研發實場化學-生物串聯除硫化氫系統及次世代定序法研究氧化亞鐵硫桿菌突變株之基因體及轉錄體

Abstract

Acidithiobacillus ferrooxidans 是一種嗜酸性自營菌，可將亞鐵離子氧化為鐵離子以獲得電子及能量。本研究使用鐵離子將硫化氫氧化，並以此菌作為鐵離子的生物再生系統，串聯反應器可藉由鐵/亞鐵離子之循環再生來進行養豬場沼氣除硫之應用。文獻中及本實驗室原有的「載體式」串聯除硫反應器在高硫化氫負荷量下，會有硫沈澱物堵塞等缺點。因此，本研究新設計了「噴灑式」串聯除硫系統，能夠可靠與長期使用於沼氣除硫用途。在第一代「噴灑式」系統的實驗室規模操作中，我們測試了不同硫化氫進流速度以及液體噴灑壓力對於硫化氫去除率之影響。此測試證明該系統能在高硫化氫負荷量下 (~ 1,300 g-S m-3 h-1) 具有高去除率 (RE > 90%)，且不會有硫沈澱物堵塞。「噴灑式」系統經放大後以野生株 CP9 進行 356 天之實場除硫測試，項目包含了突增負荷及停工試驗。在氣體滯留時間 216 秒的條件下，可達 94.8% 除硫效率及 64 g-S m-3 h-1 的負荷移除能力。此系統的快速復原能力證實了經過突增負荷及停工試驗後，不會對系統能力造成長期損害。在第二代「噴灑式」串聯除硫系統中，我們改良了化學槽及儲存槽的連接方式，提升最佳反應效能並快速有效將硫沈澱物移除。在實驗室規模操作中，我們測試了噴灑液滴及化學槽體積對於硫化氫去除率之影響。在第二代系統經放大後以高效能突變株 W3進行 500 天之實場除硫測試，最佳操作參數為氣體滯流時間 73 秒時，系統有 90% 除硫效率及 302 g-S m-3 h-1的負荷移除能力。除硫沼氣經 30 kW 發電機發電測試，以含 70% 甲烷之沼氣在 220 LPM 流速下進入發電機可得最大輸出功率 27.6 kW，此條件下的熱能使用效率達 26.4%。第二代反應器菌株 W3 之最大鐵氧化效率約為 CP9 之 2 倍，藉由化學槽體積及連接方式改良，使氣體滯留時間為原本三分之一即可處理約 5 倍的進流負荷。 再者，我們也分析了 A. ferrooxidans 野生株 CP9 及突變株 W3 之差異，本研究以次世代定序系統分析 CP9 及 W3 之基因體序列及不同培養條件下的全基因表現。CP9 及 W3基因體定序後，兩者皆有 88.4% (3309 個基因中的 2829 個基因) 的序列可比對到已知基因體的 A. ferrooxidans ATCC 23270。此外，在 W3 基因體上找到 310 個不同於 CP9 基因體之鹼基對，其中 288 個位於密碼子 (coding region) 上。以基因功能 (GO term) 分析這些突變基因的功能種類，發現與全基因分析後獲得的群組種類類似，此現象符合 W3 為隨機突變下獲得之突變株。在轉錄體分析方面，六個樣本分別有 80.3% - 81.9% 的資料可比對到已知的基因。此外，本研究首次以 NGS 的方法進行 A. ferrooxidans在不同能量來源 (硫及鐵) 下的大規模基因表現分析，對此菌在硫代謝與電子傳遞系統中提供新證據。研究結果中我們觀察到 sre 轉錄組的基因會產生將元素硫還原為硫化物的酵素 sulfur reductase，並在硫培養條件下有 2–4 倍增量表現。在鐵培養條件下，A. ferrooxidans 細胞內的硫化物來源為亞硫酸物，實驗觀察到 cysJ 及 cysI 在此條件下有約 8 倍增量表現，其酵素 sulfite reductase 可能是此催化反應的關鍵蛋白。除此之外，由結果可分析出共有十個基因，在 DNA層次為突變基因，也分別在四個條件下具有增量表現。其中的 glcF 較為特殊，其表現出的 glycolate oxidase 可將 A. ferrooxidans 在固碳作用中產生的毒性副產物 glycolate，進一步代謝為無毒的 glyoxylate。此基因只在 W3 的 lag phase 中相對於 log phase 有約 3 倍增量表現，但在 CP9 中，glcF於此兩 phases 並無明顯表現差異。本研究同時以 qPCR 確認 glcF 在 W3 菌株中有此現象，發現 lag phase 表現量為 log phase 之 6 倍，進一步證實此現象，由此我們推論在 lag phase 具 8 倍以上快速生長能力的 W3，可能是因為此基因引起之增強解毒能力的結果。

The acidophilic and autotrophic Acidithiobacillus ferrooxidans oxidizes ferrous iron into ferric iron to obtain the electrons and reducing power. The ferric iron was used to be an oxidant for H2S elimination and A. ferrooxidans was immobilized in the bioreactor for the ferric iron regeneration. This combined and renewable system was applied for the biogas purification in this study. The former “carrier-based” reactors are unadvisable as H2S elimination system because they are prone to sulfur blockage problems under heavy loading operations. Therefore, the “scrubbing-style” reactor was designed for robust biogas purification under pilot-scale long-term operation. In the type I “scrubbing-style” system of laboratory scale study, various H2S inlet flow rates and spray pressures were used to evaluate the H2S removal efficiency (RE) in the chemical reactor. The high H2S RE (> 90%) demonstrated that this scrubbing-style system performed well under heavy loading (~ 1,300 g-S m-3 h-1) without severe sulfur blockage. In the scaled-up application, the type I system using A. ferrooxidans CP9 was operated for consecutive 356 days for biogas purification, including shock loading and shutdown tests. The system achieved an average RE of 94.8% with an elimination capacity (EC) value of 64 g-S m-3 h-1 under EBRT 216 s. In addition, the system recovered quickly in the shock loading and shutdown tests without permanent damage; however, the solid sulfur still caused blockage after the long-term operation. In the type II “scrubbing-style” system, the design of the connection between the chemical absorbers and the storage tank was improved to elevate the removal efficiency and quickly remove the sulfur solid. In laboratory scale study, the effects of droplet size and column size on the optimal H2S removal were characterized. In the scaled-up application, the type II system using the high growth rate strain W3 was operated for 500 consecutive days for biogas purification. The optimal conditions were an average RE of 90% with an EC of 302 g-S m-3 h-1 under EBRT 73 s. In the power generation test with 30 kW biogas generation, the maximum power output was 27.6 kW and the maximum thermal efficiency was 26.4% at a biogas supply rate of 220 litter per minute (LPM) using 70% CH4. The W3 strain in the type II system showed approximately 100% higher maximum iron oxidation rate than the CP9 in the type I system. Furthermore, only 34% EBRT was required for the type II system to deal with the 5 folds H2S loading higher than the type I system. To further characterize the differences between A. ferrooxidans CH9 and W3, their genome and transcriptome were subjected to next-generation sequencing (NGS) analysis. The results show 88.4% of the sequenced genomes (2829 of 3309 genes) from CP9 and W3 were assembled by mapping to the reference ATCC 23270 genome. Moreover, 288 mutated paired bases were located on the 79 coding sequences (CDS), whereas 22 paired bases were located on the non-coding region of the W3 genome. The gene ontology (GO term) analysis showed similar hit term distributions for both mutant genes and total genes, which indicates that the mutation rate in each specific class is size-related and randomly mutated. In the NGS transcriptomic analysis, the total qualified paired-end sequencing reads from six samples mapped to the reference genome ranged from 80.3% to 81.9%. Also, this study is the first time to apply NGS in the differential expressed gene analysis of a different energy source for A. ferrooxidans. Collection of sulfur metabolism–related genes from the NGS data provided new evidence of candidate genes that encode key enzymes involved in unidentified pathways. For example, the sreABCD protein encoded in the sre operon was highly expressed under sulfur-growth conditions (fold change (FC) = 2–4), which was considered responsible for reducing sulfur into sulfide. Moreover, cysJ and cysI are highly expressed under iron-growth conditions (FC = 8). Thus, these genes encode proteins that catalyze the reduction of sulfite into sulfide and could involve in the only pathway for sulfide production under such conditions. Furthermore, 10 genes were found in the mutant W3 with differentially expressed under four various conditions. In particular, glcF was highly expressed (FC = 2.7) during the lag phase rather than in the log phase of the mutant W3; however, this gene was minimally expressed during both the lag phase and the log in the CP9 strain. The fold change was also examined by quantitative polymerase chain reaction (qPCR) and shows the evidence that glcF in W3 was highly expressed (FC = 6.1) in the lag phase than in the log phase. The glycolate oxidase encoded by the glc operon could catalyze the conversion of glycolate into innocuous glyoxylate in A. ferrooxidans carbon metabolism. Therefore, highly efficient detoxification could be account for the 8.5 folds higher growth rate during the lag phase of the mutant W3 than that of the CP9.