Controlled cell disruption: A comparison of the forces required to disrupt different micro-organisms

Cell disruption has been achieved using different types of equipment. Originally, yeast cells were broken by grinding with Kieselguhr and sand (Buchner, 1897) and, since then, different techniques have been developed to obtain both cell envelopes and intracellular material. The methods have included subjecting a suspension of cells to a source of ultrasonic vibration, shaking with glass beads, in addition to subjecting cell suspensions to a shear stress gradient achieved by a large reduction in pressure through a very short orifice (Rogers & Perkins, 1968; Wright, Edwards & Jones, 1974).

Milner, Lawrence & French (1950) have described an apparatus which disrupted chloroplasts, yeast and Escherichia coli by means of a shear stress gradient and measured the pressure required to obtain different amounts of chloroplast disruption. Although this disrupter can be used for the quantitative measurement of the forces required to break cells, it cannot conveniently handle large volumes as it is designed to process small single samples.

A cell disrupter has been developed which can measure the forces required to disrupt eukaryotic and prokaryotic cells. In addition, it operates a continuous process and can, therefore, disrupt large volumes of cells (Sharpe, 1975). This apparatus can subject cells to extremely high shear forces causing disruption. The shear forces are set up when the suspension is released under controlled high pressure through a short orifice. If the applied pressure is altered, the shear forces are simultaneously changed and the degree of cell disruption can be compared under different known and repeatable conditions.

Two properties of the cell disrupter have been demonstrated. Firstly, it has been used to compare the conditions required to disrupt several species of micro-organisms and, secondly, the apparatus has been used to disrupt cells by a continuous process of both large and small volumes.

The apparatus for disrupting cells in suspension consists of an orifice formed from a conical seating with a hardened ball loaded into it which allows a flow of fluid only after a given pressure has been exceeded (Sharpe, 1975). The liquid passing through the disrupter obeys laminar flow conditions (Patel & Head, 1968). The disrupter is now manufactured and supplied by Stansted Fluid Power Limited, Stansted, England.

Fig. 1 shows a diagrammatic section through the valve. The ball may be loaded either by the compression of a stiff spring or by an air-loaded piston, allowing the loading on the ball and hence the pressure drop to be continuously adjusted.

Fig. 2 shows the inlet funnel (a), high-pressure cylinder assembly (b) and disrupter valve (c) which are made in high-grade stainless steel and may be autoclaved if required. The components of the disrupting valve are designed to be disposable and may be replaced at will. The fluid containing the cells in suspension is pumped through the orifice against the loaded ball valve by a suitable electrically or pneumatically operated high-pressure pump. For the 1 · 5-mm ball valve used for these experiments, pressures up to 2 · 7 × 10⁸ Pa (4 × 10⁴ Ib/in.², 2 · 76 × 10⁵ kN m⁻⁵) were investigated. Either a 1-hp (7 · 6 × 10⁶ W) electrical pump made by Stansted Fluid Power Limited, England capable of processing 10 l./h or a smaller bench mounted pneumatic pump (also made by Stansted) which can handle up to 4 l./h were used for the experiments described.

The work done in shearing the suspension fluid and hence, the cells may be calculated from the applied pressure and the instantaneous flow or measured as the temperature rise of the fluid across the apparatus using a thermocouple in the inlet and outlet streams. For a given valve size and fluid there is a linear relationship between the applied pressure or the shear stress gradient, and the fluid temperature rise. The relationship is not affected by the flow rate, allowing the temperature rise to be used as a continuous monitor of the shear stress gradient and hence, the forces acting on the cells. The fluid temperature rise was used to calibrate and standardize the two sets of apparatus. For the valve used for these tests 6 · 8 × 10⁶ Pa (1 × 10⁶ Ib/in.², 6 · 9 × 10³ kN m⁻²) pressure drop produced 1 °C rise in the temperature of water.

The organisms used were Escherichia coli 114, Staphylococcus aureus, Oxford, Lactobacillus casei MTX/r, Bacillus subtilis 168, Streptococcus faecalis A.T.C.C. 9790, Clostridium perfringens N.C.T.C. 8237, Streptococcus zooepidermicus, Saccharomyces cerevisiae N.C.Y.C. 239, Aspergillus fumigatis, Fusarium sp. and Chlorella.

E. coli and 5. aureus were maintained on nutrient agar purchased from Oxoid Limited, London, England, and grown on Oxoid no. 2 nutrient broth. L. casei was maintained on Hedley Wright agar (Wright, 1933) and grown on medium described by Dunlap, Harding & Huennekens (1971). B. subtilis was maintained on a tryptone yeast extract agar derived from L-medium (Karamata & Gross, 1970) and grown on casein hydrolysate-yeast extract medium (CHY) (Hughes, 1968). S. faecalis was maintained on Hedley Wright agar containing 5 % horse blood and grown on CHY medium. 5. cerevisiae was maintained on a yeast plate medium (Mortimer & Hawthorne, 1969) and grown on the same medium with agar omitted. Cultures of Cl. perfringens, S. zooepidermicus, A. fumigatis, Fusarium sp. and Chlorella were provided.

Cell disruption was demonstrated qualitatively by phase-contrast microscopy. The organisms were photograhed before and after being processed through the disrupter. The detailed morphological damage caused by disruption was observed by electron microscopy, using positive staining with 1 % uranyl acetate or negative staining with 2 % sodium phosphotungstate, a service provided by Dr Wyrick and Dr Burdett.

The extinction at 260 and 280 nm of soluble material released from the cells was also used as an estimate of cell disruption. The overnight cultures (1 · 5 1.) were centrifuged at 6000 g for 20 min at 5 °C and resuspended in 0 · 05 M sodium phosphate buffer, pH 6 · 8 (1 ·0 1.). Samples (10 ml) of this suspension taken before and after passing through the disrupter at increasing applied pressure 0 to 27 × 10⁸ Pa (0 to 4 × 10⁴ Ib/in.², 2 · 76 × 10⁶ kN m⁻²), were centrifuged at 3000 g for 30 min and the supernatant solutions were diluted 10-fold with 0 · 05 M sodium phosphate buffer, pH 6 · 8, for measurement. A sample of the soluble material from B. subtilis was prepared by adding lysozyme (1 mg) to the bacterial suspension (10 ml). After incubation at 37 °C for 18 h, centrifugation and dilution, this sample was compared with those which had passed through the disrupter.

A quantitative estimate of cell disruption was made by viable counts. A viable count was performed on each culture before disruption, after passing through the apparatus with no applied pressure and also at increasing applied pressures. Serial 10-fold dilutions were prepared for each sample and the viability in 2 successive dilutions tested. Samples of E. coli and S. aureus were diluted with nutrient broth and the colony-forming units were counted on nutrient agar. The samples of L. casei, S. faecalis and B. subtilis were diluted in a minimal salt medium with glucose omitted (Davis & Mingioli, 1950) and the colony-forming units were counted on Hedley Wright agar. S. cerevisiae was also diluted in a minimal salt medium and the colony-forming units were counted on yeast-plate medium. Counts for any given dilution were made on duplicate plates. Two dilutions were routinely sampled giving effectively four counts which were averaged.

Clear evidence was obtained by phase-contrast and electron microscopy that a variety of micro-organisms can be disrupted after subjection to a high shear stress gradient in the disrupter. Fig. 3 shows a selection of organisms which have been disrupted at different pressures. Chlorella, which has a thick cellulose wall, was disrupted at 4·8 × 10⁷Pa (7 × 10³ lb/in.², 4 · 8 × 10⁴ kN m⁻²). Filamentous fungi such as A. fumigatis and Fusarium could be disrupted at 6 · 8 × 10⁷ Pa (1 × 10⁴ lb/in.², 6 · 9 × 10⁴kN m⁻²), and unicellular fungus, 5. cerevisiae was disrupted at 1 · 5 × 10⁸ Pa (2 · 26 × 10⁴ lb/in.², 1 · 56 × 10⁵kN m⁻²). Smaller micro-organisms such as E. coli, L. casei, B. subtilis, Cl. perfringens, S. faecalis, S. zooepidermicus and S. aureus may also be disrupted. Electron micrographs of B. subtilis and 5. aureus (Fig. 4) showed that the cells were broken by shearing the outer membrane. In the case of the Bacillus this was not restricted to a specific region of the rod.

The u.v. absorption in the supernatant fluids increased considerably after passing cultures through the cell disrupter. Fig. 5 A shows the results for B. subtilis. In this instance it was possible to compare the absorption of the different samples which had passed through the disrupter with the supernatant fluid remaining after a culture was treated with lysozyme. The results demonstrate that disruption is almost complete at 9 · 5 × 10⁷ Pa (1 · 4 × 10⁴ lb/in.², 9 · 65 × 10⁴ kN m⁻²), when the absorption of the cell lysate is compared with the supernatant derived from the bacterial suspensions passing through the cell disrupter. This comparison is not possible with the other species used, as they are not susceptible to lysozyme, but as shown with 5. aureus (Fig. 5B), it was always possible to demonstrate an increase in absorption with a concomitant increase in the applied pressure.

The percentage viability remaining in a cell suspension was determined with respect to increasing applied pressure. It is assumed that any viable cells are intact and undamaged while those that are not viable have been disrupted. It was observed that the cells were not all broken at any one setting of the disrupter. The pattern found for all the micro-organisms tested was that little or no disruption occurred until a certain level of applied pressure was reached, and that subsequently the proportion of cells disrupted increased rapidly with increasing disrupter pressure until almost all the cells were disrupted. Fig. 6 shows the proportion of cells disrupted at different applied pressures for each organism tested. Disruption followed a sigmoid curve in each case.

Distinctly different applied pressures were required to disrupt half the population in each culture; E. coli (Gram-negative rods, 2 · 4 × 0 · 5 μm) required 1 · 5 × 10⁶ Pa (2 · 2 × 10³ Ib/in.², 1 · 52 × 10⁴ kN m⁻²), B. subtilis (Gram-positive rods 1 · 5–3 × 0 · 5–0·8 μm) required 2 · 4 × 10⁷ Pa (3 · 6 × 10³ Ib/in.², 2 · 4 × 10⁴ kN m⁻²), L. casei (pleomorphic Gram-positive rods, varying in length up to 4 × 0 · 4–0 · 7 /tm) required 3 · 1 × 10⁶ Pa (4 · 6 × 10³ Ib/in.², 3 · 1 × 10¹ kN m⁻²), S. faecalis (Gram-positive ovoid cocci 1 · 02 μ m in diameter) required 1 · 5 × 10⁸Pa (2 · 26 × 10⁴ Ib/in.², 1 · 56 × 10⁶kN m⁻²), S. aureus (Gram-positive cocci, 1 /rm in diameter) required 1 · 9 × 10⁸ Pa (2 · 86 × 10⁴ lb/ in.², 1-9 × 10® kN m⁻²), and S. cerevisiae (oval yeast cells 7–12 × 5–8 μm) required 1 · 5 × 10⁸ Pa (2 · 21 × 10⁴ Ib/in.², 1 · 56 × 10⁶ kN m⁻²). The applied pressure required to achieve complete disruption in all cultures tested was beyond the capability of the apparatus, although greater than 95 % disruption of all cultures was achieved at 2 · 7 × 10⁸ Pa (4 · 0 × 10⁴ Ib/in.², 2 · 76 × 10⁵ kN m⁻²).

From the few species tested it seems probable that the pressure required for disruption is related to the shape of the organism, the rods being more easily disrupted than the cocci (Fig. 6). In addition to the shape, the composition of the cell wall also determines the ease with which an organism is disrupted as may be seen from the fact that E. coli, a Gram-negative rod, although similar in size to B. subtilis (a Grampositive rod) is disrupted more easily. The force required to disrupt 50 % of the cells in a culture seems dependent both on the shape of the organism and on the structure of the outer membranes.

In addition to the species specificity which might be anticipated on the basis of shape and cell wall composition, all the cells in an unsynchronized culture are not disrupted at one given pressure. When the percentage of the individual cells disrupted is plotted against applied pressure, a characteristic sigmoid curve is obtained for each micro-organism.

To determine whether cell disruption is a random or non-random process, a cell suspension was passed through the disrupter at a relatively low applied pressure which disrupted 20% of the cell population. This suspension was then passed once more through the disrupter, beginning again at zero pressure. It was observed that passing the cells through the disrupter a second time at a relatively low pressure did no-increase the number of disrupted cells. The number of viable cells decreased only when the applied pressure was increased beyond the initial maximum (Fig. 7). From this it seems most probable that the process of cell disruption is not random but is determined by some physical property of the cell population.

The theory of thin shells (Goldenveizer, 1961) suggests that the cells of a giver culture are disrupted on the basis of their size, the large cells being disrupted first, the amount of disruption at a given applied pressure being related to the number of cells having a given size or volume. If the forces required to disrupt a cell are inversely related to the physical dimension of the cells, then it should follow that the statistical distribution of disrupted cells reflects the statistical size distribution.

Kubitschek (1969) measured the individual cell volumes of E. coli in an unsynchronized growing culture and found a large distribution of cell volumes at any given time. If the number of cells of a given volume is plotted against the logarithm of that volume, it can be shown that the cultures of E. coli have a Log Normal volume distribution (Fig. 8). Similarly, it can be shown that cell viability when plotted against the Log of the applied pressure also follows a Log Normal distribution (Fig. 9). Theoretically, this suggests that the statistical distribution of disrupted cells reflects their statistical volume distribution.

It is evident that there are several factors which affect the ease with which a cell can be disrupted. Among these are cell shape, the composition of the outer membrane and, possibly, the cell volume. The advantage of the apparatus described is that different pressures can be accurately reproduced, making it possible to determine the relative forces required to disrupt each type of cell.

We wish to thank colleagues at the School of Pharmacy, University of London and at the National Institute for Medical Research, London, for the valuable help in this work.