An ancient conserved role for prion protein in learning and memory

Abstract

The misfolding of cellular prion protein (PrP^C) to form PrP Scrapie (PrP^Sc) is an exemplar of toxic gain-of-function mechanisms inducing propagated protein misfolding and progressive devastating neurodegeneration. Despite this, PrP^C function in the brain is also reduced and subverted during prion disease progression; thus understanding the normal function of PrP^C in healthy brains is key. Disrupting PrP^C in mice has led to a myriad of controversial functions that sometimes map onto disease symptoms, including a proposed role in memory or learning. Intriguingly, PrP^C interaction with amyloid beta (Aβ) oligomers at synapses has also linked its function to Alzheimer's disease and dementia in recent years. We set out to test the involvement of PrP^C in memory using a disparate animal model, the zebrafish. Here we document an age-dependent memory decline in prp2^-/- zebrafish, pointing to a conserved and ancient role of PrP^C in memory. Specifically, we found that aged (3-year-old) prp2^-/- fish performed poorly in an object recognition task relative to age-matched prp2^+/+ fish or 1-year-old prp2^-/- fish. Further, using a novel object approach (NOA) test, we found that aged (3-year-old) prp2^-/- fish approached the novel object more than either age-matched prp2^+/+ fish or 1-year-old prp2^-/- fish, but did not have decreased anxiety when we tested them in a novel tank diving test. Taken together, the results of the NOA and novel tank diving tests suggest an altered cognitive appraisal of the novel object in the 3-year-old prp2^-/- fish. The learning paradigm established here enables a path forward to study PrP^C interactions of relevance to Alzheimer's disease and prion diseases, and to screen for candidate therapeutics for these diseases. The findings underpin a need to consider the relative contributions of loss- versus gain-of-function of PrP^C during Alzheimer's and prion diseases, and have implications upon the prospects of several promising therapeutic strategies.

Keywords: Alzheimer; Anxiety; Learning; Novel object recognition; Targeted Mutagenesis; Zebrafish.

Fig. 1.

Prion mutant prp2^−/− zebrafish develop normally and display no overt phenotypes during adulthood. (A) A young adult (∼1-year-old) prp2^+/+ fish is pictured on top, while a young adult prp2^−/− fish is pictured on the bottom. (B) Zebrafish Prp2 is conserved with mammalian PrP^C at the protein domain level. Both have a signal peptide (S), a repeat region (R; though the repetitive region in zebrafish is longer and less patterned than the octarepeats in mammals), a hydrophobic domain (H) and are attached to the cell surface by a GPI anchor (G). Like mammalian PrP^C, zebrafish Prp2 also has putative N-linked glycosylation sites (N) and a disulphide bond (S–S) within its C-terminus. The zebrafish prp2 ua5001 allele has a 4 base pair deletion (frameshift), which produces an early stop codon and a putative nonsense protein lacking all these conserved domains. Further, the prp2 gene product is greatly reduced in abundance, in prp2^−/− mutant zebrafish, including in the brain tissue of adult fish (Fleisch et al., 2013). (C) Restriction fragment length polymorphism (RFLP) assay used for genotyping zebrafish at the prp2 gene locus. There is an Mva cut site in the wild-type prp2 sequence that is not present in the mutant (ua5001 allele) prp2 sequence, leading to a smaller band when the PCR product from prp2^+/+ zebrafish is digested (54 bp) compared to from prp2^−/− zebrafish (71 bp). bp, base pair.

Fig. 2.

Zebrafish lacking prion protein exhibited minor reductions in memory in an age-dependent fashion. Prp2^−/− fish showed a trend towards an age-dependent decline in familiar object preference with the object preference test. (A) Flowchart summarizing the sequence of events in the object preference test. (1) Fish were first habituated to a tank of the same size as the testing arena (the holding tank). (2) Fish were then netted and moved to the testing tank containing two identical objects (F) for the 10-min training phase. (3) Fish were then moved back to the holding tank for a 1- or 5-min period (memory retention interval), during which time one of the familiar objects (F) in the testing tank was replaced with a novel object (N). (4) Finally, fish were placed back into the testing tank for the 10-min object preference test. (B,C) Sample heat maps of 1-year-old prp2^+/+ fish and prp2^−/− fish that displayed object preference during the test phase. Top down view of the test tank, wherein fish can swim around the novel object (N) and/or the familiar object (F). Warm colours (yellows and reds) in the heat map indicate this individual fish spent more time near the familiar object, which was interpreted herein as indicating the fish remembered this object from its earlier training phase (see Materials and Methods and assumptions in Discussion). Scale bar: 3.5 cm (the approximate size of an adult zebrafish). (D,E) Sample heat maps of 3-year-old prp2^+/+ fish and prp2^−/− fish. This representative prp2^+/+ fish (D) exhibited familiar object preference during the test phase, while this example prp2^−/− fish (E) did not (quantified across multiple replicates below). (F) 1-year-old prp2^−/− zebrafish displayed familiar object preference following a 1-min retention interval, while the 1-year-old prp2^+/+ fish did not, as revealed by the D1 index of object preference (# indicates significant difference from 0 at P<0.05 using the Wilcoxon Signed Rank test; n=13 prp2^+/+ fish, n=28 prp2^−/− fish). (G) 3-year-old prp2^+/+ fish displayed familiar object preference after a 1-min retention interval while 3-year old prp2^−/− fish did not (D1 discrimination index, # indicates significant difference from 0 at P<0.05 using the Wilcoxon signed rank test; n=16 fish/genotype). (H) Zebrafish lacking prion protein (prp2^−/−) displayed a notable, though not statistically significant, reduction in familiar object preference when comparing between ages as measured by D1 (values replotted from Fig. 2F,G).

Fig. 3.

Zebrafish lacking prion protein (prp2^−/−) displayed an age-dependent decline in memory as revealed by the D2 discrimination index. (A) 1-year-old prp2^−/− zebrafish displayed familiar object preference following a 1-min retention interval, while the 1-year-old prp2^+/+ fish did not, as revealed by the D2 index of object preference (# indicates significant difference from 0 at P<0.05 using the one sample t-test; n=13 prp2^+/+ fish, n=28 prp2^−/− fish). (B) 3-year-old fish of both genotypes (prp2^+/+ and prp2^−/−) failed to show object preference following a 1-min retention interval using the D2 discrimination index (n=16 fish/genotype). (C) Zebrafish lacking prion protein (prp2^−/−) displayed a small, though not statistically significant, reduction in familiar object preference with age as measured by D2 (values replotted from Fig. 3A,B).

Fig. 4.

Zebrafish lacking prp2 exhibited an age-dependent difference in object appraisal. 3-year-old prp2^−/− fish spent more time in close proximity to the novel object than 1-year-old prp2^−/− fish in the NOA test. (A) Amongst 1-year-old fish, there was no significant difference between genotypes (prp2^+/+ and prp2^−/−) in time spent in the object (centre) zone (n=14 prp2^+/+ fish, n=29 prp2^−/− fish). (B) Time spent in the object (centre) zone was significantly greater for the 3-year-old prp2^−/− fish than for the 3-year- old prp2^+/+ fish (*P<0.05 with one-tailed Mann–Whitney test, n=16 fish/genotype). (C) 3-year-old prp2 ^−/− fish spent a significantly greater period of time in the object (centre) zone than 1-year-old prp2^−/− fish (*P<0.05 with the Mann–Whitney test; values replotted from Fig. 4A,B).

Fig. 5.

There were no detectable differences in anxiety between 3-year-old prp2^+/+ and prp2^−/− fish or age-related changes in anxiety among prp2^−/− fish using the novel tank diving test. All groups of fish displayed the typical bottom dwelling response to tank novelty. (A) The 1-year-old prp2^+/+ fish spent more time in the bottom zone and less time in the top and middle zones than age-matched prp2^−/− fish (*P<0.05 with the unpaired t-test; n=14 prp2^+/+ fish, n=29 prp2^−/− fish). (B) Among 3-year-old fish, there were no significant differences between genotypes in time spent in the top zone, middle zone, or bottom zone with the unpaired t-test (n=11 prp2^+/+ fish, n=10 prp2^−/− fish). (C) 1-year-old prp2^−/− fish and 3-year old prp2^−/− fish spent a comparable proportion of time in the bottom zone (values replotted from Fig 5A,B).