National Chiao Tung University Thesis for the Degree of Master
國立交通大學碩士學位論文
Roles of the Two-component System Response
Regulator RcsB in Klebsiella pneumoniae CG43
雙分子系統反應調節因子 RcsB 在克雷白氏
肺炎桿菌 CG43 中所扮演的角色
分子醫學與生物工程研究所 碩士班
Student: Yenxi Tay (0157103) 學生：鄭燕曦
Advisor: Hwei-Ling Peng, PhD. 指導教授：彭慧玲博士
June, 2013
i
Acknowledgements
(謝誌)
感謝我的指導教授彭慧玲老師，每週老師都會安排每個成員單獨 跟老師討論，這個部分讓我收穫許多，除了可以練習自己去推導一些 分子調控途徑之外，還練習怎麼設計實驗來解決我們想要探討的問題。 我覺得這樣的練習讓我漸漸的被訓練成一個有能力在研究一個課題的 過程中，找出需要改進的地方，並且在遇到問題或是瓶頸的時候，我 也會以積極的態度去找出問題，進而有邏輯的去思考再去解決問題。 彭老師除了在我們做研究的當兒，是我們的老師，教導我們、幫助我 們在學習的路上讓我們在生科這個領域有所成長之外，很多時候，她 更是扮演家長的角色，老師時常就像我們在學校的母親一樣，時時刻 刻關心著我們的近況，希望我們在做研究的路上也要為自己的未來好 好做規劃，並且從學校畢業之後，也會是個有能力照顧好自己、成熟 的年輕人。能夠遇上這樣一位不可多得的好老師，心中充滿了無限感 激。 感謝清華大學張晃猷老師以及張老師實驗室的夥伴，在實驗上 給予諸多的建議及幫助。 更加感謝張老師和交通大學梁美智老師，在百忙之中撥冗擔任 我的口試委員，用心的修正我論文中的錯誤及指點我實驗中未曾想過 的盲點。 感謝實驗室的大家，因為有你們，這本論文才可以順利完成， 也因為你們，這兩年的生活才會如此多彩多姿，在每一個實驗卡關的ii 時候，我才可以在你們的鼓勵之下順利突破。感謝溫柔細心的靜柔學 姊，總是像我的姐姐一樣，不管在實驗上或是生活上一直給我鼓勵和 支持，更是教會我很多事情；感謝哲充學長，讓我們學會很多待人處 事的技巧；感謝我的好拍檔冠男，沒有我們如此精準的默契，我想我 們鐵定不能這麼順利的畢業且讓人稱羨；感謝亦師亦友的子祥，我們 倆總是會有不一樣的話題，我想我們三個臭皮匠的緣份一言難盡，短 短的幾天就讓我們造就了今天的生死之交；感謝細心又總是考慮周全 的偉豐，雖然你總是表現得默默的，但是你總是為大家付出很多，希 望以後可以看到你的植物王國；感謝活潑可愛的珍儀，雖然你的課很 多，我們碰面的時間不多，但是總覺得我們有時候會有莫名的默契， 很懂對方是怎麼樣的人，真的很開心很幸運可以認識你這個活潑大方 又很有愛心的好女生；感謝單純可愛的蕙瑜，雖然我時常以虧你為和 你溝通的方式，但我真的覺得你是個很有想法、很有原則的一個人， 要繼續保持這樣單純又善良的心，畢業後我們要繼續連絡喔，朋友； 感謝嬌小又能幹的俐君，每次都讓我覺得你又知性又很有智慧，緣份 讓我們在台灣相遇，只可惜我們相處的時間總是嫌少，我們一拍即合， 一直也覺得很幸運可以跟你成為朋友！keep in touch! 感謝文靜又聰 明的豐碩，雖然不知道你會不會看到我的感謝，但是真的很開心時常 有你這個學弟和我們一起討論實驗、互相切磋，一起出遊真的很開心； 感謝搞笑卻認真的家睿，雖然每次喜歡虧你在狀況外，但卻被你對於 實驗的認真心態很是感動，博班生涯也許時常會遇到瓶頸，但是一定 要勇敢的把它完成，加油！
iii 還有感謝吳東昆老師、廖光文老師、楊昀良老師、曾慶平老師、 黃兆祺老師以及各位老師實驗室的成員，無私的提供我支援，給予我 在碩士生涯的指點與實驗上的幫助。 最重要的是，要感謝我親愛的家人，因為你們的支持與鼓勵， 我才能有今天這個碩士學位。感謝媽媽在我成長的道路上，讓我可以 不斷的往自己的嚮往的方向，不管在甚麼領域都可以自由發揮。感謝 弟弟，在我隻身在外求學的時候，擔起照顧媽媽的責任，讓我這 5 年 來可以專注於課業。特別感謝鄭暠一家人，在論文定稿這段期間的照 顧，如果沒有你們就不會有這本我人生重要的著作。還要感謝鄭暠， 一直在身邊支持和陪伴我，讓我疲憊、無助的時候，仍然可以堅持著， 並全力衝刺論文和碩士學位的最後關鍵，直到終點。 最後，我也要感謝我自己，成功並且順利的以一年的時間完成 碩士學位，並且在過程中不斷的堅持、永不放棄，最後才有今天的自 己。畢業後的日子，一定還有更多的挑戰和值得期待的未來等著我去 面對、等著我去創造，我會繼續努力、一直提醒自己莫忘初中，直到 夢想實現！ 燕曦
iv
論文摘要
當細菌感受到外界環境壓力後，雙分子調控系統（Two-component system）的偵測激酶 RcsC 和 RcsD 將訊號傳遞給下游之反 應調控蛋白 RcsB，磷酸化該蛋白的第 56 個氨基酸天冬氨酸鹽 （Aspartate）並使 RcsB 活化。 過去研究顯示，RcsB 與 RcsA 形成雙 合體來調控細菌莢膜多醣體的基因表現。在肝膿瘍克雷白氏肺炎桿菌 分離株 CG43 中，rcsB 基因的缺失可導致莢膜多醣體的生合成、第三 型線毛的表現和抗酸能力明顯下降。本論文中，我們擬探討 RcsB 的 磷酸化是否會影響莢膜多醣體、第三型線毛和抗酸能力相關基因的表 現。本研究首先構築了兩個定點突變的 RcsB，分別為 RcsB-D56A 和 D56E。D56A 模擬無法接受磷酸根的 RcsB；而 RcsB-D56E 則是模擬持續磷酸化狀態的 RcsB。CG43S3ΔrcsB 回補 pRK415-RcsB-D56A 和 pRK415-RcsB 結果相近，均提高莢膜多醣體的 生合成和細菌的抗酸能力。相對地，回補 pRK415-RcsB-D56E 則恢復 了第三型線毛的主要單元蛋白 MrkA 的表現量。此結果暗示著 RcsB 的磷酸化影響此蛋白和 DNA 或是其他蛋白質的結合，進而決定它的 調控目標。我們以西方點墨分析發現 RcsB-D56A 正向調控酸逆境伴 隨蛋白，而啟動子活性分析顯示其對抗酸相關的蛋白 YfdX 的為轉錄 層面的影響。同時，我們藉大腸桿菌來大量表現 RcsB 的 N 端並純化 之後免疫兔子後取得多株抗體，有趣的是，我們發現此多株抗體只能 偵測 RcsB 和 RcsB-D56A 卻無法辨識 RcsB-D56E，此暗示 RcsB-D56E 殘基改變可能影響其結構而致失去抗體辨識的抗原決定部位 (epitope)。v 不過，西方墨點法顯示 RcsB 可受酸誘導表現，而此抗體將可以用於 免疫共沉澱找出 RcsB 協同作用的蛋白質。這些結果暗示著 RcsB 的磷 酸化狀態對於克雷白氏肺炎桿菌 CG43 的各種致病方式的調控扮演著 非常重要的角色，並且可能會是其致病的關鍵點，往後 RcsB 的協同 作用蛋白的發現將可以幫助我們進一步去更瞭解 RcsB 的調控路徑。
vi
Abstract
The two-component system (TCS) response regulator RcsB is activated by phosphorylation at Asp56 upon receiving stress signals transferred from the sensor kinase RcsC and RcsD. RcsB has been demonstrated to be required for CPS (capsular polysaccharide) gene expression through heterodimer forming with RcsA. In Klebsiella pneumoniae CG43, a liver abscess isolate, deletion of rcsB reduced the CPS expression, type 3 fimbriae pilin MrkA production, and acid stress response. This study, investigates if RcsB phosphorylation plays a role in regulating CPS production, MrkA expression, and acid stress response. Two site-directed mutants RcsB-D56A which is unable to accept the signal transferred, and RcsB-D56E, a phosphorylation mimetic form of RcsB, have been generated. Comparing with CG43S3ΔrcsB[pRK415], CG43S3ΔrcsB[pRK415-RcsB-D56A] and CG43S3ΔrcsB[pRK415-RcsB] exhibit an increased the CPS production and acid survival rate. By contrast, CG43S3ΔrcsB[pRK415-RcsB-D56E] restored the MrkA production. These findings suggest that the phosphorylation status of RcsB affects its DNA binding or protein binding activity thereafter determines its regulation targets. Western blot analysis showed that RcsB-D56A positively regulates
vii
the acid stress chaperone YfdX expression and promoter activity assay suggested that the RcsB-mediated regulation is at the transcriptional level. In addition, the N-terminal region of RcsB was overexpressed in E. coli and the protein was purified to immunize rabbit for polyclonal antibody generation. The raised antibody could recognize RcsB and RcsB-D56A but not RcsB-D56E. This result implies that the amino acid D56E substitution of RcsB alters the protein conformation, and hence the loss of the specific epitope for the antibody. Western blot analysis also revealed that RcsB was acid inducible and co-immunoprecipitation analysis would be feasible for the identification of the RcsB interacting proteins. The findings indicated that the phosphorylation status of RcsB might be the key of the regulation of different virulence factors in K. pneumoniae CG43, and further identification of the RcsB-interacting protein will help toward understanding the regulatory pathway of RcsB.
viii
Table of Contents
Acknowledgements ... i
Abstract ... vi
List of Tables... xii
List of Figures ... xiii
Abbreviations ... xv
1. Introduction ... 1
1.1 Klebsiella pneumoniae ... 1
1.2 Infections of K. pneumoniae ... 1
1.3 Virulence Factors of Klebsiella pneumoniae ... 4
1.3.1 Capsular Polysaccharide (CPS) and lipopolysaccharide (LPS) .... 4
1.3.2 Adhesion Factors ... 5
1.3.3 Stress Responses ... 7
1.4 Two-Component System (TCS) ... 9
1.5 Regulator of Capsule Synthesis (Rcs) System ... 10
1.6 The Response Regulator -- RcsB and RcsB-dependent Regulation ... 12
1.6.1 Role of RcsB phosphorylation ... 12
1.6.2 RcsB interacting protein ... 13
1.7 Previous Studies - RcsB-dependent regulation in K. pneumoniae ... 16
1.8 Specific Aims ... 17
ix
2.1 Plasmids, primers, bacterial strains and growth conditions ... 20
2.2 Construction of the site-directed mutants RcsB ... 21
2.3 Measurement of bacterial growth ... 22
2.4 Extractions and Quantification of CPS ... 22
2.6 Biofilm Formation Assay ... 24
2.7 Yeast-cell Agglutination ... 25
2.8 Disc Diffusion Assay ... 26
2.9 Acid Stress Response ... 26
2.10 Measurement of Promoter Activity ... 27
2.11 RcsB N-terminal region antiserum preparation ... 29
2.12 Statistical Methods ... 30
3. Results ... 31
3.1 Roles of phosphorylation of RcsB in K. pneumoniae CG43 ... 31
3.1.1 Generation of specific point-mutation on RcsB ... 31
3.1.2 RcsB D56A as well as RcsB retarded the growth of K. pneumoniae CG43 ∆rcsB ... 32
3.1.3 RcsB D56A as well as RcsB restored the production of capsule polysaccharide in K. pneumoniae CG43 ∆rcsB ... 33
3.1.4 Complementation with the phosphorylated RcsB restored the expression of type 3 fimbriae in K. pneumoniae CG43 ∆rcsB ... 34
x
3.1.5 RcsB D56E increases the expression of type 1 fimbriae in K.
pneumoniae CG43 ∆rcsB ... 36
3.1.6 RcsB D56A as well as RcsB promoted oxidative stress resistance of K. pneumoniae CG43 ∆rcsB... 38
3.1.7 RcsB D56A as well as RcsB restored the acid stress resistance activity under shaking cultured condition... 39
3.1.8 Effect of different phosphorylation states of RcsB on yfdX gene expression. ... 40
3.2 Different roles of phosphorylated and unphosphorylated RcsB. ... 41
3.3 Identification of the RcsB interacting protein under acidic condition. ... 43
3.3.1 RcsB interacting proteins in the regulatory pathway of CPS synthesis are different from that involved in regulating the acid stress resistance. ... 44
3.3.2 Generate an antibody specific to RcsB N-terminal region ... 44
3.3.3 Expression of RcsB under acidic growth conditions ... 45
4. Discussion ... 47
5. Perspectives ... 51
5.1 To study whether the RcsB interacting with RcsA, RmpA and RmpA2 is in phosphorylated or unphosphorylated form ... 51
5.2 To investigate the effect of RcsB phosphorylation on acid stress response ... 51
5.3 To identify the phosphorylated and non-phosphorylated RcsB interacting proteins ... 52
xi
xii
List of Tables
Table 1. Bacterial strains used in this study ……….….... 75
Table 2. Plasmids used in this study ……….………...…... 77
xiii
List of Figures
Fig. 1. Generation of the specific point-mutation on RcsB ……….…... 80 Fig. 2. Growth curve analysis ……….… 81 Figure 3. Colony morphology comparison …... 82 Figure 4. Analysis of capsule polysaccharide (CPS) synthesis …... 83 Fig. 5. Complementation analysis for the expression of type 3
fimbriae………...…… 84 Fig. 6. Complementation analysis for the expression of type 1
fimbriae ………...……… 86 Fig.7. Effect of RcsB phosphorylation on the oxidative stress
response ……….………...… 88
Fig. 8. Effect of RcsB phosphorylation on the acid stress survival ....… 89 Fig. 9. Model of different roles of phosphorylated and unphosphorylated
RcsB ……….………...…… 92 Fig. 10. RcsB phosphorylation influences on PyfdX promoter activity …. 93
Fig. 11. SDS-PAGE analysis of the expression of the recombinant RcsB N-terminal region in E. coli BL21(DE3) and the result of protein purification ……….………...…… 95 Fig. 12. Purity assessment of the purified recombinant RcsB-N and the
specificity analysis of the antiserum of RcsB-N ………... 97 Fig. 13. RcsB expression under acidic growth condition ……....…... 99
xiv
Appendix I. Sequence alignment of Klebsiella pneumoniae, Salmonella
typhi, Escherichia coli and Erwinia amylovora. …...…… 101
Appendix II. Deletion of rcsB significantly reduced CPS production ... 102 Appendix III. Deletion of rcsB slightly decreased the expression of type 3
fimbriae ………...……...……….….. 104 Appendix IV. Deletion of rcsB decreased the acid stress response ... 105
xv
Abbreviations
Asp bp(s) CPS CFU EDTA EMSA ESBL ETC HK HTH HV IP IPTG kb kDa LB LPS Aspartate base pair(s) capsular polysaccharide colony forming unitethylenediaminetetraacetic acid electrophoretic mobility shift assay extended-spectrum β-lactamase electron transport chain
histidine kinase helix-turn-helix hypermucoviscosity immunoprecipitation
isopropyl 1-thio-β-D-galactopyranoside
kilobase(s) kilodalton(s) Luria-Bertani lipopolysaccharide
xvi µg µL mL mM µM NDM-1 ng OD ONPG PAGE PAI PCR PLA Rcs RR rpm SDS TCS X-gal microgram microliter milliliter millimolar micromolar
New Delhi metallo-β-lactamase 1 nanogram
optical density
o-nitrophenyl-β-D-galactopynoside
polyacrylamide gel electrophoresis pathogenecity island
polymerase chain reaction pyogenic liver abscess
regulator of capsule synthesis response regulator
revolutions per minutes sodium dodecyl sulfate two-component system
1
1. Introduction
1.1 Klebsiella pneumoniae
Klebsiella pneumoniae is a Gram-negative bacterium. It is facultative
anaerobic, non-motile, rod-shaped, heavily encapsulated, which belongs to the Enterobacteriaceae that is under the gamma subdivision of the phylum of Proteobacteria which include genus such as Escherichia, Salmonella,
Shigella and Yersinia [1]. The Klebsiella spp. are ubiquitous in nature, with
two plausible habitats: the surface water, soil, sewage, plants [2-6] and the mucosal surfaces of the animal host, such as human respiratory tract and intestinal tract [7].
1.2 Infections of K. pneumoniae
K. pneumoniae is an opportunistic pathogen, which usually caused
infections found in immuno-compromised individuals who are hospitalized and suffered from severe underlying diseases, such as diabetes mellitus or chronic pulmonary obstruction. In respect of bacteremia caused by
2
nosocomial Gram-negative pathogens, Klebsiella is second only to
Escherichia coli [8]. K. pneumoniae usually causes variety of diseases
including suppurative lesions, bacteriemia, urinary tract infections, pneumonia, and sometimes life-threatening septic shock [7, 9-12]. Reported carrier rates in hospitalized patients are 77% in the stool, 19% in the pharynx and 42% on the hands of patients [13, 14]. However, such high rates of nosocomial colonization were likely due to the use of antibiotics rather than the factors with delivery of care in the hospital [15, 16], especially the extended-spectrum β-lactamase (ESBL)-producing strains [17-21]. Carbapenems are considered to be the preferred agents for treatment of serious infections caused by ESBL-producing K. pneumoniae because of their high stability to β-lactamase hydrolysis and observed retained susceptibility of ESBL producers [22], however, carbapenems-resistant K. pneumoniae have been reported worldwide since 2000s [23-26]. In 2008, an emergence of multidrug resistant superbug, so-called NDM-1 was first detected in K. pneumoniae isolate from a Swedish patient of Indian origin traveled to New Delhi. A carbapenem-resistant K.
pneumoniae strain bearing the novel gene was identified from the Sweden
3
classes of antibiotics, including β-lactams, fluoroquinolones, and aminoglycosides, but most were still susceptible to colistin [27].
During the last decade, K. pneumoniae infections causing community-acquired primary pyogenic liver abscess (PLA) have also become an emerging disease receiving increasing attention. Distinct from the classical PLA, which is a complication of intra-abdominal or biliary tract infections resulting from multiple aerobic and aerobic bacterial strains [28], PLA caused by primary infection of K. pneumoniae as a single pathogen is often cryptogenic and complicated with metastatic lesions [29-31]. Most cases have been reported from Taiwan [32], which were associated with a distinct invasive syndrome in liver abscess, meningitis and endophthalmitis. Attempts have been made to identify the risk factors for K. pneumoniae infections associated with abscess-formation, especially diabetes mellitus is the most tightly associated with K. pneumoniae PLA among the host factors.
4
1.3 Virulence Factors of Klebsiella pneumoniae
The factors that contributing to K. pneumoniae pathogenicity that had been identified, includes the capsular polysaccharide (CPS), lipopolysaccharide (LPS), adhesion factors, iron acquisition systems and some other stress responses.
1.3.1 Capsular Polysaccharide (CPS) and lipopolysaccharide
(LPS)
Clinically isolated K. pneumoniae usually produces large amounts of CPS and therefore forms large glistering colonies with viscid consistency, e.g. the isolation that used in study K. pneumoniae CG43, an isolation from Chang Gung Memorial Hospital, which causes liver abscess in diabetes mellitus patients. As a major virulence factor, CPS acts to protect the bacteria from phagocytosis, killing by polymorphonuclear granulocytes and bactericidal serum factors [33-35]. Besides physical hindrance to fimbrial binding, the role of Klebsiella CPS in mediating the bacterial resistance to antimicrobial peptides has also been reported [36, 37]. There are 77 serotypes of CPS had been identified, reported that K1 and K2 CPS are the most virulent to mice [38]. Moreover, the hypermucoviscosity (HV)
5
phenotype of K. pneumoniae isolates resulting from a profound expression of CPS has also been correlated with the development of invasive syndrome [39]. The LPS O-antigen and the lipid-A are major component of Gram-negative bacterial cell walls and establishment of septic shock [33, 40]. The O-antigen is thought to play a role in resistance to complement killing [40] and to contribute to bacteremia as well as lethality during murine pneumonia infections [41]. The lipid-A (endotoxin) region is reported as the primary inflammatory component of LPS because of specific and sensitive recognition by the innate immune system [42].
1.3.2 Adhesion Factors
A number of adhesins have been suggested as potential virulence factors, including type 1 [43-45] and type 3 fimbriae [46-48], non-fimbrial adhesion CF29K [49] and KPF28 [50, 51], were associated with the initial attachment and subsequent colonization of K. pneumoniae in the respiratory and urinary tract [52]. Type 1 fimbriae have been most extensively studied in E. coli, and the corresponding structures of K.
6
regulation [44, 53-55]. The fimbrial rod consists of the major subunits FimA and the minor subunits FimI, FimF and FimG. However, several reports indicated that K. pneumoniae poorly expresses type 1 fimbriae in vitro [44, 55, 56], the thick capsule of K. pneumoniae has been shown to impede the activity of type 1 fimbriae and also to retard the assembly of type 1 fimbrial subunits from periplasm to cell surface [55, 57, 58], suggesting a cross-regulation of the expression of fimbriae and capsule for an efficient infection.
Several studies have also demonstrated an important role for type 3 fimbriae in biofilm formation on biotic and abiotic surfaces [45, 47, 59-64]. Biofilm are recognized as surface-attached bacteria embedded in a self-produced matrix, composed mainly of polysaccharide, but also containing proteins and nucleic acids [65]. Biofilm formation promotes encrustation and protects the bacteria form hydrodynamic forces of urine flow, host defenses and antibiotics [66]. The ability of bacteria to form biofilm on medical devices is believed to play major role in development of nosocomial infections, including the catheter-associated urinary tract infections, which is frequently caused by K. pneumoniae [7, 66-68]. In addition, type 3 fimbriae mediate adhesion to epithelial cells, from the
7
respiratory and urinary tracts and extracellular matrix proteins, such as collagen V, in vitro [62, 69-73]. Type 3 fimbriae are encoded by
mrkABCDF operon [46, 74, 75]. MrkA and MrkF are the major and minor
subunits, respectively, which constitute the fimbrial rod and facilitate biofilm formation [46, 47].
1.3.3 Stress Responses
Besides those bacteria structural components that suggested being pathogenic, there are some mechanisms of K. pneumoniae that responsible for stress resistances, in order to survive under variety of environmental conditions, such as different pH condition, different concentration of oxygen, changes of osmotic pressure etc. Surviving under acidic condition in human stomach is the prior ability for Enterobacteriaceae to cause further infection. Therefore, before entering into the intestinal tract, which is in high pH value, there must be some specific mechanisms for K.
pneumoniae to cope with the low pH stomach condition, which composed
of gastric acid. One of the members of Enterobacteriaceae, E. coli had been identified to have 4 clusters of acid resistant system, whereas acid
8
resistant mechanism of K. pneumoniae not yet been identified completely. Even so, reported in 2007, RcsB is responsible for the glutamate-dependent acid resistant system in E. coli [76], which also found in K. pneumoniae CG43 in our study. Suggesting that RcsB might plays role for the acid stress response in K. pneumoniae CG43.
Oxidative stresses threaten bacteria by damaging DNA, protein, cell membrane, and affect bacterial growth and replication. During infection, bacteria usually living at a condition with full of oxidative stress, including intracellular oxidative stress i.e. oxygen gas produced by the electron transport chain (ETC), environmental existing oxidative stress i.e. environmental redox products or come from other competing pathogen during infection, and oxidative stress producing during phagocytosis occurred in host cell to attack bacteria [77]. Under oxidative stress, bacteria have several ways to cope with, such as oxidant-scavenger regulators and some antioxidant enzymes. Previously our lab members had suggested that
sodA, sodB, sodC, katE, katG and rpoS, which are reported as oxidative
resistant related genes clearly in E. coli, are also involved in the oxidative resistant event in K. pneumoniae CG43.
9
1.4 Two-Component System (TCS)
All organisms must respond to changing environmental conditions quickly and efficiently. While higher order organisms like plants and animals have adaptive responses that include complex behaviors such as migration or adaptation over generations, lower order and unicellular organisms including bacteria and yeast must sense and respond to changes their environmental using alternative strategies. Bacterial adaptation occurs on the level of individual genes and proteins; the level of global regulons; the whole-cell level (via motility or sporulation); or at multicellular level (via quorum sensing and biofilm, formation) [78]. TCSs are found ubiquitously in prokaryotes, archaea, fungi, yeast, and some plants. In bacteria, TCS responsible to sense, respond, and adapt to a wide range of environments, stresses, and growth conditions [79]. In the prototypical two-component system, comprised of sensor histidine kinase (HK) and their cognate response regulator (RR) substrates, this signaling system have been implicated in mediating the response of bacteria to a wide range of signals and stimuli, including nutrients, cellular redox state, changes in osmolarity, quorum signals, antibiotics, and more [80].
10
Most of the TCS signal transduction mechanisms prevailing to control gene expression in bacteria. Typically, a signal transduction pathway is initiated when the sensor histidine kinase stimulated by stress signals. The sensor kinase catalyzes its autophosphorylation and then subsequently transfers the phosphoryl group to a response regulator, which can then effect changes in cellular physiology, often by regulating gene expression [81].
1.5 Regulator of Capsule Synthesis (Rcs) System
The Rcs system, first identified by its role in the transcriptional regulation of the genes for capsular polysaccharide in E. coli, is proving to be an unexpected complex example of such a system. Besides, several bacteria such as Salmonella typhi [82], K. pneumoniae [83], the plant pathogenic bacteria Erwinia amylovora [84, 85] and Pantoea stewartii [86], the biosynthesis of CPS is controlled by the Rcs regulatory system [81], implicating that the regulating mechanism is highly conserved among these pathogenic bacteria . The Rcs system is composed of RcsA, RcsB, RcsD, and RcsF. RcsF is an outer membrane protein that is responsible for
11
receiving extracellular stimulation. RcsC acts as an inner membrane sensor kinase in this system, with a periplasmic domain [87]. RcsD (previously called YojN) is a phosphotransfer protein, which presented adjacent to RcsC at inner membrane, it transfers phosphoryl group from sensor kinase RcsC to response regulator RcsB in this system. In phosphorelays of this system, transfer of phosphate from His to Asp is conserved, phosphate travels from His to Asp to His to Asp [81]. Generally, phosphorylation and dephosphorylation of the response regulator, in turn, change the activity of this protein, frequently by modulating its ability to bind to DNA and act as a regulator of transcription. There are plenty of studies showing that Rcs system is presented in several members of Enterobacteriaceae bacterial regulatory mechanisms. In E. coli, Rcs system regulates CPS production [87], cell division [88], motility [89], type 1 pili [90] and glutamate-dependent acid resistant [76]. In Salmonella, Rcs system involved in the regulation on the gene that encoding capsule synthesis [91], virulence factors related genes [92, 93]. In K. pneumoniae, regulation of Rcs system on CPS production was reported clearly [94-96], however, suggesting other regulations still pending on further investigation.
12
1.6 The Response Regulator -- RcsB and RcsB-dependent
Regulation
RcsB is a typical cytoplasmic activator for CPS biosynthesis [87]. According to the sequence alignment, it is highly conserved (~90% sequence identity) among different species, such as E. coli, S. typhi, E.
amylovora and K. pneumoniae [85]. It has been proved that RcsB is
activated by membrane-bound sensor RcsC and RcsD through phosphotransferring to a highly conserved Asp residue in the N-terminal domain of RcsB [87, 97]. RcsB is composed of two conserved motifs: an N-terminal domain for receiving phosphate group at the 10th, 11th, 56th Asp residues as well as interacting with other regulator proteins; the C-terminal as an effector domain, which is a DNA binding motif that places this RcsB in the LuxR family of helix-turn-helix proteins [87].
1.6.1 Role of RcsB phosphorylation
The RcsB protein contains a LuxR-like helix-turn-helix DNA binding domain. Several members of the LuxR family are response regulators (RR), which act as transcriptional activators or repressors. As receiving phosphate group from the upstream kinase is the way for signal reception
13
of RcsB, phosphorylation status of RcsB determines downstream gene expression. This can be exemplified by the regulation of CPS production in
E. coli, phosphorylated form of RcsB stimulates expression of the cps
operons [98]. Previous study showed that, Salmonella biofilm development depends on the phosphorylation status of RcsB [99]; the phosphorylated RcsB inhibits biofilm development, while unphosphorylated RcsB induces. Rcs system responsible for wide range of regulation among various species of bacteria besides those had been mentioned above, it is able to receive lot of different stimulation, and then transfer the signals to downstream in order to regulate on corresponding genes.
1.6.2 RcsB interacting protein
RcsB is a global regulator allows multiple different regulatory inputs into its activation of specific targets, providing a simple mechanism for regulation of subsets of the RcsB-regulated genes. Interaction of RcsB with coregulator, will affect its binding to target genes. There are several types of RcsB-dependent regulation among the family of Enterobacteriaceae. The first type is exemplified by the regulation of cps in E. coli and related capsule genes in the plant pathogens E. amylovora and P. stewartii [81]. In these cases, RcsA is required to form heterodimer with RcsB. A necessary
14
site for RcsB and RcsA action was mapped approximately 100 nt upstream of the transcription start site had been demonstrated in Erwinia amylovora [100]. A consensus site for this binding has been determined from mutagenesis studies. Additional sites are present upstream of the rcsA promoter of E. coli, Salmonella, Klebsiella, which is autoregulated by RcsA and RcsB [101, 102]; together, these have been used to define an “RcsAB box” with a consensus sequence of TaAGaatatTCctA. In vitro studies demonstrate a tenfold higher binding for RcsA/RcsB than for RcsB alone at the regulatory sites for capsule synthesis [103]. Although the auxiliary regulator has been studied most thoroughly is RcsA, a number of other proteins have been found to cooperate with RcsB to stimulate transcription at specific promoters. The colanic acid, which is the main component of capsule, has not been identified as a virulent factor in mammals, related E. coli and Klebsiella species use RcsB to regulate synthesis of capsules that are important for virulence [104-106]. In some cases, these organisms have evolved somewhat different helper functions for RcsB. In Salmonella typhi, while a functional rcsA gene is present, the virulence-associated Vi antigen is synthesized in a process that is dependent on RcsB and independent of RcsA. The TviA protein, encoded
15
by the first gene of the ten-gene Vi cluster, is necessary for activation [82, 107]. Although no DNA binding has yet been demonstrated for TviA, its ability to suppress an rcsB missense mutation in the DNA binding domain may suggest that it interacts with RcsB and stimulates its ability to bind its DNA site [82]. For targets that depend on other auxiliary protein, the conditions of its induction, and the sensitivity of the target to RcsB and the helper protein will determine the expression pattern. In E. coli O157:H7, RcsB interacts with GadE, which is a central regulator involved in the glutamte-dependent acid resistance system AR2, to mediate acid resistance in stationary phase [108]. Some more in E. coli, another LuxR-type family transcriptional regulator, BglJ, interact with RcsB to form heterodimer that presumably bind upstream of the bgl promoter, and then relieves the H-NS (histone-like nucleoid-structuring protein) -mediated repression of bgl operon [109]. In conclusion, RcsB is a regulator that requires auxiliary proteins while acting, and can act with many different such auxiliary proteins that activated by specific signals, hence, providing the corresponding output.
16
1.7 Previous Studies - RcsB-dependent regulation in K.
pneumoniae
Previous studies in our lab showed that, RcsB regulatory function is involved in the CPS biosynthesis [94, 110], type 3 major pilin MrkA production [111] and acid stress resistance [112] in K. pneumoniae CG43.
The understanding about the RcsB-dependent regulation on CPS biosynthesis is more thoroughly among the other RcsB-dependent regulation in K. pneumoniae. As known, RcsB requires auxiliary protein RcsA to regulate CPS biosynthesis in E. coli as well as. Moreover, synthesis of the K2 capsule in depends on a large virulence plasmid pLVPK, which has rendered an ~1,000 fold decrease in mouse virulence in curing [113]. pLVPK encodes RmpA and RmpA2, both are transcriptional activators that had been demonstrated to interact with RcsB on the regulation of the CPS biosynthesis [114, 115]. It had been showed that, the level of K. pneumoniae CG43 CPS production significantly reduced in the absence of RcsB (Appendix II) [110].
Type 3 fimbria is the adhesion component that involved in the biofilm formation of K. pneumoniae. The major subunit MrkA is usually used as
17
the indicator for the type 3 pilin expression. In the western blot that incubated with the MrkA polyclonal antibodies, the level of MrkA expression reduced in the RcsB deletion mutant K. pneumoniae CG43 (Appendix III) [111]. During infection, an enteric bacterium has to cope with the acidic condition in human stomach before it reaches the intestinal tract. Therefore, high number of the survived K. pneumoniae is required after passing through the stomach. An experiment demonstrated that, the acid survival rate of RcsB deletion mutant K. pneumoniae CG43 became lower comparing to the wild type K. pneumoniae CG43 (Appendix IV) [112]. Hence, indicating that RcsB is an important global regulator in K.
pneumoniae CG43, which involved in the regulation of several virulence
factors.
1.8 Specific Aims
Phosphorylation of RcsB has been demonstrated to be required for its regulatory function on bacterial cps gene expression [98, 116]. We intend to investigate if phosphorylation of RcsB affects its regulation on the suspected RcsB-regulated gene expressions. RcsB consists of N terminal
18
receiver domain and C-terminal DNA-binding domain. The N-terminal phosphoreceiver domain belongs to the CheY family and REC superfamily [117], and the highly conserved aspartate residue at position 56 (residue position is based on the amino acid sequence of K. pneumoniae RcsB) has been proposed as a phosphorylation site in those CheY family regulators [118]. An aspartate-glutamate substitution at the position 56 of RcsB also leads to constitutive gene expression, implying this position is a potential phosphorylation site [116, 119]. Two site-directed mutations RcsB-D56E and RcsB-D56A will be generated to mimic the phosphorylated form of RcsB and the RcsB that is unable to accept phosphate group, and then examine whether these two mutations affect the phenotypes of K.
pneumoniae CG43.
As mentioned, RcsB outputs the received signal by stimulating the cps operon expression as a transcription factor [87]. Hence, a LacZ reporter system will be set up to confirm that, RcsB affects yfdX (which suggested to be involved in the acid stress response) expression at the transcriptional level. Furthermore, electrophoretic mobility shift assay (EMSA) will be demonstrated to investigate whether RcsB regulates the transcription of
19
expression by binding to promoter as a hetero-/homodimer, while acting with different auxiliary proteins allows RcsB activates different specific target genes. Co-immunoprecipitation (Co-IP) will be done to identify the auxiliary protein, which cooperates with RcsB to regulate the target gene that is stimulated under specific condition.
20
2. Materials and Methods
2.1 Plasmids, primers, bacterial strains and growth conditions
Bacterial strains and plasmids used in this study are listed in Table 1 and Table 2, and the primers used are listed in Table 3. K. pneumoniae CG43, a clinical isolate of serotype K2, is high virulent to mice [120]. E.coli and K. pneumoniae strains were generally propagated at 37°C in
Luria-Bertani (LB) broth. M9 minimal medium was also used in some specific assay. Bacterial growth was assessed by measuring the absorbance of optical density at 600 nm (OD600). The antibiotics used include ampicillin
(100 µg/ml), chloramphenicol (35 µg/ml), kanamycin (25 µg/ml), tetracycline (5 µg/ml) and streptomycin (500 µg/ml). Microaerobic culture for the MrkA expression was added with mineral oil (M5310) which was purchased from Sigma.
21
2.2 Construction of the site-directed mutants RcsB
pHY121 that carrying rcsB and its approximately 600 bp adjacent regions was used as the template. The plasmids with site-directed mutations were constructed by quick change method. pHY121 was amplified with the complementary primer sets rcsB-D56E(＋)/rcsB-D56E(－) , wc30/wc31 and wc32/wc33 encompassing the mutation site (167th bp for D56A and 168th bp for D56E of pHY121insert) by using PfuUltra II Fusion HS DNA polymerase (Agilent Technologies) to generate mutant alleles of rcsB with the D56E or D56A mutations. The PCR product was resolved on an agarose gel, recovered, treated with DpnI for 2 hr to remove the template plasmid and transformed into E. coli JM109. The plasmid, pHY121*, carrying the mutation allele encoding RcsB (RcsB*, D56E and D56A mutations) was then prepared from the transformant colony and confirmed by sequence analysis. The mutated fragment RcsB-D56E and RcsB-D56A were subcloned into pRK415 to yield pYX001 and pYX002, respectively. pYX001 and pYX002 were then individually mobilized from E. coli S17-1 λpir to the K. pneumoniae CG43S3 ∆rcsB strain by conjugation, and the subsequent selection was performed as described above. Each site-directed mutation in K. pneumoniae was confirmed by DNA sequencing.
22
2.3 Measurement of bacterial growth
Cultures of the parental strain K. pneumoniae CG43S3, along with rcsB deletion mutant strains, ∆rcsB[pRK415-RcsB], ∆rcsB[pRK415-RcsB-D56E], ∆rcsB[pRK415-RcsB-D56A] complement strains were grown overnight in LB medium. Twenty microliters of ovenight LB cultures of K.
pneumoniae strains was used to inoculate 4 ml of LB broth. The cultures
were incubated at 37°C with shaking, and the optical density was recorded as the absorbance at 600 nm at the indicated time points. Values were the average and standard deviation from triplicate samples from one of the three independent trials.
2.4 Extractions and Quantification of CPS
Bacterial CPS was extracted using the method described [121]. Briefly, 500 µl of overnight grown bacterial was mixed with 100 µl of 1% Zwittergent 3-14 (Sigma-Aldrich, Milwaukee, WI, USA) in 100mM citric acid (pH 2.0) and incubated at 50°C for 20 min. After centrifugation, 250 µl of the supernatant was transferred to a new tube, and the CPS was precipitated with 1 ml of absolute ethanol. The pellet was dried, dissolved
23
in 200 µl de-ionized water, and then 1200 µl of 12.5 mM borax in H2SO4
was added. The mixture was vigorously mixed, boiled for 5 min, cooled down and then 20 µl of 0.15% 3-hydroxydiphenol (Sigma-Aldrich, Milwaukee, WI, USA) was added. The absorbance at 520 nm was measured, and the glucuronic acid content was determined from a standard curve of glucuronic acid and expressed as µg per 109
CFU or 1010CFU.
2.5 Western Blot Analysis
Samples for Western blot analysis were prepared by obtaining whole bacterial lysates, cells were grown in LB medium at 37°C for overnight. Each culture was harvested for 400 µl, after centrifugation, cells were resuspended in 100 µl of de-ionized water, and then were boiled at 100°C for 10 min. Quantification was taken to equalized the concentration of all cultures, and added with 2 × protein dye that containing 20% of DTT (Dithiothreitol) to boil again at 100°C for 10 min. Proteins were separated on SDS-polyacrylamide gels (13.5% to 5%). For Western blotting, proteins were transferred onto PVDF (Polyvinylidene fluoride) membranes by electroblotting. When blotting was completed, add blocking buffer which
24
contains 5% milk in TBS buffer (20 mM Tris-HCl, 50 mM NaCl, pH 7.5) and blocked for 16 hr at 4°C. Probing was carried out with diluted antibodies: anti-GAPDH, anti-RcsB N-terminal, anti-MrkA, anti-FimA or anti-YfdX at a ratio of 1: 10000 for 2 hr at room temperature. Before incubating the blot in second antibody, wash the membrane twice by TBST buffer (TBS buffer with 0.1% tween-20) for 10 min per time, and then wash by TBS buffer for 10 min to remove the non-specific binding of first antibody. Secondary hybridization was incubated with diluted anti-rabbit IgG conjugated with alkaline phosphatase at a ratio of 1: 5000 for 1 hr at room temperature. Bound ligands were detected by using the alkaline phosphatase staining method, which contain alkaline phosphate buffer (150 mM Tris-HCl, 5 mM MgCl2 6H2O, pH 9.5), BCIP (50 mg/ml of
5-bromo-4-chloro-3-indolyl phosphate in 95% dimethylformamide) and NBT (50 mg/ml of p-nitro-blue tetrazolium chloride in 70% dimethylformamide).
2.6 Biofilm Formation Assay
Overnight grown bacteria were diluted 1: 100 in LB broth supplemented with appropriate antibiotic and then inoculated into each well
25
of a 96-well microtiter dish (Orange Scientific) for statically incubation at 37°C for 12 hr and 48 hr After removal of the bacteria, the plate was washed by de-ionized water once, and 150 µl of 1% (w/v) crystal violet was added to each well. The plate was incubated at room temperature on an orbital shaker for 30 min, and then washed three times. The dye was solubilized in 1% (w/v) SDS, and absorbance at 595 nm was determined.
2.7 Yeast-cell Agglutination
Agglutination of yeast Saccharomyces cerevisiae AH109 was carried out as described [122]. Briefly, bacteria (~108 CFU/ml) were suspended in saline (0.85% of NaCl) with or without 5% mannose and then mixed with 1% of yeast (Sigma-Aldrich) on a glass slide. After 5 min incubation at room temperature on an orbital shaker, agglutination of yeast caused by bacteria could be assessed.
26
2.8 Disc Diffusion Assay
Bacteria were overnight grown with 1: 20 dilution in LB broth supplemented with appropriate antibiotic and then incubated at 37°C until OD600 reached 0.3- 0.4. Spread 100 µl of each strain of culture on M9 agar
plate, and then place a nitrate disc on the center of the plate. Add 5 µl of 10 mM paraquat on the nitrate disc and place the agar plate by staying the agar at the bottom in 37°C incubator for overnight.
2.9 Acid Stress Response
Bacterial resistance to acid challenge was determined essentially as previously described [123]. Overnight grown bacteria were diluted 1: 20 in LB broth supplemented with an appropriate antibiotic and then incubated at 37°C until OD600 reached 0.4 - 0.6. All strains were further divided into two
groups of tubes, while each tube containing 1 ml of bacteria culture. Both groups of culture was centrifuged and resuspended in 1 ml of pH 4.4 LB medium for a 1 hr adaption period before the acid challenge. After adaptation, each culture from one of the group was centrifuged and resuspended in 1 ml of pH 3.0 M9 medium for 45 min. After the acid
27
challenge, 100 µl of each culture in the second group was immediately removed for serial tenfold dilution in saline, and 100 µl of the 10-6
diluted sample was plated onto LB agar plates at what was considered to be the initial time point (t0). The group with pH 3.0 acid challenge was removed
100 µl immediately and serially diluted, 100 µl of each culture to the 10-6 dilution was plated onto LB agar plates and incubated at 37°C overnight. The survival rates of each strain were then calculated by dividing the number of colonies with acid challenge by the number of colonies at t0.
2.10 Measurement of Promoter Activity
A laboratory stocked plasmid, placZ15-PyfdX, which is the promoter
selection plasmid placZ15 that inserted with the yfdX promoter region in front of a promoter-less lacZ gene [124]. placZ15-PyfdX was then
mobilized from E. coli S17-1 λpir to K. pneumoniae wild type and ∆rcsB strains by conjugation. β-galactosidase activity was determined as previously described [124]. In brief, overnight cultures were diluted 1: 20 in LB broth supplemented with appropriate antibiotic and then incubated at 37°C until OD600 reached 0.8 - 1.0. Harvest 1 ml of each
28
culture in a tube, and wash with 1 ml of Z buffer (60 mM Na2HPO4, 40
mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 50 mM β-mercaptoethanol)
for twice. All cultures were resuspended after washed, 100µl of each culture was taken to mix with 900 µl of Z buffer, 17 µl of 0.1% SDS and 35 µl of chloroform, mixture was shaken vigorously. After incubation in 30°C water bath for 10 min, the reaction was initiated by adding 200 µl of 4 mg/ml ONPG (o-nitrophenyl-β-D-galactopyranoside) (Sigma-Aldrich). Upon the yellow-colored appearance, the reaction was terminated by adding 500 µl of 1 M Na2CO3. OD420 of each strain was
recorded and the β-galactosidase activity was expressed as Miller units
(1 Miller unit = 1000 ∗ OD420
(𝑡∗𝜐∗OD600); OD420 is the absorbance of the
yellow o-nitrophenol; t = reaction time in minutes; v = volume of culture assayed in milliliters; OD600 = reflects optical density of bacteria
culture) [125]. Each sample was assayed in triplicate, at least three independent experiments were carried out. The data shown were calculated from one representative experiment and shown as the means and standard deviation from triplicate samples.
29
2.11 RcsB N-terminal region antiserum preparation
The N-terminal domain of rcsB was PCR amplified using primers Yen001 and Yen002 and cloned into the PCR cloning vector yT&A (Yeastern Biotech, Taiwan). The NdeI/SalI fragments form the resulting plasmid (yT&A-RcsB N-terminal) were then cloned individually into pET30a (Novagen, Madison, Wis, USA) to yield pET30a-RcsB N-terminal allowing the in-frame fusion to the N-terminus His-tag. Plasmid pET30a-RcsB-Nterminal was then introduced into E. coli BL21 (DE3) (Invitrogen, USA) for overexpression of the His6::RcsB N-terminal recombinant protein.
The recombinant protein was over-produced from the mid-log phased culture (OD600 around 0.4 - 0.6) by induction with 1 mM IPTG
(isopropyl-1-thio-β-D-galactopyranoside) for 4 hr at 37°C. The proteins were then
purified from total cell lysate by affinity chromatography using His-Bind resin (Novagen, Madison, Wis). After purification, the eluent was dialyzed against 1× protein storage buffer (10 mM Tris-HCl pH 7.5, 138 mM NaCl, 2.7 mM KCl, and 10% glycerol) at 4°C overnight. RcsB N-terminal region antiserum was prepared by immunizing a New Zealand white rabbit with 2.5 mg of the purified His6::RcsB N-terminal recombinant protein and the
30
2.12 Statistical Methods
The results of the biofilm-forming activity and β-galactosidase activity assays were derived from a single experiment that was representative of three independent experiments. Each sample was assayed in triplicate and the data were presented as the mean ± standard deviation (SD). Differences between groups were evaluated by a two-tailed Student’s t-test. P-values less than 0.05 were considered statistically significant difference.
31
3. Results
3.1 Roles of phosphorylation of RcsB in K. pneumoniae CG43
3.1.1 Generation of specific point-mutation on RcsB
As a LuxR family response regulator, a common post-translational modification of phosphorylation at a conserved Asp in the N-terminal receiver domain performed by a sensor histidine kinase (HK) or by cytoplasmic phosphodonors, acetyl-phosphate. Phosphorylation of the conserved Asp in the N-terminal receiver domain of response regulator activates the protein by inducing conformational changes which facilitate interaction of the response regulator with the target DNA [117, 126]. Furthermore, some response regulators have been shown to autophosphorylate in vitro in the presence of acetyl phosphate [127-129], which has been proposed to be global signal in bacteria [130]. Three aspartate residues at position 10th, 11th, 56th are highly conserved in an N-terminal phosphorylation motif and Asp56 has been proved as a
32
phosphorylated site (Appendix I) [81]. As shown in Fig. 1, to determine the role of the conserved D56 on RcsB regulation in K. pneumoniae CG43, the site-directed mutants with single amino acid substitution of the D56 to alanine (D A) to prevent it from phosphorylation [131, 132], or to glutamte (D E) to mimic the phosphorylated state [133-136], were created. The mutated sequences were identified by the Tri-I Biotech, Inc. to confirm the single nucleotide substitution. The point-mutated RcsB was then individually introduced into K. pneumoniae CG43 ∆rcsB mutant strain.
3.1.2 RcsB D56A as well as RcsB retarded the growth of K.
pneumoniae CG43
∆rcsB
To investigate the effect of RcsB phosphorylation state on bacterial growth, K. pneumoniae CG43 wild type, ∆rcsB mutant and ∆rcsB complement strains were grown in LB medium for 24 hours. The optical density at 600 nm of each bacteria culture was recorded every hour. As shown in Fig. 2, the growth rate of K. pneumoniae CG43 ∆rcsB [pRK415-RcsB] and K. pneumoniae CG43 ∆rcsB [pRK415-RcsB-D56A] were significantly lower comparing to the wildtype K. pneumoniae CG43. We
33
assumed that K. pneumoniae CG43 with high level of RcsB or unphosphorylated RcsB expression directs the carbon metabolic pathway to CPS biosynthesis and, therefore resulting of the retarded growth rate.
3.1.3 RcsB D56A as well as RcsB restored the production of
capsule polysaccharide in K. pneumoniae CG43 ∆rcsB
K. pneumoniae CG43 is a highly encapsulated virulent strain [120]. In
order to verify whether complementation with the phosphorylated or unphosphorylated RcsB restores the production of CPS, [pRK415-RcsB-D56E] and [pRK415-RcsB-D56A] were complemented into ∆rcsB mutant, and the amounts of CPS produced were compared with that of ∆rcsB[pRK415]. All of the strains were streaked on LB agar for comparing
their colony morphology, the colony surface of ∆rcsB[pRK415-RcsB-D56E] appeared to be less mucoid than that of ∆rcsB[pRK415] (Fig. 3A), and the degree of viscosity was reduced significantly as determined by a string test [114], which refers to the ability to form a string when the bacterial colony was picked with toothpick. The CPS-deficient phenotype is evident as assessed using sedimentation assay and the amount of K2 CPS
34
produced. In the sedimentation test, all of the samples were centrifuged at 4,000 rpm for 5 min. As shown in Fig. 3B, the ∆rcsB[pRK415-RcsB-D56E] as well as ∆rcsB[pRK415] could be rapidly precipitated. In the CPS quantification analysis, the content of main substance, which is the glucuronic acid, was extracted for quantification. As shown in Fig. 3C, ∆rcsB[pRK415] and ∆rcsB[pRK415-RcsB-D56E] had reduced at least 50% of the CPS contents in comparing with ∆rcsB[pRK415-RcsB] and ∆rcsB[pRK415-RcsB-D56A]. The amount of CPS could be restored by RcsB-D56A but not by RcsB-D56E, suggesting that the regulation of RcsB on CPS biosynthesis may require the unphosphorylated form of RcsB.
3.1.4 Complementation with the phosphorylated RcsB
restored the expression of type 3 fimbriae in K.
pneumoniae CG43
∆rcsB
Type 3 fimbriae are 4 nm wide and 0.5-2 µm long surface organelles that are originally characterized in Klebsiella strains by their ability to mediate mannose-resistant agglutination of tannic acid-treated human
35
erythrocytes (MR/K haemagglutination) [75, 137]. K. pneumoniae type 3 fimbriae are encoded by mrkABCDF gene cluster [74], several studies have demonstrated an important role for type 3 fimbriae in biofilm formation on biotic surfaces [45, 138]. The ability of bacteria to form biofilm on medical devices is believed to play a major role in development of nosocomial infections. In addition, type 3 fimbriae have been demonstrated to mediate bacterial attachments to several cell types including tracheal epithelial cells, renal tubular cells, extracellular matrix proteins, and components of basement membranes of human lung tissue [69, 72, 73]. According to the previous study, the level of type 3 fimbriae expression reduced in the RcsB deletion mutant K. pneumoniae CG43. As shown in Fig. 4A, the Western blot hybridization using MrkA antiserum revealed that the expression of the type 3 major pilin, MrkA was increased approximately 1.53 fold in the ∆rcsB[pRK415-RcsB-D56E], while ∆rcsB[pRK415-RcsB-D56A] expressed a small amount of MrkA (approximately 39%). As type 3 fimbriae is the adhesion component that involved in the biofilm formation of K. pneumoniae, the biofilm formation assay was performed in the 96-plastic well and glass tubes. Consistently, ∆rcsB[pRK415-RcsB-D56E] exhibited a higher level of biofilm-forming activity than
∆rcsB[pRK415-36
RcsB-D56A], as assessed by quantitative measurement with crystal violet staining. However, the biofilm-forming activity assessed by glass tube exhibited an inverse effect. Thus, ∆rcsB[pRK415-RcsB-D56A] was not a biofilm former compared to the ∆rcsB[pRK415-RcsB-D56E]. This suggested that, the expression of type 3 fimbriae is regulated by the phosphorylated form of RcsB, and the phosphorylated state of RcsB maydetermine the adhesion of K. pneumoniae CG43 on different surfaces.
3.1.5 RcsB D56E increases the expression of type 1 fimbriae in
K. pneumoniae CG43
∆rcsB
Type 1 fimbriae are approximately 7 nm wide and 1-2 µm long surface organelles found in virtually all members of the family
Enterobacteriaceae [139, 140]. They are expressed by fimACDFGHIK
gene cluster [141] and well known for the ability to bind to mannose-containing structures on host cells and extracellular matrix. Bacteria expressing type 1 fimbriae are able to cause mannose-sensitive agglutination of yeast cells or erythrocytes (mannose-sensitive haemagglutination, MSHA) from guinea pig. Previous study showed that,
37
the deletion of mrkA increased the expression of type 1 fimbriae, while the overproduction of FimB recombinase repressed MrkA expression. Together, both indicated that the expression of type 3 fimbriae and type 1 fimbriae are regulated in a coordinate manner in K. pneumoniae [142]. Western blot analysis using anti-FimA serum, Fig. 5A revealed that, FimA did not express in K. pneumoniae CG43 wild type strain, but express only in ∆rcsB mutant and ∆rcsB[pRK415-RcsB-D56E]. The activity of type 1 fimbriae was assessed using mannose sensitive yeast agglutination analysis. Each strains of bacteria culture was added with 1% of yeast for reaction in a single well of 24-well plate, and also another reaction with 5% of mannose for the mannose competition assay. As shown in Fig. 5B, agglutination could be observed in the well with ∆rcsB mutant and ∆rcsB[pRK415-RcsB-D56E]. In the mannose-sensitive agglutination activity assay (Fig. 5C), mannose could inhibit the agglutination activity of ∆rcsB mutant and ∆rcsB[pRK415-RcsB-D56E]. The result indicates that, RcsB phophorelay status affects the expression of type 1 fimbriae, and therefore effect on the adhesion of K. pneumoniae CG43.
38
3.1.6 RcsB D56A as well as RcsB promoted oxidative stress
resistance of K. pneumoniae CG43 ∆rcsB
Once entering to the host, bacterial pathogens must circumvent the attack of reactive oxygen species produced by the immune cells. Therefore, to examine whether RcsB phosphorelay status could play role in the regulation of oxidative stress response, each overnight grown strain was refreshed cultured until OD600 around 0.3-0.4, and 100µl of each culture
spread on LB agar plate supplemented with tetracycline. A nitrocellulose disc containing 10µM of paraquat, a superoxide generator, as then placed onto the bacterial lawn. All plates were incubated at 37°C for 16 hr., the diameters of the inhibition zone were measured and recorded. As shown in Fig. 6, the inhibition zones on the LB agar plate with the ∆rcsB[pRK415-RcsB] and ∆rcsB[pRK415-RcsB-D56A] were smaller in comparing to the other strains. This indicated that the unphosphorylated RcsB might be able to restore the oxidative stress resistivity of K. pneumoniae CG43.
39
3.1.7 RcsB D56A as well as RcsB restored the acid stress
resistance activity under shaking cultured condition
Previous study revealed that, rcsB deletion decreases the ability of glutamate-dependent acid resistance in E. coli dramatically [76]. However,
K. pneumoniae is unable to survive in pH 2.5 M9 medium. Therefore, all K. pneumoniae CG43 cultures were undergoing adaption at pH 4.4 in LB
medium according to the related study [123],. The acid resistance of K.
pneumoniae CG43 was determined after 90 min challenge with pH 3.0 in
M9 medium. The result in Fig. 7A showed that the ∆rcsB[pRK415-RcsB] and ∆rcsB[pRK415-RcsB-D56A] complementation strains restored the bacterial survival rates significantly, while ∆rcsB mutant strain that transformed with RcsB-D56E only recovered slightly level of acid survival rate. Previous study showed that, the level of rcsB expression reported by the promoter activity analysis of PrcsDB is not affected by the alteration of
pH value in LB shaking culture. Using bioinformatics tool, 6 RcsB-dependent genes were identified. One of these genes, namely yfdX had been reported to endode a chaperone required for the acid stress response. The ability of acid resistance was significantly decreased in K. pneumoniae CG43 ∆yfdX mutant [112], indicating that yfdX is involved in the acid
40
resistant system. As shown in Fig. 7B and Fig. 7C, the acid survival rate and the western blot hybridization with anti-YfdX revealed that, both ∆rcsB[pRK415-RcsB-D56E] and ∆rcsB[pRK415-RcsB-D56A] complementation strains restored YfdX expression under microaerobic incubation. However, in highly aerated culturing condition, only the ∆rcsB[pRK415-RcsB-D56A] complementation strain was able to restore the expression of YfdX. It is hence that, the unphosphorylated RcsB is required for the acid resistance, while in the microaerobic environment, acid resistant system could be regulated neglecting the phosphorelay status of RcsB in microaerobic environment.
3.1.8 Effect of different phosphorylation states of RcsB on
yfdX gene expression.
Due to there are several RcsB binding boxes among the yfdX promoter region were predicted previously [112], we aimed to determine whether unphosphorylated or phosphorylated RcsB could regulate the yfdX transcription. The promoter region of yfdX gene was transcriptionally fused to the reporter plasmid, carrying a lacZ reporter gene. This recombinant
41
plasmid (PyfdX::lacZ) was then transformed into K. pneumoniae strains
CG43S3∆lacZ and ∆rcsB∆lacZ individually. The promoter activity measurements shown in Fig. 10A reveal that, the deletion of rcsB reduced the activity of PyfdX::lacZ. As shown in Fig. 10B, the complementary effects
of RcsB are different under different oxygen level of cultivation. Under static culture (microaerobic), all the RcsB complementary strains up-regulated the PyfdX activity comparing to ∆rcsB mutant strain, while
RcsB-D56E complemented strain was unable to recover PyfdX activity during
shaking culture (sufficient oxygen). This suggests that the PyfdX activity is
mediated by RcsB in acidic condition, whereas this regulation manner is determined by the phosphorylation state of RcsB only when K. pneumoniae is cultured with sufficient oxygen.
3.2 Different roles of phosphorylated and unphosphorylated
RcsB.
According to the results of the phenotype assays, we proposed a model of RcsB to show the likely regulatory pathway on different phenotypes or virulence factors in K. pneumoniae CG43. Urinary catheters
42
are standard medical devices utilized in both hospital and nursing home settings, but are associated with a high frequency of catheter-associated urinary tract infections (CAUTI). In particular, biofilm formation on the catheter surface by K. pneumoniae causes severe problems. Previous study showed that, type 1 and type 3 fimbriae expressed by K. pneumoniae enhance biofilm formation on urinary catheters in a catheterized bladder model that mirrors the physico-chemical conditions present in catheterized patients [143]. In this study, our result presentedRcsB-D56E increased the expression of type 1 fimbriae major pilin FimA, and type 3 fimbriae major pilin MrkA, suggesting that constitutively phosphorylated RcsB might induces the biofilm development via positively regulating the expression of type 1 and type 3 fimbriae. As known, mrkH, mrkI, and mrkJ are the regulatory genes of type 3 fimbriae, and a few RcsB binding boxes could be found on the putative promoter region of mrkH and mrkI. It is hence we propose that RcsB might regulates type 3 fimbriae via interacting with these regulatory proteins. Curli fibers are a major adhesin factor to surfaces, and also affect cell aggregation and biofilm formation in many enterobacteria, such as Salmonella and pathogenic E. coli strains [144-147]. Expression of both curli fibers and cellulose depends on the CsgD protein,
43
a response regulator of the LuxR family. The CsgD is also a positive regulator for the expression of type 3 fimbriae [148]. If RcsB interacts with CsgD for regulating the type 3 fimbriae expression remains to be investigated.
The complementation with RcsB-D56A recovered and/or increased CPS production, oxidative stress and acid stress responses, which suggesting that, the unphosphorylated form of RcsB plays positive role for these genes expression. Therefore, we proposed the unphosphorylated RcsB is important in the stress resistance response regulation that protects the bacteria in K. pneumoniae CG43 from stress damages.
3.3 Identification of the RcsB interacting protein under acidic
condition.
44
3.3.1 RcsB interacting proteins in the regulatory pathway of
CPS synthesis are different from that involved in
regulating the acid stress resistance.
The interacting proteins interacting with RcsB in the regulation of the CPS synthesis are known in K. pneumoniae. Here, we intended to examine whether the RcsB interacting proteins for the CPS synthesis regulation are also required for the acid stress resistant regulation. As shown in Fig. 10C, neither ∆rcsA, ∆rmpA nor ∆rmpA2 mutant strain shows the same effect on the YfdX expression as that of ∆rcsB in Western blot analysis using anti-YfdX serum. The result indicates that the RcsB on the regulation of the CPS synthesis and acid stress resistant require different interacting proteins.
3.3.2 Generate an antibody specific to RcsB N-terminal region
In order to capture the RcsB interacting protein in co-immunoprecipitation assay, the first 343 nucleotides of rcsB was fused into an expression plasmid pET30a, and the resulting recombinant plasmid pET30a-RcsB-N (Table 2) was used to transform E. coli BL21 (DE3). The expression condition of the recombinant RcsB N-terminal was then45
analyzed. As shown in Fig. 11A, an IPTG-induced overexpression of the His6-RcsB-N could be observed. The purified His6-RcsB-N of
approximately 17 kDa (Fig. 11B) was used to immunize rabbit for raising anti-RcsB antibody. The specificity of anti-RcsB-N antibody was tested by a 10000-fold dilution at room temperature. As shown in Fig. 12, the anti-RcsB-N antibody could specifically bind to the recombinant RcsB N-terminal polypeptide.
3.3.3 Expression of RcsB under acidic growth conditions
The RcsB-N antiserum was generated from the rabbit with 13 boosts of immunization. The RcsB has to be highly expressed for the isolation of the RcsB interacting protein by co-immunoprecipitation. According to Fig. 13, RcsB was induced in K. pneumoniae CG43 under acid treatment. This can be observed for ∆rcsB[pRK415-RcsB] and ∆rcsB[pRK415-RcsB-D56A], but not in ∆rcsB[pRK415-RcsB-D56E]. Nevertheless, Coomassie blue staining of the SDS-PAGE samples as shown in Fig. 13B revealed that the protein expression of ∆rcsB[pRK415-RcsB-D56E] was similar to
46
∆rcsB[pRK415-RcsB-D56A] when the bacteria cultivated under static culture. This implies that the raised antibody could not recognize the acid induced expressed RcsB-D56E protein.
47
4. Discussion
The complementation analysis of rcsB deletion showed a similar trend in most of the phenotype assays between the activity of RcsB and RcsB-D56A. Both RcsB and RcsB-D56A decrease the growth, while increase the CPS production, acid stress survivals, oxidative stress response, biofilm formation on glass tube. By contrast, RcsB-D56E increased the production of type 1 and type 3 major pilin. The increased expression of CPS, which causing energy consumed, probably renders the bacteria slow growth rate. The acid survival analysis revealed that oxygen level of the growth condition is an important factor for the RcsB-dependent acid stress resistance and this might be closely dependent on the acid stress response-related protein, YfdX. The promoter activity assay further confirmed that RcsB regulates YfdX expression at the transcriptional level. In summary, we found that the phosphate signaling relay on RcsB confers both positive and negative regulation on the downstream genes.
RcsB had been studied for several regulations in other
