基因合成精胺酸脫亞胺酶對一氧化氮合成酶媒介之神經毒性影響

Abstract

一氧化氮（nitric oxide, NO）在人體內不但涉及許多生理功能，亦與神經疾病相關，例如：帕金森氏症、阿茲海默症、及腦缺氧所致的神經損傷等，因此，適當地調控一氧化氮之重要性不容小覷。基因合成精胺酸脫亞胺酶 (recombinant arginine deiminase, rADI) 因催化精胺酸 (L-arginine) 分解為瓜胺酸 (L-citrulline)的反應，而消耗一氧化氮合成酶的唯一受質。本研究分別活化誘導型一氧化氮合成酶(inducible nitric oxide synthase, iNOS)及神經型一氧化氮合成酶(neuronal nitric oxide synthase, nNOS)，以探討rADI對於一氧化氮合成酶所媒介的神經毒性之影響。 於敝實驗室之前所建立的小神經膠質細胞和神經細胞的共同培養系統中，以2 μg/mL 的脂多醣 (lipopolysaccharide, LPS) 及1 ng/mL的γ-干擾素(interferon-γ, IFN-γ)誘導iNOS，以OX-42抗體區別該共同培養系統中的小神經膠質細胞和神經細胞，以annexin V及7-AAD試劑偵測凋亡及壞死之細胞，並以Griess method測量NO產生量。結果發現rADI可降低LPS/IFN-γ所誘導之NO產生（由88.8±3.6 μM 減為7.2±0.4 μM），將凋亡的神經細胞比率由30.5±2.2% 減為12.1±3.0%，壞死的神經細胞比率由11.3±1.0% 降至5.6±1.2%，並引起45.6±1.3%小神經膠質細胞凋亡。而rADI對神經的保護作用，不只是在與LPS/IFN-γ同時給予時能達到效果，在LPS/IFN-γ之後8小時內給予rADI，亦可達神經保護之效。 此外，我們利用NMDA於神經細胞中活化nNOS，以螢光方法測量NO產生量。即使未加入NMDA，就已有些基礎NO產生；在精胺酸存在下，NMDA誘導了39.71±3.9%之NO產生；而vinyl L-NIO（nNOS選擇性抑制劑）可抑制NMDA所誘導之NO產生，1400W（iNOS選擇性抑制劑）則無此效果；我們也發現NMDA在我們的系統中不會造成神經毒性。為瞭解精胺酸缺乏對於受NMDA刺激之神經細胞有無影響，我們將細胞分別置於含有精胺酸之緩衝液、缺乏精胺酸之緩衝液、含有精胺酸但事先加入rADI之緩衝液中，以NMDA誘導，結果發現只在含有精胺酸之緩衝液中，NMDA能誘導出NO。無論是在哪種緩衝液中、無論有否加入NMDA，細胞的粒線體膜電位都沒有差異，但是當我們將細胞培養於事先加入rADI的培養液中，以NMDA刺激24小時，細胞的存活率稍微減少至83.34±3.5%；若我們補充精胺酸至培養液中，細胞的存活率沒有減少。 在缺氧小鼠的實驗中，我們於再灌流後的2小時給予不同濃度的rADI，發現20 mU/mouse 的rADI明顯減少大腦皮質的損傷區域；然而，當我們增加樣本數，卻未發現神經保護的效果。rADI的安定性、劑量及給藥時間皆可能為影響rADI於缺氧小鼠之實驗結果的重要因素。 綜合上述，本研究之結論為：iNOS 所產生的NO會造成神經毒性，而rADI可保護神經細胞免於LPS/IFN-γ所造成之神經毒性；相反地，nNOS所產生的NO在我們的實驗系統中並不會對神經細胞造成傷害，短暫地於該系統中給予rADI對神經細胞並無影響，而長時間給予rADI則會減少神經細胞的存活率。

Nitric oxide (NO) is not only involved in a wide range of physiological functions, but also associated with neurodegenerative disorders, such as Parkinson’s, Alzheimer’s diseases, and cerebral ischemia. Therefore, modulation of NO is of vital importance. Recombinant arginine deiminase (rADI), which catalyzes the conversion of L-arginine to L-citrulline and ammonia, depletes L-arginine, the sole substrate of nitric oxide synthase (NOS). In this study, to investigate the effect of rADI on NOS mediated neurotoxicity, iNOS and nNOS were activated respectively. A neurons and microglia co-culture in which iNOS mediated neurotoxicity was previously established in our laboratory by treating 2 μg/mL lipopolysaccharide (LPS) and 1 ng/mL interferon-γ (IFN-γ) to BV2 (a murine microglial cell line) and SH-SY5Y (a human neuroblastoma cell line) co-culture. NO production was measured by the Greiss assay. OX-42 antibody distinguished microglia from neurons. Cell apoptosis and necrosis was analyzed by annexin V and 7-AAD simultaneous staining. rADI recovered LPS/IFN-γ induced neuronal apoptosis (from 30.5±2.2% to 12.1±3.0%) and necrosis (from 11.3±1.0% to 5.6±1.2%) with attenuated NO production (from 88.8±3.6 μM to 7.2±0.4 μM) and increased microglial apoptosis (from 13.0±1.0% to 45.6±1.3%). The neuroprotective effect could not only be obtained when treated at the same time with LPS/IFN-γ, but also post LPS/IFN-γ. nNOS was activated in SH-SY5Y by N-methyl-D-aspartic acid (NMDA), and NO was measured by fluorometric method. In the absence of NMDA, there was already basal constitutive NO production. In the presence of L-arginine, 1 mM NMDA induced 39.71±3.9% NO production, which could be abolished by vinyl L-NIO (a selective nNOS inhibitor), but not by 1400W (a selective iNOS inhibitor). However, neither brief (1 hr) nor prolonged (24 hr) NMDA exposure decreased neuronal mitochondrial membrane potential and viability. To understand the effect of L-arginine deprivation on NMDA stimulated neurons, the cells were treated with NMDA in L-arginine free and rADI pretreated L-arginine containing buffer, and we found NMDA induced NO only in L-arginine containing buffer. There was no difference on mitochondrial membrane potential between all treatment groups. When the cells were exposed to NMDA in rADI pretreated medium for 24 hr, the cell viability slightly decreased to 83.34±3.5%. The cell viability was recovered when L-arginine was replenished to the rADI-pretreated medium. Mice were suffered from middle cerebral artery occlusion (MCAO) to mimic transient ischemia. 20 mU rADI, which was shown to ameliorate infarct volume during dose titration, was given to the mice at 2 hr after reperfusion. However, no substantial neuroprotective effect of rADI was obtained as we enlarged the sample size. Administration timepoint, dose, as well as stability of rADI may be key factors that influence the result of ischemic mice. In summary, NO produced via iNOS was cytotoxic to neurons, and rADI protected neurons from LPS/IFN-γ induced neurotoxicity in the co-culture. In contrast, NO produced via nNOS was not toxic to neurons in our in vitro model. Brief exposure of rADI had no effect on NMDA stimulated neuronal mitochondrial membrane potential, but prolonged L-arginine deprivation by rADI reduced neuronal viability.