14th European Conference on Mixing Warszawa, 10-13 September 2012
DISRUPTION OF YEAST CELLS WITH ULTRASOUND J. Bałdygaa, M. Jasińskaa, M. Dzięgielewskab, M. Żochowskaa a
Warsaw University of Technology, Faculty of Chemical and Process Engineering, ul. Waryńskiego 1, 00-645 Warsaw, Poland
b
Industrial Chemistry Research Institute, ul. Rydygiera 8, 01-793 Warsaw, Poland
[email protected]
Abstract. Ultrasonic cell disruption has been investigated numerically and experimentally. The CFD model was used to predict spatial distribution of acoustic pressure in the systems equipped with ultrasonic disintegrator. Effects on cell disruption of suspension volume, horn tip position, tank diameter and amplitude of ultrasound waves were simulated. In experiments the suspension of yeast (Saccharomyces cerevisiae) in water was subject to the ultrasonic disruption. Experiments were carried out using ultrasonic processor operating at the ultrasound frequency of 20 kHz. The results have shown that the rate of release of protein is directly proportional to the square of the amplitude. It has been shown that the main mechanism of cell disruption is related to cavitation phenomena. The release of protein by cells was most efficient when the ultrasonic horn tip had been located either very close to the bottom of the reactor or far from the bottom as predicted by CFD modelling. Effects of other geometric parameters on protein release were in agreement with the CFD model predictions as well. Keywords: CFD, cell disruption, sonication, Saccharomyces cerevisiae, ultrasonic disintegrator, yeast
1. INTRODUCTION Yeasts are microscopic, eukaryotic, heterotrophic fungi. The most known and widely applied yeast, Saccharomyces cerevisiae is traditionally applied in the production of alcoholic beverages, industrial alcohol and glycerol; it is also used itself for baking and as addition to animal feed [1]. Yeast is used as well for recombinant protein production and as an expression system. Investigations of yeast cells provide insight in genetic, biochemical, and drug discovery research. Yeast cells were intensively investigated, so a complete genomic sequence is available for researchers. They grow rapidly on simple media; amino acids are synthesized from inorganic acids and sulfur containing salts, whereas carbon is often taken from organic media, usually wastes such as molasses, fruit pulps, milk whey. Moreover, yeasts can grow to high density [2]. Yeasts are single cells of 5 to 10 µm size and of spherical or oval size; they are surrounded by a thick, rigid, enduring cell wall, which makes difficult extraction of the mentioned above desired intracellular products [3]. One of main problems related to yeast utilization and research is thus a choice of a proper method for lysis of cells and extraction of proteins. One can use enzymatic digestion or hydrolysis but they can affect proteins. Another possibility is to use mechanical disruption, applying such processes as bead milling, high pressure treatment or sonication. The most popular is bead milling with glass beads but this method involves harsh conditions, which may result in protein denaturation and decreasing of yield. In this work we consider the method of sonification for cell lysing. This method is usually used to treat small samples; it is fast, efficient and easy to apply [4]. 25
2. MODELLING OF ACOUSTIC PRESSURE FIELD In the sonication process cell walls are disrupted and thus cells disintegrated by very high shear forces that are induced by the collapsing cavitation bubbles. The ultrasonic field propagates in the liquid medium from the emitter surface by pressure waves that periodically expand and compress the medium, which can create locally the above mentioned transient cavitation microbubbles. It is well known that the acoustic pressure distribution depends on the system geometry, medium and sonication cell properties, and sonication power. Schematic diagram of the set-up applied in experiments and geometry of the sonication cell are presented in Figure 1.
piezoelectric transducer
temperature probe
converter and control module
H
horn tip yeast suspension
dH
cooling bath
HT
D
DT
Figure 1. Schematic of the experimental set-up and geometry of the sonication cell.
The distribution of acoustic pressure can be obtained by solving the wave equation, 1 ∂2 p ∇2 p − 2 2 = 0 (1) c ∂t where p is acoustic pressure and c is speed of sound. Assuming that the acoustic pressure is G G time harmonic, p ( r , t ) = P ( r ) ⋅ eiωt , one gets from eq.(1) the Helmholtz equation ∇ P+ 2
ω2
P = ∇2 P + k 2 P = 0
(2) c with ω being an angular frequency, and k the wave number, k = ω c . The instantaneous pressure can be then calculated from the field of the pressure amplitude, P: G G G p ( r , t ) = ℜ ⎡⎣ P ( r ) ⎤⎦ cos (ωt ) − ℑ ⎡⎣ P ( r ) ⎤⎦ sin (ωt ) (3) 2
To solve eq. (2) one needs the boundary conditions. We assume P=0 at water-air interface (infinitely soft boundary), P=PA at the emitter surface, and ∂P ∂n = 0 for infinitely rigid walls. Computations were performed using the free software Elmer [5]. Boundary conditions P=PA
green
P=0
blue
∂P = 0 red ∂n
Figure 2. Boundary conditions applied in simulations.
26
PA, the t pressuree amplitudee at the emittter surface that is calcuulated from PA = ρ L Aωc = 2 ρ L cJ s
(4)
where A is the em mitter ampliitude and Js representts the ultrasonic intennsity defined d as the power per p unit of emitter e surfaace. In whatt follows wee present reesults of sim mulations. We start from investigatioon of effectt of the disttance from the cell boottom, D, to o emitter surface,, on propagaation of ultrrasonic wavves. Results of simulations are pressented in Fiigure 3.
Figure 3. Effect of thhe distance frrom the cell bottom to em mitter surfacee, D, on spattial distribution of pressure amplitude, P [Pa]; V=900cm3, DT=45 mm, dH=13 mm, f=20 kHz, k Amp=62 µm.
e al., [6] tthat the deccrease in Resuults of simuulations agree with obbservations of Klima et ultrasonnic intensityy when incrreasing the distance fro om the emiitter surfacee to the celll bottom (from 6 to 20 mm m) can be reeversed to the t intensity y increase due to mulltiple reflecctions by G 2 increasiing D. Noticce that the local values of the ultraasonic intennsity are J = P ( r ) ( 2 ρ L c ) . Effeects of cell diameter, DT, and celll volume, V, V on distribbution of accoustic presssure are presenteed in Figurees 4 and 5 reespectively..
Figure 4. 4 Effect of thhe cell diameeter on spatiaal distribution n of pressuree amplitude, P [Pa]; V=70 0cm3, D=15 mm, dH=13 mm m, f=20 kHzz, A=62 µm.
Figure 5. Effect of thhe cell volum me on spatial distribution of pressure amplitude, a P [Pa]; DT=50 0 mm, D=15 mm, dH=13 mm m, f=20 kHzz, A=62 µm
27
Also effects of the amplitude of oscillation of the emitter surface were simulated. Results of simulations show similar distributions of the acoustic pressure for different values of the amplitude, with minimum and maximum pressure proportional to the amplitude at the emitter surface. For example for A = 24.8 µm the minimum and maximum pressures were equal to -2.8 MPa and 4.60 MPa, for A = 62 µm, -5.2 MPa and 11.5 MPa, for A = 124 µm, -10.4 MPa and 23.0 MPa respectively. Presence of bubbles (results not shown here) complicates this simple interpretation. 3. EXPERIMENTAL INVESTIGATIONS
Experimental investigations were carried out in the system presented in Figure 1 using ultrasonic processor operating at an ultrasound frequency of 20kHz. The suspension of yeast (Saccharomyces cerevisiae) in water was subject to the ultrasonic disruption. Two strains of Saccharomyces cerevisiae were applied. An influence of process conditions on efficiency of the ultrasonic disruption was investigated by measuring the amount of soluble protein released. Concentration of proteins released to the aqueous phase from the disintegrated yeast cells was measured using the Lowry method [7]. Results of experiments are presented in what follows: Effect of the horn tip position and its amplitude on the protein release is shown in Figure 6 for strain 1 of Saccharomyces cerevisiae. 140
protein concentration [mg protein/g yeast d.w.]
A = 62 μm A = 49.6 μm
120
100
80
60
40 5
10
15
20
D [mm]
25
30
35
Figure 6. Effect of horn tip position and its amplitude on protein concentration; V=75cm3, DT=48 mm, dH=13 mm, f=20 kHz, process time=180 s, strain 1 of Saccharomyces cerevisiae.
Figure 7 shows effect of the horn tip position and the diameter of the sonication cell on the protein release. protein concentration [mg protein/g yeast d.w.]
90
DT = 45 mm DT = 50 mm DT = 53 mm DT = 57 mm
80
70
60
50 0
10
20
D [mm]
30
40
50
Figure 7. Effect of horn tip position and sonication cell diameter on protein concentration; V=90cm3 DT=48 mm, dH=13 mm, f=20 kHz, A=62 µm process time=180 s, strain 2 of S. cerevisiae
Effect of the suspension volume on the protein release for two strains of Saccharomyces cerevisiae is presented in Figure 8. 28
140
protein concentration [mg protein/g yeast d.w.]
strain 1 strain 2
120
100
80
60
40 45
50
55
60
65
70
75
sample volume [cm3]
80
85
Figure 8. Effect of the suspension volume and strain of Saccharomyces cerevisiae on protein concentration; DT=48 mm, D=15 mm, dH=13 mm, f=20 kHz, A=62 µm, process time=180 s.
4. DISCUSSION AND CONCLUSIONS
Results of experiments agree with simulation results: comparison of Figure 3 with Figures 6 and 7 shows that the variation in ultrasonic intensity observed when increasing the distance from the emitter surface to the cell bottom, D, results in similar behavior of the efficiency of protein release. Minimum of intensity for D close to 20 mm observed for DT=50 mm results in minimum release of protein for D in a range between 15 mm and 20 mm. Further, effect of sonication cell diameter (Fig. 7) agrees well with results of simulations presented in Figure 4. Results of simulations of the effect of increase of the suspension volume that are presented on Figure 5 show relative decrease of the volume active for cell disruption. This phenomenon is well observed on Figure 8, where monotonic decrease of protein concentration with increasing suspension volume is observed. Release of protein in time can be described using classical relation reported by Doulah [8] (5) C = Cmax ⋅ ⎡⎣1 − exp ( − Kt ) ⎤⎦ where K represents the protein release constant. Figure 9a shows that the form of eq. (5) agrees with experimental data, which enables identification of the protein release constant. 0.4
300
strain 1 strain 2
250 0.3
200 K [min-1]
protein concentration [mg protein/g yeast d.w.]
350
150 strain 1, A = 49.6 μm strain 1, A = 74.4 μm strain 1, A = 99.2 μm strain 2, A = 62 μm strain 2, A = 99.2 μm
100 50
0.1
0
a)
0
10
20
30
40
time [min]
50
60
70
0.2
80
0
b)
20
40
60
80
100
120
140
A [μm]
Fig. 9. (a) Protein release versus time, (b) Protein release constant versus amplitude of emitter surface oscillations: V=50cm3, DT=50 mm, D=15 mm, dH=13 mm, f=20 kHz; periodic operation: 5s on, 5s off.
Dependence of the protein release constant on the amplitude of oscillations of the emitter surface, A, is presented in Figure 9b. One can see that K is roughly proportional to A2. In fact from [8] one has β K ∝ ( Pa − Pa0 ) (6)
29
where Pa is the power input and Pa0 represents the cavitation threshold power. From presented results we get β 1( ±0.05 ) , which agrees with results for Acetobacter peroxydans [9] and E. coli [4], whereas Doulah reports β 0.9 for yeast. Dependence shown in Figure 9b can be explained by substituting for the power 1 2 Pa = S ρ L c (ω A ) (7) 2 where S represents the surface of emitter. Then, for β = 1 , eq.(6) takes the form (8) K = α S ρ L cω 2 ( A − A0 ) where A0 represents the cavitation threshold amplitude for considered emitter, and α is the constant of proportionality. Notice that whereas maximum release from both considered strains is different (300 and 350 mg/g yeast d.w. as shown in Fig. 9a), the protein release constants take very similar values for both strains. One can conclude that the CFD modeling can be very helpful for identification of optimal process conditions for protein release from yeast cells. This is very well observed when geometry of the sonication cells varies. We have found that especially position of ultrasonic horn and suspension volume affect significantly protein release. In any case there is a very good qualitative agreement between predictions of CFD and experimental data. Experimental data showing effects of process conditions on protein release can be interpreted using the concept of protein release constant. Dependence of this constant on the emitter amplitude agrees very well with theory. The values of the protein release constant are very similar for both considered yeast strains. However, the strains differ significantly in maximum protein release, which means that they are really different. All this shows that rather mechanical phenomena than details of yeast cells structure control the process of protein release. 2
5. REFERENCES
[1] Bailley J.E., Ollis D.F., 1986. Biochemical Engineering Fundamentals. McGraw-Hill Book Company, New York. [2] Rai M., Padh H., 2001. “Expression Systems for Production of Heterologous Proteins”, Current Science, 80(9), 1121-1128. [3] Zhang N., Gardner D.C., Oliver S.G., Stateva L.I., 1999. “Genetically Controlled Cell Lysis in the Yeast Saccharomyces cerevisiae”, Biotechnol. Bioengin., 64, 607-615. [4] Feliu J.X., Cubarsi R., Villaverde A., 1998. “Optimized Release of Recombinant Proteins by Ultrasonication of E. coli Cells”, Biotechnol. Bioeng., 58, 536–540. [5] Elmer 6.0 http://www.csc.fi/elmer [6] Klima J., Frias-Ferrer A., Gonzales-Garcia J., Ludvik J., Saez V., Iniesta J., 2007. “Optimization of 20 kHz sonoreactor geometry on the basis of numerical simulation of local ultrasonic intensity and qualitative comparison with experimental results”, Ultrasonic Sonochemistry, 14, 19-28. [7] Lowry O.H., Rosebrough N.J., Farr A.L, Randall R.J.,1951. “Protein measurement with the folin phenol reagent”, J. Biol. Chem., 193, 265-275. [8] Doulah M.S., 1976. ”Mechanism of disintegration of biological cells in ultrasonic cavitation”, Biotechnol. Bioengin., 19, 649-660. [9] Kapucua H., Gulsoy N., Mehmetoglu U., 2000. “Disruption and protein release kinetics by ultrasonication of Acetobacter peroxydans cells”, Biochemical Bioengineering Journal, 5, 5762.
30
