Characterization of microorganisms using Raman tweezers

Characterization of microorganisms using Raman tweezers Ota Samek *a, Zdeněk Pilát a, Alexandr Jonáš a , Pavel Zemánek a , Mojmír Šerýa , Jan Ježek a, Silvie

Bernatová a, Ladislav Nedbalb, Martin Trtílek c aInstitute of Scientific Instruments of the AS CR, v.v.i., Academy of Sciences of the Czech Republic,

Kralovopolska 147, 61264 Brno, Czech Republic; bInstitute of Systems Biology and Ecology of the AS CR, v.v.i., Academy of Sciences of the Czech

Republic, Zámek 136, 37333 Nové Hrady, Czech Republic; cPhoton Systems Instruments, Drásov 470, 664 24 Drásov, Czech Republic

ABSTRACT

The ability to identify and characterize microorganisms (algae, bacteria, eukaryotic cells) from minute sample volumes in a rapid and reliable way is the crucial first step in their classification and characterization. In the light of this challenge related to microorganisms exploitation Raman spectroscopy can be used as a powerful tool for chemical analysis. Raman spectroscopy can elucidate fundamental questions about the metabolic processes and intercellular variability on a single cell level. Moreover, Raman spectroscopy can be combined with optical tweezers and with microfluidic chips to measure nutrient dynamics and metabolism in vivo, in real-time, and label free. We demonstrate the feasibility to employ Raman spectroscopy-based sensor to sort microorganisms (bacteria, algae) according to the Raman spectra. It is now quite feasible to sort algal cells according to the degree of unsaturation (iodine value) in lipid storage bodies. Keywords: Raman spectroscopy, bacteria, algal cells, lipids, iodine value

1. INTRODUCTION It has been shown in several studies that Raman microspectroscopy is capable of rapid identification and discrimination of biological samples including medically and biotechnology relevant microorganisms (bacteria, yeast, and algae). This experimental technique employs a laser beam that is focused with a microscope objective in order to excite and collect Raman scattering spectra from a small volume of the sample. A typical Raman spectrum contains a wealth of information indicative of the cellular content of nucleic acids, proteins, carbohydrates, and lipids. Such a spectrum functions as a cellular ‘fingerprint’ and serves as a sensitive indicator of the physiological state of the cell. Raman spectra thus enable to differentiate cell types, actual physiological states, nutrient conditions, and phenotype changes. In principle, Raman spectroscopy requires measurement times on the order of minutes, and sample preparation can be short and extremely economical: the same substrates, which are used in hospitals for quick infection checks (growing bacteria colonies in Petri dishes), can be directly used. A deep and detailed understanding of mechanisms that regulate biological processes including the response of cells to external stimuli requires studying the physiological state of bio-systems at the level of a single cell. Only in this way, it is possible to observe directly the cell-to-cell heterogeneity which is otherwise masked by the average of a cell ensemble obtained from traditional measurement approaches. In case one wishes to compare the results of the single-cell measurements with traditional techniques, single-cell data can be statistically processed to yield population averaged values. Optical trapping represents an elegant approach for conducting Raman microspectroscopic measurements on individual cells under relevant environmental conditions. *[email protected]; phone 420 541 514 253; fax 420 541 514 402; www.isibrno.cz

Invited Paper

Optical Trapping and Optical Micromanipulation VIII, edited by Kishan Dholakia, Gabriel C. Spalding, Proc. of SPIE Vol. 8097, 80970F · © 2011 SPIE · CCC code: 0277-786X/11/$18 · doi: 10.1117/12.893420

Proc. of SPIE Vol. 8097 80970F-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/12/2013 Terms of Use: http://spiedl.org/terms

With an optical trap, living cells suspended in a liquid cultivation medium can be immobilized in solution using the forces generated by a tightly focused laser beam. When the wavelength of the trapping light is appropriately chosen, the light is not absorbed in the cell and the confinement is non-invasive. Since both optical tweezers and Raman microspectroscopy rely on the use of a strongly-focused laser beam, it is straightforward to combine both techniques using a single beam for both trapping and spectroscopy. This combination - usually termed Raman tweezers – has found numerous applications especially in the fields of analytical and physical chemistry and it currently starts to be applied in the cell and molecular biology. In connection with a microfluidic chip platform that allows a fast, controlled exchange of the cell suspension fluid Raman tweezers can form a base of a unique system for cell micromanipulation and sorting. Raman spectroscopy is a very attractive alternative that derives the sorting signal from detecting vibrations of chemical bonds of molecules naturally present in the cells. After acquiring a spectrum from an optically trapped cell, the data is analyzed and the cell is subsequently sorted by active micromanipulation with the optical trap. Consequently, cell sorting can be a completely non-invasive process and the sorted cells can be used for further cultivation and analysis. Furthermore, Raman tweezers implemented in a microfluidic chip can serve to study the dynamics of the response of an individual cell to a controlled external stimulus or stress factor.. In our study we mainly focus on the exploitation of fatty acids within the two types of microorganisms – bacteria and algae. This means that in algae cells the degree of fatty acid unsaturation - which is the main parameter that determines their application potential whether they could be used as fuels or as dietary supplements or for pharmaceutical raw materials – can be estimated. For bacterial cells the potential of Raman spectroscopy to indicate presence of the main fatty acid families occurring in the cell membranes was studied. So far, Raman applications in microbiology have aimed mostly at detecting medically relevant organisms 1-3. A database of Raman spectral features of biologically relevant molecules that facilitates assignment of the most prominent Raman bands observed in living cells was recently published4. The use of Raman spectroscopy for the detection and identification of important, basic molecules in biological samples has been summarised in recent reviews 5-7. However, Raman spectroscopy of algal cells is complicated by presence of pigments. Thus, applications performed by other groups are so far restricted to only a small number of algal species 8-10. In our experiments the Raman spectra were collected from different algae species cultivated in the stationary phase for a prolonged period, inducing an overproduction of lipids. We have shown that these algal species display significantly different unsaturation of their storage lipids, which is important finding for the biofuel production and food industry. An example is Trachydiscus minutus which contains a high amount of highly unsaturated fatty acids 11. The other algal samples under investigation was Botryococcus sudeticus and Chlamydomonas sp. These two algal species exhibits relatively low level of unsaturation and have been for a long time in the focus of an intense research. The intensities of the Raman spectral peaks that correspond to the saturated (CH2) and unsaturated (C=C) carbon bonds in lipid molecules were used to estimate the degree of unsaturation (iodine value) in the lipid bodies similarly to the references 12-15. The degree of unsaturation was quantified using the iodine value (IV). Iodine value is widely applied in biofuel and food industry16. We believe that the detection of the iodine value for algal species, based on Raman microspectroscopy, has a great potential for testing/production of higher generation biofuels and possibly for production of dietary supplements. The bacteria used in our experiment were only Staphylococcus species cultured on Mueller-Hinton (MH) agar plates containing no antibiotic. The plates were incubated overnight at 37 degrees. Bacteria were analysed in order to check that it is possible to reveal contribution of fatty acids to the overall statistical variance using chemometrics techniques. In this paper we demonstrate that the combination of identification algorithms based on Raman spectra and the active sorting switch could well provide a potentially useful sensor for fast microorganisms identification and sorting in microfluidic chip. This sensor can be employed in real-word applications (bioreactor, cultivation pond or microbiological laboratory) as a base for the microorganismsl selection and sorting.

2. MATERIALS AND METHODS 2.1 Organisms and cultivation conditions

Botryococcus sudeticus Lemmermann, CCALA 780 (VAZQUEZ-DUHALT/UTEX 2629), and Trachydiscus minutus (Bourrelly) Ettl, CCALA, were obtained from the Culture Collection of Autotrophic Organisms, CCALA (Institute of Botany, Academy of Sciences of the Czech Republic). T. minutus was cultivated in 50% Šetlík- Simmer medium17 in



100 mL air-bubbled batch cultures. The cells were harvested in early stationary phase. Chlamydomonas sp. and Botryococcus sudeticus were cultivated in 150 mL Erlenmeyer flasks in BBM medium at room temperature in daylight at a laboratory window with occasional manual mixing. The cells were harvested at late stationary phase. The long term cultivation in the stationary phase was observed to induce the deposition of storage lipids in algal cells18. For the sake of simplicity, and to simulate a “real-word” hospital environment, the bacterial samples were analyzed directly on the Petri culture dishes, on which the bacterial colonies were cultivated. For the microfluidic experiments a solution was prepared from the cultured colonies.

2.2 Raman microspectroscopy

Experiments employing Raman spectroscopy based sensor with microorganismss were carried out using a home-built experimental system based on a custom-made inverted microscope frame19,20 attached to the active sorting switch21. Recorded spectra were processed using in-house written software routines that include also the identification and chemometrics algorithms (for sorting of cells) implemented in Matlab software (MathWorks). 2.3 Spectrum processing and analysis

In our experiments with algae, we determine the ratio of unsaturated-to-saturated carbon-carbon bonds in algal lipid molecules obtained from Raman spectra20. We employed two spectral peaks at 1,656 cm−1 (cis C=C stretching mode proportional to the amount of unsaturated C=C bonds) and at 1,445 cm−1 (CH2 scissoring mode proportional to the amount of saturated C-C bonds). We found these peaks free of any significant interference or overlaps with Raman signals of other cellular components 20. Both of these peaks are well documented as being very strong in Raman spectra of lipids20. From the ratio C=C/CH2, mass unsaturation can be estimated using the calibration curve 20 plotted on the basis of the published iodine values for the fatty acids20,22. Thus, it is possible to directly convert the measured values of C=C/CH2 to the iodine values for a given sample. In order to extract quantitative information from the experimentally obtained spectral data, we adopted the Rolling Circle Filter (RCF) technique for background removal20,23. Bacterial species were analyzed using suitable chemometrics algorithms (written in-house) required when analyzing a full range of clones. Because unsaturated fatty acids are only minor constituent in bacterial species under our study, we could not employ the procedure as used for the algal samples. Instead, we focused on fatty acid marker for C-C stretching all-trans skeletal vibration at 1064 cm-1.

3. RESULTS AND DISCUSSION

3.1 Sorting of microorganisms The presented Raman spectroscopy based sensor employs an active sorting switch21 and combines the versatility of Raman tweezers24-,26 instrumentation with the automation potential delivered by the microfluidic technology. A focused laser beam is used as the sorting mechanism. Laser power of the trapping and sorting beams can be independently regulated so that maximal performance of Raman tweezers and sorting switch are achieved. An example of our feasibility study for microorganisms includes the two bacterial species, for more details see Figure 1.

3.2 Calibration of Iodine Value against spectroscopic data

In order to link the experimentally observed Raman signal intensities from algae samples (C=C/CH2) to the actual iodine values (IV), we performed a series of Raman spectroscopic measurements of pure fatty acids with varied degree of unsaturation to construct a calibration curve20. Consequently, such calibration curve can be employed to relate the Raman spectroscopic data (C=C/CH2) to the corresponding iodine values of the studied sample and also to the number of C=C bonds for algal samples - see Figure 2 and 3, respectively.



Finally, we would like to note that our study should be viewed as a feasibility study (proof of principle study) than as fully developed recognition/sorting system ready for routine laboratory use. In order to achieve this goal, more extensive investigations with larger selection of samples is required. In bacterial species fatty acids are one of the most important building blocks of cellular materials. In bacterial cells, fatty acids occur mainly in the cell membranes. Membrane fatty acids consist of two major families i) straight-chain fatty acids family (palmitic, stearic, and arachidic acids); and ii) branched-chain fatty acid family (iso- and anteiso-fatty acids). Total amount of straight-chain fatty acids is about 15%-20% and branched-chain fatty acids is about 80%. As was mentioned above unsaturated fatty acids are nonessential for bacteria with the branched-chain membrane system (Staphylococcus epidermidis). Thus, unsaturated fatty acids are minor fatty acid constituent and are not considered in our Raman study. The related Raman spectra for saturated fatty acids are currently being investigated by different recognition/chemometric algorithms.

Figure 1 Schematic diagram of the experimental setup for the Raman microscopy based sensor. Here, microorganism A is Staphylococcus epidermidis and Microorganism B is Staphylococcu aurerus. Bacterial cells are flown into a homemade microfluidic chip which was constructed as a fork with one input and two output channels. Bacteria cell on its way through the channel is trapped and its spectra recorded using Raman tweezers. In the next step, the obtained Raman spectra are analysed using the identification algorithms. In case of algae samples the algae cells are analysed using algorithm based on the iodine value calibration curve 20. Consequently, depending on input data and limits set by the in-house written software routines, the cells can be sorted to the two output channels using the active sorting switch employing focused laser beam.



Figure 2 Calibration curve for estimating the iodine value from the Raman data for the three algal species mentioned in the text. The calibration data can be fitted well using a parabolic function (shown at the bottom of the Figure). Raman spectrum of EPA (Eicosapentaenoic acid, Omega-3 fatty acid) used for constructing the calibration curve for iodine value estimation is shown at the top part of the Figure. In this spectrum Raman bands used to calculate iodine value are indicated – CH2 (saturated fat indicator) and C=C (unsaturated fat indicator).

4. CONCLUSIONS We have demonstrated that it is possible - using the algorithm based on analysis of Raman spectra - unambiguously determine the important parameters of fatty acids, such as iodine value and the average number of carbon-carbon double-bonds in algal cells. To assess whether Raman spectroscopy could differentiate between different bacterial strains feasibility study was performed to indicate presence of the main fatty acid families occurring in the bacterial cell membranes - branched-chain and straight-chain fatty acids. Consequently, we have performed a feasibility study to sort the microorganisms in a microfluidic system using the active sorting switch. We believe that Raman spectroscopy combined with optical tweezers and with microfluidic chips will be of significant assistance to research groups involved in system biology, microbiology, and genetic engineering.



Figure 3 Dependence of the recorded Raman spectra on the number of double bonds C=C. The last calibration point for EPA is indicated. From this calibration curve the average number of C=C bonds within particular algal species and fish oil (here only for comparison) can be estimated. Data also fit extremely well to a parabolic function (shown at the top left corner). Lipid mixtures within the included species exhibit average double bonds per lipid molecule-equivalent of about 1 – 2.5.

ACKNOWLEDGEMENTS

The authors acknowledge the support for O. Samek by a Marie Curie Reintegration Grant (PERG 06-GA-2009-256526). This work also received support from the Ministry of Education, Youth, and Sports of the Czech Republic together with the European Commission, Czech Science Foundation, and Ministry of Industry and Trade of the Czech Republic (projects ALISI No. CZ.1.05/2.1.00/01.0017, GA P205/11/1687 and FR-TI1/433).

REFERENCES

[1] Maquelin, K., Kirschner, C., Choo-Smith, L.P., van den Braak, N., Endtz, H.P., Naumann, D. and Puppels, G.J.,

“Identification of medically relevant microorganisms by vibrational spectroscopy,” J. Microbiol. Meth. 51, 255-271 (2002).

[2] Harz, M., Rosch, P., Peschke, K.D., Ronneberger, O., Burkhardt, H., Popp, J., “Micro-Raman spectroscopicidentification of bacterial cells of the genus Staphylococcus and dependence on their cultivation conditions,” Analyst 130, 1543-1550 (2005).



[3] Samek, O., Al-Marashi, J.F.M. and Telle, H.H., “The potential of Raman spectroscopy for the identification of biofilm formation by Staphylococcus epidermidis,” Laser Phys. Lett. 7, 378-383 (2010).

[4] De Gelder, J., De Gussem, K., Vandenabeele, P. and Moens, L., “Reference database of Raman spectra of biological molecules,” J. Raman Spectrosc. 38, 1133-1147 (2007).

[5] Notingher, I., “Raman spectroscopy cell-based biosensors,” Sensors 7, 1343-1358 (2007). [6] Downes, A. and Elfick, A., “Raman spectroscopy and related techniques in biomedicine,” Sensors 10, 1871-

1889 (2010). [7] Movasaghi, Z., Rehman, S. and Rehman, I.U., “Raman spectroscopy of biological tissues,” Appl. Spectrosc.

Rev.42, 493 – 541 (2007). [8] Heraud, P., Beardall, J., McNaughton, D. and Wood, B.R., “In vivo prediction of the nutrient status of

individual microalgal cells using Raman microspectroscopy,” FEMS Microbiol. Lett. 275, 24-30 (2007). [9] Heraud, P., Wood, B.R., Beardall, J. and McNaughton, D., “Effect of pre-processing of Raman spectra on in

vivo classification of nutrient status of microalgal cells.” J. Chemometrics 20, 193-197 (2006). [10] Huang, Y.Y., Beal, C.M., Cai, W.W., Ruoff, R.S. and Terentjev, E.M., “Micro-Raman spectroscopy of

algae:Composition analysis and fluorescence background behavior,” Biotechnol. Bioeng. 105, 889-898 (2010). [11] Řezanka, T., Petránková, M., Cepák, V., Přibyl, P., Sigler, K. and Cajthaml. T., “Trachydiscus minutus, a new

biotechnological source of eicosapentaenoic acid,” Folia Microbiol. 55, 265-269 (2010). [12] Tran, H.L., Hong, S.J. and Lee, C.G., “Evaluation of extraction methods for recovery of fatty acids

fromBotryococcus braunii LB572 and Synechocystis sp. PCC 6803,” Biotechnol. Bioprocess Eng. 14, 187-192 (2009).

[13] Bailey, G.F. and Horvat, R.J., “Raman spectroscopic analysis of the cis/trans isomer composition of edible vegetable oils,” J. Am. Oil. Chem. Soc. 49, 494-498 (1972).

[14] Sadeghi-Jorabchi, H., Hendra, P.J., Wilson, R.H. and Belton, P.S., “Determination of the total unsaturation in oils and margarines by Fourier Transform Raman spectroscopy” J. Am. Oil. Chem. Soc. 67, 483-486 (1990).

[15] Ozaki, Y., Cho, R., Ikegaya, K., Muraishi, S. and Kawauchi, K., “Potential of near-infrared Fourier Transform Raman spectroscopy in food analysis,” Appl. Spectrosc. 46, 1503-1507 (1992).

[16] Schober, S. and Mittelbach, M., “Iodine value and biodiesel: Is limitation still appropriate?” Lipid Techn. 19, 281-284 (2007).

[17] Setlik, I., “Contamination of algal cultures by heterotrophic microorganisms and its prevention,” Ann. Rep. Algol. Trebon , 89-100 (1966).

[18] Greenspan, P., Mayer, E.P. and Fowler, S.D., “Nile red: A selective fluorescent stain for intracellular lipid droplets,” J. Cell Biol. 100, 965-973 (1995).

[19] Jonáš, A., Ježek, J., Šery, M. and Zemánek, P., “Raman microspectrometry of optically trapped micro- and nanoobjects,” Proc SPIE 7141, 714111 (2008).

[20] Samek, O., Jonáš, A., Pilát, Z., Zemánek, P.,Nedbal, L., Tříska, J., Kotas, P. and Trtílek, M., “Raman microspectrometry of individual algal cells: sensing unsaturation of storage lipids in vivo,” Sensors 10, 8635-8651 (2010).

[21] Šerý, M., Pilát, Z., Jonáš, A., Ježek, J., Jákl, P., Zemánek, P., Samek, O., Nedbal, L., and Trtílek, M.,“Active sorting switch for biological objects,” Proc SPIE 7762, 776210 (2010)

[22] Ham, B., Shelton, R., Butler, B. and Thionville, P., “Calculating the iodine Value for marine oils fatty acidprofiles,” J. Am. Oil. Chem. Soc. 75, 1445-1446 (1998).

[23] Brandt, N.N., Brovko, O.O., Chikishev, A.Y. and Paraschuk, O.D., “Optimization of the rolling-circle filter for Raman background subtraction,” Appl. Spectrosc. 60, 288-293 (2006).

[24] Jonaš, A., and Zemánek, P., "Light at work: The use of optical forces for particle manipulation, sorting, and analysis", Electrophoresis 29, 4813-4851 (2008).

[25] Samek, O., Zemánek, P., Jonaš, A., and Telle, H.H., “Characterization of oil-producing microalgae using Raman spectroscopy”, Laser Phys. Lett. 8, 701-709 (2011)

[26] Dholakia, K., MacDonald, M.P., Zemanek, P., and Cizmar: T., "Cellular and colloidal separation using optical forces", Meth. Cell Biol. 82, 467-495 (2007).



Characterization of microorganisms using Raman tweezers

Documents

Transcript of Characterization of microorganisms using Raman tweezers