The influence of different pH on the electrophoretic behaviour of Saccharomyces cerevisiae modified by calcium ions

The effect of a different pH on Saccharomyces cerevisiae cells modified with calcium ions was investigated by the capillary zone electrophoresis technique. For the identification of the wild strain of S. cerevisiae, the ribosomal nucleic acid sequencing and internal transcribed spacer sequencing as well as spectrometric approach were applied. The potentiometric titration and Fourier transform infrared spectroscopy have shown the occurrence of active functional groups such as carboxyl, amine/hydroxyl, phosphate/hydrogen phosphate groups on the surface of native yeast cells. Moreover, the spectroscopy study in a medium infrared range was carried out to identify the functional groups of yeast cells that participate in calcium ions binding interaction. Furthermore, the microscopic and spectrometric analysis shows that the pH value of the calcium ions solution has a significant effect on the intensity yeast cells clumping. Additionally, the impact of yeast cell clumping on the electrophoretic behaviours was examined. The modification of surface functional groups by calcium ions significantly affected the efficiency of electrophoretic separation. However, these changes did not affect the accuracy of S. cerevisiae identification by MALDI equipment with BioTyper platform. These results form the analytical solution for coupling of electrophoresis and MALDI-TOF MS technique.

Capillary zone electrophoresis (CZE) is a well known and widely used technique enabling the separation, identification and electroanalysis of yeast or bacteria 1 . Despite such widespread use of CZE, this technique is in the process of evolution and requires the development of new methods for sample preparation. In recent years a lot of interest has been focused on the possibility of applying this technique to electroanalysis to provide the identification of wild stains of microorganisms in diagnostics laboratory. However, the electrophoretic analysis of such complicated systems as microorganisms can cause many difficulties associated with uncontrolled cell aggregation and adhesion to the inner surface of the capillary 2,3 . Over the years, several strategies aimimg at the elimination of the phenomenon of aggregation and adhesion, such as the addition of poly (ethylene oxide) (PEO) to buffer solution 1,4 and inner capillary surface modification were proposed 5 . The new approach involves changes of functional groups on the microbial surface by divalent metal ions resulting in aggregation of cells controlled to a certain extent 6 . The most widely used thermodynamic model for the description of the aggregation phenomenon is DLVO (Derjaguin, Landau, Verwey, Overbeek) theory 2 . The physicochemical characterization of the microorganisms surface using instrumental analysis techniques is necessary to understand their behaviour during the electrophoretic analysis and to explain the mechanisms responsible for this process 1,2,4-6 .
One of the oldest eukaryotic microorganisms classified as fungi are yeasts. There are about 1500 species of yeasts and some of them, such as Saccharomyces cerevisiae, are used in brewing beer or baking 7 . Moreover, they are significant tools in molecular and cell biology -intensive studies concerning their genome or cell structure allowed to discover \e.g. many proteins involved in the cell cycle, signalling or metabolic pathways. S. cerevisiae can also be called a model organism because they are small with a short generation time and can be easily cultured 7 . The yeasts cell wall structure is composed of polysaccharides (glucans, chitin and mannans; about 85%) and proteins (about 15%) [8][9][10] . In glucans, glucose residues are bound through β-1,3 and β-1,6 bands, with other glucose molecules. Chitin is a linear polymer of β-1,4-linked N-acetylglucosamine (GlcNAc) and it forms microfibrils which are stabilized by hydrogen bonds. In mannans α-1,4glycosidic bonds occur 11 . Among them there are, mannoproteins and cell-wall proteins (CWPs). Mannoproteins are linked to polysaccharide non-covalently and CWPs are branched by β-1,6 or β-1,3 linkages of glucans 12 . Moreover, such a construction of the yeast cell wall results in there being many functional groups on its surface such as phosphate, carboxyl and amino groups. Under appropriate conditions of pH (above pK a value), these groups are deprotonated, which allows their interaction with positive charged metal ions which results in cells flocculation 13 . Many metal ions such as Mg 2+ , Mn 2+ , Zn 2+ , Cd 2+ or Co 2+ have been described as yeast cell aggregation inducers. However, calcium cations are known as the most effective in flocculation promotion 14,15 . It is caused by the chemistry of calcium ions, especially in contrast to D-block elements, which did not create the aggregates in pH greater than 7 14,15 . The hypothesis of calcium -binding suggests that calcium ions cause cells aggregation by binding mostly carboxyl groups present on the yeast surface. The second theory is related to the lectin-like mechanism. According to this theory, zymolectins on the yeast cells are binding specifically to mannose residues on the adjacent cell. In this conditions, calcium ions are necessary to provide zymolectins correct conformations 16 .
Once the yeast cell has a complex wall structure and is capable to induce flocculation, the Saccharomyces cerevisiae can be considered a model of (bio)colloids. The (bio)colloids are characterized by the complex surface topography and the heterogeneous distribution of the electric charge 6,17 . The term "(bio)colloids" comes directly from the term "colloid". "(Bio)colloids" are considered as the biologically heterogeneous systems which are intermediate between the real solutions and suspensions. In (bio)colloids, the particles of the dispersed phase are called microns and their diameter ranges from 0.2 μm to 0.1 mm 17 . (Bio)colloids are classified into lyophobic -repulsive dispersion medium and lyophilic -showing affinity of the dispersed phase to the dissipation phase. On the surface of (bio)colloids, a double electric layer is formed. The first one is a layer, irreversibly related to the real surface of the (bio)colloid called a chemisorption layer. The second layer is formed by reversible physical adsorption. The kinetic stability of the (bio)colloid is determined by the existence of a surface charge -zeta potential 6,17 .
The aim of this study was to investigate the relationship between the heterogeneity and charge of the microorganism surface as well as behaviour of tested (bio)colloids during electrophoretic analysis. Furthermore, the impact of the cells surface modification by calcium ions on clumping of (bio)colloids and on the effectiveness of the electrophoretic mobility was examined. Saccharomyces cerevisiae was used as a model of (bio)colloids. The novel approach of the microbial sample preparation for the CZE analysis may constitute a significant contribution to the future use of this technique in diagnostics laboratory. Moreover the spectrometric analysis of yeast modified by calcium ions may be a foundation for the coupling of capillary electrophoresis and MALDI-TOF MS analysis to eliminate the preconcentration problem of microbiological samples.

Results and Discussion
Identification of Saccharomyces cerevisiae strain. The 28 S rDNA and ITS sequencing results revealed that the investigated yeast strain was identified as Saccharomyces cerevisiae [Accession number: MG012794] with 100% (539/539) identity and sequence overlap with the most similar Saccharomyces cerevisiae strain KBP:Y-4511 (28 S) or Saccharomyces cerevisiae strain Hitachi-SUY1 (ITS) (Fig. 1A). Moreover, the molecular identification results were successfully confirmed by the analysis performed by the MALDI Biotyper platform at a species cutoff score value of 2.08 (Fig. 1B). Classification and identification of the microorganisms are based on proteomic fingerprinting using the high-throughput matrix assisted laser desorption/ionization time-of-flight mass spectrometry. Specifically, the MALDI Biotyper CA System estimates highly abundant proteins which occur in every microorganism. The specific patterns of those proteins are applied to the credible and precise identification of a certain microbe by matching it against the FDA-cleared library. According to the manufacturer of the BioTyper software (Bruker Daltonics), scores of >2.0 allow for microorganisms identification at species level 18 . Characteristics of the surface properties of Saccharomyces cerevisiae. All of the yeast cell wall components such as polysaccharides, proteins, lipids and sterols are considered to constitute a system consisting of the core and the specific surface functional groups 9 . For polysaccharides, the core is a hexagonal carbon ring and the main surface groups are hydroxyl groups (−OH). In the case of phospho-manann, there are additional (di)hydrogen phosphate (H 2 PO 4 − /-HPO 4 2− ) groups 19 . Each of them undergoes protonation and deprotonation reactions giving the electric charge on the cell surface.
In order to identify the surface functional groups of S. cerevisiae, potentiometric titration was performed. The analysis was carried out under anaerobic conditions due to the fact that yeasts conduct alcoholic fermentation resulting in the emission of carbon oxide(IV).
The pH range of these analysis ranged from 3 to 12. In order to confirm the viability of the test cells at such extremely low and high pH values, the microscopic analysis was carried out. Supplementary Table S1 shows the percentage of living and dead yeast cells at different pH and the Supplementary Fig. S2 shows the microscopic images of yeast at different pH. The results acquired indicate that in the entire pH range, yeast cells show quite good viability. Moreover, its cell wall is not damaged and cells are integrated. Figure 1C shows the titration curve of the S. cerevisiae. The pK a of yeast functional groups was calculated and were as follows: pK 1 = 4.53 ± 0.43, pK 2 = 10.24 ± 0.24, pK 3 = 7.67 ± 0.38 and pK 4 = 5.70 ± 0.59. The obtained pK a values can be attributed to the protolytic equilibrium of the surface functional groups. The results show that the carboxyl (pK a 2-6), amine/hydroxyl (pK a 8.5-10.6) and phosphate/hydrogen phosphate (pK a 5.5-7.2) groups have the dominant impact on the surface of S. cerevisiae 20  In order to determine the surface charge of S. cerevisiae, the zeta potential measurement was performed. Figure 1D shows the relationship between the zeta potential and the pH value of the yeast suspended in NaNO 3 (5 mM) solution. The zeta potential value for the yeast in the pH at range 2-6 decreases from −3 to −18 mV, while for pH above 7 its value ranges from −19 to −20 mV. Our experimental data highlighted that with the increasing pH of the suspending solution, the absolute value of the zeta potential rises. Moreover, the zeta potential value greater than ±20 mV indicates the dispersion stability of the system 21 . This suggests that tested (bio)colloid is stable at pH higher than 8.
The knowledge of the main functional groups on the yeast cell surface and its pK a values as well as of its dispersion stability allows the selection of sample preparation conditions for further analysis. Following the obtained results the pH for yeast cells modification should be equal or higher than 6, because in these conditions most of the functional groups begin to deprotonate and the system shows quite good stability of dispersion. However, the microscopic study indicates yeast cell incubation at pH higher than 9 and the change of the structure of the cells. Cytoplasm was significantly more vaculated than in cells in lower pH and its viability started to decrease ( Supplementary Fig. S2). For this reason pH 6, 8 and 9 were chosen to continue the study of cells modification by calcium ions.

Characteristics of the cells surfaces properties after its modification with calcium ions. Ca 2+
ions are considered to be one of the most important factors enabling aggregation of S. cerevisiae cells as a result of calcium-bridging of neighbouring cells by the binding surface-carboxyl groups 16 or lectin-like mechanisms of flocculation in which Ca 2+ enables the lectins to acquire a suitable active conformation and thus interact with carbohydrate residues of α-mannans present on the cell surface 22 . Moreover, regardless of cells aggregation mechanism, the pH value of culture medium may significantly affect the flocculation process since its changes alter the ionization of the functional residues present on the cell surface, mainly the carboxyl group, thus modifying the availability of Ca 2+ 23,24 . In order to confirm the surface modification of S. cerevisiae with Ca 2+ ions, the FTIR analysis was performed. According to Ojeda et al. 25 the FTIR analysis is an inexpensive and rapid tool to determine the chemical properties of microbial cells and also to identify the surfaces functional groups of microbes.
In order to eliminate the possibility that the observed changes on the FT-IR spectra result only from different pH condition, the FT-IR spectra of yeast at different pH without calcium ions were recorded. Supplementary  Fig. S3 presents the FT-IR spectra of incubated S. cerevisiae at different pH without the calcium solution. These results indicate that a change only in the pH of the cells environment did not significantly affect the resulting FT-IR spectrum. Figure 2 shows the FTIR spectra of S. cerevisiae unmodified and modified by calcium ions at different pH. Comparing the spectra of the unmodified and modified yeast, many important changes can be observed. After the surface modification, a new signal at v = 722 cm −1 appeared -it may derive from P = O vibrations from phosphate/hydrogen phosphate groups. There is also a shift of signal from v = 1652 cm −1 to v = 1644 cm −1 which is associated with the vibrations of the carboxyl and amino group 17,19,26 . These changes can indicate the creation of bonds between the calcium cation and particular functional groups responsible for this vibration. Moreover, as a result of the modification, merging signals at v = 1715 cm −1 and v = 1460 cm −1 present in the spectrum of unmodified yeast can be observed. Another characteristic band appears at v = 1377 cm −1 and derives from the stretching vibration of C-N from surface proteins. These changes are probably related to the complexation reactions between the calcium cation and O-and N-donors of proper functional group 26,27 . The obtained data prove that the S. cerevisiae exhibits the ability of the sorption of divalent metal cations such as calcium ions 27 . Furthermore, it can be observed that the most significant changes occur in the spectrum of the yeast modified by 5 mM Ca(NO 3 ) 2 solution at pH = 9, which is caused by the total deprotonation of -COO − , -OPO 3 2− and -NH 2 groups. These results are also confirmed by the zeta potential measurement in this study -the highest dispersion stability of the system was observed at the pH range from 8 to 10 21 . Therefore, under these conditions, the tested (bio)colloid is characterized by a large active surface and the highest availability of functional groups able to bind calcium ions. This indicates that the most effective biosorption process occurs in these conditions. The results indicate that the surface modification of (bio)colloid has a significant impact on its electrophoretic mobility. This phenomenon was also observed by Pomastowski et al. 28 in the case of bacteria. They stated that the change in the electrophoretic mobility of bacteria cells and the reduction of repulsive forces have resulted in clumping of cells and a signal amplification. Moreover, after the modification, the sharpening of the peaks and improvement of the shape of the base line on the electropherogram as well as the reduction in the number of aggregates and the improvement of the reproducibility can be observed; it was evidenced by a significant reduction in the relative standard deviation value in comparison to unmodified yeast. Another interesting observation is the effect of pH on modified (bio)colloid behaviour during the electrophoretic analysis. With the increase in the pH of medium in which sorption was conducted, the increase in the electromigration time of modified yeast cells and as a decrease in the electrophoretic mobility (38.1 ± 2, 27.8 ± 1 and 24.5 ± 1 [10 −5 cm 2 Vs] for pH 6, 8 and 9, respectively) were observed (Fig. 3). However, these effects were not observed for the yeast cells incubated at different pH but without the presence of calcium ions, as demonstrated in Supplementary Fig. S4.
The highest reproducibility (RSD < 5%) of the electrophoretic analysis was obtained for S. cerevisiae which was modified with Ca(NO 3 ) 2 solution at pH = 9. This phenomenon is probably connected with the total deprotonation of surface functional groups, which is also confirmed by the results of the spectroscopic and potentiometric analysis. The total deprotonation of active functional groups at pH = 9 causes the high dispersion stability (zeta potential measurements) and high availability of ion binding sites at the starting point of the sample preparation stage, which determines the creation of stable electrokinetic clumping during the yeast cells electroanalysis. A high availability of the deprotonated functional groups on the surface of S. cerevisiae allows binding of a large amount of calcium ions which promote the formation of aggregate systems.
Moreover, the study of the size of modified and non-modified yeast cells at different pH ( Supplementary  Fig. S5) indicates that the change occurring only in the pH of the incubation does not have a significant impact on the size distribution of the examined cells, while the modification of cells by calcium causes the unification of their size. This phenomenon may also contribute to the improvement of their electrophoretic separation.
Microscopic and spectrometric studies of modified and non-modified yeasts were performed in order to highlight the changes in the molecular profile of (bio)colloids under different pH conditions.
The MALDI-TOF MS spectra of yeast incubated at different pH (6, 8 and 9) without calcium ions are present in Supplementary Fig. S6. As it can be observed the incubation of S. cerevisiae at different pH does not significantly affect their molecular profile in comparison to the control -S. cerevisiae sample taken directly from growth media. On the obtained spectra only slight changes can be observed. What may be observed is the appearance of signals at about 2776 and 5853 m/z as well as the disappearance of signals at 8128 m/z on the spectra of yeast at different pH in comparison to the control sample. Moreover, it can be noticed that the spectra of the yeast incubated at pH 9 are the most similar to control. There are a number of signals which are not present on the spectra of the yeast incubated at pH 6 and 8 e.g. at 4451, 5178, 7101 and 9433 m/z. This result suggests that the cells incubation at pH 9 causes the smallest changes in the S. cerevisiae molecular profile. Figure 4 shows the MALDI spectra and microscopic images of S. cerevisiae suspended in calcium solution at pH 6, 8 and 9 and Table 1 presents signals that appear and disappear in spectra for individual samples compared to the control sample.
The MALDI spectra of native yeast cells and yeast incubated with calcium solution at different pH show obvious changes. On the MALDI spectra of yeast modified by calcium at pH 6, the new signals e.g. at 3213, 5004 and 7319 m/z were registered. Moreover, in comparison with the control, some signals e.g. m/z = 3023, 5781 and 8418 disappeared or move to higher m/z values (Fig. 4A). The same phenomenon can be observed in case of yeast modified by calcium ions at pH 8 and 9 (Table 1). However, the largest changes in the molecular profile of S. cerevisiae were observed after its incubation in Ca(NO 3 ) 2 at pH 9 (Fig. 4C). It can also be observed that the modification with calcium ions leads to the disappearance of the same signals regardless of the incubation pH value e.g. at 5137 and 6636 m/z. Moreover, the same signals disappeared on the spectra of yeast modified in lower values of pH (6 and 8) (e.g. at 3424, 3834 and 3876 m/z) as on the spectra of native yeast. A similar relationship can be seen for the spectra at pH 8 and 9. The some signals present on the native yeast spectra disappear (e.g. at 4570, 4987 and 5483 m/z) in this condition and one signal not present on the control spectra has appeared (3268 m/z). The cluster analysis of MSP (Main Spectrum Profile) spectra of control and yeasts incubated in the calcium solution at pH = 6, 8 and 9 shows the significant differences between samples (Fig. 5A). The largest number of similarities were observed for a native sample and yeast incubated in Ca(NO 3 ) 2 solution at pH = 6. The greatest change in molecular profiles, as compared to the control sample, was observed for the yeast incubated in calcium solution at pH = 9.
The spectrometric study shows the differences in the MALDI spectra of the incubated yeast cells in calcium solutions in comparison with the control sample (unmodified yeast). The changes were observed at the molecular level of ribosomal, cytosol and membrane proteins. It is known that the presence of the Ca 2+ ions in the culture medium may strongly influence the expression of proteins in S. cerevisiae cells, mainly ribosomal or ribosome-associated proteins as well as proteins related to the cell wall. High concentrations of Ca 2+ leads to    Table 1. Summarizes of signals observed on MALDI-TOF MS spectra (Fig. 4), where (−) mean disappearance of signal in comparison to control and (+) mean appearance of signal in comparison to control. calcium ions at pH 6, 8 and 9 is probably associated with the decrease in the expression level of H/ACA ribonucleoprotein complex subunit 3 which occurs in nucleus and is responsible for the RNA binding process. In turn, the disappearance of signal at 7041 m/z on the yeast spectra incubated with Ca 2+ at pH 6 and 8 indicates the cessation of the production of one of the mitochondrial proteins; in the same condition the appearance of signals at 5821 and 3203 m/z derivating from cytochrome protein was noted. Moreover, many changes on each spectra of modified yeast in comparison to spectra of native yeast may be observed, which indicates a change in the expression of membrane proteins. There can also be seen a disappearance of signals at 6509, 6249 and 3643 m/z at pH 6, 8 and 9, respectively as well as the appearance of new signals at 6752 m/z at pH 8 and 6974 m/z at pH 9.
The MALDI-TOF MS technique is a new approach enabling for a quick and accurate identification of microorganisms. However, this method requires a relatively high concentration of the microbial sample. According to Pomastowski et al. 28 , less than 10 4 cells per spot is not enough to generate the spectrum and the optimal number of cells amounts 10 5 -10 9 per spot. For this reason, an extremely important aspect is the development of an effective method of the concentration of microbiological samples so that it does not affect matching of the obtained spectrum with the reference spectra of the microorganisms deposited in the database. The precontamination method of modifying microbial cells by calcium ions by the capillary electrophoresis presented in this work may be considered an analytical solution of this problem. Obtained results indicate that the modification of yeast cells by calcium ions at strictly defined pH leads to its effective electrofocusing by the capillary electrophoresis. Moreover, this sample modification does not affect the identification degree. The score for modified and non-modified yeast was about 2 (using BioTyper database), which ensures unambiguous identification.
The microscopic observation shows that the highest aggregation of the yeast cells was achieved for the system, where solution at pH = 9 was used. In the case of the calcium solution at pH = 6, cluster formation wasn't observed (Fig. 4A). Moreover, microscopic results prove that the yeast surface changes with a calcium ion affect the aggregates formation. A similar phenomenon was observed by Dziubakiewicz et al. for the bacterial cells 6 . As a result of the bacteria surface charge modification by calcium ions, the cells created a compact aggregates, whereby, fewer high-intensity signals on the electropherograms were observed. They attributed this score to the bridging effect of calcium ions between bacterial cells. According to the well-known theory of dicationic bridges, carboxyl and amine surface groups are involved in the formation of aggregates. Partially deprotonated carboxyl groups and protonated amino groups are not able to form bridges between (bio)colloids. After yeast centrifugation from calcium salt solution and suspended in TB buffer, probably borate ions associate and/or complex adsorbed calcium ions. The created ions of TB buffer respectively associate components of the microorganisms surface. In turn, yeast suspended in a calcium solution at pH = 8 form aggregates with part-compact structure (Fig. 4B).
Deprotonated carboxyl groups allow for the creation of a cationic bridge between the surfaces of the yeast cell wall (Fig. 5B). After suspending the system in TB buffer, the metal cations previously adsorbed on the (bio) colloids surface, are associated and/or complexed by suitable buffer ions. In turn, yeast suspended in the calcium nitrate solution at pH = 9 forms aggregates of compact structure (Fig. 4C).
Deprotonated carboxyl surface groups allow for the creation of a series of the cationic bridges between (bio) colloids surfaces (Fig. 5B). After transfering the system to the TB buffer, the metal cations previously adsorbed on the (bio)colloid surface are then associated and/or complexed by suitable buffer ions. Based on the above considerations, it can be seen that with the increasing pH the buffer growth of (bio)colloid particles weight (cells clumping) is observed. Therefore, when the size and masses of (bio)colloids are increasing, the electrophoretic mobility of (bio)colloids is decreasing.

Conclusion
Capillary zone electrophoresis allows for the determination of a variety of biological systems such as bacteria, yeast and fungi. However, the complexity of the microorganism surface morphology forced us to carry out a series of research on the physicochemical properties in order to interpret the phenomena occurring at the interface of (bio)colloids. The characterization of microorganisms surface functional groups and its modification allowed us to develop a effective method for the electrofocusing of yeast cells. The use of potentiometric titration and the FTIR spectroscopic analysis allowed for the identification and determination of an amount of dominant surface functional groups of yeast, namely carboxyl (16%), amino/hydroxyl (16%) and phosphate/ hydrogen phosphate (68%) groups. The measurement of surface charge was another step to broaden the knowledge of S. cerevisiae electrochemistry. The knowledge of the surface charge and dominant surface functional groups of S. cerevisiae led to the selection of appropriate analysis conditions in order to eliminate the adverse effects of yeast adhesion to the capillary wall and uncontrolled clumping of native species. For this purpose, yeast surface was modified with calcium salt solution at different pH. Moreover, microscopic, spectrometric and spectroscopic studies performed for non-modified and modified S. cerevisiae have shown that under specified conditions the tested system tends to aggregate. This phenomenon explains the stability of (bio)colloid subjected to the biosorption process in electrophoretic separation conditions. In addition, it was observed that along with the increasing pH of modifier the weight and surface of particles of colloidal system increases, which results in the reduction of the (bio)colloid electrophoretic mobility. Obtained results indicate that the proposed new sample preparation approach of wild microorganisms strains may be a foundation for the application of capillary electrophoresis in diagnostics labolatory in the future. Moreover, the results show that the different pH and modification of the cells by calcium ions influence the molecular profiles of yeast cells but do not affect the identification quality by the MALDI-TOF MS equipment with the BioTyper database. Such results may provide a basis enabling for coupling of capillary electrophoresis and the MALDI-TOF MS analysis.

Materials and Methods
Materials. All solvents and materials were purchased from Avantor (Gliwice, Poland) and Sigma Aldrich (Poznan, Poland). Ultra-pure water was purified using the Milli-Q RG system (Millipore Intertech, Bedford, MA, USA).

Culturing and identification of Saccharomyces cerevisiae cells isolated from baker's yeast.
Pure culture of Saccharomyces cerevisiae cells was obtained from baker's yeast (Gliwice, Poland) by a serial dilution method using a sterile saline solution (0.87%). 100 µl of 10 −5 dilution of yeasts was smeared on the YPD (Yeast Peptone Dextrose) agar plates (Sigma Aldrich, Poznan, Poland) and incubated for 24 h at 37 ± 0.5 °C. Subsequently, single colonies were transferred on the YPD slants, incubated overnight at 37 ± 0.5 °C and stored at 4 ± 0.5 °C. In the experiment overnight cultures of yeast cells were transferred into the YPD Broth (Sigma Aldrich, Poland) and cultivated for 24 h at 37 ± 0.5 °C.
In order to identify the S. cerevisiae strain used in this work, the DNA isolation from yeasts cultures using Extractme ® DNA Yeast Kit (BLIRT S.A., Gdansk, Poland) was performed, which was followed by the amplification of two genome regions by PCR reaction -(1) 28 S using 28S_NL1: 5-GCATATCAATAAGCGGAGGAAAAG-3 and 28S_NL4: 5-GGTCCGTGTTTCAAGACGG-3 primers as well as (2)  3 (Cerko Lab System, Gdansk, Poland with PT1000 probe and ERH-13 electrode with Cell software). Another sample prepared in the same way was titrated with NaOH (0.1 M) to pH 11. pKa values were calculated by Protofit 2.1rev1 free software licensed under the Gnu General Public License (GPL) 33 .

Preparation of yeast cells modified by calcium ions at different pH and yeast cells at different pH.
S. cerevisiae samples were suspended in 5 mM Ca(NO 3 ) 2 adjusted to the desired pH (6, 8 and 9) by the addition of saturated Ca(OH) 2 . After 60 minutes of incubation (20 ± 0.5 °C), the suspension was centrifuged (21 ± 0.5 °C, 9000 rpm, 15 min). The resulting precipitate was washed twice with water to remove unbound calcium ions. The same procedure was used to obtain the control sample of yeast at different pH, but instead 5 mM Ca(NO 3 ) 2 , water adjusted to desired pH using 0.1 mM NaOH solution was used. As for the control sample, yeast taken directly from media were used.
Spectrometric and microscopic analysis and determination of yeast cell size. In case of the microscopic analysis, the modified and non-modified S. cerevisiae samples were stained with acridine orange (0.12 μg/mL) for 5 min in the dark, then centrifuged at 4000 rpm (5 min). The supernatant was eliminated to remove the unbound dyes and yeast cells were detected through the green fluorescence using the Zeiss Axiocom D1 fluorescence microscope. For the spectrometric analysis the sample of modified and non-modified S. cerevisiae was analysed according to the previously described methodology 28 . The size of non-modified and modified S. cerevisiae cells at different pH was determined using the Mastersizer 3000 (Malvern).
Capillary zone electrophoresis analysis. 50 mg of modified and non-modified S. cerevisiae were transferred to the outlet TB buffer tris(hydroxymethyl)aminomethane (TRIS) C TRIS = 4.5 mM, boric acid (BA) C BA = 50 mM; pH = 7.98) as the inlet buffer TBH (TRIS C TRIS = 4.5 mM, BA C BA = 50 mM, hydrochloric acid C HCl = 3.31 mM; pH = 7.31) was used. Directly before the CZE analysis the obtained yeast sample was placed in an ultrasonic bath (2 min) (Ultrasonic Cleaner USC600THD, Radnor, Pennsylvania). Before the electroanalysis of yeast the EOF (t d = 1.23 min) was determined by thiourea injection using the pressure mode (10 mbar, 10 s). The CZE analysis was performed using PA 800 plus (Beckman Coutner system, Brea, CA, USA) equipped with a DAD detector with the use of fused silica capillaries (id = 75 μm; L tot = 33.5 cm; L eff = 25 cm; Composite Metal Services, Shipley, UK). The yeast samples were injected in the capillary using a pressure mode (20 mbar, 8 s). The analysis were performed at a constant voltage = 15 kV and at the temperature 23 ± 0.5 °C. The samples were detected at λ = 214 nm. The CZE analysis was performed in pseudo-isotachophoretic mode 6 .