PDF《www.nature.com/npjscilearn》

1 National Institute of Child Health and Human Development, Bethesda, MD, USA;

2 Sano?-Genzyme, Framingham, MA, USA;

3 GlaxoSmithKline, R&

D China, Shanghai, China and

4 School of Medicine, Tsinghua Univ., Beijing, China. Correspondence: B Lu ([email protected]) Received

2 December 2015;

revised

18 February 2016;

accepted

6 March

2016 www.nature.com/npjscilearn Published in partnership with The University of Queensland proBDNF? When does tPA/plasmin cleavage of proBDNF take place during the course of L-LTP? During our study of BDNF regulation, we unexpectedly found that L-LTP in hippocampal CA1 synapses can be divided into two temporally distinct stages: an induction (I) and a maintenance (II) stage, based on their sensitivity to the inhibition of tPA/plasmin and protein synthesis inhibitors. We show that both stages require BDNF, but each stage engages distinct underlying mechanisms. Conversion of secreted proBDNF to mBDNF by tPA/plasmin protease system is required at stage I, but not stage II. In contrast, mBDNF might function as a product of activity-dependent translation, responsible for the maintenance of L-LTP at stage II, but not stage I. By using a membrane impermeable inhibitor for tPA and a membrane permeable inhibitor for FIN/PC1, we demonstrated that extra- cellular cleavage of proBDNF is required for stage I, whereas intracellular cleavage is required for stage II. These results provide important insights into the molecular mechanisms underlying L-LTP and long-term memory. RESULTS BDNF is required at both induction and maintenance stages of L-LTP Field recordings were performed in hippocampal slices from wild-type and conventional BDNF heterozygous mutant (BDNF +/ ? ) mice. L-LTP was induced by applying '

12-theta burst'

stimulation (or long TBS, l-TBS in brief) to Schaffer collateral afferents of CA1 synapses.5 Consistent with previous studies,5,35 a severe impairment in L-LTP was observed in slices derived from the BDNF+/ ? mice (Figure 1a). To determine whether there is a speci?c time window in L-LTP that is sensitive to BDNF regulation, we applied exogenous BDNF (200 ng/ml) to BDNF+/ ? (control) slices at two different stages during L-LTP. Application of BDNF

60 min before to

5 min after l-TBS failed to rescue L-LTP in BDNF+/ ? slices (Figure 1a). Field excitatory postsynaptic potential (fEPSP) measured

3 h after l-TBS in control 101.8 ± 8.6% of baseline, but 115.5 ± 12.7% in '

+ BDNF'

group (Student'

s t-test, P = 0.408). Similarly, BDNF applied

5 min after l-TBS until the end of the recording did not rescue L-LTP in BDNF+/ ? slices (Figure 1b, control, 101.8 ± 8.6%;

+BDNF, 105.1 ± 5.8%, P = 0.788). However, perfusion of BDNF throughout the entire experiment completely rescued L-LTP in BDNF+/ ? mice (Figure 1c, control, 101.8 ± 8.6%;

+BDNF, 155.7 ± 9.7%, Po0.005). Acute application of BDNF to wild-type hippocampal slices has also been shown to potentiate basal synaptic transmission.36,37 This requires high concentration of BDNF perfused at a fast rate (4 ml/min).38 Application of BDNF at a lower perfusion rate (?2 ml/min) in our hands did not elicit a change in basal synaptic transmission (data now shown). S........