PrPC knockdown by liposome-siRNA-peptide complexes (LSPCs) prolongs survival and normal behavior of prion-infected mice immunotolerant to treatment

Abstract

Prion diseases are members of neurodegenerative protein misfolding diseases (NPMDs) that include Alzheimer's, Parkinson's and Huntington diseases, amyotrophic lateral sclerosis, tauopathies, traumatic brain injuries, and chronic traumatic encephalopathies. No known therapeutics extend survival or improve quality of life of humans afflicted with prion disease. We and others developed a new approach to NPMD therapy based on reducing the amount of the normal, host-encoded protein available as substrate for misfolding into pathologic forms, using RNA interference, a catabolic pathway that decreases levels of mRNA encoding a particular protein. We developed a therapeutic delivery system consisting of small interfering RNA (siRNA) complexed to liposomes and addressed to the central nervous system using a targeting peptide derived from rabies virus glycoprotein. These liposome-siRNA-peptide complexes (LSPCs) cross the blood-brain barrier and deliver PrP siRNA to neuronal cells to decrease expression of the normal cellular prion protein, PrPC, which acts as a substrate for prion replication. Here we show that LSPCs can extend survival and improve behavior of prion-infected mice that remain immunotolerant to treatment. LSPC treatment may be a viable therapy for prion and other NPMDs that can improve the quality of life of patients at terminal disease stages.

Fig 1. LSPCs target the majority of neuronal cells and deliver PrP^C siRNA to the brain when injected intravenously.

A) Flow cytometry of neuronal cells labeled with BAR-224, an anti-PrP^C antibody, and RVG-9r reveals that 98% of cells in the brain display PrP^C and nAchRs on their surfaces. Ninety-six percent of those cells express both PrP^C and nAchRs. A smaller proportion of kidney cells (78%) express both PrP^C and nAchRs. B) In vivo live imaging revealed that RVG-9r increased LSPC delivery to the brain two minutes to ten days after intravascular injection compared to siRNA and peptide complexes without liposomes (SPCs). We controlled for background fluorescence using PBS-injected mice. C) Flow cytometry analysis 24 hours after fluorescent LSPC injection revealed siRNA delivery to 47% of brain cells and 15% of kidney cells (black histograms) compared to PBS-injected controls (gray histograms). Data are representative of at least three independent experiments.

Fig 2. Intravenous treatment of naïve wild-type mice with PrP LSPCs decreased PrP^C mRNA and cell surface PrP^C protein at various time points.

FVB (n = 7), CD-1 (n = 6) and C57Bl/6 mice (n = 5) were injected with LSPCs intravascularly. We assessed PrP^C mRNA and cell surface PrP^C expression via ddPCR and flow cytometry, respectively, in brain and kidney cells in three mice at each of the indicated time points after treatment. LSPCs reduced PrP (A) mRNA and (B) protein expression 20–50% in brains and kidneys of wild -type mice for up to 21 days after a single injection. Error bars indicate 95% confidence interval. * p<0.05, ** p<0.01, *** p<0.001, repeated measures ANOVA.

Fig 3. Repeated LSPC treatment prolongs survival of and decreases PrP^Res accumulation in a subset of prion-infected mice.

Wild type mice were injected intraperitoneally with RML-5 prions and treated intravenously, intranasally, or both with 1578 PrP^C siRNA LSPCs or 1672 PrP^C siRNA LSPCs. We observed no significant differences in survival times among different routes or siRNA used, so we present compiled data here. See S3 Fig for survival curves for each delivery route. A) We observed no significant difference between control infected untreated mice (dashed black line, n = 9) and prion-infected, LSPCs-treated mice (solid black line, n = 19). However, a subset of treated mice positively responded to LSPC treatment (both solid red lines, n = 6, one shown extending from solid black line of all infected treated mice, the other as its own group), surviving significantly longer than non-responders and untreated mice (*p<0.05, survival analysis). Three of five uninfected treated mice (dotted black line) also died unexpectedly. All infected mice harbored PrP^Res (B and C). All samples were treated with 50 μg/mL PK. Red asterisks denote samples from responder mice, 5 of 6 of which appeared to contain relatively less PrP^Res. Double asterisks denote responder #4. We detected no PrP^Res in uninfected (un)treated mice (D) All samples except lane 1 were treated with 50 μg/mL PK. Black lines to the left of each blot indicates the 25 kD molecular weight marker. NBH, normal brain homogenate. RML, Rocky Mountain Lab prion strain. Samples in all lanes were digested with 50 μg/mL Proteinase K, except NBH.

Fig 4. LSPC treatment decreases neuropathology in prion-infected mice.

Immunohistochemistry (IHC) of brain sections through the (A) cerebellum and (B) hippocampus in infected and uninfected mice reveal that (A) while we observed no difference in cerebellar PrP^Res accumulation among infected untreated, non-responder or responder mice, both GFAP stain intensity, a measure of astrogliosis, and vacuolation, a hallmark of prion-mediated neurodegeneration, were significantly decreased in LSPC responder mice. (B) In the hippocampus, we observed decreased PrP^Res, GFAP and vacuolation in responder and non-responder mice compared to infected untreated mice. We detected no PrP^Res or vacuolation but significant GFAP expression in both cerebellum and hippocampus of uninfected treated mice compared to uninfected untreated mice. Boxed areas indicate magnified areas in the panel directly below. Scale bars, 100 μm. Quantitation of PrP^Res and GFAP is expressed as relative pixel intensity ± 95% CI per mm². Vacuolation scores indicate number of vacuoles ± 95% CI per mm². IHC images are representative of at least three mice per group. We collected data from at least three non-consecutive slides from at least two animals from each group. *p < 0.05, **p < 0.01, compared to scores from infected untreated mice, except where indicated otherwise.

Fig 5. Prion-infected LSPC-treated mice have improved behavior scores compared to infected, untreated mice.

We instituted longitudinal burrowing and nesting tests to assess behavioral changes in prion-infected mice starting four weeks before the first LSPC treatment. A) PrP siRNA LSPC treatment prolonged normal burrowing behavior in infected mice. B) We observed even more significant improvement in nesting behavior in infected mice treated with PrP siRNA LSPCs. Error bars indicate 95% confidence intervals. *p<0.05, ** p<0.01, *** p<0.001, two-way ANOVA.

Fig 6. Significant anti-RVG-9r IgG titers detected in mice repeatedly treated with LSPCs, but not in responder mice, coincide with loss of PrP^C knockdown.

A) Total IgG levels, measured using an ELISA assay, against RVG-9r. We detected significant IgG titers against the RVG-9r peptide in uninfected treated and non-responder mice and one responder (#4) mouse. Error bars indicate 95% confidence intervals. (B) We detected significant anti-RVG-9r IgG titers in mice beginning after the fourth LSPC treatment (B) that coincide with de-repression of PrP mRNA (C) and protein (D) expression. * p < 0.05, ** p < 0.01, one-way and two-way ANOVA.

Fig 7. Limited LSPC treatment extends survival times and normal nesting behavior in an accelerated prion disease mouse model.

(A). Survival of prion infected TgA20 mice was significantly extended with just one PrP siRNA LSPC treatment given 40 DPI (dashed line), and extended further with two (dashed dotted line, given 40 and 60 DPI) and three (dotted line, given 20, 40 and 60 DPI) treatments over infected untreated mice (solid black line). All uninfected treated mice appeared normal and survived over 100 days after three treatments (solid gray line). (B) LSPC treatment reduced prion replication up to 90%. All mice were infected with 10⁶ LD₅₀ units of RML-5 prions. The graph shows equivalent prion titers based on time to terminal disease. (C) Representative immunoblot showing infected, LSPC treated mice appeared to harbor less PrP^Res in their brains compared to brain samples from infected untreated mice, although densitometry revealed no significant differences (p = 0.074), except in uninfected treated mice, which harbored no PrP^Res. All samples except lane 1 were treated with 50 μg/mL PK. (D). IHC revealed no PrP^Res and little to no GFAP reactivity in uninfected mice treated three times with LSPCs (first column). While we observed no difference in cerebellar (Cb) PrP^Res accumulation in infected mice untreated (second column) or treated (third column) with LSPCs, we did observe significantly less PrP^Res in the hippocampus (Hp) of infected, treated mice compared to infected, untreated mice. We also observed significantly less GFAP and vacuoles overall in brains of infected and uninfected, treated mice compared to infected untreated mice. Scale bar, 100 μm. Quantitation of infected treated samples (third column) show values for Cb/Hp. All other values are combined Cb and Hp scores. We collected data from at least three non-consecutive sections from at least two animals from each group. (E) Combined nesting scores of all treatment groups revealed prolonged, significantly improved nesting behavior in treated mice (dashed line with squares) compared to infected untreated mice (dotted line and triangles). Uninfected treated mice sustained normal behavior for the duration of the study (solid gray line). *p < 0.05, **p < 0.01, ***p < 0.001 compared to values from infected untreated mice.