Altered hippocampus synaptic function in selenoprotein P deficient mice

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.


Background
Selenium (Se) is a naturally occurring micronutrient that is essential for several known major metabolic pathways, including; thyroid hormone metabolism [1][2][3] and antioxidant defense systems [4,5] in both humans and rodents. Dietary selenium can exist as selenomethionine, selenocysteine, selenate or selenite [6] and is incorporated as selenocysteine into a subset of specific selenium-dependent proteins (selenoproteins) [7]. Of particular interest are the selenoproteins involved in oxidative stress, such as the glutathione peroxidase enzymes (classical GPX-1, gastrointestinal GPX-2, plasma GPX-3, phospholipid hydroperoxide GPX-4) and the thioredoxin reductase 1 and 2 (TR) [8,9]. The dietary intake of selenium has a delicate balance between the harmful effects of excessive selenium uptake leading to selenium toxicity and the damaging effects on selenoprotein function during selenium deficiency. Interestingly, the CNS appears to be resistant to fluctuations in selenium and can maintain stable levels despite near complete depletion of dietary selenium uptake [10]. This suggests that the process of selenium transport needed for selenocysteine-protein incorporation is important for normal CNS function.
Approximately 60% of selenium in plasma is present as selenoprotein P [11]. This protein differs from other selenoproteins in that it incorporates up to 10 Se atoms per molecule in the form of selenocysteine as opposed to the single selenocysteine incorporated in other selenoproteins [12]. Selenoprotein P is abundant throughout the body suggesting that one function is to serve as a primary transporter in systemic selenium delivery [13,14]. This is especially evident in the CNS where selenoprotein P levels can be maintained independent of plasma selenium [15]. However, genetic ablation of selenoprotein P results in reduced, but not a commensurate decrease in CNS-associated selenium levels, suggesting that other selenium proteins compensate for the selenoprotein P deficiency and supporting the hypothesis that basal selenium levels are essential for the brain and have a priority for available selenium [3,16]. Sepp1(-/-) mice fed a selenium-deficient diet show severe motor dysfunction associated with neuron degeneration, which can be prevented by supplementation of high dietary selenium [16][17][18].
Reduced dietary selenium can have significant effects on levels of selenoproteins involved in oxidative stress, such as glutathione peroxidases, thioredoxin reductases and methionine sulfoxide reductases [19,20]. Selenium, through the incorporation into selenoproteins, provides protection from reactive oxygen species (ROS)-induced cell damage [21]. Because oxidative stress, and subsequent production of ROS, has been implicated in neurodegenerative disorders, such as Alzheimer's disease, Parkinson's disease and Duchenne muscular dystrophy [22], there may also be a role for selenium in these disorders. In the present study we examined the consequences of selenoprotein P deficiency on cognitive capacity and synaptic function with a focus on the hippocampus, an area of the CNS intimately involved in learning and memory processes. Sepp1(-/-) mice demonstrate no overt behavior phenotype, but were found to have a subtle disruption in acquisition of spatial learning and memory. In contrast, synaptic transmission is altered and short-and long-term synaptic plasticity is severely disrupted in area CA1 of hippocampus. Interestingly, we found that when Sepp1(+/+) mice were fed a low Selenium diet (0 mg/kg), they too exhibited altered synaptic transmission and synaptic plasticity. Our observations suggest an important role for both selenoprotein P and dietary selenium in overall proper synaptic function.

Animals
Littermates obtained from Sepp1(+/-) crosses were genotyped as described previously [16]. This line has been back-crossed to C57BL6/J ten times. Animals were fed a Torula yeast-based diet supplemented with either 0 mg Se/kg or 1 mg Se/kg in the form of sodium selenite [18] and housed in a 10/14 hr light/dark cycle. Mice used for behavior were fed a diet of 1 mg Se/kg. Behavior testing was performed with 3-5 month old males and females during the light cycle. No gender-dependent differences were observed in any of the tests, therefore data were combined for graphical presentation and statistical analysis. All animal testing procedures were approved by the Institutional Animal Care and Use Committee of Vanderbilt University and followed the NIH guidelines for the care and use of laboratory animals.

Rotorod
Starting speed for the rod began at 4 rpm and increased to 40 rpm over a 5 min period. Latency to fall was recorded as the time in which the mouse fell off the rod or failed to continue walking on the rod after two revolutions. The mice were given four trials per day with an inter-trial interval of 1 hour for two consecutive days. Data represent the mean ± SEM. Student's t test was used for statistical analysis with p < 0.05 as significance criteria.

Open field
Total distance traveled in the open field chamber (27 × 27 cm) during 15 min under standard room lighting conditions was assayed. Mouse activity was measured by 16 photoreceptor beams on all sides of the chamber connected to an Activity Monitor program (Med Associates, Inc). Zone analysis was performed to determine the time spent in the center of the chamber (designated by an 8 cm × 8 cm region) compared to the perimeter. Data represent the mean ± SEM. Student's t test was used for statistical analysis with p < 0.05 as significance criteria.

Elevated plus maze
The elevated plus maze apparatus consisted of two opposing open arms (30 cm × 5 cm) and two opposing closed arms (30 cm × 5 cm × 15 cm) connected by a central square platform and positioned 40 cm above the ground. Mice were placed in the open arms facing the closed arms at the beginning of the 5 min session. Data represent the total time spent in the various locations ± SEM. Student's t test was used for statistical analysis with p < 0.05 as significance criteria.

Conditioned fear
Training for contextual and cued fear conditioning consisted of a 2 min exploration period, followed by two conditioned stimulus (CS)-unconditioned stimulus (US) pairings separated by 90 sec (tone: 85 dB white noise, 30 s duration; foot shock intensity: 0.5 mA, 2 s duration). Context tests were performed in the conditioning chamber 2 hrs, 24 hrs, and 1 week post training. Cue tests were performed in a dissimilar chamber 2 and 24 hours post training; baseline freezing was monitored before presentation of the tone (85 dB white noise, 3 min duration). Freezing was defined as the absence of movement for 2 sec and measured objectively by a motion monitoring program (Med Associates, Inc). Data represent the mean ± SEM. Student's t test was used for statistical analysis with p < 0.05 as significance criteria.

Water maze
Training for the hidden platform version of the Morris water maze consisted of four trials (60 s maximum; intertrial interval, 60 min) each day with the starting location changing for each trial to avoid quadrant bias. On days 5, 7 and 10, a probe trial was administered before training by removing the platform and monitoring the animals' trajectory over 60 sec. The visible platform test was performed 2 hours after the last training session on day 10. A new platform was placed in the quadrant opposite from where they were trained. Above the submerged platform was a large red flag. The mice were placed in the opposite quadrant and latencies to find the new platform were recorded. Animal swim paths were objectively recorded with a video tracking system (HVS Image Analyzing VP-200). Data represent the mean ± SEM. Student's t test linear regression and χ 2 tests were used for statistical analysis with p < 0.05 as significance criteria.

Normal hippocampal morphology in Sepp1(-/-) mice
The generation and metabolic characterization of the Sepp1(-/-) mice has been described previously [16]. The Normal hippocampal morphology in Sepp1(-/-) mice following experiments focus on hippocampus-dependent learning and memory, as well as direct measurement of synaptic transmission and plasticity in area CA1 of the hippocampus. Therefore we performed a gross morphologic examination of the CNS in general, and hippocampus specifically, in order to identify any potential differences due to developmental abnormalities. We find that Sepp1(+/+) and Sepp1(-/-) fed 1 mg Se/kg and Sepp1(+/+) fed a 0 mg Se/kg diet does not alter the gross morphology of the CNS in general (data not shown) or the size and organization of the hippocampus ( Figure 1). In addition, chronic depletion of selenium from the diet of adolescent mice does not result in gross morphological changes in the hippocampus. These observations also suggest that any hippocampus-dependent behavioral or electrophysiological phenotypes we observe are not likely the result of gross morphological abnormalities in the hippocampus, but rather altered Sepp-or selenium-dependent metabolic or signaling processes.

Sepp1(-/-) mice show motor coordination defects but normal levels of locomotor activity and anxiety
Because it has been shown that selenoprotein P plays a role in maintenance of selenium levels in the brain and that it may be linked to certain neurodegenerative disor-Sepp1(-/-) mice exhibit normal associative conditioned fear learning and memory ders, we examined potential Sepp1(-/-) behavioral phenotypes in several well-characterized behavioral paradigms. These tests also serve as necessary controls when elucidating the role of selenoprotein P in cognitive processes. Both Hill et al (2004) and Schweizer et al (2004) performed rotorod experiments to test overall locomotor function in Sepp1-deficient mice and found reduced motor coordination [18,22]. We performed a modified version of their protocol to include motor learning over time. For both days of training, we observed that Sepp1(-/-) mice had shorter latencies to fall off the accelerating rotorod than their littermate controls (Figure 2A). We also found a statistically significant increase in latencies over the two day period for both genotypes. In agreement with previous reports, we found motor coordination is compromised in Sepp1(-/-) mice and we extend those findings to include that motor learning is preserved in the absence of selenoprotein P.
Previous research of selenium deficiency in mice revealed reduced open field activity and that the mice spent less time in the center of the field indicating increased anxietylike behavior [25]. Thus, we performed the same open field test to assay general locomotor activity and anxiety. In comparing the total distance traveled during a 15 min exposure to the chamber we found no statistical difference between the Sepp1(-/-) mice and Sepp1(+/+) ( Figure 2B). Zone analysis revealed that both genotypes spent equivalent amounts of time in the center of the chamber and the greatest amount of time in the surrounding areas, nearest to the walls of the chamber ( Figure 2C). These results indicate that, unlike dietary selenium deficiency, the Sepp1(-/-) mice have normal exploratory tendencies and behave similar to their littermate controls when introduced to a novel open field.
A more sensitive analysis of anxiety can be achieved with the elevated plus maze. Both genotypes demonstrated a similar pattern of behavior when presented with the elevated plus maze. Equivalent amounts of time were spent in the open arms of the maze with the preponderance of time spent in the closed arms ( Figure 2D). These observations are consistent with the open field analysis and suggest that genetic elimination of selenoprotein P does not have an adverse effect on overall anxiety-like behavior.

Sepp1(-/-) mice exhibit normal associative fear learning and impaired spatial learning
To measure associative fear learning and memory in Sepp1(-/-) mice, we used an aversive stimulus (unconditioned stimulus-US, mild foot shock) paired with an auditory cue (conditioned stimulus-CS, 84 dB tone). During training, mice showed similar freezing rates when presented with the two CS-US pairings over the 7 min period ( Figure 3A). To assess short-term memory retrieval, mice were subjected to context and cue tests 2 hrs post training. No difference was observed between the genotypes in either short-term memory tests ( Figure 3B and 3C). Longterm associative memory formation was assayed 24 hrs after training. Again the Sepp1(-/-) mice were indistinguishable from the Sepp1(+/+) in both the context and cue tests ( Figure 3B and 3C). Finally, to assess hippocampusdependent long-term memory retention, we placed the mice back in the training context 1 week post training for 8 min. No difference in freezing while in the context was observed between the Sepp1(-/-) and Sepp1(+/+) mice ( Figure 3D). Taken together, these results suggest that selenoprotein P is not essential for contextual or cued associative learning and memory.
The disruption of spatial learning during selenium deficiency has been previously reported [25]. Despite the lack of a phenotype in our associative fear conditioning paradigm, we performed a standard spatial learning test using the Morris hidden platform water maze paradigm. For this study, both Sepp1(-/-) and Sepp1(+/+) mice were trained to locate a platform submerged in a circular pool filled with opaque water. Distal visual cues were positioned outside the pool and the latency to escape via the platform was recorded for 4 trials per day. By training day 3, a statistical difference in latency between genotypes emerged ( Figure 4A). Moreover, when linear regression was performed on the individual trials for days 2 and 3, the slope of the line corresponding to Sepp1(-/-) mice latencies was not significantly different from 0 (slope = 0.5), indicating that during those days of training the animals did not remember the location of the hidden platform ( Figure  4B). As expected, the slope of the line corresponding to Sepp1(+/+) mice latencies was significantly less than 0 (slope=-3.2), indicating spatial learning of the location of the hidden platform was occurring ( Figure 4B). A probe trial was given on day 5 and no difference in the time spent searching the target quadrant was observed between the genotypes ( Figure 4C). Training resumed on day 6 and again a difference in escape latency was observed ( Figure  4A). To test whether an increased time between training and probe trial would reveal a genotype-dependent difference in long-term memory retention, we trained on day 7, then again on day 10 and performed a probe trial. We found no genotype-dependent difference in the escape latencies ( Figure 4A), or in the time spent searching the target quadrant (data not shown). A visible platform test was performed 2 hrs after training on day 10. Latency to find the flagged platform located opposite of training platform was recorded. Both genotypes showed similar latencies to find the new platform ( Figure 4D p = 0.70). Taken together, these results suggest that Sepp1(-/-) mice have a subtle, yet significant spatial learning deficit that is not the result of impaired visual acuity, swimming ability, or motivation and can be overcome with continued training.

Sepp1(-/-) mice exhibit altered basal synaptic transmission and short-term plasticity in area CA1 of the hippocampus
We next sought to determine if the spatial learning deficit observed in Sepp1(-/-) mice correlated with a defect in synaptic plasticity. To address this question, we examined the physiologic properties of these mice in area CA1 of the hippocampus, a brain area necessary for normal spatial learning and memory formation. First, we assessed Schaffer collateral-CA1 synaptic connectivity by examining the general input-output of synaptic responses. We stimulated the presynaptic fibers with increasing stimulus intensities and recording the evoked CA1 field excitatory postsynaptic potentials (fEPSPs) in stratum radiatum at Schaffer collateral synapses. As stimulus intensity increased, the Sepp1(-/-) slices exhibited significantly greater fEPSP slopes than the wild type slices ( Figure 5A), suggesting an increase in afferent activation. To determine Sepp1(-/-) mice exhibit impaired acquisition of spatial learning and memory slopes than slices obtained from the Sepp1(+/+) ( Figure  5B). This result suggests that Sepp1(-/-) animals have enhanced synaptic transmission in area CA1 of the hippocampus.
The enhanced basal synaptic transmission suggests that synapses in Sepp1(-/-) mice might have an improved ability to undergo synaptic plasticity. Paired-pulse facilitation (PPF) is a form of short-term plasticity. When two pulses at a short interpulse interval are given to the afferent pathway, the postsynaptic response to the second stimulus is increased when compared with the first response. This phenomenon is understood to be due to residual calcium in the presynaptic terminal that facilitates neurotransmitter release upon the second stimulation resulting in the subsequent increase in the post synaptic response [26]. We tested PPF by delivering paired stimuli at intervals ranging from 20 to 300 ms apart. The Sepp1(-/-) slices showed significantly reduced facilitation to the pairedpulse stimulation at intervals of 20, 40 and 120 ms (Figure 5C). Taken together with the input-output relationship data, these results suggest that changes in both presynaptic (altered neurotransmitter release) and postsynaptic properties (enhancement of fEPSP slopes) result from a loss of selenoprotein P.
A type of long-term plasticity exhibited by the Schaffer collateral-CA1 pathway is long-term potentiation (LTP). Two trains of high frequency stimulation (HFS: 100 Hz, 1 sec) spaced 20 s apart faithfully induces LTP in slices obtained from wild type mice characterized by a transient and highly robust post-tetanic increase in the slope of the fEPSP that is followed by a sustained and less robust increase ( Figure 5D). Surprisingly, we found that this stimulation paradigm failed to elicit LTP in the Sepp1(-/-) slices. Strikingly, not even a modest transient post-tetanic increase was observed. In fact, we observed the opposite; Sepp1(-/-) slices responded to the HFS with a robust transient depression of the fEPSP slope ( Figure 5D). We performed multiple HFS (2 trains of 100 Hz stimulation, separated by 20 sec repeated 4 times with each pairing given 5 minutes apart) on Sepp1(-/-) slices to determine if a stimulation threshold was preventing LTP induction. We found that this HFS protocol did not induce potentiation greater than our standard two train 100 Hz protocol (data not shown). This suggests that the LTP deficit in Sepp1(-/-) mice is independent of the amount of HFS presynaptic input, but does not rule out the possibility of occluded LTP.

Sepp1(+/+) mice feed a selenium deficient diet recapitulate the Sepp(-/-) phenotype
One apparent role of selenoprotein P is to deliver selenium to the CNS. Hill et al (2004) reported that when Sepp1(-/-) mice were fed a diet with a normal level of sele-nium (0.25 mg Se/kg), they had a 43% reduction in selenium levels in the brain [18]. When Sepp1(-/-) are fed a selenium-deficient diet (0 mg Se/kg), they exhibit severe neurological abnormalities and typically do not live past 6 weeks of age, whereas their wild type littermates will survive even with approximately 50% reduction in selenium levels [18]. To test whether the observed Sepp1(-/-) alterations of synaptic plasticity were due to the deletion of the selenoprotein P gene during development or reduction of selenoprotein P and selenium levels in the adult brain, we examined basal synaptic transmission and synaptic plasticity in Sepp1(+/+) mice that were fed a selenium-deficient diet (0 mg Se/kg; designated Sepp1(+/+ 0Se)). Importantly, selenium depletion through dietary restriction results in the absence of selenocysteine and subsequent early translation termination at the site of selenocysteine incorporation of all the selenoproteins, including selenoprotein P [27]. Therefore, Sepp1(+/+ 0Se) have reduced selenium in the CNS coupled with a decrease of selenoprotein P [16]. Similar to the Sepp1(-/-), hippocampal slices obtained from Sepp1(+/+ 0Se) mice exhibit enhanced postsynaptic responses to fixed intensities of increasing stimulation when compared to the Sepp1(+/+) mice fed 1 mg Se/kg (designated Sepp1(+/+ 1Se); Figure 6A). In figure 6B, the Input-Output relationships for the Sepp1(+/+ 0Se) show similar presynaptic depolarization to Sepp1(+/+ 1Se) because the fiber volley amplitudes were indistinguishable.
The similarity we saw between Sepp1(-/-) mice and Sepp1(+/+) 0Se mice in enhanced basal synaptic transmission does not extend to pre-synaptic testing by pairedpulse facilitation. Unlike the Sepp1(-/-) slices, the Sepp1(+/ +) slices, regardless of diet, exhibit normal PPF at all interpulse intervals ( Figure 6C). Taken together with the previous observation that Sepp1(-/-) slices exhibit reduced PPF, these results suggest that the absence of selenoprotein P during development may account for the defect in PPF.
LTP also is altered in slices obtained from Sepp1(+/+) 0Se mice. Similar to the Sepp1(-/-) slices, Sepp1(+/+) 0Se slices do not exhibit LTP in response to HFS. However, unlike the Sepp1(-/-) slices, Sepp1(+/+) 0Se slices do not respond to HFS with a robust transient depression of the fEPSP slope ( Figure 6D), suggesting that the combination of presynaptic and postsynaptic dysfunction in Sepp1(-/-) mice may underlie their significant post-tetanic depression. Moreover, this observation paired with the PPF data suggests that a chronic reduction in selenium during adulthood does not recapitulate completely the alterations in synaptic function that result from genetic abolition of selenoprotein P.

Discussion
The exact role(s) of selenoprotein P in tissues throughout the body are unknown, but since its discovery in 1977 several unique characteristics of the protein have been identified (for review see [28]). First, selenoprotein P incorporates 10 selenocysteine residues in the primary protein [12]. In contrast, other identified selenoproteins exist with the incorporation of a single selenocysteine. Second, selenoprotein P is an abundant plasma protein found to be made in most tissues [13,14]. High concentrations of selenoprotein P mRNA are in the liver, and the liver secretes selenoprotein P into plasma and interstitial fluid [29]. Finally, selenoprotein P contains two distinct protein domains. The first domain consists of the N-terminal domain residue up to just before the second incorporated selenocysteine (AA 1-240). This domain contains a heparin binding motif and exhibits modest peroxidase activity [30][31][32]. The second domain extends from the second selenocysteine to the remaining protein and contains the other ~nine selenocysteines. Taken together, this suggests that selenoprotein P serves the essential role of transporting necessary selenium to other tissues in the form of selenocysteine, while maintaining a distinct domain that may serve in tissue-specific targeting, in particular the CNS.
Interestingly, dietary Se restriction of Sepp1(-/-) mice maintains relatively high concentrations of selenium in the CNS (Sepp1(-/-) 86 ± 12 ng Se/g, Sepp1(+/+) 99 ± 27 ng Se/g) compared to other tissues that often show 50-90% reduction in Se content (kidney, testis, liver) [16]. Thus, the CNS appears to demonstrate a priority for selenium that goes beyond the capability for selenoprotein Pdependent selenium delivery. Furthermore, Se deficiency induced through dietary restrictions results in disruption of motor function and memory formation [16,25], but these observation may be a result of neuronal cell death in the CNS. Thus, the present studies were performed in order to better define the contribution of selenoprotein P to synaptic function beyond selenium delivery [17].
We find distinct behavioral phenotypic similarities and differences between Sepp1(-/-) mice and selenium deficiency. For example, Sepp1(-/-) show normal overall activity and no change in anxiety. In contrast, selenium deficiency in mice have reported decreased activity in the open field test and increased anxiety demonstrated as decreased entry to the center of the field [25]. This suggests that cerebellar function is especially sensitive to reduced selenium, which is supported by the deficit in balance and coordination determined by the cerebellumdependant rotorod test. Spatial learning assessed with the Morris hidden platform test is disrupted in both Sepp1(-/-) mice and selenium-deficient mice, although the deficit in Sepp1(-/-) mice is subtle and can be overcome with con-tinued training. The increased latency time for days 3, 4 and 6 suggest a learning deficit in acquisition, but the lack of a defect following the probe tests on day 5 suggests that the retention of memories once formed are unaffected. The subtlety of this learning deficit is more apparent in light of normal associative fear conditioned learning in Sepp1(-/-) mice compared to Sepp1(+/+).
Seeing no abnormal phenotype for the hippocampaldependant fear conditioning test, we did not expect to see any unusual electrophysiology results. However, the most surprising results of these studies were seen in the electrophysiologic characterization of synaptic function in area CA1 in the hippocampus. Sepp1(-/-) mice show a significantly higher output with a given stimulus compared to Sepp1(+/+). The increase in the slope of the fEPSP at a given stimulus without a change in the fiber volley amplitude suggests that the defect may reside with postsynaptic function. This is supported by the severe deficit of LTP in Sepp1(-/-) mice, which is a predominately postsynapticdependent phenomenon. However, decreased PPF at short interpulse intervals is characteristic of presynaptic dysfunction. Interestingly, our results from Sepp1(+/+) mice on a 0 mg Se/kg diet show altered synaptic transmission and reduced LTP similar to that of Sepp1(-/-) mice. Thus, it is likely that selenoprotein P deficiency has no deleterious developmental consequences, but reductions in selenium, whether through the disruption of selenoprotein P gene or dietary selenium restriction, results in discernable differences in synaptic function. Regardless, it is interesting that these mice exhibit a very severe defect in synaptic plasticity across Schaffer collateral synapses of the hippocampus, yet show only subtle defects in the Morris hidden platform water test and normal associative fear conditioned learning, two distinct hippocampusdependent behavioral processes.
What might be the mechanism underlying the changes in synaptic function? We can speculate on at least three putative scenarios. First, absence of selenoprotein P may result in diminished anti-oxidant capacity resulting in reduced LTP and disruption of presynaptic responses. However, most research shows that inhibition of antioxidants leads to profound memory disruption as well as disruption of LTP [33,34], a phenotype we don't see in our Sepp1(-/-) mice. Second, the reduced selenium content in the CNS may be disrupting the specific function of a selenoprotein involved in synaptic plasticity. Homology studies have identified a number of selenium containing proteins, but none of these are obvious candidates as plasticity proteins. Finally, selenoprotein P may be acting as a signaling molecule through an as yet undetermined receptor. Receptor-dependent endocytosis of selenoprotein P would be a likely mechanism for selenium delivery to cells of the CNS. Numerous signal transduction pathways intimately involved in synaptic plasticity and learning and memory rely on ligand-induced endocytosed receptors [35][36][37]. Selenoprotein P deficiency, through direct or indirect interactions, may impact an important signaling system.
The present studies establish that selenoprotein P deficiency results in subtle spatial learning deficits and severe synaptic plasticity defects. It is difficult to discern whether this is due to selenoprotein P itself, or the loss of selenium transport to the CNS. However, the use of adult Sepp1(+/ +) mice maintained on a 0 mg Se/kg diet suggests that the predominance of synaptic dysfunction in Sepp1(-/-) mice is not a result of developmental abnormalities. Future studies will help define whether selenoprotein P is acting as a signaling molecule in the CNS or simply facilitates the systemic distribution of necessary selenium.