探討南台灣第二型登革病毒兩波連續流行中不同病毒族群替代、持續傳播及動態變異與流行趨變的關係

Abstract

登革病毒是一蟲媒病毒，會引發感染者自無症狀、發燒等病毒症候群(viral syndrome)至輕微的登革熱(dengue fever, DF)與嚴重的出血、休克(dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS)與死亡，影響全球公共衛生甚鉅。南台灣每年的登革熱流行多起因於不同血清型登革病毒自他國境外傳入。2001-2003年的登革熱跨年流行是由第二型登革病毒(DENV-2)所引起，造成兩波流行[各有217位登革熱與10位登革出血熱及5311位登革熱與252位登革出血熱(含21位死亡)]，是台灣自1987至該年的最大與最嚴重疫情。為何這波疫情如此大而趨嚴重?在大規模的跨冬疫情中的病毒動態變化、維持傳播與流行病學特徵趨變是一重要問題。然而，2001年跨越冬季現象的病毒機制、提高流行幅度和臨床嚴重度的因素尚不清楚。因此，本研究提出一假說，認為在同一流行中的不同第二型登革病毒分離株的流行潛力(epidemic potential)有所差異，且此與其宿主內的病毒變異分群之差異及其在人群中的傳播力與地理擴散度有相關性，因此研究目的有四: (一)藉由病毒親源關係與胺基酸分析，了解在特定位點具特殊胺基酸的不同第二型登革病毒次變異群(viral variants)在此連續二年流行中的動態演化及變異；(二)利用高通量次世代定序(next generation sequencing, NGS)分析，追蹤新病毒群(emerging newly dominant strain)的消長，並探察宿主內(Intra-host)與宿主間(Inter-host)的病毒變異分群之動態變化和差異；(三)探討病毒跨冬傳播的機制，尤其渡冬的病毒群之跨越兩波流行之角色、演化率、傳播率(R) 和在宿主體內的病毒變異數；及(四)探究三病毒次群(viral subpopulations)的病毒特性及表徵上隨著整體流行過程的改變。此研究是首度整合病毒與其抗體、臨床表徵及流行病學，徹底探察在流行時空變化中的病毒與流行趨變。 在目的(一)與(二)，首先分析自2001-03年收集的104位登革病患血漿中抽取的第二型登革病毒序列，據蒙地卡羅馬可夫算法 (Markov Chain Monte Carlo，MCMC）分析病毒套膜(E)蛋白與病毒多蛋白的整段基因(polyprotein open reading frame, ORF) 和保守胺基酸殘基推斷的親緣關係，鑑定三病毒群組(Ia，Ib和II組)。親緣關係分析顯示此三組病毒的共同祖先均與2001年菲律賓分離株在親源樹中群聚一起(極為相近)。此外，針對病毒E蛋白基因進行的高通量次世代定序，分析77位登革病患血漿中的病毒變異分群之基因序列，確認了2001年開始流行的病毒，至2002年大流行以及到2003年整體的兩波大流行具有共同的病毒祖先。 在目的(三)與(四)上，藉由病毒分離率得知Ia組病毒群在第一時期的2001年（75.8％，n = 29）占主導地位，但至第二時期的2002年2-9月（85.19 %, n=54）和第三時期的2002年的10-12月（95.24%, n = 21）改為II組病毒群主導，而Ib組是在整個流行中扮演維持傳播的角色，在此三時期各為24.14％至14.81％及4.76%，這意味著Ib有過冬的延續流行作用。藉由MCMC推估有效再生數（Effective Reproductive Number，簡稱R），結果顯示Ib組病毒群在2001-2003年期間較其他兩組病毒群的傳播力為低，且其R值在2002年初上升（峰值接近1.0），表示其雖在冬季持續傳播，但傳播力遠低於Ia及 II兩病毒群組(R的峰值分別為3.0及2.0左右)；另可由Ib群組與較Ia和II群組病毒有較低的核苷酸取代率，且又具有較低的宿主內變異之兩綜合結果相一致而再度證實登革熱病毒Ib群組在流行期間保持低度傳播，並維持越冬角色；但有趣的是雖此三病毒群組感染的患者體內血中之病毒量相似；但地理分佈整合分析，發現凡登革病人體內具較高病毒變異數者易集中於空間流行病學顯示的病例聚集區，而在整個流行期間扮演維持病毒傳播角色的Ib群組病毒卻分布於病例聚集處之外圍區。更進一步比較自感染性克隆所產生的三組病毒株在白線斑蚊細胞(C6/36)、非洲綠猴腎細胞(Vero)及乳倉鼠腎細胞(BHK-21)的複製能力不盡相同，尤其Ib及II兩組病毒的複製力在病毒生長曲線的前期均較Ia組快速，第一波病毒顆粒產生的時間也較早；此顯示隨著流行時間拉長，Ib與II病毒變異在流行中已增其複製的能力。 本篇論文研究兩波連續的登革病毒跨冬流行，明顯看出登革病毒在流行中的動態變異，尤其在同一地流行過長而造成許多病例聚集的地區，在流行過程中更易篩選出不同性徵的病毒。未來公共衛生監測應採NGS法而更關注低度傳播與易趨變的登革病毒次群轉為優勢，以減少隨後流行致病毒傳播力提高而增流行的嚴重度。未來研究方向可進行此三組病毒群組在蚊子及動物模型中的傳播及病毒性狀探究，尤其病毒變異分群在人、蚊不同宿主與特殊流行病學條件下自然傳播中的變異競爭及協同作用之各角色為何。

Human infection with dengue virus (DENV), an arbovirus, results in a wide range of clinical spectrum from asymptomatic, viral syndrome to mild dengue fever (DF) and severe hemorrhage, shock [dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS)], and death, posing a global health threat. Annual outbreaks of dengue in southern Taiwan have mostly caused by different serotypes of dengue viruses (DENVs) imported from other countries. The 2001-2003 epidemic caused by serotype 2 (DENV-2) resulting in two waves [217 DF and 10 DHF cases and 5311 DF and 252 DHF cases (including 21 deaths) in the 1st and 2nd wave, respectively] had been the largest in Taiwan till 2003 since 1987. What're factors involved in this DENV-2 epidemic that would cause to increase epidemic scale and severity (with more DHF % and deaths) remained unclear, particularly the dynamics in virus change during the over-wintering period and the mechanism maintaining viral transmission that correlate well with changing epidemiological characteristics. We thus proposed a hypothesis that epidemic potentials of different DENV-2 isolates varied through the epidemic and would correlate with their differences in viral quasispecies within the host, transmissibility in human population and spatial dispersion patterns. Therefore, this study had four specific aims: (1) to understand changes of DENV-2 isolated from the two waves during 2001-2003 by phylogenetic and amino acid analyses and monitoring dynamic changes of different viral subpopulations with unique signatures of amino acids at specific sites through the epidemic process; (2) to search for viral sub-populations and to monitor the rise and fall of emerging newly dominant strain from an individual (Intra-host) to human population level (Inter-host) (i.e. dynamic changing through the epidemic by using next generation sequencing; (3) to characterize the maintenance role of transition strain during two waves of epidemics with evolution rate, transmission rate (reproductive number, R), and variation numbers within host; and (4) to analyze the virological and phenotypic differences of the three specific viral sub-population of DENV-2 viral variants through the entire epidemic process. This study is the first attempt to fully understand temporal and spatial changes in viral and epidemiological characteristics by integrating dynamic changes of DENV, anti-DENV antibody, clinical manifestations and tempo-spatial changes in epidemiological attributes. To study aim #1 and #2, using the 104 DENV-2 sequences sampled from dengue patients’ plasma samples collected during 2001-2003, three virus groups, group Ia, Ib and II, were identified according to the phylogenetic analyses inferred by the Bayesian Markov Chain Monte Carlo (MCMC) method on envelope gene and polyprotein open reading frame (ORF) and conserved amino acid residues. Phylogenetic analyses show that a common ancestor of these groups clustered with the 2001 Philippine DENV-2 isolate. In addition, deep sequencing in 77 plasma samples focusing on viral E gene confirms the ancestral role of the viruses in 2001 epidemic wave contributing to the larger epidemic beginning in 2002 to 2003. To investigate the aim #3 and #4, the DENV-2 group Ia dominated in the 1st period in 2001 (75.8%, n=29) but replaced by group II dominated during the 2nd periods in Feb. –Sept., 2002 (85.19%, n= 54) and the 3rd periods in Oct. to Dec., 2002 (95.24%, n=21). Whereas Ib maintained 24.14% to 14.81%, and 4.76% from the 1st to the 2nd and 3rd periods, respectively, implying the important role of Ib in over-wintering. Phylogenetic tree integrated with effective reproductive number (R) analyses show the Ib viruses had lower transmissibility throughout 2001-2003 than the other two DENV-2 groups and sustained viral transmission across the winter with its R resurged (peaking close to 1.0) in early 2002, much lower than the peaked R values (around 3.0 for Ia and 2.0 for II). Additionally, the DENV-2 group Ib presented a lower substitution rate and lower intra-host variants than the group Ia and II, corresponding the maintaining role of group Ib in overwint4ering. Although group Ia, Ib and II shared similar viral titers in patients’ plasma, they showed different geographical distributions (Ia, and II with high variants located in the case clusters, whereas group Ib located in the outer proximity of case clusters throughout the period). Furthermore, using infectious clone to construct similarly-pure clone of the Ia, Ib and II, the replication of three viruses groups was significantly different in both C6/36, Vero, and BHK-21 cells groups Ib and II had higher titer in the early period of viron production and earlier time to burst the first virions, indicating increasing of Ib and II in viral replication efficiency through the extending of epidemic waves. In summary, the dynamics of changes of DENV-2, particularly in longer epidemic waves and higher dengue clusters help to select the viral variants with better adaptation. Future effort needs to use NGS to search for DEMV with low transmissibility and increasing dominance to minimize epidemic severity. Future research effort needs to use mosquito and animal models to study the selection mechanisms of these groups Ia, Ib and III and find out the roles of competition and complementation in nature.