Biotechnol. frog. 1995, 11, 127-132
127
ARTICLES Protective Effects of Polymer Additives on Animal Cells Exposed to Rapidly Falling Liquid Films Jianyong Wu$ Andrew J. DaugulisJ Peter Faulkner,*and Mattheus F. A. Goosen*’+ Department of Chemical Engineering and Department of Microbiology and Immunology, Queen’s University at Kingston, Kingston, Ontario, Canada K7L 3N6
The protective effects of polymer additives on insect cells against fluid mechanical damage was investigated in a falling film-flow device. The falling liquid film creates rapidly moving air-liquid interfaces and high fluid shear stress, mimicking the characteristics of a bursting bubble in aerated cell culture. The additives tested included a group of surface-active polymers, (i.e., Pluronic F68,poly(ethy1ene glycol)s, and Tween 80) and a group of viscosity-enhancing polymers (i.e., dextrans, methylcellulose, and (carboxymethy1)cellulose). We found that methylcellulose, which was previously considered a viscosity-enhancing polymer, actually had significant surfaceactive properties. All of the surface-active polymers exhibited significant protective effects, with Pluronic F68 and the higher molecular weight poly(ethy1ene glycol), PEG 20M, providing the best protection. In contrast, the viscosity-enhancing polymers, with the exception of methylcellulose, showed little or no protection for insect cells in the film flow. All of the protective polymers had surface-active properties, even though some of them did not change the surface tension in the actual insect cell medium. There was no correlation between the protective effect a n d the changes in liquid viscosity a n d surface tension due to the polymer additives. The level of protection was shown to be dependent upon the type of polymer, its concentration in the culture medium, and the polymer molecular weight. We concluded that the mechanism of protection of these surface-active polymers was through interaction of the polymer molecules with the cell plasma membranes: a fast-acting biological mechanism.
Introduction For mass-producing animal cells and their products, suspension culture systems, such as stirred-tank and airlift bioreactors, have the advantages of relatively uniform conditions throughout the reactor as well as simple scaleup. Growth of animal cells in suspension culture systems, however, can be inhibited by harmful fluid mechanical stresses arising from mechanical agitation and bubble aeration. Animal cells are very sensitive to mechanical stresses due to the lack of a rigid cell wall. Therefore, there is often a need for the cells to be protected in suspension culture systems where mechanical forces may become significantly deleterious to the cells. It has been known for many years that the presence of certain chemical additives in the suspension culture medium can effectively reduce animal cell damage. Among the most widely recognized protective medium additives are various polymers such as Pluronic F68 (a block copolymer of poly(oxyethy1ene) and poly(oxypropylene)) and methylcellulose (a cellulose ether). The mechanism of protection by medium additives against animal cell damage still is not fully understood. It has been proposed that the protective effect may be associated with two basic mechanisms, one physical and one biological (Michaels et al., 1991). The physical
* Author to whom correspondence should be addressed. +
Department of Chemical Engineering.
t Department of Microbiology and Immunology.
8756-7938/95/3011-0127$09.00/0
mechanism refers to interactions of the additive with the culture fluid, thereby reducing the level and frequency of mechanical forces received by the cell. The biological mechanism, on the other hand, pertains to the interaction of the additive with the plasma cell membrane and/or cell metabolic events, enhancing the resistance of the cell itself to mechanical stress. However, it still is not clear which of the mechanisms is in effect under given conditions. In addition, the specific types of interaction between the additives and the cells or the fluid require further investigation. According to their physicochemical properties, polymer medium additives can be classified as surface-active and viscosity-enhancing polymers. Surface-active polymers or surfactants are characterized by having both hydrophobic and hydrophilic components in their molecules. This molecular structure allows the polymers to adsorb spontaneously a t air-liquid interfaces in a solution to lower the overall free energy of the system (Myers, 1991). On the other hand, the viscosity-enhancing effect of polymers is mainly caused by the large size of the dissolved polymer (Allcock and Lampe, 1990). Polymer molecules in solution form coiled chains filled with many solvent molecules. Each of the polymer coils and the absorbed solvent molecules form a colloid-like bead with a volume much larger than that of the actual polymer molecule. These colloid beads exert great resistance to the flow of adjacent liquid layers with different velocities,
0 1995 American Chemical Society and American Institute of Chemical Engineers
128
Biotechnol. frog., 1995, Vol. 11, No. 2 Side View
+pI&
I
Front View
t
lil, _L$
Wires
Peristaltic
-
Bulk Liquid
Figure 1. Falling film-flow device for the study of animal cell damage at well-defined moving air-liquid interfaces.
resulting in an increase in viscosity relative to that of pure solvent. In a previous study, we constructed a falling film-flow device to expose insect cells to well-defined rapidly moving air-liquid interfaces (Wu and Goosen, 1994). Severe cell damage was detected in the film-flow device due mainly to the mixing shear stress produced by the falling film as it rapidly merged with the bulk liquid a t the bottom. In the present study, we employed the filmflow device for screening the protective effects of various polymer additives against cell damage a t rapidly moving air-liquid interfaces. A number of surface-active polymers (i.e., Pluronic F68, poly(ethy1ene glycol), and Tween 80) and viscosity-enhancing polymers (Le., methylcellulose, (carboxymethyl)cellulose, and dextran) were tested. Through these experiments, we intended to clan@whether additive protection was a result of polymer-cell interaction or was due to the changes in media physicochemical properties (e.g., increase in viscosity andor decrease in surface tension).
Materials and Methods Insect Cells and Maintenance. Insect Sf-21 cells were cultivated in shake flasks with the medium IPL-41 supplemented with 5-10% fetal bovine serum (FBS) (Gibco Laboratories, Grand Island, NY). The shake flasks were 125-mL Erlenmyer flasks on a rotary shaker shaking a t 110 rpm, with each flask containing 15-20 mL of cell suspension. The cell density in the shake flasks could reach more than (8-10) x lo6 cells/mL a t a population doubling time (PDT) of about 24 h. For the film-flow tests, the cells were harvested from the shakeflask culture during the exponential growth phase, with a density of 6 x lo6 cells/mL and a viability of no less than 90%. Film-Flow Device for the Study of Air-Liquid Interfacial Cell Damage. The falling film-flow device has been described in detail elsewhere (Wu and Goosen, 1994). In brief, it consists of a thin slot and two vertical guide wires. The liquid (cell suspension) flows through the slot and forms a thin sheet of liquid film falling down between the two guide wires into the bulk liquid a t the bottom (Figure 1). The film has a n interfacial area of 36 cm2 (2 x 18, width x length), and its thickness at the bottom is about 0.02 cm. Liquid circulation in the device was achieved with a peristaltic pump with x 3/8 Nalgene VI grade plastic tubing (Nalge Co., Rochester, NY).The total liquid volume in the device was 140 mL, and the volumetric flow rate was 260 mumin. To account for cell damage caused by pumping, agitation,
and flow through tubing, control experiments were conducted by removing the film-flow section from the device. Other conditions, such as temperature, pH, and dissolved oxygen, maintained constant during the test, were not considered to be significant factors affecting the cells. The medium IPL-41 containing 5% FBS was used throughout the film-flow experiments. To test the effect of a polymer on insect cells in the film flow, we added the polymer to the cell suspension either a t the beginning of each run or during the run. Samples (about 150 pL) were taken every 10-30 min for cell viability and lysis assays. Cell viability was determined by trypan blue dye exclusion with a hemocytometer, and cell lysis by the release of lactic dehydrogenase (LDH) from the cells. The viability and lysis values presented in the results were the fractions of viable cells remaining and lysed cells a t a given time, respectively. The details of cell viability and the lysis determination were presented previously (Wu and Goosen, 1994). Preparation of Polymer Additives. The polymer additives tested in this study were divided into two groups on the basis of their physicochemical properties: a surface-active group and a viscosity-enhancing group. In the surface-active group were several nonionic surfactants: Pluronic F68, poly(ethy1ene glycol) (PEG), and Tween 80 (polyoxyethylenesorbitan monooleate). The viscosity-enhancing group included the cellulose derivatives, methylcellulose (MC) and (carboxymethy1)cellulose (CMC) and the polysaccharides (dextrans). MC and PEG were purchased from BDH (Poole, England). All other chemicals were from Sigma Chemical Co. (St. Louis, MO). Stock solutions of the viscosity-enhancing polymers were prepared by autoclaving the dry powders at 121 "C for 20 min, then dissolving them in IPL-41 medium a t concentrations of 5-10 times those required in the flow tests. The surface-active polymers were dissolved without autoclaving in deionized water and then filtersterilized through 0.2-pm membranes to obtain 10% stock solutions. The final concentrations of polymer additives in the flow tests varied from 0.1% to 0.4% (wlv), except those of dextrans which were as high as 5%. These are the most common concentrations found in the literature for similar polymer additives (Murhammer and Goochee, 1990a,b; Michaels et al., 1991; Croughan et al., 1989; McQueen and Bailey, 1989). For most of the polymers, the molecular weights (averages) were provided by the chemical suppliers, except for the cellulose derivatives, MC and CMC. The molecular weights of MC and CMC were determined through the measurement of intrinsic viscosity (Allcock and Lampe, 1990). Viscosity of the liquid medium (IPL4115% FBS with or without polymer additives) was determined a t 27 "C with a n Ubbelohde viscometer, and surface tension was determined with a capillary rise surface tension apparatus (Fisher Scientific Co., Pittsburgh, PA). The effect of polymer additives on cell growth was tested in both static tissue culture flasks and shake-flask suspension.
Results Effects of Polymers on Medium Properties and Cell Growth. The surface-active polymers Pluronic F68 and Tween 80 had a significant effect on the surface tension of the IPL-41 medium supplemented with 5% FBS, reducing the surface tension by about 20-25% (Table 1). The other surface-active polymers, PEG 20M and PEG 4000, had a negligible effect on the surface tension. All of the viscosity-enhancing polymers in-
129
Biotechnol. Prog., 1995, Vol. 11, No. 2 Table 1. Effects of Polymer Additives on the Viscosity and Surface Tension of Insect Cell Medium IPL-41 Supplemented with 6%FBS" surface molecular concentration viscosity tension polymer type weight (% w/v) (cP) (dyn/cm) 64.4 none (control) 0 0.93 Surface-Active Polymers 50.5 Pluronic F68 8400 0.1 0.99 0.2 1.02 47.9 51.6 Tween 80 1300 0.1 0.93 60.5 PEG 4000 4000 0.2 0.96 60.5 0.4 0.98 60.8 PEG 20M 20000 0.4 1.12 Viscosity-Enhancing Polymers 63.6 MC, low DS 515000 0.1 1.48 CMC, high viscosity 186000 0.1 1.44 0.15 63.4 1.81 CMC, low viscosity 42000 0.15 1.16 0.45 1.59 0.5 1.71 63.4 dextran 35600 35600 2.5 1.41 5.0 2.02 dextran 150000 150000 1.0 1.21 1.85 1.72 2.4 2.03 63.1 a Abbreviations: FBS, fetal bovine serum; PEG, poly(ethy1ene glycol); MC, methylcellulose; CMC, (carboxymethy1)cellulose;DS, degree of substitution.
creased the medium viscosity considerably. By definition, the viscosity of a Newtonian fluid is a constant equal to the ratio of shear stress to the rate of shear (i.e., q = z/y). However, solutions of many high molecular weight polymers are non-Newtonian fluids, and the zly ratio is not constant. Therefore, we also measured the z vs y curves with a capillary viscometer (device not shown) for the polymer solutions used in our study. It was found that most of the polymers had a constant viscosity (straight lines in t vs y plots), except for the CMCs. The CMC solutions showed a slight deviation from Newtonian behavior, i.e., having a viscosity dependent on the rate of shear (pseudoplastic). Therefore, the viscosity in Table 1for CMCs may be taken as the apparent viscosity under the given experimental conditions. Cell growth in static or suspension culture was not significantly influenced by the polymer additives, except for Tween 80 a t 0.1%. The population doubling times (PDT) of the Sf-21 cells at the exponential phase with the non-growth-inhibiting polymers were 23-25 h (repeated a t least twice), which was similar to the value for the control in IPL-41/10% FBS medium without polymer additives. Tween 80 was found to be toxic to the insect cells, causing rapid cell lysis; the cells in IPL-41 medium containing 0.1% Tween 80 showed an apparent specific death rate of 0.028 h-l in the shake flasks. In the filmflow device, this toxicity may be enhanced due to a higher oxygen transfer rate a t the air-liquid interface. The cell death detected in medium containing Tween 80, therefore, could be partly attributed to the toxicity of the additive. Protective Effects of Polymers on Cells in the Film Flow. The loss of insect cell viability in the filmflow device followed first-order death rate kinetics, represented by ln(X/X,J = - k a , (X/Xo is cell viability a t time t , k d is the cell death rate constant or specific death rate, and N is the accumulated number of passes through the film-flow device from time 0 until time t). Plots of logarithmic cell viability vs time were linear, with slopes dependent on the medium additive (Figure 2). The protective effect of polymer additives on insect cells in the film-flow device was indicated by a decrease in the
0.5% CMC Low Viscosity
0.01
0
30
60
90
120
150
180
Time (min)
Figure 2. Decrease in death rate of insect cells in the filmflow device with various polymer medium additives.
cell death (or lysis) rate, in comparison with the control (IPL-41/5%FBS without any polymer additive; Table 2). For example, the addition of 0.1% methylcellulose resulted in a decrease in the specific death rate from 1.89 x to 0.11 x Since the polymers were added just before the tests, and there was no pretest incubation of cells in the polymer-containing medium, polymer protection against cell damage in the film flow had to be the result of an immediate action. Furthermore, the fastacting nature of polymer additive protection was clearly exhibited in experiments with the polymers added to the cell suspension during a test run. This immediately slowed down cell death and lysis (e.g., with Tween 80; Figure 3). The degree of protection by a polymer additive was found to be mainly dependent upon the type of polymer. Among all of the surface-active polymers tested, Pluronic F68 (0.2%) and PEG 20M (0.4%) provided the strongest protection, decreasing the cell death rate by a factor of more than 20 (Table 2). Methylcellulose (0.1%) also provided strong protection, reducing the specific death rate by a factor of 18 (from 1.89 x to 0.11 x The other viscosity-enhancing polymers (CMCs and dextrans), however, had little or no protective effect on cells in the falling film-flow device. Surface-active polymers PEG 4000 and Tween 80 also had higher protective effects than the viscosity-enhancing polymers (except MC). The protective effect of a polymer additive may also depend on its concentration in the medium, as well as on polymer molecular weight. With Pluronic F68, for example, an increase in its concentration from 0.1% to 0.2% enhanced the degree of protection, reducing the specific death rate from 0.2 x to a negligible level. This is in agreement with the result of Murhammer and Goochee (1988) for insect cells in a n air-lift bioreactor. The effect of polymer molecular weight on protection was shown between the two poly(ethy1ene glyco1)s (i.e., PEG 4000 and PEG 20M; Table 2). The higher molecular weight polyethylene glycol, PEG 20M, with a molecular weight of 20000, provided a much better protective effect than PEG 4000 (MW -4000). When the surfactants Pluronic F68 and Tween 80 were added to the cell suspension during the film-flow tests (e.g., with Tween 80, a t t = 45 min; Figure 31, a slight increase in cell viability was observed in the measurement (comparing the viability a t t = 60 min and t = 45 min in Figure 3). The same phenomenon was also observed earlier by Murhammer and Goochee (1988) with Pluronic F68. They suggested that the surfactant polymer interacts with the cell membrane and inhibits the uptake of trypan blue by the cells, thus resulting in a n
Biotechnol. Prog., 1995, Vol. 11, No. 2
130 Table 2. Protective Effect of Polymer Additives on Insect Cells in the Falling Film-Flow Devicea specific relative concentration death rateb death (%, w/v) k d f s ( x IO2) rate polymer type Surface-Active Polymers none (control) 0 1.89 f 0.25 1.0 Pluronic F68 0.1 0.22 f 0.06 0.12 0.2 0.0 & 0.0 0.0 PEG 4000 0.4 0.46 f 0.10 0.24 PEG 20M 0.4 0.0 f 0.0 0.0 Tween 80 0.1 1.35 f 0.10 0.71 Viscosity-Enhancing Polymers MC 0.1 0.11 & 0.05 0.06 CMC, low viscosity 0.5 1.71 & 0.19 0.90 CMC, high viscosity 0.15 2.26 f 0.24 1.2 dextran 35600 2.5 1.82 f 0.25 0.96 1.0 5.0 1.90 0.20 dextran 150000 1.0 1.85 f 0.15 0.98 1.24 2.0 2.19 f 0.23 ~
*
a Relative death rate is the ratio of the death rate with additive to that without the additive under the same flow conditions. s is the standard deviation of k d from a t least two duplicate runs (Le., n 2 2).
1.2
'- k
* No Tween 80
-
9
*With Tween 80
Add 0.1%Tween 80
30
60
90
120
I 150
Time (min)
Figure 3. Decrease in cell death and lysis immediately after the addition of surfactant Tween 80 (at 0.1%) during a filmflow test, indicative of a fast-acting protection mechanism.
overestimation of the number of viable cells. Another interesting phenomenon was that Tween 80, although toxic to insect cells, could still provide short-term protection against mechanical damage to the cells in the film flow (Table 2 and Figure 3). The addition of 0.1% Tween 80 to the medium in the film-flow device reduced the specific cell death rate by almost 30% (i.e., from 1.89 x to 1.35 x Table 2). Without the toxic effect, a more significant reduction in the cell death rate would have been observed.
Discussion Polymer Physical Properties and Protective Effects. As stated early in this paper, polymer additives protect animal cells either through interaction with the fluid by reducing the intensity of mechanical stress on the cells (i.e., physical mechanism), or through interaction with the cells by strengthening the cells (i.e., biological mechanism). Most polymer medium additives can have one of the two major physical effects on the culture fluid, an increase in liquid viscosity (mainly by the viscosity-enhancing polymers) and a decrease in liquid surface tension (by the surface-active polymers). Our experimental results, however, showed no apparent correlation between the protective effect of polymers and their effects on fluid physical properties (Tables 1 and 2). With the surface-active polymers, for example, Pluronic F68 and PEG 20M both showed a very strong
protective effect, although the former had a stronger effect on medium surface tension than the latter. On the other hand, the protective effect of methylcellulose, MC, does not appear to be associated with a n increase in media viscosity. Higher concentrations of CMCs and dextrans certainly made the medium much more viscous than did 0.1% MC. MC, however, showed a protective effect many times stronger than that any of the other viscosity-enhancing polymers. With insect cells subjected to well-defined laminar shear stress, Goldblum et al. (1990) also suggested that the protective effect of M e r e n t types of methylcellulose was not associated with their abilities to increase medium viscosity. With the surface-active polymers, there is a possibility that protection may arise from their interaction with the air-liquid interface. However, we have shown previously that cell damage in the falling film-flow device is mainly caused by the mixing she'ar stress in the region where the falling film merges with the bulk liquid (Wu and Goosen, 1994). This mixing stress has little relation to the state of the air-liquid interface. Therefore, it is unlikely that polymer interaction with the interface would account for much of the protective effect observed in the film-flow device. Polymer Molecular Structure and Interaction with Animal Cells. Since the protective effect of these surface-active polymers and MC cannot be attributed to their physical effects on the culture fluid, it may have been primarily due to a polymer-cell interaction mechanism. This will be a hnction of the polymer molecular structure. The interaction between surfactants and biological membranes was detected many years ago in the studies of solubilization of membrane proteins with biological detergents (small amphipathic molecules, surfactants). The nature of the interaction was shown to be mainly hydrophobic (Helenius and Simons, 1975). All biological membranes, including the plasma membranes of animal cells, are composed of a phospholipid bilayer embedded with proteins (Alberts et al., 1989). Proteins in the cell membranes or in solution have hydrophobic patches or crevices on their surfaces providing the binding sites for surfactants. When a detergent is mixed with cell membranes, the hydrophobic ends of the detergent molecules bind to the hydrophobic regions of the membrane proteins. At low surfactant concentrations (well below the lytic level), the binding may affect only cell membrane properties such as an increased membrane resistance to osmotic, mechanical, and acid lysis and changes in membrane permeability (Seeman and Weistein, 1966). A further increase in surfactant concentration leads to a n increased adsorption to, and penetration into, the cell membrane by the surfactant molecules. Cell lysis will occur upon a massive increase in the binding of detergent molecules to the membrane a t sufficiently high concentrations (Helenius and Saderlund, 1973). Tween 80 tested in our study is one of those biological detergents. Therefore, the cell lysis caused by Tween 80 was in fact a result of its strong interaction with the cell membrane, although this interaction has not been proved to be directly responsible for its short-term protection in the film flow. It appears that this surfactant, on the onehand, causes negative biological andlor physiological effects to the cells (cell-lysing effects), and on the other hand, its binding to the cell membranes can strengthen the cells mechanically. This may be a plausible explanation for the cell-lysing and mechanical protective effects of Tween 80 on the insect cells observed in our film-flow experiments. Many surface-active polymers (nonionic type) have
131
Biotechnol. Prog., 1995, Vol. 11, No. 2 0- CHS -C-
I
I
H a.Methylcellulore
O-CH&OOH C --
I
I
H b.Carboxymethyl cellulose
Figure 4. Molecular difference between methylcellulose and (carboxymethy1)cellulose.
been assessed for their protection of animal cells (mainly hybridoma and Sf insect cells) against fluid mechanical damage associated with sparged or entrained air bubbles in agitated and aerated culture vessels. These include Pluronic polyols (Murhammer and Goochee, 1990a,b), various molecular weight PEGS, and a poly(viny1alcohol) (PVA) (Michael and Papoutsakis, 1991). Some of the Pluronic poly01 and PEG surfactants were found to cause cell lysis or cell growth inhibition, suggesting their interaction with animal cell membranes, as occurred for the Tween 80 tested in our work. By analogy, those noncell-growth-inhibiting and cell-protecting surfactants, with molecular structures similar or identical to the celllysing surfactant polymers, must also have the tendency for hydrophobic binding with the cell membrane. Protective Effect of Cellulose Derivatives. It is quite interesting to notice the distinct protective effect of methylcellulose compared to that of (carboxymethyl)cellulose on the cells in the film-flow experiments (Table 2). Both MC and CMC are cellulose ethers prepared by treating alkaline cellulose compositions with alkyl halides. MC is obtained from the reaction of alkaline cellulose with methyl chloride, and CMC with sodium chloroacetate (Ott, 1943; Allcock and Lampe, 1990). The only difference in the molecular structure between the two ether celluloses is that MC has an alkyl ether side group (-O-CH& while CMC has a carboxymethyl ether side group (-O-CH&OOH) (Figure 4). This must be responsible for the differences in their physical and chemical properties. Methylcellulose is soluble in cold water, and its solubility decreases with temperature, while CMC is soluble in both cold and hot water. It is believed that the solubility of cellulose ethers is due to the unetherified hydroxyl groups made available for hydration by the alkyl ether groups spacing the individual cellulose chains (Ott, 1943). The alkyl ether group in MC tends t o be hydrophobic, which offers an opportunity for the molecule to bind to cell membranes. If this is true, the interaction of the viscosity-enhancing MC molecules with the cell membrane will be the same as that of surface-active polymers. With CMC, both the residual hydroxyl and the substituent carboxymethyl groups are hydrophilic. Hydrophobic interaction between cell membrane and polymer molecules in this case is, therefore, not possible. To determine whether MC but not CMC has surfaceactive properties, we carried out surface tension measurements of aqueous solutions of MC and CMCs. The result turned out to be just what we had expected (Table 3): MC did cause a significant decrease in the liquid surface tension, showing appreciable surface-active properties; in contrast, CMC, low- or high-viscosity grade, did not show any effect on the surface tension. This means that methylcellulose, which generally has been known as a viscosity-enhancing polymer, actually has strong surface-active properties. These surface-active properties may have made it distinct from other viscosity-enhancing polymers in protecting animal cells from fluid mechanical damage. The fact that only the surface-active polymers, including MC, provided significant cell protection in the film-
Table 3. Surface Tension of Aqueous Solutions of Cellulose Derivative Polymem, MC and CMC (2' = 27 O C ) surface concentration tension standard number polymer (%, w/v) (dydcm) deviation of tests pure water (control) 0 72.2 0.0 4 MC 0.05 64.4 0.7 4 0.1 64.7 0.6 6 CMC, low viscosity 0.1 72.2 0.0 4 CMC, high viscosity 0.05 72.2 0.0 4 0.1 72.2 0.0 4
flow device strongly suggests that the protective effect of these surface-active polymers was attributed, a t least in part, to their surface-active properties. Surface-active polymers, having amphipathic molecular structures (i.e., both hydrophobic and hydrophilic regions), may have enhanced the mechanical strength of animal cells through hydrophobic interaction with membrane proteins (Murhammer and Goochee, 1990a). The practical implication of this mechanism is that in the search for new protective medium additives, we should pay more attention to those polymers with amphipathic molecular structures, and meanwhile avoid those that can bind strongly (and irreversibly) to the membrane proteins so as to impair membrane function or even cause cell lysis. A further question concerns why the polymer-cell membrane interaction can strengthen the cells. Possible explanations may be (1)that the interaction leads to a change in cell membrane structure, resulting in a membrane with a higher mechanical strength, and (2) that the polymer molecules bound to the membrane form a protective layer (together with water molecules) surrounding the cell (Kilburn and Webb, 1968).
Conclusions The protective effects of various surface-active and viscosity-enhancing polymers on Sf-21 insect cells against air-liquid interfacial and fluid shear damage in a falling film-flow device were shown to be dependent mainly on polymer type (molecular structure), polymer concentration in the medium, and polymer molecular weight (chain length). The level of protection offered by the polymer additives had no correlation with changes in medium physicochemical properties, surface-tension, and viscosity. On the basis of these results, we concluded that the protection was primarily a result of interaction between polymer molecules and the cell plasma membrane (i.e., fast-acting biological mechanism). The strong protective effect of many surfactant polymers, including methylcellulose, on mammalian and insect cells may be related to their amphipathic molecular structure, having the tendency to bind their hydrophobic ends to the hydrophobic areas of plasma membrane proteins. However, such interaction may also cause cell growth inhibition and cell lysis as in the case of Tween 80, which has a strong interaction with cell membranes. On the other hand, viscosity-enhancing polymers, such as (carboxymethy1)celluloses and dextrans, showed little or no protective effect in our film-flow experiments. This may have been due to weak or no interaction of these polymers with the Sf-21 cells. Acknowledgment This work was funded by a Strategic Grant from the Natural Sciences and Engineering Research Council of Canada. Literature Cited Alberts, B.; Bray, D.; Lewis, J.; Raff, M.; Roberts, K.; Watson, J. D. Molecular Biology of the Cell; Garland Publishing, Inc.: New York, 1989; pp 286-287.
Biotechnol. Prog., 1995, Vol. 11, No. 2
132 Allcock, H. R.;Lampe, F.W. Contemporary Polymer Chemistry, 2nd ed.; Prentice Hall: Englewood Clifts, NJ, 1990;pp 379394. Croughan, M. S.;Sayre, E. S.; Wang, D. I. C. Viscous reduction of turbulent damage in animal cell culture. Biotechnol. Bioeng. 1989,33,862-872. Goldblum, S.;Bea, Y.-K.; Hink, W. F.;Chalmers, J. Protective effect of methylcellulose and other polymers on insect cells subjected to laminar shear stress. Biotechnol. Prog. 1990,6, 383-390. Handa-Corrigan, A.; Emery, A. N.; Spier, R. E. Effect of gasliquid interfaces on the growth of suspended mammalian cells: mechanisms of cell damage by bubbles. Enzyme Microb. Technol. 1989,11, 230-235. Helenius, A.; Soderlund, H. Stepwise dissociation of the semliki forest virus membrane with Triton X-100.Biochim. Biophys. Acta 1973,307,287-300. Helenius, A.; Simons, K. Solubilization of membranes by detergents. Biochim. Biophys. Acta 1975,415,29-79. Kilburn, D.G.; Webb, F. C. The cultivation of animal cells at controlled dissolved oxygen partial pressure. Biotechnol. Bioeng. 1968,10,801-814. McQueen, A.; Bailey, J. E. Influence of serum level, cell line, flow type and viscosity on flow-induced lysis of suspended mammalian cells. Biotechnol. Lett. 1989,8, 531-536. Michaels, J. D.;Papoutsakis, E. T. Polyvinyl alcohol and polyethylene glycol as protectants against fluid-mechanical injury of freely suspended animal cells (CRL8018).J. Biotechnol. 1991,19,241-258.
Michaels, J. D.; Peterson, J.; McIntire, L. V.; Papoutsakis, E. T. Protection mechanisms of freely suspended animal cells (CRL8018)from fluid-mechanical injury. Viscometric and bioreactor studies using serum, Pluronic F68 and polyethylene glycol. Biotechnol. Bioeng. 1991,38,169-180. Murhammer, D. W.; Goochee, C. F. Scaleup of insect cell cultures: protective effects of Pluronic F-68.Bio I Technology 1988,6,1411-1418. Murhammer, D. W.; Goochee, C. F. Structural features of nonionic polyglycol polymer molecules responsible for the protective effect in sparged animal cell bioreactors. Biotechnol. Prog. 1990a,6, 142-160. Murhammer, D.W.; Goochee, C. F.Sparged animal cell bioreactors: mechanisms of cell damage and Pluronic protection. Biotechnol. Prog. 199Ob,6, 391-397. Myers, D. Surfaces, Interfaces, and Colloids Principles and Applications; VCH Publishers, Inc.: New York, 1991. Seeman, P.; Weistein, J. Erythrocyte membrane stabilization by tranquilizers and antihistamines. Biochem. Pharmacol. 1966,15,1737-1752. Wu, J.; Goosen, M. F.A. Investigation of air-liquid interfacial damage of animal cells in a falling film-flow device. Biotechnol. Tech. 1994,8, 111-115. Accepted September 9,1994.@ Abstract published in Advance ACS Abstracts, October 15, 1994. @
