可見光飛秒雷射對細菌活性降低之研究

全文

(1)國立台灣師範大學物理研究所博士論文. 可見光飛秒雷射對細菌活性降低之研究 Study of the Bactericidal Effects of a Visible Femtosecond Laser on Escherichia coli. 指導教授：徐鏞元博士研究生：呂杰翰撰中華民國壹百零三年九月 1.

(2) 摘要. 關鍵字: 可見光飛秒雷射、細菌活性降低、衝擊受激拉曼散射、去氧核醣核酸鬆弛. 在近年來的研究中，可見光飛秒雷射被發現可應用於降低廣泛種類的微生物活性，其作用機制亦被發現與功率密度與雷射脈衝寬度有關，然而對於飛秒雷射與細菌間的交互作用而言，詳細的作用機制與理論基礎上仍有懸而未解的問題，在這篇論文中，將探討經雷射照射後大腸桿菌（Escherichia coli ）的細胞膜表面性質與完整性，細菌的新陳代謝率以及質體去氧核醣核酸的構形變化，我們的研究結果顯示當受到 60 分鐘的雷射照射後，細菌將出現細胞質洩漏、蛋白質聚集的現象，以及細胞膜的物理性質改變，同時亦能觀察到一受到雷射功率密度倚變的超螺旋質體去氧核醣核酸之弛豫現象。而在 10 分鐘的短時間雷射照射下，細菌的有氧葡萄糖細胞呼吸率可在細胞質洩漏並未被觀測到的情況下損失 75%的活性，針對細胞呼吸電子傳遞鍊的進一步測試中，氧化還原酶的測試結果顯示飛秒雷射對於不同種類的酶與輔酶分子造成程度不一的破壞，細菌經雷射照射後，該迅速產生的呼吸抑制效應被認為在細菌活性降低的早期過程中扮演著重要的角色。. 2.

(3) Abstract Key words: Visible femtosecond laser, bacteria inactivation, ISRS, DNA relaxation. Visible femtosecond laser is shown to be capable of selectively inactivating a wide spectrum of microorganisms in a power density and pulse width dependent manner. However, the mechanism of how visible femtosecond laser affects the viability of bacteria is still elusive. In this thesis, the cellular surface properties, membrane integrity, metabolic rate and plasmid DNA conformation of Escherichia coli (E. coli) irradiated by a visible femtosecond laser with different power density and exposure time were investigated. Our results showed that femtosecond laser treatment for 60 minutes (min) led to cytoplasmic leakage, protein aggregation, and alternation of the physical properties of E. coli cell membrane. A power density dependent genetic damage from laser induced relaxation of supercoiled plasmid DNA was observed as well. In comparison, a 10 min exposure of bacteria to femtosecond laser irradiation induced an immediate reduction of 75% of the glucose-dependent respiratory rate, while the cytoplasmic leakage was not detected. Results from enzymatic assays showed that oxidase and dehydrogenases involving in E. coli respiratory chain exhibited divergent susceptibility after laser irradiation. This early commencement of respiratory inhibition after a short irradiation is presumed to play a dominant effect on the early stage of bacteria inactivation.. 3.

(4) Contents CHAPTER 1 INTRODUCTION. 11. CHAPTER 2 THEORY. 15. 2.1 PRINCIPLE OF IMPULSIVE STIMULATED RAMAN SCATTERING. 15. 2.2 STRUCTURE OF E. COLI. 20. 2.3 AEROBIC RESPIRATORY OF E. COLI. 21. 2.4 COMMON ENZYMES AND CO ENZYMES IN RESPIRATORY CHAIN. 23. 2.4.1 Dehydrogenases. 23. 2.4.2 Quinone. 24. 2.4.3 Cytochrome bo oxidase. 24. CHAPTER 3. 25. MATERIAL AND METHOD. 25. 3.1 SETUP OF FEMTOSECOND LASER INACTIVATION SYSTEM. 25. 3.2 ATOMIC FORCE MICROSCOPY (AFM). 26. 3.2.1 Poly-l-lysine Mica Preparation and Bacteria Immobilization. 26. 3.2.2 Air mode AFM. 27. 3.2.3 Liquid mode AFM. 27. 3.3 BACTERIA PREPARATION; PROTEIN & PLASMID EXTRACTION. 28. 3.3.1 Bacteria Incubation and Viability Assay. 28. 3.3.2 Membrane Fraction Preparation. 28. 3.3.3 Soluble Protein Extraction. 29. 3.3.4 Plasmid DNA Extraction. 29 4.

(5) 29. 3.4 ELECTROPHORESIS 3.4.1 Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis. 29. 3.4.2 DNA Agarose gel electrophoresis. 30. 3.5 QUANTIFICATION OF CELL LEAKAGE AND FLORESCENCE IMAGING. 30. 3.5.1 Spectroscopic quantification. 30. 3.5.2 Florescence imaging. 30. 3.6 RESPIRATION ASSAYS. 31. 3.6.1 Glucose Dependent Respiratory Assay. 31. 3.6.2 Oxidase Assays. 31. 3.6.3 Dehydrogenase Assays. 32. CHAPTER 4 EXPERIMENTAL RESULTS. 33. 4.1 INACTIVATION OF E. COLI BY 415NM FEMTOSECOND LASER. 33. 4.2. FEMTOSECOND LASER ALTERS THE SURFACE PHYSICAL PROPERTY OF BACTERIA 34 4.3. FEMTOSECOND LASER CAUSED LEAKAGE OF CELLULAR SUBSTANCES. 37. 4.4. FEMTOSECOND LASER ALTERS BACTERIAL PROTEIN EXPRESSION PROFILE. 40. 4.5. SHORT TIME LASER EXPOSURE AFFECTS BACTERIAL RESPIRATORY RATE. 43. 4.6. EFFECTS OF THE LASER ON MEMBRANE-ASSOCIATED RESPIRATORY ENZYMES. 45. 4.7 FEMTOSECOND LASER INDUCED PLASMID DNA RELAXATION. 48. 4.7.1 Assignment of the Bands in Electrophoresis. 48. 4.7.2 Laser Effect on the Plasmid DNA pCR II-TOPO. 49. 4.7.3 Laser Effect on the Plasmid DNA pBluescript. 52. 4.7.4 Laser Effect on the Plasmid DNA pUC 19. 53. 5.

(6) CHAPTER 5 DISCUSSION. 55. 5.1 NON-GENETIC DAMAGE. 56. 5.2 GENETIC DAMAGE. 59. CHAPTER 6. 67. CONCLUSIONS. 67. REFERENCE. 71. 6.

(7) List of Figures Figure 1. Pictorial demonstration of how the protein subunits are disintegrated by ISRS effect.. 19. Figure 2. Structure of the E. coli envelope. 21. Figure 3. Pictorial model of the respiratory enzyme located on the cytoplasmic membrane of bacteria.. 22. Figure 4. Schematics of the femtosecond laser microorganism inactivation system. 26 Figure 5. Increasing visible femtosecond laser power density or exposure time reduced E. coli cell viability.. 33. Figure 6. Cell topography change after laser irradiation.. 35. Figure 7. Air mode AFM topographical images. 35. Figure 8. Liquid mode AFM mechanical properties mapping; Young’s modulus and Adhesive force. 37. Figure 9. Exposure to femtosecond laser caused E. coli cellular nucleic acid leakage. 38 Figure 10. Exposure to femtosecond laser caused E. coli cellular nucleic acid leakage. 39 Figure 11. Exposure to femtosecond laser caused E. coli cellular nucleic acid and protein leakage.. 39. 7.

(8) Figure 12. The absorption spectra of supernatant from E. coli suspension irradiated by laser for 10 min under the power density of 1.7 GW/cm2.. 40. Figure 13. Exposure to visible femtosecond laser caused soluble protein aggregation. 41 Figure 14. Exposure to visible femtosecond laser (415nm) caused membrane protein aggregation.. 42. Figure 15. Exposure to visible femtosecond laser (830nm) caused membrane protein aggregation.. 43. Figure 16. Immediate reduction of cell respiratory rate followed short exposure to femtosecond laser.. 44. Figure 17. Effect of the brief (10min) femtosecond laser on dehydrogenase. 46. Figure 18. Effect of the brief (10min) femtosecond laser on oxidase. 46. Figure 19. Effect of the long (1hour) femtosecond laser on dehydrogenase.. 47. Figure 20. Effect of the long (1hour) femtosecond laser on oxidase.. 48. Figure 21. Assignment of the band profile in DNA gel electrophoresis.. 49. Figure 22. 415 nm Laser induced DNA relaxation in pCR II plasmid.. 50. Figure 23. 830 nm Laser induced DNA relaxation in PCR II plasmid.. 51. Figure 24. Damping effect on the laser induced DNA relaxation in pCR II plasmid. 51 Figure 25. Power density dependent on the laser induced DNA relaxation in pBS plasmid.. 52. Figure 26. Power density dependent on the laser induced DNA relaxation in pUC 19 plasmid.. 54 8.

(9) Figure 27. Dependence between plasmid size and threshold power. 60. Figure 28. Examination on the specific cleavage site on pBS plasmid. 63. Figure 29. A pictorial model of the interaction between femtosecond laser and E. coli. 68 Figure 30. A concept picture of the visible femtosecond laser treatment.. 9. 70.

(10) List of Tables Table 1. Summary of the cell properties acquired by AFM ......................................... 36 Table 2.List of the detrimental effects induced by femtosecond laser ......................... 55 Table 3. Summary of the conformation transitions induced by the irradiation of femtosecond laser. ........................................................................................................ 64. 10.

(11) Chapter 1 Introduction Research efforts focusing on finding alternative antibacterial therapeutics have been raised due to the side effects and persisting antibiotic resistance of the clinically adopted antibiotics [1, 2]. Existing disinfection technologies, for example, irradiation of ultraviolet light, gamma-ray, microwave absorption, pharmaceutical treatments, and photodynamic therapy all have limit of application when treating the infection caused by antibiotic resistance bacteria. UV light and gamma ray are effective bactericidal technique. However, UV and gamma irradiation have no selectivity between harmful bacteria and tissue cells because the high energy photon cause cross linking or dissociation of chemical bonds along its propagation path. Another drawback of these two radiation treatments is the risk of genetic mutation both on the infected tissue and bacteria strains since the photon energy is so high that could break the covalent link within nucleic acids. A strategy to overcome the genetic mutation induced by high energy photon is to make use of the microwave resonant absorption of the capsid of microorganisms, which is typically in the range of 10 GHz to 1 THz [3-7]. Maneuvering the property of microwave absorption can effectively transfer the microwave energy to the vibrational energy of microorganism capsid. However, most pathogenic bacteria are waterborne or coexist with water, and water severely absorb microwave energy in this frequency range. The application of photodynamic therapy technology on eliminating pathogenic microbial serves as a good example of a successful tactic in killing antibiotic-resistant 11.

(12) bacteria by using a combination of photosensitizer and low intensity visible light[8, 9]. However, side effects such as allergic dermatitis and porphyria have been reported and the carcinogenic potential of certain photosensitizers has been discovered [9, 10]. In recent years, an emerging pathogen elimination method using a visible femtosecond laser had been reported to be capable of selectively inactivating a wide spectrum of microorganisms, including bacteria, enveloped and nonenveloped viruses on a wavelength and pulse width dependent manner [11-22] Maneuvering of this laser technology has several advantages over the other existing pathogen elimination treatments. First, it is environmentally friendly, because no introduction of chemical or nanoparticle based-reagents is needed. Moreover, previous reported experimental data shown that there exists a therapeutic window in laser power density between 1 GW/cm2 and 10 GW/cm2 which allows the inactivation of a variety of pathogens while leaving mammalian cells unharmed. Second, it is a general disinfection strategy for a variety of viral and bacterial pathogens since the technology has been demonstrated to be effective on the enveloped and non-enveloped, single-stranded, double-stranded DNA or RNA viruses and grampositive and gram-negative bacteria [13]. Third, its antimicrobial efficacy is likely as effective for the original strain as for the mutated ones since damage caused to microorganisms through vibration of the mechanical structures likely would not be immune to mutation of cell surface receptors. On the contrary, with the pharmaceutical treatment, a new drug is usually needed to fight with mutated strain of microorganisms. Thus the approach would be applicable not only to the raw strains but also to anti-biotic ones of bacteria [13]. Previous study on M13 bacteriophage and murine norovirus, which are nonenveloped viruses, showed that the capsid proteins of the viruses were disintegrated by. 12.

(13) the laser irradiation [11, 22]. On the other hand, for the enveloped viruses, the experimental results indicated that the aggregation of capsid proteins and the subsequent inhibition of capsid function was the cause of the inactivation [12]. The fundamental effect a femtosecond laser exerted on these two kinds of viruses can both be attributed to the disruption of hydrogen bonds and/or hydrophobic interactions through the Impulsive Stimulated Raman Scattering (ISRS) [11, 12]. In nowadays, pathogenic bacteria have raised a public health challenge because of the development of antibiotic resistance. Previous research has suggested that the relaxation of supercoiled plasmid DNA caused by femtosecond laser treatment can induce genetic damage, resulting in the inactivation of Salmonella typhimurium [14]. Nevertheless, the time dependency and the inactivation mechanism of bacteria by the femtosecond laser irradiation remained unclear. In this study, the cellular surface properties, membrane integrity and metabolic rate of Escherichia coli (E. coli) irradiated by a visible femtosecond laser with different exposure time and excitation power density were investigated. Our results demonstrated that the surface physical properties such as the stiffness (Young’s modulus) and the adhesive force of the treated bacteria were altered when E. coli exposed to a femtosecond laser for 60 min. In addition, cytoplasmic leakage, protein aggregation, genetic damage, and shrinkage of cell volume as well as respiratory inhibition were observed with the femtosecond laser treated bacteria as well. These results suggest that membrane protein structure alteration due to protein aggregation may be the cause of cytoplasmic leakage and cell volume reduction of the laser-treated bacteria. In comparison, a 10 min exposure of bacteria to femtosecond laser irradiation induced an immediate reduction of 75% of the glucosedependent respiratory rate, while the cytoplasmic leakage was not detectable. Further enzymatic activity assays showed that the oxidases and dehydrogenases involving in E. coli respiratory chain demonstrated divergent susceptibility after a short femtosecond. 13.

(14) laser irradiation. This compromised respiratory enzymes functions may play a role in the early stage of bacteria inactivation by the visible femtosecond laser. A characterization of the dependence of plasmid DNA relaxation process on the excitation power density is conducted. Our observation shown the threshold power density that trigger the relaxation of supercoil structure is correlated with the length of the irradiated DNA. The relaxation of supercoiled DNA is believed to induce genetic error in the cellular replication process leading to the long term reduction of viability.. 14.

(15) Chapter 2 Theory 2.1 Principle of Impulsive Stimulated Raman Scattering. The normal modes is the vibration that objects oscillates around its equilibrium geometry with resonant vibration frequencies. For biomolecules, the global collective motion is of the order of 10 GHz to 1 THz [3, 5, 7]. In previous researches, the feasibility of exciting the collective motion of an influenza virus through the microwave waveguide is demonstrated [23, 24]. However, in most physiological states, the pathogenic microorganisms coexist with water, which is not transparent to this wavelength regime of electromagnetic wave. Alternatively, the normal mode can be excited by choosing a visible or infrared ultrafast laser with the pulse width to be equal to or shorter than the oscillation period of the specific mode of the targeted molecule. The advantage of using the visible or infrared laser is the laser energy would neither be absorbed by water nor the tissues in the infected area comparing to the microwave electromagnetic wave. Once the Stoke-shifted frequency of the normal mode is included in the spectral content of the pulse laser, the biomolecules in the exposure area are able to be brought into resonance. In this section, the principle of ISRS and its effect onto a microorganism system (virus) will be introduced concisely. A comprehensive introduction of both the experimental result and theoretical model of this topic has been published as a review paper. [25]. 15.

(16) Considering a simplified model of the vibrational normal mode represented by a normal coordinate labeled Q. The dispersion in the index of refraction is ignored. Also, the incident electric field is assumed not depleted or attenuated by the stimulated scattering. The equation of motion for Q can be written as [25]:  2Q Q 2  2   0 Q  F (t ) 2 t t. (1). Where γ is the damping constant, ω0 is the resonant angular frequency of vibration, and F (t) is the impulsive force caused by the external electrical field of the pulse laser. The external electrical field ⃑⃑⃑⃑ 𝐸𝐿 induces a polarization on the excited molecule due to the polarizability α of the molecule as   P(Q, t )  NEL. (2). The Taylor expansion of the polarizability α can be written as 1 2.  (Q)   0   ' 0 Q   ' ' 0 Q 2  .... (3). Where α0 is the zero order term contributing to the Rayleigh scattering,     Q is the first order term which results in the first Raman scattering  Q  0.  ' 0 Q  . process, and. 1 1   2  ' ' 0 Q 2   2 2 2  Q.  2  Q is the second term. 0. On the other side, the force F(t) can be represented as F (t )  . U (Q, t ) Q. (4). And the potential energy U(Q,t) can be written as.   U (Q, t )   12 P(Q, t )  EL (t ). (5). 16.

(17) Substituting (2), (3), (4), (5) into (1) while keeping up to the first order term in polarizability. We will get.  2Q Q 1  2 2  2  0 Q  N ( ) 0 E L 2 t t 2 Q. (6). The right hand side of equation (6) is the external driving force of the normal mode vibration. As the force is linear proportional to the first derivative of polarizability with respect to normal coordinate Q, it gives the explanation why the process is called impulsive Raman scattering. For excitation by a single-beam ultrashort laser having a pulse width of τL and 𝑡2. intensity 𝐼(𝑡) = 𝐼0 𝑒. −( 2 ) 𝑡 𝐿. , the displacement of the vibration can be shown to be. Q(t )  Q0 e t sin0 t . (7). Of the most importance, Qo is the maximum amplitude of the displacement away from the equilibrium produced by ISRS and is given by. Q0 (t ) . 2 2  n  ' 0 L I 0 e  ( 0  L / 4 ) 2 cK 0 0. . (8). Here I0 is the peak intensity of the excitation laser, n is the index of refraction of light within the medium, c is the speed of light, and Kε0 is the dielectric permittivity of the medium. As shown in equation (7), the displacement induced by ISRS is proportional to the intensity of the excitation laser and Raman scattering cross section. When the laser power density is sufficiently large, the resonant amplitude could be strong enough to produce irreversible displacement of some atoms, leading to the damage of the weak links such as hydrogen bonds or hydrophobic interaction related to a low frequency vibration.. 17.

(18) The presence of the term e  (0 L / 4 ) indicates that the pulse width τL has to be 2 2. chosen so that ω0τL <1 in order to produce a significantly large amplitude. In particular, the pulse width of cw (continuous wave) laser is infinitely large. Hence the exponential term in (7) approaches to zero, which suggests that cw laser cannot excite the resonant vibration of the molecules. Therefore, an ultrafast laser that has the pulse width of subpicosecond to femtosecond is able to produce a non-vanishing e  (0 L / 4 ) term and a 2 2. large atomic displacement. This mechanical impact could coherently excite the Raman vibrational modes of a viral capsid of M13 bacteriophage. As shown in the figure 1(a), the viral capsid is composed of alpha helix subunit that is connected with each other by hydrophobic contact or Van der Waal force. When the laser energy brings molecules into oscillation, the resonant vibration amplitude may exceed the limit of bond length (figure 1(b)), leading to the damaging or disintegrating of the capsid (figure 1(c)).. 18.

(19) Figure 1. Pictorial demonstration of how the protein subunits are disintegrated by ISRS effect.. 19.

(20) 2.2 Structure of E. coli. Escherichia coli is a Gram-negative, facultative anaerobic, bacterium that often appears in the intestine of warm-blooded organisms. Generally speaking, most of the E. coli strains are harmless, but some serotypes are pathogenic strains that can cause diarrhea. One of the most notorious strain is O157:H7, which induces bloody diarrhea and kidney failure. The transmission of E. coli is via the fecal-oral route and the outbreak was often related to the distribution of contaminated foods. According to the report of WHO, the antibiotic resistance of E. coli to the fluoroquinolones, which is the most used widely antibacterial drugs for the oral treatment or urinary tract infection is very widespread [2]. The emerging of antibiotic resistant E. coli strains will consume more health-care resources and induce a significant public health problem in nowadays and near future. E. coli cell is typically rod-shaped and about 2 μm in length and 1μm in diameter. The E. coli cell is surrounded by a membrane. This membrane acts as a barrier to hold nutrients, proteins, DNA and other essential components of the cytoplasm within the cell. Cell envelope of E. coli is composed of cytoplasmic membrane, cell wall, and outer membrane (figure 2). The outer membrane includes lipid layer, porin protein and lipopolysaccharide (LPS), which is also regarded as the endotoxin of the bacteria. LPS has been implicated in many cell activities, such as surface adhesion, interaction with other cell and bacteriophage, and protecting the membrane from chemical attack. Cell wall is a rigid, mesh like layer made of peptidoglycan, which is a polymer composed of sugars and amino acids, located in the periplasmic space between the two selective permeable membranes. The cell wall is considered to be related to the resistance of antibiotics and the maintenance of the rod shape of bacteria. Size of the periplasmic space depends on the physiological environment and may constitute to 40% of the total. 20.

(21) cell volume. In addition, it includes solutes such as ions and proteins, which are essential to a variety of functions from nutrient binding, substance transport, electron transport, and the generation of proton gradient. The cytoplasmic membrane has large content of proteins, which are important and responsible for biological activities. Some of the proteins are enzymes involved in the metabolism activity of the bacteria and are crucial to the cell energy production.. Figure 2. Structure of the E. coli envelope. 2.3 Aerobic Respiratory of E. coli. Aerobic respiration is the combination of a series of reduction reactions of the respiratory enzyme molecules. Electrons enter the reaction from an electron donor and are transferred to the terminal electron acceptor, O2. The process generate an electrical potential across the cytoplasmic membrane which is called proton motive force. The. 21.

(22) potential difference play a key role in the synthesizing ATP through the transmembrane ATP synthase. The E. coli aerobic electron transport chains consist two categories of respiratory enzymes, dehydrogenases and oxidases [26, 27]. Electron donor, such as NADH, releases electrons into electron transport chain via a specific dehydrogenase which only reacts with the certain substrate. Cytochrome oxidase accept the electrons from dehydrogenase via a mediate electron carrier and reduces the terminal acceptor O2 to H2O. The mediate electron carrier is the cofactor ubiquinone (Q), which is a lipophilic molecule buried in cytoplasmic membrane. A pictorial model of the respiratory chain is depicted in figure 3. The overall reaction can be represented as DH + O2 → D + H2O Where D is the electron donor.. Figure 3. Pictorial model of the respiratory enzyme located on the cytoplasmic membrane of bacteria. For example, when one molecule of glucose is metabolized via the glycolysis to pyruvate, two molecules of NADH are formed. NADH then serves as the electron donor in the respiratory electron transport chain.. 22.

(23) 2 NADH + 2 H+ + O2 → 2 NAD+ + 2 H2O The pyruvate enters to the TCA cycle and further oxidizes to CO2. Full TCA cycle will provides additional NADH that donates more electrons into respiration chain. Since E. coli was the dominant microbe in the large intestine of human where most of the nutrition have been absorbed in the upper digest tract. The persistence of E. coli in the gut reflects the high efficiency with which E. coli exploit carbon sources from the environment that only has limited nutrition. The bacterium can use a variety of electron donors and its respiratory system can adapt to environmental and bioenergetics demands. Therefore, most dehydrogenases in respiratory chain show induced expression in response to metabolic needs effected by the environment.. 2.4 Common Enzymes and Co Enzymes in Respiratory Chain. 2.4.1 Dehydrogenases. NADH is the most common electron donor in the respiratory chain. NADH dehydrogenase catalyzes the reaction that electrons are transferred to quinone. The membrane enzyme simultaneously forms a proton electrochemical gradient through pumping protons from cytoplasm to the periplasmic space. Succinate dehydrogenase participates in aerobic respiratory chain by catalyzing the transport of electron from succinate to quinone. The enzyme plays a role connecting the TCA cycle to the electron transport chain.. 23.

(24) Lactate dehydrogenase is an enzyme that catalyzes the oxidation of lactate. The electron from the lactate transport to quinone via the subunit of lactate dehydrogenase and subsequently transport to the cytochrome oxidase. Glycerol 3-phosphate dehydrogenase catalysis the oxidation of glycerol-3phosphate to dihydroxyacetone phosphate. The membrane enzyme enables E. coli make use of Glycerol 3-phosphate dehydrogenase as electron donor in electron transport chain.. 2.4.2 Quinone. E. coli and related enteric bacteria synthesize three different quinones, ubiquinone and the naphtho-quinones menaquinone and demethylme-naquinone. The ubiquinone is the predominate quinone to transport electron within the cytoplasmic membrane in the aerobic respiration. It serves as a mediator between dehydrogenases and terminal oxidases and has a midpoint potential in the oxidoreduction reaction in the electron transport chain. 2.4.3 Cytochrome bo oxidase. The cytochrome bo oxidase mediates the electron oxidation of ubiquinol with the electron reduction of oxygen. This oxidase is most abundant in the cell grown in an aeration condition. This enzyme is an integral transmembrane protein which generates a proton motive gradient by consuming protons in the cytoplasm.. 24.

(25) Chapter 3 Material and Method 3.1 Setup of Femtosecond Laser Inactivation System. Optical setup of femtosecond laser system is presented schematically in figure 4. The excitation source employed in this work was a diode pumped mode-locked Tisapphire laser. The laser produced a continuous train of about 70 fs pulses at a repetition rate of 80 MHz. The output of the second harmonic generation system at wavelength of 415 nm was used to irradiate the sample. It has a maximum average power of about 250 mW and a pulse width of full-width at half maximum (FWHM) ≅ 100 fs. An achromatic lens was used to focus the beam into a glass vial containing bacteria suspension. An estimation of the focused spot size is made by knife edge method. The peak power density is defined as that at the tightest focused region formed by the focusing lens. For example, a laser spot of 30μm and average power of 100mW gave a peak power density of 1.7GW/cm2. The power densities varies from ≅ 20 MW/cm2 to ≅ 4.3 GW/cm2 were achieved by adjusting the average power with a calibrated variable attenuator. In our setup, the active volume that laser could interact efficiently with bacteria is a cylinder having approximately 30μm in diameter and 1.4mm in length. A magnetic stirrer was placed within the vial and driven by a rotated permanent magnet below to ensure the laser fluence is evenly deposited into the solution. The assays were conducted after the exposure of laser for several minutes to hours. All the experimental results reported here are obtained at T = 20◦C and with single beam excitation. 25.

(26) Figure 4. Schematics of the femtosecond laser microorganism inactivation system. The arrow denotes the propagation of the laser.. 3.2 Atomic Force Microscopy (AFM). 3.2.1 Poly-l-lysine Mica Preparation and Bacteria Immobilization. In order to image bacteria with AFM, a reliable immobilization method is required to ensure the cell is stable under the interaction with tip force. Immobilization 26.

(27) of bacteria on poly-l-lysine treated mica is demonstrated to be a reliable method for AFM both in air and liquid [28, 29]. We followed the protocol of preparing substrate proposed by Bolshakova et al. with slight modifications[29]. 1ml of 0.01% Poly-llysine solution (P4707 Sigma-Aldrich) was dropped on the fresh cleave mica (SPI supplies) and spread evenly on the surface. The mica discs were then put in an oven for 2 hours in 60。C. To immobilize bacteria, 20 μl of bacteria suspension were placed on mica and incubated for 20 minutes at room temperature. The mica carrying immobilized bacteria were immediately placed in a liquid cell with 0.5 ml PBS for AFM investigation.. 3.2.2 Air mode AFM. The tapping mode AFM in air is performed by operating Solver P47H atomic force microscope of NT-MDT. Cell topography of the E. coli cell was scanned with PPP-SEIHR probe from Nanosensor, which had a spring constant of 15 N/m and resonant frequency of 130 kHz.. 3.2.3 Liquid mode AFM. The liquid mode AFM in this study was collected by using Bruker DNP-10 cantilever, with a nominal spring constant of 0.06 N/m and tip diameter of 2 nm. The microscope was operated in Peak Force QNM mode which enabled the simultaneous measurement of topography imaging and the quantification of biomechanical properties, such as adhesive force and Young’s modulus. The scan rate was 0.5 Hz and the maximum applied force was limited to 1 nN to prevent the plastic deformation of the bacteria structure. Data processing and the extraction of mechanical properties are done with NanoScope Analysis from Bruker. The nonlinear part in the extraction force-. 27.

(28) distance curve was fitted with DMT model to obtain Young’s modulus of the cell membrane. The intercept between the no-contact line and the minimum point is defined as the maximum adhesive force between the tip and cell surface. Topographical images are presented with contrast adjustment and background removal with the open source software Gwyddion [30].. 3.3 Bacteria Preparation; Protein & Plasmid Extraction 3.3.1 Bacteria Incubation and Viability Assay The TOP10 E. coli strain (Invitrogen) was utilized in this study. A single bacteria colony picked from LB agar plate was incubated in 4 ml liquid LB at 37。C for 16 hours, then subcultured by a 1:50 dilution in fresh LB liquid to mid-log phase (O.D. =0.6) and harvested by centrifuge at 5000× g for 1 min. Cells were washed twice and then suspended in 1 ml phosphate buffered saline (PBS) and loaded into glass vials for the subsequent irradiation. The irradiated bacteria were diluted in series, spread on LB agar plates and incubated at 37。C for 16 hours. The numbers of viable bacteria were counted as the numbers of colonies formed on LB plates.. 3.3.2 Membrane Fraction Preparation. The membrane fractions were prepared by sonication overnight incubated bacterial cells with cell disruptor following the protocol used by Lönnerdal B.et al [31]. Bacteria cells were suspended in 30 mmol l-1 Tris-HCl buffer ( pH 8.0 ), and sonicated on ice with a cell disruptor for five cycles for 30 sec at intervals of 30 sec, at a power set at 3. After 20 min if incubation at 37°C, the unbroken bacteria were removed by low speed centrifugation at 800×g for 5 min at 4°C. At the last step, membrane fractions were separated from cytoplasmic material by centrifugation at 3000× g for 30 min at 28.

(29) 4°C. The membrane will be suspended into Tris-HCl buffer for subsequent assay. The protein concentration was determined by Bradford assay (Bioshop). 3.3.3 Soluble Protein Extraction. Total proteins of bacteria were extracted with B-PER protein extraction reagents (Thermo scientific). The overnight incubated cells were pelleted by centrifugation at 5000× g. Prior to the addition of reagent, 2 µL of lysozyme and 2 µL of DNase I per 1 mL of B-PER were added. The soluble proteins were separated from the insoluble proteins by 15000×g centrifugation after incubation for 15 min at room temperature. Protein concentration was determined by Bradford assay.. 3.3.4 Plasmid DNA Extraction. Overnight incubated E. coli was used in the plasmid preparation process. The isolation and purification of the plasmid DNA is conducted using the EasyPure Plasmid DNA Mini Kit (Bioman).. 3.4 Electrophoresis. 3.4.1 Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE). Solutions of laser-treated or control soluble proteins containing equivalent quantities were boiled in reducing loading buffer and separated on a 4-20% gradient express PAGE gel (Genscript). Protein bands were visualized with Coomassie blue staining (Bioman scientific). 29.

(30) 3.4.2 DNA Agarose gel electrophoresis. Solution of laser-treated or control plasma DNA containing equivalent quantities were mixed with loading buffer and separated on a 1% agarose gel containing Ethidium Bromide. DNA bands were visualized with a UV florescence image system.. 3.5 Spectroscopic quantification of cell leakage and florescence imaging. 3.5.1 Spectroscopic quantification. Irradiated or control bacteria were precipitated by centrifuge at 5000g for 10 minutes. The supernatant was then carefully pipetted out for colorimetric measurement using NanoDrop 1000 spectrophotometer (Thermo scientific).. 3.5.2 Florescence imaging. The control or laser-treated bacteria were stained with propridium iodide (PI. 710400 KPL) for 5 min and then observed using a florescent microscope after several washes. PI is a dye that specifically bound to DNA and excluded from viable bacteria thus the excited florescence could be used to quantify the permeability change after laser treatment. All the images were collected with Leica DM RE microscope. The quantification and contrast adjust were made using image processing software Gwyddion.. 30.

(31) 3.6 Respiration Assays. 3.6.1 Glucose Dependent Respiratory Assay. Respiration rates in cell suspensions were determined at 20◦C by measuring oxygen consumption rate using a Clark type dissolved oxygen sensor and interface controller (CoachLab II+, CMA). Bacteria suspension in PBS buffer were irradiated by laser for 10 minutes with a range of peak power density, or irradiated with constant power for different exposure time. Oxygen consumption rate was recorded as mg·l1. ·min-1 oxygen consumed after the adding of glucose in the suspensions.. 3.6.2 Oxidase Assays. Chemicals such as Decylubiquinone (D7911), Sodium DL-lactate (71720), DLα-Glycerol phosphate magnesium salt hydrate (17766), β-Nicotinamide adenine dinucleotide, reduced disodium salt hydrate (N8129), Sodium succinate dibasic hexahydrate. (s2378),. phenazine. methosulphate. (P9625),. and. 2,6-. Dichlorophenolindophenol sodium salt hydrate (33125) used in the assays were purchased from Sigma Aldrich. The assays on oxidase and dehydrogenase activities were conducted following the procedures used by Lönnerdal B.et al [31]. Prepared membrane fractions suspended in 50 mmol l-1 Tris-HCl were irradiated for 10 min, at a protein concentration of 0.1 to 0.5 mg ml-1, then were chosen to examine the enzyme activities in respiratory chain. The oxidases activities were determined by measuring the oxygen consumption rate after the addition of 20 μl 2 mol l-1 sodium succinate, 1 mol l-1 DL-lactate or 0.1 mol l-1 NADH, 30 μl of 2 mol l-1 DL-glycerol-3-phosphate,. 31.

(32) and 10 μl of 77.5 mmol l-1 reduced DB in methanol. Results are expressed as mg·l1. ·min-1 per milligram of protein.. 3.6.3 Dehydrogenase Assays. The reduction of the dichlorophenol-indophenol indicator was monitored by recording the absorbance at 600nm to determine the activities of dehydrogenases. 1 ml of the prepared bacterial membrane fractions suspended in 50 mmol l-1 Tris-HCl, pH 8.0, at a protein concentration of 50±100 μg ml-1, 16 μl 4.65 mmol l-1 DCIP and 25 μl 65 mmol l-1 PMS were added. The reaction was initiated by the addition of 10 μl 2 mol l-1 sodium succinate, 2 mol l-1 DL-lactate or 2 mol l-1 DL-glycerol-3-phosphate. For the determination of NADH dehydrogenase activity, PMS was omitted from the mixture and the reaction was initiated by the addition of 10 μl 0.1 mol l-1 NADH.. 32.

(33) Chapter 4 Experimental Results 4.1 Inactivation of E. coli by 415nm femtosecond laser. Previous report had demonstrated that femtosecond laser having a wavelength of 425 nm, pulse width of 100 fs and the laser power density greater than or equal to 50 MW/cm2 can inactivate the M13 bacteriophages [18]. In our study, we tested the effect of femtosecond laser on the E. coli viability utilizing a 415 nm, 100 fs laser source. As can be seen in Figure 5, the load reduction of bacteria depends on the laser power density (Fig. 5A) as well as the laser exposure time (Fig. 5B). In general, the longer the exposure times and the larger the laser power densities result in greater load reduction. We have found that a load reduction as large as 3 in log10 scale of bacteria viability was observed after 1 hour laser irradiation.. Figure 5. Increasing visible femtosecond laser power density or exposure time reduced E. coli cell viability. Log-reduction factor is plotted as a function of (A) peak power density (B) exposure time for the laser irradiated bacteria under 1.7 GW/cm2.. 33.

(34) 4.2. Femtosecond laser irradiation alters the surface physical property of bacteria. To investigate the membrane integrity and cell topography after 1 hour laser irradiation, we adopted AFM to evaluate the cellular volume, height, rigidity and adhesiveness of the bacteria before and after irradiation. As can be seen in figure 6(a), untreated E. coli cells are appeared in rod shape. In comparison, the topology of laser irradiated bacteria looked similar to the control (figure 6(b)) except that their cell volume and height were significant decreased (Table 1). Specifically, the volume of cell shrinks from 1.78±0.85 to 0.79±0.35 μm3, and the average cell height decreases from 516±99 to 394±98 nm. It is worthwhile to notice that the changes of cell topography are only observed in the liquid mode AFM images. Topographical images of E. coli scanned in air exhibited identical features in both control and irradiated groups (figure 7(a) and 7(b)). The measurement in cell width, length, volume, and roughness of laser treated bacteria all shown indistinguishable difference than the untreated E. coli (Table 1).. 34.

(35) Figure 6. Cell topography change after laser irradiation. Liquid mode AFM topographical images of (A) untreated bacteria (B) laser irradiated bacteria.. Figure 7. Air mode AFM topographical images of (A) untreated bacteria (B) laser irradiated bacteria.. 35.

(36) Table 1. Summary of the cell properties acquired by AFM The mechanical properties of bacterial surface were probed by AFM tip and the results demonstrated that there are considerable differences between the irradiated and control. In the modulus mapping image of the laser treated bacteria (figure 8(b)), the large contrast between cells and substrate indicates the significant difference of non-linear deformation while comparing to the control (figure 8(a)). Similarly, the adhesive force mapping of the treated E. coli shown a modified force distribution than control (figure 8(c) and 8(d)). As can be seen in Table 1, the Young’s modulus of control (1.27±0.46 Mpa) is significantly lower than the Young’s modulus of the irradiated bacteria (3.00±0.72 Mpa), and the adhesive force of the control (0.33±0.13 nN) is notably higher than the adhesive force of the irradiated one (0.20±0.12 nN).. 36.

(37) Figure 8. Liquid mode AFM mechanical properties mapping; Young’s modulus of (A) untreated bacteria (B) laser irradiated bacteria. Adhesive force of (C) untreated bacteria (D) laser irradiated bacteria.. 4.3. Femtosecond laser irradiation caused leakage of bacterial cellular substances. To investigate the cause of viability reduction after 1 hour laser treatment, we adopted florescence imaging and absorption spectroscopy to detect the leakage of cellular substances of the laser-irradiated bacteria. The bacteria exposed to 1hr laser irradiation (figure 9(m)-(r)) shown a strong florescence from PI intercalating nucleic acids relative to the untreated bacteria (figure 9(a)-(f)). The counts of the bacteria with definite and distinguishable florescent images in a 2020 μm2 frame was found dependent to the laser irradiation time (figure 10). Number of PI stained bacteria increased from 1.6±1.4 (control) 15.5±2.5 of the bacteria exposed to laser for 60 min. 37.

(38) In addition, no significant change of membrane permeability between bacteria exposed to 10min laser irradiation and the control samples (figure 9(g)-(l)). The number of PI stained bacteria exposed to 10 min laser irradiation was 5.0±2.0 per frame.. Figure 9. Exposure to femtosecond laser caused E. coli cellular nucleic acid leakage. Represent florescent images of PI stained (A-F) untreated bacteria and the bacteria exposed to 10 min laser irradiation (G-L) and the bacteria exposed to 1 hr laser irradiation under the power density of 1.7 GW/cm2 (M-R) are shown.. 38.

(39) Figure 10. Exposure to femtosecond laser caused E. coli cellular nucleic acid leakage. The counts of the detectable bacteria florescence signals are presented in histogram.. The optical density of the supernatant of the centrifuged bacteria suspension at the wavelength of 260 nm and 280 nm shown in figure 11 corresponds to the absorption of nucleic acids and proteins, respectively. The power density dependent leakage of nucleic acids and proteins is shown in the figure inset, which indicates that the laserinduced cellular substance leakage increases as the power density increases. These results strongly suggest that the integrity of the cell membranes was compromised after 1 hour laser treatment. On the contrary, the supernatant of the bacteria exposed to 10 min laser irradiation shows no definite cytoplasmic leakage (figure 12).. Figure 11. Exposure to femtosecond laser caused E. coli cellular nucleic acid and protein leakage. Optical density of the media from the irradiated bacteria at 260nm. 39.

(40) () and 280nm (). The inset is the enlargement of the segment of the optical density curve between 0 to 4.3 GW/cm2. Figure 12. The absorption spectra of supernatant from E. coli suspension irradiated by laser for 10 min under the power density of 1.7 GW/cm2. No cell leakage from the irradiated bacteria is detected.. 4.4. Femtosecond laser irradiation alters bacterial protein expression profile. To investigate whether 1 hour laser treatment of E. coli changed the total protein expression profile, total soluble proteins from control and treated bacteria were separated and visualized using SDS-PAGE technique [12]. As indicated in the dotted rectangular box in figure 13, while majority of the protein bands from the irradiated groups showed reduced intensity, the high molecules weight protein aggregations appeared on the top of the electrophoresis wells, and the intensity of the aggregated protein band increases as the power density increases. This result demonstrated that, similar to the visible femtosecond laser treated virus [12], protein aggregation occurred. 40.

(41) after bacteria were exposed to laser irradiation. Similar protein aggregation effect was also observed on the membrane fraction exposed to laser irradiation (figure 14). Although the protein expression of the membrane proteins is slightly different from the total soluble protein, the high molecule weight aggregation appeared on the top of the gel as well. It suggested that the femtosecond laser driven protein aggregation might be a general effect to the soluble proteins and membrane proteins.. Figure 13. Exposure to visible femtosecond laser caused soluble protein aggregation. Image from coomassie blue stained SDS-PAGE containing total soluble proteins extracted from control (lane 2) or laser irradiated (lane 3 and 4) E. coli. Protein sizes are indicated by protein size marker (unit in KDa) in Lane 1.. 41.

(42) Figure 14. Exposure to visible femtosecond laser caused membrane protein aggregation. Image from coomassie blue stained SDS-PAGE containing membrane protein extracted from control (lane 2) or laser irradiated (lane 3) E. coli. Protein sizes are indicated by protein size marker (unit in KDa) in Lane 1.. In figure 15, the expression profile of the soluble protein irradiated by the 830nm femtosecond laser was shown comparing to the untreated protein, and the protein irradiated by 415nm laser as positive control. There is no detectable change between the proteins irradiated by infrared laser between the untreated samples, which indicated the laser wavelength was a crucial factor on laser driven aggregation.. 42.

(43) Figure 15. Exposure to visible femtosecond laser caused membrane protein aggregation. Image from coomassie blue stained SDS-PAGE containing membrane protein extracted from control (lane 2) and laser irradiated by 415nm laser (lane 3) and by 830nm (lane 4). E. coli. Protein sizes are indicated by protein size marker (unit in KDa) in Lane 1.. 4.5. Short time femtosecond laser exposure affects bacterial respiratory rate. Both previous reports [12] and this study demonstrated that protein aggregation occurs after microorganisms were irradiated with femtosecond laser for 1 hr or longer.. 43.

(44) However, whether shorter femtosecond laser irradiation time affects bacterial normal physiological activities is not known. To investigate the effect of shorter laser irradiation on bacteria, we measured the oxygen consumption rates for the control and laser-irradiated bacteria because glucose-dependent aerobic respiration rate can directly reflect the bacterial physiological state. In figure 16(a), the respiratory rates are plotted as a function of the exposure time of laser irradiation at a constant peak power density of 2.8 GW/cm 2. The respiratory activity descends rapidly despite a relatively low laser fluence applied. An over 50% decrease of the respiration rate is reached in less than 10 minutes of irradiation. The dependence of respiratory rate to peak power density is exhibited in figure 16 (b). The respiration is quickly suppressed when exposing to laser with peak power larger than 0.2 GW/cm2. Therefore, the aerobic respiratory rate is immediately affected after a brief femtosecond laser irradiation. Figure 16. Immediate reduction of cell respiratory rate followed short exposure to femtosecond laser. (A-B) The oxygen consumption rate of the control () and the irradiated sample () are plotted as a function of exposure time at a constant peak power density of 2.8 GW/cm2 (in A), or as a function of power density (in B). 44.

(45) 4.6. Effects of the femtosecond laser on membrane-associated respiratory enzymes. To understand whether quick inactivation of membranes-associated respiratory enzymes through laser-induced molecular vibration in E. coli (ISRS effect) [12] accounts for the reduction of respiratory rate, the DCPIP (dichlorophenol-indophenol) oxidoreduction reaction assays were performed with the membrane fractions extracted from irradiated and control bacteria. Utilizing different electron donors to trigger the reduction of DCPIP by each corresponding dehydrogenases, our results indicated that the dehydrogenases from the irradiated groups (denoted as open legends) could not efficiently reduce the DCPIP as their counterparts from the control groups (filled legends) did (figure 17(a)-17(b)). Interestingly, while the succinate, DL-lactate, and glycerol-3-phosphate dehydrogenases are 41.3±2.6%, 42.1±2.6% and 20.1±7.4% reduced respectively, 87.6±3.5% of the NADH dehydrogenase activities from the irradiated bacteria remained (figure 17(a)-17(b)). In comparison, the oxygen consumption rates from the irradiated groups (denoted as open legends) were greatly inactivated comparing to their counterparts from the control groups (filled legends), where 61.4±8.9% of the succinate; 63.9±5.7% of glycerol-3-phosphate, 61.2±7.1% of the DL-lactate, and 46.5±4.8% of the NADH oxidase activities were reduced respectively (figure 18(a)-18(b)). In addition, the inactivation of terminal oxidase activity measured using Decylubiquinone (reduced DB), a ubiquinone analogue, was 69.5±7.2% to the control (figure 18(a)-18(b)). The notable decrease of the terminal oxidase activity indicated a disruption of electron transport chain at the level of using the reduced equivalents of the quinone.. 45.

(46) Figure 17. Effect of the brief (10min) femtosecond laser irradiation on respiratory enzymes. The absorbance at 600nm were plotted as the function of time in (A), where untreated groups were denoted as filled symbols and the laser irradiated groups were denoted as open ones. The inactivation of the dehydrogenase activities with respect to different substrates tested were summarized in (B).. Figure 18. Effect of the brief (10min) femtosecond laser on respiratory enzymes. In figure (A), the dissolved oxygen concentrations were presented as a function of time and the inactivation of oxidase activities were plotted in (B) with substrates tested.. 46.

(47) As the exposure time increases to 1hour, the inactivation of the dehydrogenases activity raise to a certain high degree except NADH dehydrogenase. After the exposure of laser, the inactivation of dehydrogenase activity are 78±3%, 64±12% and 46±5% for succinate, glyceraol-3-phosphate and DL-lactate respectively while the inactivation of NADH dehydrogenase activity was only 18±9% (figure 19(a)-19(b)). The percentage of the oxidases inactivation are 97±6% for succinate; 80±8% for glycerol-3-phosphate, 76±12% for DL-lactate, 89±10% for NADH, and 65±5 (figure 20(a)-20(b)). Our experimental results shown a strong inhibition of the electron transport in respiratory chain with all the substrates tested after 1 hour laser irradiation.. Figure 19. Effect of the long (1hour) femtosecond laser on respiratory enzymes. The absorbance at 600nm were plotted as the function of time in (A), where untreated groups were denoted as filled symbols and the laser irradiated groups were denoted as open ones. The inactivation of the dehydrogenase activities with respect to different substrates tested were summarized in (B).. 47.

(48) Figure 20. Effect of the long (1hour) femtosecond laser on respiratory enzymes. In figure(C), the dissolved oxygen concentrations were presented as a function of time and the inactivation of oxidase activities were plotted in (D) with substrates tested.. 4.7 Femtosecond Laser Induced Plasmid DNA Relaxation. 4.7.1 Assignment of the Bands in Electrophoresis. The plasmid DNA pCR II Topo (pCR), pBluescript (pBS), and pUC 19 (pUC) were examined by agarose gel electrophoresis to investigate the conformation changes after exposing to femtosecond laser irradiation. In the figure 21, three distinguishable bands from untreated pCR II plasmid DNA were observed. The band with the strongest intensity can be assigned as supercoiled form unambiguously. The EcoRI restricted DNA was shown in the lane 2 of the gel, and the single isolated band is composed only of linear DNA digested from restriction enzyme. Comparing these two lane, we could assign the three bands as supercoiled, linear, and circular DNA in the sequence of mobility.. 48.

(49) Figure 21. Assignment of the band profile in DNA gel electrophoresis. Image from EtBr stained gel electrophoresis containing untreated pCRII plasmid (lane 2) and the linear DNA digested by EcoRI (lane 3). DNA sizes are indicated by DNA size marker (unit in kDa) in Lane 1. The three bands in lane 2 can be assigned as open circular, linear, and supercoiled bands from up to down.. 4.7.2 Laser Effect on the Plasmid DNA pCR II-TOPO. As previously reported by Tsen et al., the femtosecond laser could induce the supercoiled plasmid DNA relaxation of E. coli [14]. In our experimental result, we observed that the relaxation process is dependent on the peak power density of femtosecond laser (figure 22 (a)). A fitting result of the electrophoresis band intensities was plotted in the figure 22 (b). When the power density of laser was beyond the threshold of 0.52 GW/cm2, the intensity of the circular band in electrophoresis increased and therewith supercoiled band intensity dropped. At 0.69 GW/cm2, the. 49.

(50) population of the open circular came to 74±18% and became the dominant DNA conformation.. Figure 22. Laser induced DNA relaxation in pCR II plasmid. (a)Image from EtBr stained gel electrophoresis containing control (lane 2) and laser irradiated plasmid by 415 nm laser (lane 3-6). DNA sizes are indicated by DNA size marker (unit in KDa) in Lane 1. (b) Fitting result of the band intensity is plotted vs. laser power density.. Interestingly, the relaxation effect would not take place on the linear DNA pre-digested by EcoRI (figure 23 (a)) and on the DNA irradiated by the infrared laser with the wavelength of 830 nm (figure 23 (b)). The result indicated the relaxation process could be triggered only when certain criteria were met. As shown in our observation, the wavelength of the excitation laser and the conformation of the plasmid could be the decisive factors.. 50.

(51) Figure 23. Laser induced DNA relaxation in PCR II plasmid. (a)Image from EtBr stained gel electrophoresis containing untreated linear DNA (lane 1) and laser irradiated linear plasmid by 415nm laser (lane 2). (b) Image containing untreated DNA (lane 1) and laser irradiated plasmid by 830nm laser (lane 2).. To further investigate the oscillator behavior of the supercoiled DNA molecules, glycerol was added to the DNA solution to increase the viscous damping of the medium. The fraction of supercoiled DNA increased with the concentration (lane 3 to 5 in figure 24(a)) of glycerol under the same laser fluence. As the concentration of glycerol increased from 10% to 30%, the survived supercoiled DNA as well increased from 9±8% to 24±6% of the total DNA (figure 24(b)).. Figure 24. Damping effect on the laser induced DNA relaxation in pCR II plasmid. (a)Image from EtBr stained gel electrophoresis containing control (lane 2) and laser irradiated plasmid by 415nm laser in the present of glycerol (lane 3-5). DNA sizes are indicated by DNA size marker (unit in KDa) in Lane 1. (b) Fitting result of the band intensity is plotted vs. glycerol concentration.. 51.

(52) 4.7.3 Laser Effect on the Plasmid DNA pBluescript. The agarose gel of electrophoresis of plasmid pBS was shown in the figure 25(a). Untreated plasmid and the ones irradiated by the laser power density lower than threshold were presented in the lane 2, 3, and 4 in figure 25(a) respectively. The untreated DNA expressed three distinguishable bands, where the lowest band with the strongest intensity can be assigned as the supercoiled form. The rest two components could be assigned as open circular band (upper) and linear band (lower) according to the mobility of DNA molecule. In the lane 5, 6 of the figure 25(a), two major bands from supercoiled and linear DNA are present with non-equimolar amount. The lower band is less intense than the upper band, which indicated the laser irradiation caused a supercoiled to linear transition above the threshold power density of 1.1 GW/cm2 (figure 25(b)). Similarly, pBS plasmid shown a conformation change in a power density dependent manner as pCR. However, the resulted conformation change of pBS is the relaxation from supercoiled form to linear DNA, which is different from the case we observed in pCR system.. Figure 25. Power density dependent on the laser induced DNA relaxation in pBS plasmid. (a)Image from EtBr stained gel electrophoresis containing control (lane 2) and laser irradiated plasmid by 415 nm laser (lane 3-6). Two plasmid DNA irradiated with the same laser fluence in the power density below (lane 7) and above (lane 8) (b) Fitting result of the band intensity is plotted vs. laser power density. 52.

(53) Moreover, in the 7 and 8 lane of the figure 25(a), the plasmid DNA solution were shined with the same laser fluence with different power density below/above the threshold 1.1GW/cm2 as a positive control. Plasmid in these two lanes expressed different band profiles in gel electrophoresis. As the power density is lower than the threshold, no transition of the DNA conformation appeared while the transformation took place in the sample above the threshold even though the fluence deposited in two samples is identical. 4.7.4 Laser Effect on the Plasmid DNA pUC 19. Both the control and irradiated pUC 19 plasmid were composed of three distinguishable bands in the agarose gel electrophoresis (figure 26(a)). The assignment of the three bands are open circle, linear and supercoiled in the sequence from top to bottom. After exposing to the laser irradiation above the power density threshold of 2.0 GW/cm2, the dominant band of plasmid DNA changed from supercoiled form to the linear form (figure 26(b)). It was also noticed that there was a raise in the fraction of the open circular form accompanying the major transition. The experimental data indicated that both transition could simultaneously contribute to the interaction between the plasmid DNA and femtosecond laser.. 53.

(54) Figure 26. Power density dependent on the laser induced DNA relaxation in pUC 19 plasmid. (a)Image from EtBr stained gel electrophoresis containing control (lane 2) and laser irradiated plasmid by 415 nm laser (lane 3-6). DNA sizes are indicated by DNA size marker (unit in kDa) in Lane 1. (b) Fitting result of the band intensity is plotted vs. laser power density.. 54.

(55) Chapter 5 Discussion In this study, we have observed the laser induced dehydration of cells, the alteration of bacterial membrane properties and the inhibition of respiratory metabolism led by the irradiation of visible femtosecond laser in vitro. We also have found the protein aggregation and relaxation of plasmid DNA in vivo (Please see the table 2 for the summary of the laser induced effects). In order to clarify the interaction mechanism between femtosecond laser photons and bacteria, the effects will be classified into the damages related to genetic material (genetic damage) and the damages related to the disruption of other cellular components and molecules of bacteria (non-genetic damage). Discussions toward these different categories of laser damage onto bacteria will be made in the following sections separately.. Table 2. List of the detrimental effects induced by femtosecond laser. 55.

(56) 5.1 Non-genetic damage. We have demonstrated that irradiation for 1 hour with a femtosecond laser having a wavelength centered at 415 nm, pulse width of 100 fs and the power density equal to 3.1 GW/cm2 can efficiently inhibit bacteria growth. Utilizing liquid mode AFM to evaluate the physical property of irradiated bacteria, we found that their cellular volume, height, rigidity (stiffness) as well as the adhesiveness were altered (Table 1). Utilizing permeability assay and absorption spectroscopy to detect the nucleic acid and protein amount in the media, we found that the membrane permeability of irradiated bacteria is altered (Figure 9-12). In the liquid AFM examination, the decrease of cell volume after femtosecond laser irradiation is significant. However, we did not detect the appearance of large amount of DNA and proteins in the supernatant of the irradiated bacteria, indicating that the materials leaking out could primarily be water and cellular ions. One feature worth notice is that the amount of the leaking substances is relatively low when comparing to the totally lysed bacteria. This observation indicates that the membranes can still restrain most of the large biomolecules after 1 hour laser treatment when the power density is below 4.3 GW/cm2. Moreover, in the air mode AFM, the bacteria cells were inevitably dehydrated during the scanning by air mode AFM, hence the cell shape of control and irradiated bacteria are very similar as shown by the air mode AFM. On the other side, the cell volume of the irradiated bacteria observed with liquid mode AFM shrink to about 50% of the control, which is closed to the dehydrated cell volume observed in air. This observation evidenced the aforementioned water leakage and cell dehydration of the irradiated E. coli.. 56.

(57) Although the leakage of protein and nucleic acid was not detected, the loss of ion could still impact cell metabolism. Cellular ion balance is essential for maintaining proper cellular functions, such as the control of turgor pressure and solute transport. Therefore, the increased permeability could become a detrimental effect to the normal metabolism of irradiated bacteria. Another observation from AFM examination is that although there is no observable defect or indentation on the membrane surface (figure 6(a)-6(b)), the increase of surface stiffness (Young’s modulus) suggested that an alteration of membrane structure was induced by the laser irradiation, because change of Young’s modulus reflects the conformational change of the extracellular layer and the dimensional change of periplasmic space that are sensitive to the internal structural details of cellular materials [32, 33]. Furthermore, because prior reports indicated that the decrease of the adhesion force on both Gram-negative and positive bacteria cell surfaces after antibacterial photodynamic therapy or environmental pH alteration reflects the destruction of envelope polymers and the subsequent exposure of proteins inside the envelope [32, 34], the decrease of adhesiveness of the laser irradiated E. coli in our study thus further suggested that laser irradiation induced structural alteration of the bacteria membrane. Consistent with this notion, the results from the permeability assay indicated that the integrity of the cell membranes was compromised after 1 hour laser irradiation. In our case of femtosecond laser irradiated bacteria, the alteration of membrane structural integrity could be related to the aggregation of membrane proteins. Utilizing gel electrophoresis (SDS-PAGE) to separate and visualize the membrane and total soluble protein expression of irradiated bacteria, we found that similar to the femtosecond laser treated virus [12], the high molecules weight protein aggregations appeared indeed, and the intensity of the aggregated protein band increases as the laser power density increases (figure 13-14). Because the 415 nm femtosecond laser adopted. 57.

(58) in this study lacks enough photon energy to excite the covalent bond formation in bacterial proteins [12], the disruption of weak bonds and the partial-unfolding of proteins by the ISRS-mediated coherent molecular vibrations are the possible causes of the protein aggregation observed in this study. In addition, because the permeability and the cell surface physical properties are maintained and regulated by the structural and chemical composition of cell membranes, the alteration of membrane structure caused by protein aggregation might be the cause for the releasing of the cellular substances and the instability of cellular ion concentration. The test of glucose-dependent respiration is often adopted as an assay to examine the damage of membrane structure, which eventually leads to cytoplasmic leakage and a lethal effect to bacteria [35-37]. In this study, we tested whether a shorter irradiation could affect the glucose-dependent respiratory rate. The results showed that the inhibition of respiration, the decrease of oxygen consumption, is triggered when exposing bacteria suspension to the power density greater than 0.2 GW/cm2. The respiratory rates remained on 25% to the control regardless the further increasing of power density or exposure time (figure 18(a), 18(b)). Because the laser-induced genetic damages would require a longer period of time to become effective [14], this early commencement of respiratory inhibition after a shorter exposure time might be caused by the inactivation of respiratory enzymes through laser-induced molecular vibration that blocked the energy uptake process and interrupt metabolism in E. coli. Consistently with this notion, the membrane-associated dehydrogenases and oxidases showed various degrees of inhibition (figure 16-20). Nevertheless, in the assay of dehydrogenases, the NADH dehydrogenase activity showed only 12.4% inactivation by the laser irradiation. On the other side, the NADH oxidase inactivation abruptly increased to 46.5%. The significant difference between the extents of inactivation might be explained by the selective inactivation of the quinone reduction subunit in the NADH. 58.

(59) dehydrogenase [31]. It suggests that the laser irradiation could cause local alteration on the enzyme complex through the ISRS induced protein aggregation. Among the substrate carbon sources exposed to 1hr irradiation, the activity of succinate dehydrogenase was the most susceptible and the NADH dehydrogenase was the least affected enzyme. In the assay of the terminal oxidase activity, 64% of the function of terminal cytochrome oxidase, which was believed to be cytochrome bo3 in our aerobic growth condition, was eliminated (Figure 20). Due to the large loss of dehydrogenases activities and the partial elimination of the electron transport via the oxidation of quinone, the substrate oxidases activities are all inhibited to a certain high degree. After one hour of laser exposure, the inactivation of oxidases were all above 70%, and enzyme activity is nearly completely inhibited in the case of succinate oxidase (Figure 20). The enzymes involved in electron transport chain pump proton to periplasm space and the resulted transmembrane electrochemical potential is essential to the synthesis of ATP. The laser induced interruption of enzyme activities could block the energy taken process and hence reduce bacterial viability. Also, the effect femtosecond laser deposited to the enzyme might be dependent to the species of enzymes and the subunit composition.. 5.2 Genetic damage. The femtosecond laser induced DNA relaxation and the subsequent genetic damage was firstly reported by Tsen et al in the laser irradiation experiment of the plasmid DNA from E. coli [11, 14]. This laser induced genetic damage was presumed to be the dominant detrimental factor that caused the reduction of bacterial viability. The 59.

(60) argument was used to successfully explain the difference between the log-load reduction factor between the wild type and a strain that is deficient in DNA repairing protein [14]. Additionally, the reactive oxygen species produced by visible femtosecond laser was regarded as the dominant mechanism that cause the denaturation of DNA. To obtain further insight of the femtosecond laser induced genetic damage, we have performed irradiation of various types of plasmid DNA by a visible femtosecond laser. A general characteristic of all the plasmid DNA tested in this study is the laser power density threshold of the DNA conformation transformation. With the laser power density above certain threshold, the supercoiled structure would be relaxed to the linear, open circular form, or both of them. The interesting feature is, the supercoiled structure remain stable while the laser power density used is below the threshold, even the exposure time is prolonged to ensure laser fluence is equivalent (figure 25). The other common feature of the DNA relaxation is, that the large sized plasmid seems to be more sensitive to the visible femtosecond laser pulses and is easily to be relaxed comparing to the smaller plasmid with less base pairs ( figure 27 ).. Figure 27. Dependence between plasmid size and threshold power. 60.

(61) It is already demonstrated that ISRS could induce protein structure alteration through the disturbance of the weak links within protein molecules [13]. Analogously, the weak links exist in the nucleic acid structure could also be affected by the resonant vibration caused by ISRS and leads to the conformational change. A well know low frequency collective motion in DNA molecule in the range of 10–100 cm−1, corresponding to the range of terahertz frequency (3×1011 to 3×1012 Hz) has been observed in several researches [5, 7, 38]. The visible femtosecond laser adapted in our study has significant spectral content at the frequency of the collective vibration and is able to bring the molecules into resonant oscillation. In principle, ISRS effect could result in an amplitude of the vibrational displacement in the range of of 0.01 to 1Å to a low frequency vibration mode in the condition of a moderate Raman scattering cross section and a reasonable excitation intensity [39]. In the quasi-continuity model of DNA, the two polynucleotide backbones are intertwining around the z axis and connected by the hydrogen bond spring between the complementary bases. Some of the hydrogen bonds would temporally open at thermal equilibrium. Every 20-50 intact hydrogen bond, there is an open bond that demarcate the DNA into individual segment [5]. When the molecule is brought into resonance, the vibration possess the feature of the coupled standing wave oscillators. As the coupling effect is present, the vibrational energy can be concentrated upon some of the segments at certain moments, and the effect will induce an amplified vibration amplitude within the segments. In a simplified model considering only 8 oscillators coupling proposed by Chuo et al., the amplified amplitude can increase 2.5 times than the case when no coupling is present [5]. In a realistic DNA molecule with far more oscillator segments, the amplification of the vibration becomes more significant, and such over concentrated vibration energy can lead to a quake like motion and a hole between polynucleotide. 61.

(62) chains that allows the intercalation of an external molecule or the start of DNA replication process. According to our observation, the length of plasmid molecules is correlated to the power density threshold. The possible underlying mechanism is, that the larger the plasmid size is, the more intact segments are involved into the coupling resonant vibration of DNA, thereby increase the possibility to create gigantic vibration amplitude (quake) between the polynucleotide chains. When the amplitude exceeds the strain limit of associated chemical bonding, the conformation of DNA could therefore be affected. For instance, in the case of the relaxation of supercoiled DNA into the open circular form, the resonant vibration deposit excessive energy than the strain limit of polynucleotide bond. One of the chain subsequently split off and release the elastic energy stored in the supercoiled structure, leading to the formation of the nicked circular DNA. If the vibrational energy is even concentrated onto certain segments via the coupling effect and impulse stimulation, the polynucleotide chain could break off and result in a DNA molecule with linear conformation. An intriguing question about the DNA relaxation process is if the breakage is dependent on the genetic sequence or specialized mechanical structure of the plasmid DNA. As shown in the figure 28, the untreated, laser irradiated DNA, and the EcoRI digested untreated and laser irradiated pBS plasmid DNA were examined by EtBr stained agarose electrophoresis. According to the assignment of electrophoresis bands in section 4.7.2, we can identified the predominant band in the first lane was the supercoiled DNA from the untreated plasmid. The major band from the irradiated DNA in the lane 2 shown a less mobility than the untreated, which can be assigned as linear conformation relaxed by femtosecond laser according to the EcoRI predigested DNA shown in lane 4. In the third lane, the digested DNA shown a smear along the lane with several definite bands with relative strong intensity. The presence of those DNA 62.

(63) fragments with shorter length suggests there could exist some weak points within the DNA chain that are easily detached when the vibrational energy is deposited to the molecules via ISRS. When the laser energy brings the molecules into coherent collective vibration, the antinodes of the vibrational modes are at where the maximum atomic displacement takes place. A possible explanation to the existence of the weak points is that the plasmid DNA could be cleaved at the antinodes when the molecules are excited via ISRS, and the excited plasmid DNA therefore break into small distinct fragments from the antinodes of the vibrational modes.. Figure 28. Examination on the specific cleavage site on pBS plasmid Image from EtBr stained gel electrophoresis containing untreated pBS plasmid (lane 1) ,the laser irradiated DNA (lane 2), the laser irradiated DNA digested by EcoRI (lane 3) and the untreated DNA digested by EcoRI (lane 4).. 63.

(64) Moreover, each plasmid DNA we tested in this study shows specified conformation transformation after laser exposure (Table 3) and the correlation between the resonant coupling mechanisms between the exact DNA structures is still elusive.. Table 3. Summary of the conformation transitions induced by the irradiation of femtosecond laser. Our experimental result also demonstrated the influence of damping effect on the energy transfer via the resonant coupling. As demonstrated in the eq. (7), the maximum amplitude induced by ISRS effect will be reduced and hence lower the efficiency of energy transfer. Damping effect from the viscous solvent molecules could act as cushion that absorbs a part of the energy from the laser impulse. In a realistic cytoplasm, several dissipative effects coexist with DNA. Thus the genetic damage in a living cell induced by femtosecond laser could be reduced than the purified plasmid DNA investigated in this study. To discuss the role that genetic damage plays in the E. coli inactivation, the experimental condition of the laser irradiation experiment should be carefully examined. In a laboratory condition, the generation time of E. coli is about 15-20 min in a medium which is abundant in nutrition. In an environment that lacks of nutrition, the generation time could be as long as several hours, even one day. The bacteria used for our laser treatment is suspended in a nutritionless PBS buffer solution, and the treatment is taking place at the temperature of 20-22。C. In such experimental conditions, the metabolism activity of the bacteria abates to a certain low degree, and the generation time will hence. 64.