Activity-induced targeting of profilin and stabilization of dendritic spine morphology.
Nature Neurosci. 6, 1194-1200 (2003) | |
) (Рис.1.) | Distribution of Profilin II-GFP and unmodified, soluble GFP in dendrites of transfected hippocampal pyramidal neurons. (a) In ca. 4% of neurons profilin II-GFP is highly enriched in dendritic spines (arrowheads) compared to the dendrite shaft. (b) In ca. 44% of transfected cells profilin II was evenly distributed between spines and shaft. (c) In all cells expressing unmodified GFP the soluble marker is evenly distributed throughout dendrite shafts and spines. (d) Percentages of transfected hippocampal neurons showing spine enrichment of profilin II-GFP or unmodified GFP * ++, cells showing enrichment in all spines; + cells showing enrichment in some spines; - cells showing no spine enrichment. (n = total number of cells examined for each construct) ) (Рис.2.) | Low magnification images showing profilin II-GFP distribution in dendrites of two profilin II-GFP transfected cells from independently established cultures. After 25 days in vitro each culture was exposed to 10 μM glutamate for 30 min and a live cell image captured before (0') and after (30') treatment. Scale bars: 10 μm ) (Рис.3.) | Profilin II remains concentrated in spine heads after withdrawal of the targeting stimulus. (a) Segment of a dendrite before stimulation. Letters next to boxes identify individual dendritic spines shown in detail in panels d, e and f. (b) The same dendritic segment after 30 min exposure to 5 μM glutamate showed accumulation of profilin II-GFP in spine heads. (c) 45 min after glutamate was washed out, profilin II was still strongly concentrated in spine heads. (d-f) Detail of individual spines shown in the boxed areas of panel a. The contrast has been inverted for clarity so that areas of high fluorescence appear dark. (d,e,f) spines before glutamate treatment. (d',e',f') the same spines 45 min after glutamate was removed. Scale bars: (a) 5 μ m, (d,e,f) 1 μm. ) (Рис.4.) | Stimulation procedures used to induce profilin targeting do not compromise the health of hippocampal neurons. (a) Neurons (21 DIV) were stimulated by exposure to either 5 μM glutamate, 5 μM NMDA, or recording medium lacking Mg2+. 12 hours later after treatment cells were stained with propidium iodide to stain nuclei of dead cells. Hoechst was used to stain all nuclei. (b) Diagram indicating the percentage of dead cells for each treatment. 8 regions containing 7-15 cells of 3 differently established cell cultures for each treatment were randomly chosen and analyzed (mean s.e.). Treatments with 5 μM glutamate, 5 M NMDA and zero Mg2+ do not lead to a significant increase in cell death (p > 0.1; compared to non-treated cells). By comparison, application of 100 μM glutamate, which has been shown to have excitotoxic effects (Choi, D.W. , 1988.Neuron 1, 623-634), induces cell death in most cells examined (p < 0.01; two tailed T-test). ) (Рис.5.) | Expression of CFP-G(GP5)3 destabilizes spine structures when transfected in mature neurons. Differently established cultures (n = 5) were transfected at DIV 18 with CFP-G(GP5)3 and cells were analyzed 2-4 days after transfection (n = 8). Control cells of the same cultures were transfected with YFP-actin and examined for spine morphology (a, a') Dendritic segment of a cell expressing CFP-G(GP5)3 having destabilized spine structures (b, b') Control cell expressing YFP-actin having normal spine shapes. (c, d) An independent culture in which a cell expressing CFP-G(GP5)3 has irregularly shaped spines (c) whereas a cell expressing YFP-actin has spines of normal morphology (d).  (Рис.6.) | Transgenic mice express profilin II-GFP at low levels. Expression levels of profilin II-GFP was compared between transfected hippocampal neurons and neurons derived from transgenic animals. Neurons were grown in cell culture for 3 weeks and five randomly-selected cells were analyzed for their expression level by calculating the average fluorescence pixel intensities. Neurons derived from transgenic mice have an averaged pixel intensity of 140.3 16.1 whereas transfected neurons have a average value of 375.9 147.2 (both measurements mean s.d.; arbitrary units  (Рис.7.) | (Supplementary data to Fig. 4). Low frequency stimulation induces profilin targeting to dendritic spines and simultaneously stabilizes dendritic spine morphology. Neurons were doubly transfected to express both profilin II-YFP and CFP-actin and the culture was subjected to LTD-pattern stimulation (900 pulses at 1 Hz). The distribution of profilin II-YFP before (upper panel) and 30 min after (lower panel) the onset of stimulation showing the redistribution of profilin to spine heads. (See Supplementary Video 1) |
Дендритные шипики образуют постсинаптические точки контакта для большинства возбудительных синапсов. Известно, что долгосрочные изменения в синаптической эффективности (synaptic efficacy) сопровождаются стабилизацией морфологии шипиков, однако неясно как синаптическая активность влияет на стабильность шипиков. Авторы предложили механизм, основанный на зависимом от синаптической активности таргетировании цитоскелетного регулятора шипиками.
Profilin – белок, регулирующий полимеризацию основного цитоскелетного элемента актина. Авторы исследовали распределение profilin в зависимости от синаптической активности и в зависимости от подвижности шипиков. Прежде всего, нейроны гиппокампа в культуре были обработаны глутаматом, который активировал постсинаптические N-methyl-D-aspartate (NMDA) рецепторы. Было показано, что это вело к накоплению profilin в дендритных шипиках, которое блокировалось, если клетки обрабатывали антагонистами NMDA рецепторов.
Авторы показали, что электрическая стимуляция нейронов также может вести к тому, что profilin становится мишенью шипиков. Однако эффективными оказались лишь определенные паттерны электрической активности. Интересно то, что они оказались теми же паттернами, которые вызывают долгосрочные изменения в синаптической эффективности (synaptic efficacy).
Авторы также показали, что стабилизация шипиков зависит от перераспределения profilin. Они блокировали накопление profilin в шипиках, трансфицируя нейроны конструкцией, которая кодировала цитоплазматический profilin-связанный пептид. Форма дендритных шипиков в трансфицированных клетках оставалась неправильно-продолговатой, хотя в норме они должны были подвергаться созреванию и приобретать компактную грибовидную форму.
Эти находки дают возможность моделировать стабилизацию шипиков. В соответствии с этой моделью синаптическая активность является причиной того, что profilin становится мишенью шипиков, которые, в свою очередь, способствуют полимеризации актина. И хотя profilin и актин не контролируют прямым способом синаптическую эффективность (synaptic efficacy), авторы предполагают, что они могли бы выделить синапсы, которые подвергаются долгосрочным изменениям и, таким образом, представляют важный промежуточный этап в процессе консолидации памяти. См. также: 1. Ottersen, O. P. & Helm, P. J. How hardwired is the brain? Nature 420, 751-752 (2002)
2. Martin, K. C. & Kosik, K. S. Synaptic tagging – who's it? Nature Rev. Neurosci. 3, 813-820 (2002) Supplementary Table 1 MOV file (770K)
Supplementary Video 1. Time-lapse recording of CFP-actin before stimulation began showing that dendritic spines change shape rapidly under resting conditions. Recordings are of 6 min duration at 1 frame every 15 sec.
MOV file (841K)
Supplementary Video 2. Time-lapse recording of CFP-actin 30 min later showing that dendritic spines had become morphologically stable following following LTD-pattern stimulation. Recordings are of 6 min duration at 1 frame every 15 sec.
MOV file (879K)
Supplementary Video 3. (Supplementary data to Fig. 4). Low frequency stimulation induces stabilization of dendritic spine morphology visualized by recording GFP-actin dynamics. To demonstrate that spine motility was not influenced by overexpression of YFP-tagged profilin, the experiment shown in Supplementary Videos 1 and 2 was repeated using neurons transfected with GFP-actin alone where only endogenous profilin is present. This is a time-lapse recording made prior to stimulation showing rapid actin-based motility in dendritic spines under resting conditions.
MOV file (879K)
Supplementary Video 4 (Supplementary data to Fig. 4). Low frequency stimulation induces stabilization of dendritic spine morphology visualized by recording GFP-actin dynamics. To demonstrate that spine motility was not influenced by overexpression of YFP-tagged profilin, the experiment shown in Supplementary Videos 1 and 2 was repeated using neurons transfected with GFP-actin alone where only endogenous profilin is present. This is a time-lapse recording made 30 min after stimulation with 900 pulses at 1 Hz spine showing that following stimulation actin dynamics are suppressed. Recording settings are as for Supplementary Videos 1 and 2.
MOV file (903K)
Supplementary Video 5 (Supplementary data to Fig. 4). Dendritic spine motility is arrested between 10 to 20 minutes after the onset of low frequency stimulation. This is a time-lapse recording of a GFP-actin transfected neuron made prior to low frequency stimulation (recording settings as for the first video) showing the motile character of dendritic spines under control conditions.
MOV file (1.2M)
Supplementary Video 6 (Supplementary data to Fig. 4). Dendritic spine motility is arrested between 10 to 20 minutes after the onset of low frequency stimulation. This shows that 10 min after the start of stimulation spines are still motile but motility declines progressively over the next 10 min at which time spine morphology has become stable.
MOV file (569K)
Supplementary Video 7 (Supplementary data to Fig. 6). Z-axis scan of Z-series image stack taken of the hippocampus section shown in Fig. 6a.
MOV file (1.3M)
Supplementary Video 8 (Supplementary data to Fig. 6). Rotatable image of the area shown in Fig. 6a. |