A型H1N1流行性感冒病毒紅血球凝集素之基因與抗原分析並與H3N2病毒之比較

Abstract

A型流行性感冒是一種非常重要的人類感染疾病，它對全世界的公共衛生造成極大的威脅。H1N1為A型流感中的一種亞型，在二十世紀曾發生過全球性大流行並且造成約兩千萬人的死亡。最近源自於豬的新型H1N1病毒開始感染人類並造成2009年的全世界性大流行。紅血球凝集素(Hemagglutinin，HA)為流感病毒表面的抗原性醣蛋白，在流感病毒感染或接種疫苗期間紅血球凝集素會被抗體所中和。紅血球凝集素上之突變累積會造成抗原性漂移(Antigenic drift)的發生，此時疫苗往往需要重新設計來對於下一波疫情提供足夠的保護力。然而直到2009年為止大部分對於紅血球凝集素的研究主要針對H3N2亞型，目前對於H1N1亞型紅血球凝集素的了解仍然不足。針對H1N1亞型的紅血球凝集素進行研究對於公共衛生與疫苗發展是一個重要且有高度急迫性的議題。 在本論文中，我們針對H1N1亞型之紅血球凝集素進行基因與抗原分析，並且將結果與H3N2亞型做比較。在基因層次，我們收集了一千五百二十五株H1N1之紅血球凝集素序列，並且利用亂度值(Shannon entropy)分析每個胺基酸位置之改變程度。在抗原層次，我們從近40年的流病週報(Weekly Epidemiological Record)以及相關文獻收集了二百筆HIN1之血球凝集抑制試驗值(hemagglutination inhibition，HI)，並且利用統計方法量度每個胺基酸位置之抗原性變化大小。最後並且利用決策樹方法(Decision tree)(C4.5)來預測H1N1病毒之抗原性漂變株(antigenic variants)。 透過在H1N1上進行基因與抗原分析，我們在紅血球凝集素上挑選了三十個具有重要性的胺基酸位置，觀察這些位置中，有二十六個位於表面，此外有九個位置位於抗原決定位(Epitope)上。根據基因與抗原分析的結果，本論文發現H1N1亞型之重要區域主要分為二塊，包括鄰近受體嵌合區(Receptor binding site)以及遠離受體嵌合區之抗原決定位(epitope)。相較之下，H3N2亞型之重要區域大部分皆分佈於鄰近受體嵌合區。同時透過比較H1N1與H3N2亞型，發現了許多在H3N2亞型抗原決定位上之胺基酸位置也很有可能是H1N1亞型的抗原決定位。除此之外，決策樹方法建立之模型可達到85%的預測率。由以上的結果，顯示我們的方法具有穩健之特性並且有助於了解H1N1病毒之基因與抗原性演化，並且對於設計疫苗有極大幫助。

Influenza A virus causes significant morbidity and mortality in humans. H1N1 is one of the current circulating influenza A subtypes in human. The H1N1 pandemic occurred in the early 20th century and resulted in approximately 20 million deaths in the world. Recently, the emerged swine-origin H1N1 virus has infected human population and cause the 2009 influenza pandemic. Hemagglutinin (HA), which is an antigenic glycoprotein on the surface of influenza virus, is neutralized by antibodies during infection or vaccination. Accumulation of mutations on HA can lead to antigenic drift. The emergence and spread of antigenic variants often requires a new vaccine strain to be selected before coming epidemic. Most of studies on HA focused on the H3N2 subtype. However, the genetic evolution and antigenic evolution of the HA is poor understood for subtype A (H1N1). To study the genetic and antigenic evolution of subtype A (H1N1) is an emergent issue for public health and vaccine development. In this thesis, we performed the genetic and antigenic analysis to the HA of A (H1N1) viruses. In the sequence level, we collected 1525 HA sequences and used Shannon entropy to quantify the genetic diversity of each amino acid. In the vaccine efficacy level, we collected 202 pairs of HI assays from weekly epidemiological record (WER) and publications in last 40 years. Based on the collected Hemagglutination Inhibition (HI) assays, we applied a statistical index to quantify the antigenic score of each amino acid on HA. Finally, a decision tree tool (C4.5) was used to build a model for predicting the antigenic variants of H1N1 virus. We select 30 critical positions of H1N1 hemagglutinin by the genetic and antigenic analysis. There are 26 positions on the surface of the HA and 9 positions on the H1N1 epitopes. Based on the genetic and antigenic analysis on HA, we found that there are two sites with both high genetic diversity and antigenic score in A (H1N1) virus. These two sites include one site around the receptor binding site and the other antigenic site about 45 Å distant from receptor binding site. In contrast, there is only one site, which is around the receptor binding site, have high genetic diversity and high antigenic in A (H3N2) virus. By comparing the HA of two subtypes of influenza A virus, we found that some amino acid positions locating on the antigenic sites of influenza A (H3N2) virus are potential epitope residues for influenza A (H1N1) virus. In addition, the accuracy of our model for predicting antigenic variants was 85% by using HA sequences as input. We believe that our methods are useful for the vaccine development and understanding the genetic and antigenic evolution of influenza A (H1N1) virus.