in CD14^−/− mice than in WT mice at 90 and 120 dpi. Similar changes were also observed in the immunofluorescence staining for CD68, a phagocytic marker of macrophages and microglia: CD68-positive cells increased time dependently both in WT and in CD14^−/− mice, and CD68 immunoreactivity was more prominent in CD14^−/− mice than in WT mice. We also analyzed CD45, a common leukocyte antigen expressed on all leukocytes. Ramified parenchymal microglia express lower levels of CD45 than peripheral macrophages (26, 27), while the expression of CD45 in microglia may be increased under certain pathological conditions such as HIV encephalitis (28, 29) and Alzheimer's disease (30). For the study here, expression of CD45 was clearly more upregulated in CD14^−/− mice than in WT mice at 90 and 120 dpi (Fig. 7).

Expression of anti-inflammatory cytokines.

Depletion of CD14 in the AD model mouse resulted in a reduced amyloid β (Aβ) plaque burden with the increased gene expression of the anti-inflammatory cytokine IL-10 (17). Similar to the reduced Aβ plaque burden in AD model mice lacking CD14^−/−, this study observed that accumulation of PrP^Sc in prion-infected CD14^−/− mice was delayed by 90 dpi (Fig. 3C and D). The knockout of the IL-10 gene has been shown to greatly shorten the survival time of prion-infected mice, suggesting a protective role of IL-10 in the progression of prion diseases (8). Therefore, we analyzed the anti-inflammatory cytokine expression in prion-infected CD14^−/− mice to determine if a lack of CD14 modulates the inflammatory response in the brain.

Compared to the mock-infected mice, the prion-infected WT mice showed a weak but detectable IL-10 expression from 60 to 120 dpi. Interestingly, IL-10 immunoreactivity was observed more frequently in the thalamus of CD14^−/− mice than in WT mice throughout the time of observation (Fig. 8A), and IL-10 immunoreactivity in CD14^−/− mice appeared to increase from 60 to 90 dpi but to decrease from 90 to 120 dpi. The IL-10-positive areas were quantified and statistically analyzed if more than 3 mice were available. At 60 dpi, the areas positive for IL-10 in CD14^−/− mice were significantly larger than in WT mice (P < 0.01, Student's t test) (Fig. 8B). The IL-10-positive areas in CD14^−/− mice were also larger than those in WT mice at 90 and 120 dpi. We also examined the cell type in the brains expressing IL-10 by double staining (Fig. 8C to F). Most of the IL-10 immunoreactivity was detected in CD11b- or Iba1-positive microglia (Fig. 8C and D) and some in NeuN-positive neurons (Fig. 8E), in the brains of both WT and CD14^−/− mice from 60 to 120 dpi. There was no IL-10 immunoreactivity detected in GFAP-positive astrocytes (Fig. 8F).

We also analyzed another anti-inflammatory cytokine, TGF-β (Fig. 9A), which has been reported to play a role in the suppression of progression of prion diseases (9). At 60 dpi, TGF-β immunoreactivity was observed more frequently in CD14^−/− mice than in WT mice infected with the Chandler or Obihiro strains. Although no significant difference was observed in TGF-β-positive areas between WT and CD14^−/− mice at 60 dpi (P = 0.053, Student's t test), the areas in two of three samples from the CD14^−/− mice were larger than in all samples from WT mice (Fig. 9B), suggesting this tendency. However, little difference in the TGF-β immunoreactivity was observed between WT and CD14^−/− mice after 90 dpi (Fig. 9A and B). However, there was little difference in the TGF-β immunoreactivity of WT and CD14^−/− mice after 90 dpi. Double staining showed that TGF-β immunoreactivity was detected mostly in CD11b- and Iba1-positive microglia (Fig. 9C and D) and only occasionally in NeuN-positive neurons (Fig. 9E). Very little TGF-β immunoreactivity was detected in GFAP-positive astrocytes (Fig. 9F).

Expression of proinflammatory cytokine.

We also assessed if a lack of CD14 influenced the expression of proinflammatory cytokines in the brains of prion-infected mice. Here, we focused on IL-1β, since a lack of IL-1β receptor signaling has been reported to delay the progression of prion diseases (7). Different from the situation in anti-inflammatory cytokines, more immunoreactivity of IL-1β was detected in WT mice than in CD14^−/− mice at 60 and 90 dpi, particularly in the corpus callosum, internal capsules, and cerebral peduncles (Fig. 10A). Quantitative analysis showed that IL-1β-positive areas in the internal capsule of WT mice were significantly larger than the same areas of CD14^−/− mice at 60 dpi (P < 0.05, Welch's t test) (Fig. 10B). Also, IL-1β-positive areas of WT mice tended to be larger than those of CD14^−/− mice at 90 and 120 dpi (Fig. 10B), although statistical analyses could not be carried out due to the limited number of animals (n = 2). The differences in the expression of IL-1β of WT and CD14^−/− mice appeared to be less prominent in the thalamus throughout the time of the observations (data not shown).