Isolation method of Saccharomyces cerevisiae from red blood cells based on the optically induced dielectrophoresis technique for the rapid detection of fungal infections.

Saccharomyces cerevisiae (S. cerevisiae) has been classically used to treat diarrhea and diarrhea-related diseases. However, in the past two decades, fungal infections caused by S. cerevisiae have been increasing among immunocompromised patients, and it takes too long to isolate S. cerevisiae from blood to diagnose it in time. In this paper, a new method for the isolation and selection of S. cerevisiae from red blood cells (RBC) is proposed by designing a microfluidic chip with an optically-induced dielectrophoresis (ODEP) system. It was verified by theory and experiments that the magnitude and direction of the dielectrophoresis force applied on RBCs and S. cerevisiae are different, which determine that the S. cerevisiae can be isolated from RBCs by the ODEP system. By designing the specific light images and the dynamic separation mode, the optimal operating conditions were experimentally achieved for acquiring higher purity of S. cerevisiae. The purity ranges were up to 95.9%-97.3%. This work demonstrates a promising tool for efficient and effective purification of S. cerevisiae from RBCs and provides a novel method of S. cerevisiae isolation for the timely diagnosis of fungal infections.


Introduction
Saccharomyces cerevisiae (S. cerevisiae) is the most widely related type of yeast to humans and is used to make food, such as bread and steamed buns, and wine. S. cerevisiae is also rich in vitamin B, protein and a variety of enzymes, so it is used to fabricate S. cerevisiae tablets for the treatment of indigestion. Raw materials for the production of nucleic acid derivatives, coenzyme A, cytochrome C, glutathione and various amino acids are also extracted from S. cerevisiae. In addition, its' specific subtype, Saccharomyces boulardii, is also used as a probiotic agent to prevent and to treat a variety of diarrheal diseases, such as the diseases related to clostridium difficile infection or parenteral nutrition [1]. Probiotic therapy is generally safe, however, many cases of invasive S. cerevisiae infections have been reported in immunocompromised and critically ill patients [2][3][4]. S. cerevisiae usually causes fungal infections when S. cerevisiae spreads through the bloodstream and enters infection sites [5,6]. Therefore, detecting S. cerevisiae from blood samples is a common method to achieve the timely diagnosis and early prevention of the fungal infections, where isolation and culture of the S. cerevisiae are the critical steps before the detection of S. cerevisiae in clinical. However, those are so time-consuming that miss the best diagnosis time and early prevention of the fungal diseases. Therefore, a new method for S. cerevisiae isolation with free label, time saving and high efficiency is urgently needed.
Micromanipulation technique is a promising method applied on isolating micro/nano-particles, such as biological cells, with non-contact, free damage and high precision. Dielectrophoretic (DEP) technique is a typical non-contact micromanipulation method with the advantage of simple operation and has been used to manipulate and isolate living cells [7][8][9]. However, for specific isolation operations, DEP technology requires corresponding design and manufacture of the metal electrode. Optically induced dielectrophoresis (ODEP), proposed in 2005 firstly [10], is an evolved version of DEP. It uses light images as virtual electrodes instead of real mental electrodes to generate uneven electric fields in the illuminated regions to manipulate electrically polarized particles. The combination of ODEP and microfluidic technology has been used successfully in manipulation, characterization and isolation of various particles, such as circulating tumor cells (CTCs) [11][12][13], fertilized egg cells [14], prostate cancer cells [15], melanocytes [16], gastric cancer cells [17] and multi-component micromachines [18]. In addition, ODEP has also been used to manipulate microbial cells [19] and to measure the mass and density of yeast cells [20]. Our previous work [21], [22] have explored the feasibility of separating S. cerevisiae from non-biological components. However, the main components of the sample of the invasive S. cerevisiae infection patients are RBCs. And both facts that the concentration of S. cerevisiae is much lower than that of RBCs and the physical characteristics of S. cerevisiae and RBCs are more similar make the high-efficient of high-purity S. cerevisiae from RBCs a huge challenge.
In this paper, a new method to isolate and collect S. cerevisiae form RBCs was proposed by designing a microfluidic chip integrated in ODEP system. S. cerevisiae was isolated from RBCs by ODEP system due to different amplitude and direction of the dielectrophoretic forces. In the experimental study, the separation of the mixture was achieved under the optimal working conditions of electric field frequency of 100 kHz and dielectric conductivity of 150 µS/cm. The experimental results showed that the method isolated S. cerevisiae with the high purity in no more than 5 minutes.

Structure of the ODEP system and the design of the microfluidic chip
A microfluidic structure for the separation of S. cerevisiae from its environment using ODEP system is proposed. The schematic diagram of microchannels of the microfluidic chip is shown in Fig. 1(a). A main microchannel (L: 10 mm, W: 0.6 mm, H: 0.1 mm) and a side microchannel (L: 20 mm, W: 0.4 mm, H: 0.1 mm) are designed to separate S. cerevisiae from RBCs. The optical control section of ODEP is performed in the connecting area between the two microchannels (L: 0.4mm, W: 0.6mm, H: 0.1mm), defined as the "isolation area". The sample is perfused at the end of the main microchannel, and the target S. cerevisiae are collected at the end of the right of the side microchannel. Meanwhile, the waste is collected at the left end of the side microchannel. The explosion structural diagram of the entire microfluidic chip is shown in Fig. 1(b). A PEEK connector (Layer A) is at the first layer. An etched indium-tin-oxide (ITO) glass with three holes (Layer B) is at the second layer. A double-sided tape with microchannels (Layer C, thickness: ∼80 µm) is at the third layer. At last, another ITO glass which is covered with a photoconductive material (hydrogenated amorphous silicon (α:Si-H) layer) (D layer).

Microfabrication and experimental setup of the ODEP chip
A glass drilling machine is used to mechanically drill three through holes in ITO glass with the thickness of 0.7 mm. A CO 2 laser machine (No. 4060, 31°Inc., China) is applied on carving the microchannels at the Layer C, whose raw material is a double-sided adhesive tape (No.30400, Deli Inc., China). Then, in order to increase the adhesion between the ITO glass and the amorphous silicon layer, a 10-nm-thick layer of molybdenum is sputtered onto the surface of the ITO layer. Then, the photoconductive material (a 500-nm-thick α-Si: H layer) is deposited onto the manufactured ITO glass through a PECVD process [23,24]. Finally, the Layer A and the Layer B are pasted with double-sided tape and 502 super glue. Because the material of the Layer C is double-sided adhesive tape, the Layer B, Layer C and Layer D can be pasted together. Before the connection of the Layer B, Layer C and Layer D, the Layer C should be cleaned by alcohol and blow-dried it with nitrogen to remove impurities from the laser carving, respectively.
The schematic and real diagrams of entire ODEP system including the microfluidic chip are shown in Fig. 1(c) and 1(d), respectively. The system contains syringe pump, CCD coupled microscope (Zoom 6000, Navitar Inc., Canada), digital projector (VPL-F500H, SNOY Inc., Japan), function generator (AFG3022C, Tektronix Inc., USA) and computer. High frequency AC voltage is applied on the both ITO layers of the chip by the function generator, the sample solution is injected into the microchannel by the syringe pump, and the animation projection of light images on the surface of the Layer D are generated by the digital projector. The manipulation of cells is observed by the CCD-equipped microscope.

Isolation process via the microfluidic chip
In the proposed ODEP system, light electrodes generated by light images are utilized to separate, and then isolate S. cerevisiae from the surrounding RBCs based on the different amplitude and direction of the ODEP force applied on S. cerevisiae and RBCs as schematically illustrated in Fig. 2. The whole process of S. cerevisiae isolation is divided into seven steps. At first, the mixture of S. cerevisiae and RBCs is injected into the main microchannel of the microfluidic chip through layer A via the syringe pump ( Fig. 2(a)). When the sample of cells is flow to the isolation zone, the syringe pump no longer continues to work. The function of the horizontal red light bars is used to capture S. cerevisiae and RBCs at the isolation zone and the vertical red light is to prevent the sample flowing into the side microchannel. Afterwards, when the AC voltage is applied on the chip, S. cerevisiae are attracted to the middle of the light bars and RBCs are trapped between horizontal light bars due to the different directions of the ODEP forces between S. cerevisiae and RBCs ( Fig. 2(b)). Then, the yellow light arrays inside the horizontal red light bars are applied to isolate S. cerevisiae to the right side microchannel, while the yellow light arrays between the horizontal red light bars to prevent RBCs being attracted to the right of the side microchannel ( Fig. 2(c)). The former yellow light arrays move to right at horizontal direction, while the latter yellow light arrays move in the opposite direction. These two steps are repeated again later ( Fig. 2(d-e)) to increase the purity of the isolated S. cerevisiae. Meanwhile, a vertical yellow light bar would appear at the right end of the junction to collect the isolated S. cerevisiae and prevent them being attracted to the junction. When finishing the second isolation process, the vertical yellow light bar would push the isolated S. cerevisiae to the right side of the microchannel ( Fig. 2(f-g)).

Principles of ODEP for S. cerevisiae isolation
The specific working principle of ODEP chip is as follows: when the external AC voltage is applied to the surface of ITO film on the upper and the lower glass substrates, the conductivity of hydrogenated amorphous silicon material is low in the absence of light, so there is almost no pressure drop in the solution of the layer C, so there is no space electric field in the solution. However, when the incident light illuminates on the lower surface of the ODEP chip, the hydrogenated amorphous silicon absorbs the energy of the emitted photons. When the energy of incident light photon is larger than the gap width of hydrogenated amorphous silicon, the electrons in valence band absorb the energy of a photon and transition to conduction band, and a hole is generated in valence band, thus a conductive electron hole pair is formed in hydrogenated amorphous silicon. The electron hole pair generated by incident light excitation increases the number of carriers in the hydrogenated amorphous silicon material and makes the conductivity of the hydrogenated amorphous silicon increase rapidly, that is, the photoconductive effect of the internal photoelectric effect of the hydrogenated amorphous silicon is produced under the action of incident light. As the photoconductivity of hydrogenated amorphous silicon increases rapidly, the resistance of hydrogenated amorphous silicon at the irradiation place of incident light decreases significantly, so that the applied voltage acts on the solution layer at the irradiation place of incident light, and then a spatial non-uniform electric field is generated at and around the incident light in the solution layer. The irradiation area of incident light is the virtual electrode.
Unlike metal electrode electrophoretic technology, ODEP chip controls the required nonuniform electric field, without the traditional photolithography technology to produce a metal microelectrode that generates a spatial non-uniform electric field, instead of using a photosensitive material -hydrogenated amorphous silicon. Theoretically, ODEP force applied on a manipulated cell can be expressed by Eq. (1) describing the dielectrophoresis (DEP) force [25][26][27][28]: where r, ε m , E and K(ω) represent the radius of a manipulated cell, the permittivity of the medium, the electric field strength, and the real part of the Clausius-Mossotti (CM) factor, respectively [29]. For a single shell model, the CM factor can be described as Eq. (2): where ε * p =ε p − j σ p ω , ε * m =ε m − j σ m ω , and ε * p , ε * m , σ p , σ m and ω represent the permittivity of cells, the permittivity of medium, the conductivity of cells, the conductivity of medium and the angular frequency of the electric field, respectively. The value of the real part of the CM factor, which depends on the inherent physical and electrical properties of the cell and the medium solution at a fixed AC frequency, determines the direction of polarization of the operated cell. Since the conductivity and permittivity of cells are fixed, the real part of CM factor can be changed by varying the frequencies of applied voltage and the conductivity of medium solution [30].
In theory, the direction of the dielectrophoresis force depends on the real part of the polarization factor, that is, when the real part of the polarization factor is greater than zero, the object is affected by the positive dielectrophoresis force and moves to the region of the maximum electric field intensity gradient, that is, the incident light, and then gathers at the incident light. On the contrary, the object is repelled by the action of negative dielectrophoresis and moves away from the incident light. According to the Eq. (1) and Eq. (2), when Re[K(ω)] = 0, the frequency of the applied AC voltage is the crossover frequency between positive and negative dielectrophoresis forces. Therefore, when Re[K(ω)] = 0, the relationship between the crossover frequency of manipulated cell and the conductivity of liquid can be obtained, and is expressed as Eq. (3): Figure 3 shows the relationship between the crossover frequency and the conductivity of solution of S. cerevisiae and RBCs, respectively, calculated according to Eq. (3), based on the parameters shown in Table 1 [31,32]. For each curve in Fig. 3, the upper right region represents that the cell was applied a negative ODEP force on, while the lower left region represents that the cell was subjected a positive ODEP force. Therefore, by selecting the solution conductivity and applied AC voltage frequency at the area between the two curves, the simultaneous separation of S. cerevisiae and RBCs can be achieved by using different directions of ODEP force, that is, positive ODEP force is applied to S. cerevisiae, while negative ODEP force is applied to RBCs.
where λ represents the wavelength of the illuminated light. The Eq. (4) shows that the ODEP force decreases with the increase of visible the wavelength of the illuminated light. And this is the reason why red light images are applied to separate cells, and yellow light images are applied to move the cells.

Sample preparation
In this study, we used common commercial dry S. cerevisiae, and RBCs from the volunteer's blood. Firstly, 0.1g dry S. cerevisiae were re-activated in the 20 mL 5% glucose solution at the temperature of 37 • C for 30 minutes. Then, phosphate buffer saline (PBS) was injected to the solution to increase the conductivity of the solution from 10µs/cm to 150µs/cm measured by a conductivity meter (Cond 3310, WTW Inc., Germany). The purpose of adding PBS is to ensure the ODEP force applied on the S. cerevisiae is positive and the ODEP force applied on the RBCs is negative. Afterwards, the RBCs from volunteers were centrifuged at 4°C and 1000 RPM for 5 min to remove plasma. After that, the RBCs were injected into 1mL isotonic solution and centrifuged again with the same parameters. The main function of isotonic solution is to provide osmotic pressure to maintain the normal morphology of cells. Cells can survive in isotonic solution for about an hour. The electrical conductivity of isotonic solution meets the requirement of dielectric electrophoresis force control, and has no damage to cells [34].
In order to quantitatively evaluate the performance of and RBCs separated by ODEP, a quantitative cell ratio method was used. The concentration of RBCs in each experiment was kept at about 5 × 10 7 cells/mL. RBCs with this concentration were mixed in S. cerevisiae solutions with different concentrations, and finally mixed solutions with three concentration ratios, namely, the concentration ratios of RBCs and S. cerevisiae in the mixed solution were 1:1, 5:1 and 10:1, respectively. The cell mixture of these three concentrations was placed in a ODEP chip, and the experiments of separating these two kinds of cells by ODEP were carried out.

Optimal operating conditions for S. cerevisiae isolation
Although the simulation above showed the S. cerevisiae could be separated from RBCs based on their different types of ODEP force applied on, the optimal operating parameters, including the frequency of the AC applied on the ITO glasses and the conductivity of the medium, should be set at first for the successful S. cerevisiae isolation through repeated experiments.
In order to obtain the optimal separation condition, ODEP forces applied on S. cerevisiae and RBCs were measured respectively using ODEP system with different the medium conductivity and electric fields frequency, and the results are shown in Table 2. Test samples contains S. cerevisiae and RBCs with different medium conductivities (50 µS/cm, 100 µS/cm, 150 µS/cm and 200 µS/cm) were used and electric field with various frequencies (50 kHz, 100 kHz, 150 kHz, 200 kHz, 300 kHz, 500 kHz, 1 MHz and 5 MHz) at the same voltage of 20V were applied on the ODEP chip. The results validated that S. cerevisiae or RBCs can be subjected to opposite ODEP forces with proper medium conductivity and frequency of the AC electric field. For example, when the medium conductivity is 100 µS/cm and the frequency of the electric field is 200 kHz∼500 kHz, the ODEP force applied on the S. cerevisiae is positive, while that on the RBCs is negative. The medium conductivity of 150 µS/cm and electric fields frequency of 100 kHz were chosen as the separation condition owning to opposite ODEP force applied on the S. cerevisiae and RBCs, so that S. cerevisiae and RBCs can be separated from each other. The lower electric field frequency would cause electroosmotic flow, which make the manipulated escape the operated area [35]; while the higher electric field frequency would reduce ODEP force that even not enough to manipulate the cells. In addition, the higher medium conductivity would generate heat problem that cause the cells to be inactive [36,37].

Table 2. ODEP response of S. cerevisiae and RBC cells to the medium conductivity and the electric field frequency, where an AC voltage of 20V has been applied on the ODEP chip
For the given medium conductivity and electric field frequency, the two kinds of cells subjected to ODEP force and the cell distribution after being captured were verified by repeated experiments. S. cerevisiae and RBCs with the medium conductivity of 150 µS/cm were injected into the microchannel, respectively, as shown in Fig. 4(i). The both cells have obvious differences in appearance, that is, S. cerevisiae is elliptical and RBC is flat and the diameters are ∼5 µm and ∼7 µm, respectively. Then, a same light image was projected on the microchannel, as shown in Fig. 4(ii)-(iii), and a same AC voltage of 20 V and frequency of 100 kHz was applied on the ODEP chip. The results show that S. cerevisiae are trapped inside the illuminated area by the positive ODEP force, while RBCs are captured outside the boundary of the illuminated area by the negative ODEP force. Finally, the ratios of the captured cells to the total number of the cells were counted. Figure 5 presents the capture ratio of S. cerevisiae and RBCs. The capture ratio of S. cerevisiae is calculated by dividing the number of the S. cerevisiae captured inside the illuminated area by the total number of the S. cerevisiae, while the capture ratio of the RBCs is calculated by dividing the number of the RBCs captured outside the boundary of the illuminated area by the total number of the RBCs. Furthermore, the capture ratio of S. cerevisiae and RBCs are 99.8% and 95.7% respectively, as shown in Fig. 5. Therefore, for the manipulation of a mixed sample of S. cerevisiae and RBCs, the location of the cell at illuminated area can be used as criterion for judging the cell type, that is, the cells inside the illuminated area are S. cerevisiae, otherwise they are RBCs.

Experiments on S. cerevisiae isolation
To find out the performance (e.g. the purity of isolated S. cerevisiae) of the proposed ODEP system for S. cerevisiae isolation, experimental investigations were performed used the samples with different ratios of S. cerevisiae and RBCs. Three samples of cells ratios (S. cerevisiae: RBCs) of 1:1 ( Fig. 6(a)), 1:5 ( Fig. 6(b)) and 1:10 ( Fig. 6(c)) were injected into the microfluidic chip, as shown in Fig. 6(i), and followed by the isolation process of S. cerevisiae as described in Fig. 2. When the AC was applied on the chip, S. cerevisiae were attracted to the middle of the light bars and RBCs were trapped between horizontal light bars, due to different directions of the ODEP forces between S. cerevisiae and RBCs, as shown in Fig. 6(ii). Then, S. cerevisiae were pushed into the right end of the side microchannel by the moving yellow light arrays from left to right inside the horizontal red light bars, while RBCs were captured at the isolation zone by the horizontal red light bars and were pushed to the left by moving yellow light arrays from right to left (Fig. 6(iii)). These steps were repeated again later (Fig. 6(iv)) to increase the capture ratio and purity of the isolated S. cerevisiae. Meanwhile, the isolated S. cerevisiae in the right of the side microchannel were attracted to the vertical yellow light bar at the right end of the isolation area, as shown in Fig. 6(iv). When the second isolation steps were ended, the isolated S. cerevisiae were pushed to the right end of the side microchannel by the vertical yellow light bar (Fig. 6(v)). After completing the isolation process, the isolation ratio and the purity of the isolated S. cerevisiae were evaluated by the microscopic images. The isolation efficiency is calculated by dividing the number of the S. cerevisiae isolated inside the right end of the side microchannel by the total number of the S. cerevisiae injected into the microfluidic chip, and the purity of the S. cerevisiae is calculated by dividing the number of the S. cerevisiae isolated inside the right end of the side microchannel by the total number of the cells isolated inside the right end of the side microchannel. The isolation efficiency for different cells ratios (S. cerevisiae: RBCs) of 1:1, 1:5 and 1:10 are 45.3%, 62.8% and 65.6%, and 72.3%, 92.2% and 93.6% for different separated steps, respectively, as shown in Fig. 7(a). The results show that a higher isolation ratio is obtained by repeating the separation steps twice. And further increase the separation times, the entire separation time is longer but the isolation efficiency is less improved. Furthermore, the higher isolation efficiency is achieved for the lower proportion of S. cerevisiae, and the fact is conducive to the actual application of the ODEP system in the infected patients, because the proportion of the S. cerevisiae in the patient's blood to RBCs is very small. Figure 7(b) show that the purities of twice isolation with three different cells ratios (S. cerevisiae: RBCs) of 1:1, 1:5 and 1:10 are 96.2%, 97.5%, 97.4% and 98.7%, 99.5%, 99.6% for different separated steps, respectively. Compared with the relevant previous reports of cells isolation, for example, the purity of isolated S. cerevisiae from polystyrene microspheres was up to 99.8% within 10 minutes [21], and the purity of isolated CTCs from leukocytes was up to 99% [14], the purity of the isolated S. cerevisiae from RBCs in this work was up to 99.6% in the same way, which proved that the purity of isolated cells in this work can achieve the same orders of magnitude in previous studies. The results further validated that the method proposed in this work can acquire high purity of the isolated S. cerevisiae in short time, which is significant for the rapid diagnosis of the invasive S. cerevisiae infections.

Conclusions
In summary, we proposed a novel method integrated a microfluidic chip into an ODEP system for the first time to realize S. cerevisiae isolation from RBCs with high purity and high efficiency. A microfluidic chip and the corresponding light images were specifically designed for the S. cerevisiae isolation. It is verified by theoretical analysis that adjusting the proper medium conductivity and the electric field frequency can make the S. cerevisiae and RBCs subject opposite ODEP force, which proves the feasibility of separating S. cerevisiae from RBCs by the ODEP system. And the experiments further verified the facts and also obtained better operating condition for the higher isolation performance. Furthermore, the experiments on the S. cerevisiae isolation for samples with different mixed ratio of S. cerevisiae and RBCs shows that the isolation efficiency is up to 93.4% and the purity of the isolated S. cerevisiae approximates 99.6%. This work provides a novel method to rapidly isolate S. cerevisiae from RBCs, which shortens the