Translin KO mice show deficits in a specific form of PKA-dependent long-lasting LTP
In our previous study, we found that translin knockout mice display normal basal synaptic transmission measured by paired-pulse facilitation and input-output curves. These mice also exhibit unaltered transient potentiation, induced by a single 100 Hz stimulation, that requires neither PKA activity nor protein synthesis (15). In the present study, we first tested long-lasting forms of LTP induced by spaced 4-train (four 100 Hz trains of 1 s each, delivered 5 minutes apart) or massed 4-train (four 100 Hz trains of 1 s each, delivered 5 s apart) stimulation. The latter does not depend on PKA activation, whereas the former requires postsynaptic PKA activity (28-32). Also, the massed 4-train LTP is less sensitive to protein synthesis inhibition by anisomycin as compared to spaced 4-train LTP (31). Hippocampal slices from translin KO mice showed marked impairment in spaced 4-train LTP (Fig.1B; n = 5 for each group, two-way repeated-measures ANOVA, F(1,8) = 34.43, p = 0.00038). The average of the initial fEPSP slope over the last 20 min of the recordings was reduced in slices from translin KO mice compared to slices from wildtype littermates (wildtype littermates: 176.6 ± 13.5%, n = 5; translin KO mice: 89.5 ± 9.6%, n = 5, t-test, p = 0.00037). On the other hand, massed 4-train LTP was unaltered in slices from translin KO mice as shown in Fig. 1C (n = 5 for translin KO mice, n = 5 for wildtype littermates, two-way repeated-measures ANOVA, F(1,8) = 0.923, p = 0.365). The average of the initial fEPSP slope over the last 20 min of the recordings was similar between slices from translin KO mice and wildtype littermates (wildtype littermates: 143.6 ± 8.2%, n = 5; translin KO mice: 154.6 ± 9.9%, n = 5, t-test, p = 0.364).
Next, we examined two other long-lasting forms of LTP induced by either theta-burst stimulation (TBS; 15 bursts of four 100 Hz pulses delivered at 5 Hz) or bath application of the adenylyl cyclase activator forskolin (FSK). These forms of LTP rely on increased transmitter release and require presynaptically compartmentalized PKA signaling (19, 30, 33-35). TBS-LTP was unaffected in slices from translin KO mice (Fig. 2A; n = 5 for each group, two-way repeated-measures ANOVA, F(1,8) = 0.007, p = 0.94). The average of the initial fEPSP slope over the last 20 min of the recordings was similar between slices from translin KO mice and wildtype littermates (wildtype littermates: 150.03 ± 7.8%, n = 5; translin KO mice: 151.2 ± 13.4%, n = 5, t-test, p = 0.936). Furthermore, slices from translin KO mice showed no impairment in the FSK-LTP compared to the WT mice (Fig. 2B; n = 5 for translin KO mice, n = 6 for wildtype littermates, two-way repeated-measures ANOVA, F(1,9) = 0.07, p = 0.79). The average of the initial fEPSP slope over the last 20 min of the recordings was comparable between slices from translin KO mice and wildtype littermates (wildtype littermates: 180 ± 14.3%, n = 6; translin KO mice: 180.4 ± 11.8%, n = 5, t-test, p = 0.982).
Taken together, these data suggest that translin is selectively involved in mediating the long-lasting form of LTP induced by spaced tetanic stimuli, but not in LTP induced by massed stimuli, TBS or forskolin.
Translin KO mice exhibit unaltered mGluR-LTD and protein levels of hippocampal FMRP, but show deficits in NMDAR-dependent LTD
One of the most well-studied RBPs is fragile X mental retardation protein (FMRP). Exaggerated metabotropic glutamate receptor-mediated LTD (mGluR-LTD) is a well characterized phenotype of FMRP KO mice and has been proposed as an underlying mechanism of fragile X syndrome (36-38). Because both translin/trax and FMRP mediate local protein synthesis (15, 38), we tested mGluR-LTD in hippocampal slices from translin KO mice. In contrast to the findings from FMRP KO mice, mGluR-LTD was unaffected in slices from translin KO mice (Fig. 3A; n = 5 for each group, two-way repeated measures ANOVA, F(1,8) = 0.08, p = 0.79). The average of the initial fEPSP slope over the last 20 min of the recordings was comparable between slices from translin KO mice and wildtype littermates (wildtype littermates: 75.9 ± 3.2%, n = 5; translin KO mice: 78.9 ± 3.1%, n = 5, t-test, p = 0.473). We reasoned that if translin and FMRP are functionally independent, loss of translin/trax should not cause a compensatory increase in FMRP protein levels. Indeed, Western blot analyses showed no changes in the protein levels of hippocampal FMRP in translin KO mice relative to wildtype littermates (Fig. 3B; translin KO mice: 102.5 ± 0.4%, n = 6; wildtype littermates: 100 ± 4.3%, n = 6, t-test, p = 0.36).
We then investigated NMDAR-dependent LTD induced by prolonged low-frequency stimulation (24, 25), because this form of LTD is mediated by anchored PKA signaling (39, 40) and is not affected in the FMRP KO mice (37). We found that NMDAR-LTD induced by prolonged LFS was impaired in slices from the translin KO mice (n = 7) compared to the wildtype littermates (n = 5) (Fig. 3C; two-way repeated measures ANOVA, F(1,10) = 5.342, p = 0.043). The average of the fEPSP slopes over the last 20-min recordings was significantly different between slices from translin KO mice and wildtype littermates (translin KO mice: 94.9 ± 0.3%, n = 7; wildtype littermates: 71.2 ± 0.3%, n = 5; t-test, p = 0.043).
Together, these data indicate that translin/trax and FMRP play distinct roles in hippocampal synaptic plasticity.