以蛋白質功能域間作用和微小RNA兩部分討論 A型流感病毒和人類的相互作用

The Interactions of Influenza A Virus and Human for Two Parts: Domain-domain Interactions and MiRNAs.

張馨文(Hsin-Wen Chang)

指導教授 ： 唐傳義

國立清華大學/電機資訊學院/資訊系統與應用研究所/碩士(2012年)

https://doi.org/10.6843/NTHU.2012.00043

摘要

A型流感病毒( Influenza A virus )屬於RNA病毒，遺傳物質具有較高變異性，因而其中亞型H1N1曾經引發多起世界性流感大流行。當人體受流感病毒感染後，引發急性呼吸道與肺部疾病外，也常引發其他併發症，嚴重者甚至死亡。 A型流感至今仍對人類的健康造成威脅。然而，流感病毒感染人體的機制目前仍有許多未知的部分，故本論文以流感病毒及人類宿主間交互作用做為研究主題。為瞭解流感病毒差異性與宿主細胞物質之間的關係，本研究取用致病性差異較大的三株流感病毒，分別為三株1918, 2007, 2009的季節流感病毒做為研究材料。 本論文提供二個構面來研究A型(H1N1)流感病毒，在第一構面中，針對流感病毒功能作用域提出預測影響人體蛋白的研究流程，過程中結合現存生物資料庫 (Pfam, DOMINE, KEGG) 的相關資料，建構出十個流感病毒蛋白功能作用域的關係，發現彼此間具有重疊的影響性。三株流感病毒對所影響人體的蛋白質建構出功能作用域上的路徑，說明病毒對人體蛋白質的影響過程。流感病毒功能作用域推測對人體蛋白質的影響，其預測結果與文獻資料做參照。研究結果發現雖然三株流感病毒存在遺傳變異性，所影響的人體蛋白質類似。此結果說明三株流感病毒須透過影響人體特定的蛋白質族群，便會產生致病的現象。 隨著電腦計算技術的進步，許多生物資訊軟體逐漸應用於輔助實驗預測。人體內小分子RNA(miRNA)調控體內許多作用，此外，也有文獻顯示人體內小分子RNA會對流感病毒產生影響。 在第二構面中，利用生物資訊軟體 (miRanda) 預測三株不同毒性的流感病毒受人體小分子RNA的影響，建立預測流感病毒與人體小分子RNA交互作用的研究流程。過程中，參考人體內小分子RNA在組織內的表現量，並以主要發病組織內的小分子RNA作研究重心。研究結果發現，發病組織內有共同且表現量高的小分子RNA，這些小分子RNA存在著與流感病毒基因體結合的預測結果，但三株流感病毒結合位置有些許差異。所提出的小分子RNA預測(miR-143, miR-145, miR-150 etc.)也有文獻中提到會對免疫細胞的調控造成影響。此部份研究結果顯示，所挑選小分子RNA預期與流感病毒感染過程中有一定程度的關係。 透過流感病毒及人類宿主功能區域之間的交互作用，以及流感病毒與人類小分子RNA交互作用的兩種層面。推論流感病毒對於人體宿主的影響機制是多方面的，預測感染流感病毒對人體宿主所產生的變化，可以提供相關病毒流行病學更深入的了解。由兩構面研究流程的預測結果，未來可提供給相關研究團隊進行生命科學方面的實驗驗證。

關鍵字

A型流感病毒 ； 資料探勘 ； 生物資訊數據庫 ； 功能區域之間的交互作用 ； 小分子RNA ； 流行病學

並列摘要

Influenza A virus belonged to RNA viruses that genetic materials were high variability. One of the subtypes, H1N1 has caused several world-wide and seasonal influenza pandemics. After influenza virus infected human, the infection caused acute respiratory and lung diseases. The infection often led to other complications, and even severe death cases. Influenza A viruses are still threats of human health. However, the mechanism of influenza A virus infects human is still unclear.
In this thesis, the interactions between influenza virus and human host were research topics. In order to understand the relationships between the different influenza virus and host cell materials (such as proteins, and miRNAs), the research selected the three strains influenza virus with pathogenic differences. The three 1918, 2007 and 2009 seasonal influenza virus strains were respectively research materials. This thesis provided two dimensions to study the influenza A (H1N1) virus.
In the first dimension, the functional domains of the influenza virus were used to predict human affected proteins. In the process, domain interactions for ten influenza virus proteins were constructed out with the combination of existing biological databases (Pfam, DOMINE, KEGG etc.). This result was found the three influenza viruses affected the overlapping human proteins. The domain-domain path constructions were from three influenza virus to affected human proteins, and explained the affected process of the virus on the human proteins.
Some human proteins were speculated that affected by influenza, and the predictions were referable with some literatures. Although the three strains influenza virus had genetic variability with each other, the impacts of the human proteins revealed that they were similar. These results indicated that the three strains influenza virus went through the influence of human-specific protein populations, and caused pathogenic phenomena.
With the advances of computing technology, bioinformatics tools gradually applied to the predicted experiment auxiliaries. The human small molecules, miRNAs, regulated many effect roles in vivo. In addition, some literatures showed that miRNAs had impacts on influenza virus.
In the second dimension, the bioinformatics software (miRanda) was in use to predict human miRNAs targeted affairs on three different toxicity influenza viruses, and built the researching process about influenza virus and human miRNAs interactions. In this process, the human miRNAs expression levels within the related organizations were consulted, particularly in major incidence of infected organizations.

This study found that the infected tissues had common and high expression miRNA molecules which also had some targeted results with influenza viral genomes. However, the binding sites differed with the other strains. The predicted miR-143, miR-145, and miR-150 etc. were mentioned that had regulations of the immune cells in recorded literatures. The results in this part showed the selected miRNAs were expected to be in certain relationships with influenza virus.
Through two aspects of domain-domain interactions and the miRNA-genome relations between human and influenza virus, these interactions inferred the mechanisms are manifold and associated with both.
Some forecasts were associated with virus epidemiology, and provided more in-depth knowledge. The current predictions provided further experimental verification of biological related research team.

並列關鍵字

influenza A virus ； data mining ； bioinformatics database ； domain-domain interaction ； miRNAs ； epidemiology

參考文獻

[1] Palese, P. and Shaw, M. L. in Fields Virology 5th edn, Vol. 2 (eds D. M. Knipe & P. M.Howley) Ch. 47, 1647–1689 (Lippincott Williams & Wilkins, 2007).

[2] Tumpey, TM., Basler, CF., Aguilar, PV., et al. (2005) Characterization of the reconstructed 1918 Spanish influenza pandemic virus. Science, 310:77–80.

[3] Samuel, C. E. (2006) Virus-host interaction minireview series: human immunodeficiency virus, hepatitis C virus, and influenza virus. J Biol Chem, 281, 8305–8307.

[4] Zhirnov, OP., Klenk, HD. (1997) Histones as a target for influenza virus matrix protein M1. Virology, 235: 302–310.

[5] Nagata, K,. Kawaguchi, A., and Naito, T. (2008) Host factors for replication and transcription of the influenza virus genome. Rev. Med. Virol, 18(4): 247-260.