果蠅Rpn2與Spn-F交互作用之研究

Abstract

神經系統是個複雜的網路，負責協調生物體內的各種訊號。為了因應內在發育的變化與外來環境的刺激，持續的重組對於神經迴路的精準連結非常重要，修剪(pruning)是神經重組(neuronal remodeling)的機制之一，同時也是一個不會導致細胞死亡而去移除多餘神經突觸的過程。此外，這個機制也被發現參與在神經系統的損傷與疾病中,但是調控神經修剪的分子機制目前仍不清楚。在果蠅中，第四型樹突神經細胞(class Ⅵ da neurons)會在果蠅變態過程中歷經大規模的神經樹突修剪。根據這項特性，第四型樹突神經細胞是一個很好的模式系統來了解神經修剪的分子運作機制。 之前的研究顯示在第四型樹突神經細胞中，一個稱為Ik2的絲胺酸/蘇胺酸激酶(Serine/Threonine kinase)會去磷酸化另一個含有螺旋區段(coiled-coil domain)的Spn-F蛋白，並與其交互作用去調控神經樹突修剪，然而，對於Ik2與Spn-F共同調控路徑的下游機制目前仍不清楚。在前人研究中，發現蛋白酶體(proteasome)中的一個亞單位─Rpn2，與Spn-F存在著交互作用，此外，泛素-蛋白酶體系統 (ubiquitin-proteasome system)已被證實參與在調控神經樹突修剪的機制中，再者，我們發現spn-F與rpn2兩者間存在著基因交互作用(genetic interaction)，這結果暗示著他們有可能參與在同一條樹突修剪的調控路徑中，此外也觀察到在化蛹後16小時，Rpn2突變的神經會發生樹突修剪的缺失。因此，基於上述理由，我們假定蛋白酶體是Ik2與Spn-F的下游。 首先，我們確認了這兩個蛋白質之間用來互相結合的區域，分別是Spn-F上的C端高度保留區段(SCD domain)與鄰近序列和Rpn2的N端與PC重複區段。接著，為了研究Spn-F與Ik2是否有具有調控蛋白酶體的能力，我們利用非變性聚丙烯醯胺凝膠電泳(Native-PAGE)去測量蛋白酶體的數量，同時，也使用膠內肽酶檢定(in-gel peptidase assay)來量測蛋白酶體的活性。在細胞中分別大量表現Spn-F、磷酸化位點突變的Spn-F、Ik2和失去磷酸化功能的Ik2並不會影響蛋白酶體複合物(19S和26S)的數量與26S的活性。當我們分別降低細胞中Spn-F與Ik2的表現量，蛋白酶體複合物的數量與活性同樣不受影響。此外，同時降低Spn-F與Ik2的表現量，對於蛋白酶體而言不會有加成性的影響。總結來說，我們的研究指出Ik2與Spn-F並非透過調控蛋白酶體去影響神經樹突的修剪。

Neuronal system is a complicated network that coordinates multiple signals throughout the whole body. The continuous remodeling is crucial for the precise connection for neural circuitry adapting to internal changes during development or external stimuli from environment. Pruning, one of the neuronal remodeling mechanisms, is a self-destructive process to eliminate unnecessary neuronal processes without causing cell death. In addition, pruning also occurs in response to neural injury and disease. However, the molecular mechanism of neuronal pruning is still unclear. In Drosophila, class Ⅵ dendritic arborization (da) neurons will undergo a large-scale dendrite pruning during metamorphosis. Based on their characteristics, class Ⅵ da neurons provide an ideal model to study the underlying molecular mechanism of neuronal pruning. Our studies showed that Ik2, a Serine/Threonine kinase, phosphorylates and interacts with Spn-F, a coiled-coil domain containing protein, to regulate dendrite pruning in class Ⅵ da neurons. However, the downstream effector of Ik2 and Spn-F is still unclear. Here, we identified a protein, Rpn2, which is a subunit of proteasome and has known as a Spn-F-interacting protein. Besides, Ubiquitin-proteasome system has been known to involve in the dendrite pruning process. Moreover, we found the pruning defect in da neurons of rpn2 mutant pupae at 16 h APF and also found that Spn-F has genetic interaction with Rpn2, which suggested that they may function in the same pathway in dendrite pruning. Thus, we hypothesized that proteasome is a downstream effector of Ik2 and SpnF in dendrite pruning. First, we identified the interaction region between these two proteins. Both SCD domain and nearby sequence of SpnF are required for the interaction with Rpn2. On the other hands, Rpn2 interacts with Spn-F by its N-terminal and PC repeat domain. Next, to examine whether Spn-F and Ik2 can regulate proteasome, we performed Native PAGE to measure the amount of proteasome and in-gel peptidase assay for detecting proteasome activity in S2 cells. Overexpression of Spn-F, Spn-F phopho-mutants, Ik2 and Ik2 kinase-dead mutants respectively did not affect the amount of proteasome complexes (26S and 19S) and 26S activity in S2 cells, neither did knockdown of Spn-F or Ik2. Furthermore, double knockdown of Spn-F and Ik2 did not have synergistic effects. Taken together, our studies implicated that proteasome may not be the downstream effector of Spn-F and Ik2 in dendrite pruning.