Altered hippocampus synaptic function in selenoprotein P deficient mice

Abstract

Selenium is an essential micronutrient that function through selenoproteins. Selenium deficiency results in lower concentrations of selenium and selenoproteins. The brain maintains it's selenium better than other tissues under low-selenium conditions. Recently, the selenium-containing protein selenoprotein P (Sepp) has been identified as a possible transporter of selenium. The targeted disruption of the selenoprotein P gene (Sepp1) results in decreased brain selenium concentration and neurological dysfunction, unless selenium intake is excessive However, the effect of selenoprotein P deficiency on the processes of memory formation and synaptic plasticity is unknown. In the present studies Sepp1(-/-) mice and wild type littermate controls (Sepp1(+/+)) fed a high-selenium diet (1 mg Se/kg) were used to characterize activity, motor coordination, and anxiety as well as hippocampus-dependent learning and memory. Normal associative learning, but disrupted spatial learning was observed in Sepp1(-/-) mice. In addition, severe alterations were observed in synaptic transmission, short-term plasticity and long-term potentiation in hippocampus area CA1 synapses of Sepp1(-/-) mice on a 1 mg Se/kg diet and Sepp1(+/+) mice fed a selenium-deficient (0 mg Se/kg) diet. Taken together, these data suggest that selenoprotein P is required for normal synaptic function, either through presence of the protein or delivery of required selenium to the CNS.

Figure 1

Normal hippocampal morphology in Sepp1(-/-) mice. 0.1% Toludine blue staining of the cell bodies in 40 μm hippocampal slices obtained from (A) Sepp1(-/-) mice fed 1 mg Se/kg (B) Sepp1(+/+) mice fed 1 mg Se/kg and (C) Sepp1(+/+) mice fed 0 mg Se/kg. The major regions of the hippocampus, dentate gyrus (DG), CA3 and CA1, are indicated in panel A. Magnification at 4×. Scale bar roughly 1 mm.

Figure 2

Sepp1(-/-) mice show reduced motor coordination and normal anxiety. (A) Average latency to fall off the rotorod from 4 trials for 2 consecutive days (n = 16 mice per genotype). Statistical analysis of Day 1 Sepp1(-/-):Sepp1(+/+) *p < 0.0001, Day 2 Sepp1(-/-):Sepp1(+/+) *p = 0.0008, Day 1 Sepp1(-/-):Day 2 Sepp1(-/-) #p = 0.006, Day 1 Sepp1(+/+):Day 2 Sepp1(+/+) #p = 0.01. (B) Total distance traveled during 15 min in the open field (n = 16 mice per genotype, p = 0.26). (C) Zone analysis of the time spent in the center of the open field during 15 min. (D) Time spent resting in each of the arm conditions of the elevated plus maze (n = 16 mice per genotype).

Figure 3

Sepp1(-/-) mice exhibit normal associative conditioned fear learning and memory. (A) Freezing during the two trial, paired conditioned stimulus (CS: 85dB tone represented by black line) and unconditioned stimulus (US: 0.5mA foot shock represented by the arrow) training paradigm. Both genotypes respond comparably to training (n = 16 mice per genotype). (B) Freezing during the context tests (2.5 min) performed 2 hrs and 24 hrs post training. Both genotypes learned to associate the context with the US equally well. (C) Freezing during the cued tests performed 2 hrs and 24 hrs post training. Both genotypes learned to associate the CS with the US equally well. Graph shows the % increase in freezing to the CS compared to baseline freezing before presentation of the CS. (D) Freezing during the context test (8 min) performed 1 week post training. Both genotypes exhibited long-term memory for association of the context with the US.

Figure 4

Sepp1(-/-) mice exhibit impaired acquisition of spatial learning and memory. (A) Average escape latency from 4 trials per day. Sepp1(-/-) mice showed a significantly longer latency on days 3 and 6 (n = 10 mice per genotype, Day 3 *p = 0.04, Day 6 *p = 0.04). (B) Linear regression of escape latencies per trial from days 2 to 3 (Sepp1(-/-) slope = 0.5, Sepp1(+/+) slope = -3.5). (C) Time spent searching the 4 quadrants for the hidden platform during a 60 sec probe trial on day 5 (TQ-target quadrant, OP-opposite, R-right, L-left). Both genotypes selectively searched the target quadrant. (D) Day 10 latencies to locate a visible platform in the opposite quadrant from training. Both genotypes demonstrated similar ability to locate the visible platform.

Figure 5

Sepp1(-/-) mice exhibit enhanced synaptic transmission and severe synaptic placticity defects. (A) Input-output relationship of the slope of the CA1 field excitatory postsynaptic potential (fEPSP) in response to increasing stimulation of the Schaffer collateral fibers. Slices obtained from Sepp1(-/-) mice exhibit greater evoked fEPSP slopes than wild type controls (Sepp1(-/-) n = 12, Sepp1(+/+) n = 8). Nonlinear regression zero to top analysis confirms that the curves are different (p < 0.0001). (B) Relationship between the slope of the evoked fEPSPs from panel A and the corresponding fiber volley amplitude. Sepp1(-/-) slices exhibit a greater postsynaptic response than wild type slices to similar presynaptic depolarization. Nonlinear regression zero to top analysis confirms that the curves are different, p < 0.0001) Inset: Representative traces of half the maximum fEPSP slope show a greater fEPSP slope for Sepp1(-/-) slices (red) compared to wild type slices (blue) despite similar fiber volley amplitudes (dashed box). (C) Percent paired-pulse facilitation (PPF) achieved with increasing inter-pulse intervals. Sepp1(-/-) slices have significantly reduced PPF at 20, 40 and 120 ms inter-pulse intervals (Sepp1(-/-) n = 12, Sepp1(+/+) n = 8; 20 ms *p = 0.04, 40 ms *p = 0.03, 120 ms *p = 0.01). Inset: Representative PPF traces at 20 ms (blue) and 40 ms (red) inter-pulse intervals from (1) Sepp1(+/+) slices and (2) Sepp1(-/-) slices. (D) Long-term potentiation induced by high frequency stimulation (HFS: 100 Hz, 1 sec × 2, 20 sec interval). Sepp1(-/-) slices fail to potentiate following HFS (Sepp1(-/-) n = 12, Sepp1(+/+) n = 8; ANOVA p < 0.0001). Inset: Representative traces from (1) Sepp1(+/+) slices and (2) Sepp1(-/-) slices at time pints a, b and c (a: black-first baseline recording, b: red-2 min post HFS, c: blue-60 min post HFS). All trace scale bars are 1 mV by 10 ms.

Figure 6

Adult selenium deficiency enhances synaptic transmission and inhibits long-term potentiation. (A) Input-output relationship of the slope of the CA1 field excitatory postsynaptic potential (fEPSP) in response to increasing stimulation of the Schaffer collateral fibers. Adult selenium deficiency increases evoked fEPSP slopes (Sepp1(+/+ 0Se) n = 11, (Sepp1(+/+ 1Se) n = 8). Nonlinear regression zero to top analysis confirms that the curves are different (p < 0.0001). (B) Relationship between the slope of the evoked fEPSPs from panel A and the corresponding fiber volley amplitude. Adult selenium deficiency increases the postsynaptic response to fixed presynaptic depolarization. Nonlinear regression zero to top analysis confirms that the curves are different, p < 0.0001) Inset: Representative traces of half the maximum fEPSP slope show a greater fEPSP slope for Sepp1(+/+ 0Se) slices (red) compared to Sepp1(+/+ 1Se) (blue) despite similar fiber volley amplitudes (dashed box). (C) Percent paired-pulse facilitation (PPF) achieved with increasing inter-pulse intervals. Adult selenium deficiency does not effect PPF (Sepp1(+/+ 0Se) n = 11, Sepp1(+/+ 1Se) n = 8). Inset: Representative PPF traces at 20 ms (blue) and 40 ms (red) inter-pulse intervals from (1) Sepp1(+/+ 1Se) slices and (2) Sepp1(+/+ 0Se) slices. (D) Long-term potentiation induced by high frequencystimulation (HFS: 100 Hz, 1 sec × 2, 20 sec interval). Sepp1(+/+ 0Se) slices fail to LTP following HFS (Sepp1(+/+ 0Se) n = 11, Sepp1(+/+ 1Se) n = 8; ANOVA p < 0.0001). Inset: Representative traces from (1) Sepp1(+/+ 1Se) slices and (2) Sepp1(+/+ 0Se) slices at time pints a, b and c (a: black-first baseline recording, b: red-2 min post HFS, c: blue-60 min post HFS). The same Sepp1(+/+ 1Se) from Figure 5 were compared with Sepp1(+/+ 0Se). All trace scale bars are 1 mV by 10 ms.