Quantcast
Last updated on April 24, 2014 at 1:21 EDT

High-Pressure Homogenization As a Non-Thermal Technique for the Inactivation of Microorganisms

December 17, 2006

By Diels, Ann M J; Michiels, Chris W

In the pharmaceutical, cosmetic, chemical, and food industries high-pressure homogenization is used for the preparation or stabilization of emulsions and suspensions, or for creating physical changes, such as viscosity changes, in products. Another well-known application is cell disruption of yeasts or bacteria in order to release intracellular products such as recombinant proteins. The development over the last few years of homogenizing equipment that operates at increasingly higher pressures has also stimulated research into the possible application of high-pressure homogenization as a unit process for microbial load reduction of liquid products. Several studies have indicated that gram-negative bacteria are more sensitive to high-pressure homogenization than gram-positive bacteria supporting the widely held belief that high- pressure homogenization kills vegetative bacteria mainly through mechanical disruption. However, controversy exists in the literature regarding the exact cause(s) of cell disruption by high-pressure homogenization. The causes that have been proposed include spatial pressure and velocity gradients, turbulence, cavitation, impact with solid surfaces, and extensional stress.

The purpose of this review is to give an overview of the existing literature about microbial inactivation by high-pressure homogenization. Particular attention will be devoted to the different proposed microbial inactivation mechanisms. Further, the different parameters that influence the microbial inactivation by high-pressure homogenization will be scrutinized.

Keywords High-pressure homogenization; Microbial inactivation; Non-thermal inactivation

1. INTRODUCTION

Heat treatment is one of the most commonly used preservation treatments for food and other perishable products. It is an efficient and economical process to inactivate microorganisms, but it can cause unacceptable deterioration of products with a heat labile chemical or physical structure. Therefore, over the last ten years, considerable research efforts have been directed towards the development of novel mild non-thermal preservation processes which combine efficient germ reduction with a maximal retention of the chemical and physicochemical product properties, such as the use of high hydrostatic pressure, pulsed electric field, ultra-violet light, pulsed light and high-pressure homogenization. The application of high hydrostatic pressure treatment to food processing is an attractive technology regarding safety improvement and extension of shelf-life (Hoover et al. 1989; Shigeshia et al. 1991). However, practical exploitation of this technique is limited because it is only possible to use batch or semi-batch operations (Hayashi 1989; Vachon et al. 2002).

In 1900 Auguste Gaulin invented homogenization and presented it at the World Fair in Paris with great success (French Patent no. 295.596). Since then, homogenization has been introduced in the food industry. It allows the production of dairy and food emulsions with improved texture, taste, flavor and shelf-life characteristics, especially for dairy products like milk, cream and ice cream and provides also enhanced consumer acceptance of some products in comparison with the untreated product (Dickinson and Stainsby 1988). The demand of consumers for longer shelf-life and products with better stability has led to evolutions and new developments in the homogenization technology. Higher pressures or double homogenizations have been successfully introduced (Paquin 1999). More recently, in the early 1990s, a new generation of homogenizers, referred to as high-pressure homogenizers, has been developed. This new technology-high-pressure homogenization-has a different reaction chamber geometry which can withstand pressures 10 to 15 times higher than the classical homogenizers. Different types of equipment in this category now exist, both prototype and industrial scale equipment (Burgaud et al. 1990). The development of equipment operating at increasingly higher pressure has created possibilities for new applications and product enhancements not available with lower-pressure operations. The term ‘high pressure’ is not precisely defined. In the early days, 34 MPa was considered as a high pressure for homogenization; but today, 150 MPa is not unusual. Some designs achieve even pressures of 300 MPa or more (Pandolf 1998). Even though the homogenizer may commonly be thought of as a machine to process dairy products, it is now an essential part of many industrial applications. In the pharmaceutical, cosmetic, chemical and food industries high-pressure homogenization is used for the preparation or stabilization of emulsions and suspensions, or for creating physical changes, such as viscosity changes, in products (Pandolf 1998; Paquin 1999; Floury et al. 2002). For some applications that require a high degree of dispersion, pressures up to 200 MPa or higher are currently being explored. Another application is cell disruption of yeasts or bacteria in order to release intracellular products such as recombinant proteins.

Since high-pressure homogenization can be applied for cell disruption of dense microbial cultures (Kelemen and Sharp 1979; Engler 1990; Harrison et al. 1991, Siddiqi et al. 1997; Shirgaonkar et al. 1998), it can be anticipated that high-pressure homogenization for the production of emulsions or suspensions will also cause a partial inactivation of the microbial population. In the literature there are indeed a few studies indicating this (Popper and Knorr 1990; Lanciotti et al. 1994, 1996; Guerzoni et al. 1999). This reduction of microorganisms, although not the primary purpose of the process, is also potentially interesting since it may result in an extended shelf-life and improve the microbiological safety of the processed products. As such, it may reduce the need for other process steps that are carried out for reduction of microorganisms (heat sterilization, irradiation) or reduce the required intensity of such process steps. This may influence product quality and process cost in a beneficial way.

2. PRINCIPLE OF HIGH-PRESSURE HOMOGENIZATION

A homogenizer consists of a positive displacement pump and a homogenizing valve. The pump is used to force the fluid into the homogenizing valve where the work is done (Middelberg 1995). Effluent from the homogenizer is normally chilled to minimize thermal damage to the product caused by friction heat generated due to the high fluid velocity that elevates the product temperature about 2.0 to 2.5C per 10 MPa (Engler 1990; Popper and Knorr 1990). In the homogenizing valve, the fluid is forced under pressure through a small orifice between the valve and the valve seat (Figure 1). The fluid leaves the gap in the form of a radial jet that stagnates on an impact ring (Middelberg 1995). Finally, it exits the homogenizer at low velocity and essentially atmospheric pressure. The operating pressure is controlled by adjusting the distance between the valve and seat.

3. INACTIVATION MECHANISMS OF MICROORGANISMS BY HIGH-PRESSURE HOMOGENIZATION

In spite of its apparent simplicity and widespread use for large scale disruption of microorganisms and other particles, high- pressure homogenization remains a poorly understood unit operation (Middelberg et al. 1991). The mechanism of disruption is still a matter of some debate. Several mutually interacting mechanisms have been proposed as causing disruption.

FIG. 1. The flow path through a simple homogenizing valve.

All the previous modeling approaches have been developed to describe experimental data but they have not led to the development of a predictive homogenization model. Within the past 10-15 years, some researchers have started to use Computational Fluid Dynamic (CFD) models to estimate the spatial and temporal variations of velocity, shear rate and energy dissipation rate versus operating conditions and equipment configurations (Weetman and Hutchings 1989; Brucato and Micale 1998). In each cell of the numerical grid, the equations of motion are solved by either a finite difference or a finite volume approach. It is unlikely that detailed measurements of pressure and velocities with the high-pressure homogenizer can be obtained experimentally, due to the presence of large variations in both velocity and pressure over small distances (Floury et al. 2004). Secondly, the dimensions of the homogenizing valve are very small and the operating pressure can be very high. As a consequence, it may be impossible to measure pressure or velocities in such high- pressure devices (Floury et al. 2004). To obtain estimates of these parameters, numerical studies are required. A few researchers have investigated whether CFD can help to understand flow in a high- pressure homogenizer. Kleinig and Middelberg (1996 and 1997) used CFD to model the flow in a homogenizing valve through numerical simulations. In their first study they modeled the flow only in the inlet part of the valve and the valve gap region. Then, in the second paper, they used the velocity profiles obtained at the outlet of the gap in the previous work (Kleinig and Middelberg 1996) as inlet velocity boundary condition to model the flow in the impinging region of the valve. Stevenson and Chen (1997) were the first to model the flow pattern in the entire valve \with a commercial CFD code. Miller et al. (2002), by using the Fluent software, developed a 2D CFD model of flow trough a cell disruption homogenizing valve. This model represents an improvement over previous models because of the optimized grid and the use of a more appropriate turbulence model. Based on this model they concluded that the impact distance only affected wall impact pressure. Fluid viscosity on the other hand, was found to significantly affect turbulence, pressure gradient, channel strain rate and wall impact pressure. Research done by Diels et al. (2005a) confirmed that fluid viscosity plays an important role in bacterial inactivation. These authors developed a model describing bacterial inactivation as function of homogenization pressure and fluid viscosity. In addition, they proved that between 5 and 45C, the effect of temperature on the inactivation of E. coli MG1655 by high-pressure homogenization could be entirely explained by the temperature-dependent change of fluid viscosity (Diels et al. 2004).

All of the previous studies have been conducted with an APV- Gaulin high-pressure homogenizer (Figure 2a), in which the fluid is fed axially into the valve seat, and then accelerated radially into the gap between the valve and the seat. When the fluid leaves the gap, it becomes a radial jet that stagnates on an impact ring before leaving the homogenizer. Recently Floury et al. (2004) studied the flow pattern of a new type of high-pressure homogenizers from Stansted Fluid Power Ltd (Figure 2b). In this design, the fluid streams axially under high pressure along the mobile part of the valve and flows with high velocity through the radial narrow gap formed between the valve seat and the piston, before leaving the valve seat at atmospheric pressure. Once the fluid in a Stansted Fluid Power Ltd homogenizer leaves the gap between the valve and the valve seat, it becomes a radial jet. In the APV-Gaulin homogenizer, this jet then violently collides with an impact ring before leaving the homogenizer at atmospheric pressure, and cell disruption occurs partially in this impinging jet region (Kleinig and Middelberg 1997). In the Stansted homogenizing valve in contrast, Floury et al. (2004) noticed that the fluid velocity decreased very quickly after leaving the valve gap, indicating that the fluid does not strongly impact on the valve wall. The flow was strongly fluctuating in the downstream region of the homogenizing valve, and the maximum turbulence intensity of the fluid at the exit of the valve gap increased strongly with the homogenizing pressure. Also the volume occupied by turbulent eddies increased a lot with pressure. Again, this behavior is different from that in the APV-Gaulin homogenizer, where turbulence could be changed by varying the valve seat diameter, but was independent of homogenizing pressure (Rovinsky 1994).

FIG. 2. Homogenizing valve geometries. A: APV-Gaulin valve; B: Stansted valve.

5. PARAMETERS AFFECTING MICROBIAL INACTIVATION BY HIGH-PRESSURE HOMOGENIZATION

Microbial inactivation by high-pressure homogenization has been studied in buffer systems and in real food products like milk, ice cream and orange juice (Feijoo et al. 1997; Guerzoni et al. 1999; Jean et al. 2001 ; Kheadr et al. 2002; Vachon et al. 2002; Thiebaud et al. 2003; Diels et al. 2005a). However, larges differences in inactivation are sometimes found in different studies. A possible explanation for this is that all experiments were done in different media, under different process conditions, with different microorganisms and with different equipments.

In this paragraph, the most important parameters that influence the microbial inactivation by high-pressure homogenization will be discussed. In general, these parameters can be divided into 3 different groups: process parameters, microbial-physiological parameters and parameters that are related to characteristics of the fluid.

5.1. Process Parameters

As already mentioned in earlier sections, process parameters such as pressure, temperature, number of passes and type of equipment influence the effectiveness of microbial inactivation and should be chosen well if a maximum degree of inactivation is to be achieved.

Further, the inactivation by high-pressure homogenization was generally found to increase with increasing process temperature. For example, Harrison et al. (1991) reported a 1.6-fold increase of the protein release of Alcaligenes eutrophus when the process temperature increased from 12 to 26C. Experiments done by Diels et al. (2003 and 2004) showed that an increase in process temperature resulted in a gradual increase of the inactivation of E. coli, Y. enterocolitica and S. aureus. The effect of process temperature on the inactivation was also investigated by Vachon et al. (2002). The inactivation of S. Enteritidis and L. monocytogenes increased with increasing process temperature while there was no difference observed for E. coli O157:H7. Vachon et al. (2002) explained this temperature effect by changes in the physical properties of the cell membrane due to the temperature. It is known that temperature markedly affects fluidity. Under physiological conditions lipids in biological membranes are usually in a fluid, liquid-crystalline state that provides optimum permeability and flexibility (Suutari and Laasko 1994). Vachon et al. (2002) concluded that the higher resistance to pressures at 25C than at 45 and 55C may be attributed to greater membrane flexibility at that temperature. However, the cell membrane, which affects cell membrane permeability, is not believed to be de primary site of high-pressure homogenization damage. Bacterial resistance to high-pressure homogenization is probably due to the cell wall structure, more particularly the amount of peptidoglycan. Fluid temperature, on the other hand, is inversely related to fluid viscosity, and the latter is also known to affect bacterial inactivation by high-pressure homogenization. Diels et al. (2004) demonstrated that the temperature effect on microbial inactivation could be explained by an indirect effect of fluid viscosity (see 5.3.1 below).

Process pressure and temperature are not the only process parameters that influence inactivation. Also the application of successive rounds of high-pressure homogenization can be used to increase the inactivation. Baldwin and Robinson (1994) found that the inactivation increased with increasing number of passes (N). However, they also concluded that for Candida albicans the increase in the fraction disrupted was rather small when the number of passes exceeded three. Vachon et al. (2002) investigated the effect of a cyclic high-pressure homogenization treatment on the inactivation of E. coli O157:H7, L. monocytogenes and S. Enteritidis. In milk as well as in 10 mM phosphate buffered saline the inactivation increased with increasing passes. Moroni et al. (2002) came to the same conclusion when they subjected lactococcal bacteriophages to a high-pressure homogenization treatment. Sauer et al. (1989) and Wuytack et al. (2002) also concluded that the fraction of E. coli cells disrupted was proportional to the number of passes through the disrupter. Based on all the previous findings it can be concluded that use of successive rounds of high-pressure homogenization can be a promising approach to increase the microbial efficiency of the treatment. Under the conditions used in the work of Wuytack et al. (2002), i.e. homogenization pressure up to 300 MPa and treatment temperature 25 or 45C, inactivation of the most resistant organisms (S. aureus) by a single homogenization treatment remained under one log unit, which is far insufficient for applications such as food pasteurization. The level of inactivation increased to almost 4 log units after four rounds of homogenization, and can probably be further increased by applying additional treatment rounds.

The application of higher pressures alone does not necessarily guarantee that the maximum inactivation yield will be obtained. Changes in homogenizing valve geometry can cause a significant increase in the yield at the same pressure. This emphasizes the importance of homogenizing valve design (Pandolf 1998). Several valve configurations and materials have been evaluated over the last years for their effectiveness for cell disruption. In most cases, valves having a simple flow path have been found to be the most effective, particularly the knife-edge valve (Figure 3) (Hetherington et al. 1971; Masucci 1981; Keshavarz-Moore et al. 1990). However, the efficiency of this valve type was shown to depend strongly on the sharpness of the valve edge, which decreased rapidly during extended use because microbial cell debris is apparently highly abrasive. Therefore a cell disruption valve made of wear-resistant ceramic for extended life is more and more used.

5.2. Microbial Physiological Parameters

High-pressure homogenization kills microorganisms mainly through mechanical destruction of the cell integrity. Therefore not only process parameters but also microbial parameters that affect the microbial cell strength influence microbial inactivation by high- pressure homogenization.

5.2.1. Influence of the Type of Microorganism

The most important parameter affecting cell strength is the cell wall. For this reason, this structure will be briefly discussed below for every group of microorganisms mentioned.

Bacteria. The bacterial cell wall is a unique prokaryotic structure which surrounds the cell membrane. The cell wall is very important as a cellular component, except in a few wallless bacterial species. It performs two important functions. First, structurally, the wall is necessary for maintaining the cell’s characteristic shape. Second, it prevents the cell from bursting when the difference in osmotic pressure inside and outside the cell exceeds the tensile strength of the cell membrane. Addition\ally, it forms a biologically active boundary between the bacteria and its external environment, although it is more permeable than the cell membrane and therefore does not play a major role in regulating the entry of materials into the cell. Peptidoglycan, also called murein, forms the basic structural framework of the cell wall. The basic peptidoglycan structure, similar in all bacteria, consists of linear polysaccharide chains of alternating N-acetyl-D-glucosamine (NAG) and N-acetyl-muramic acid (NAM) residues linked by β-(1-4) glycosidic bonds. The NAM residues carry a tetrapeptide of the basic structure Lalanine, D-glutamic acid, L-lysine, or di-aminopimelic acid and D-alanine. The peptide branches of parallel neighboring chains can be further cross-linked. The resulting rigid structure acts as a single macromolecular network to provide the shape and tensile strength of the cell wall. In two studies it was suggested that peptidoglycan chains are aligned perpendicular to the main axis of the bacterium (Verwer et al. 1978 and 1980). De Pedro et al. (1997) have produced electron micrographs of muramidase-treated sacculi. They found that the holes produced in the sacculi were usually oblong and the long axes were invariably oriented in the circumferential direction. Koch (1998) however, found that the information about the orientation of the glycan chains can not be drawn from electron micrograph pictures because the observed pattern may have to do with the fact that the sacculi are affixed to the substrate before they are enzymatically treated. Therefore they concluded that it is not excluded that the glycan chains might be parallel to each other in either orientation, or have an intermediary orientation. Early models proposed peptidoglycan of E. coli to be a single monolayer. Subsequent evidence has shown that it is actually multilayered (Glauner et al. 1988; Labischinski et al. 1991; Kock 1998; Vollmer and Holtje 2004) and that its thickness increases from 6.6 1.5 to 8.8 1.8 nm during the transition from exponential to stationary phase (Leduc et al. 1989; Yao et al. 1999; Matias et al. 2003). These values correspond to 2-3 and 4-5 layers of peptidoglycan, respectively. Wall strength depends on the thickness of the peptidoglycan layer and the degree of crosslinkage between adjacent polysaccharide chains (Middelberg et al. 1992; Middelberg and O’Neill 1993), and the cell wall of gramnegative bacteria differs a lot from that of gram-positive bacteria in these aspects. Gram-positive bacteria have a thicker peptidoglycan (about 40 layers) than gram-negative bacteria (1 up to 5 layers) (Madigan et al. 2000), and this contributes to the greater structural resistance to mechanical breakage in the former (Harrison et al. 1991 ; Lengler et al. 1999; Madigan et al. 2000). Gram-negative bacteria however, have another structural layer around the peptidoglycan namely the outer membrane. This is a bilayer with an inner shell containing mainly phospholipids and an outer shell containing mainly lipopolysaccharides. The outer membrane also contains proteins and is attached to the peptidoglycan by lipoproteins. It separates the peptidoglycan layer from the bulk medium environment, thereby preventing their interaction. Divalent cations (Ca^sup 2+^, Mg^sup 2+^) play an essential role in its stabilization (Engler 1985; Madigan et al. 2000). The outer membrane of E. coli can maintain the cell shape under certain circumstances (Henning 1975). In conclusion, three layers of the cell envelope require consideration in the rupture of gram-negative bacterial cells. The outer membrane protects the inner layers form direct chemical or enzymatic attack. The peptidoglycan layer provides the mechanical strength of the cell. The cytoplasmic membrane may be considered the biochemical boundary of the cell and the major player in permeability. In gram-positive bacteria, the peptidoglycan layer is not protected from the external environment by an outer membrane but provides a much greater structural strength.

FIG. 3. Homogenizing valve configuration evaluated for disruption of microbial cells (Engler, 1990). A: standard valve, B: cell rupture valve, C: grooved valve, D: knife edge valve, E: conical valve, F: ball cell disruption valve.

Kelemen and Sharpe (1979) investigated the inactivation of Escherichia coli, Streptococcus faecalis, Bacillus subtilis, Lactobacillus casei, and Staphylococcus aureus (i.e. one gram negative and four gram positive bacteria of which the latter has a coccoid shape). They concluded that the gram-negative bacterium was more sensitive to high-pressure homogenization than the gram- positive bacteria. Vachon et al. (2002) working with L monocytogenes (gram-positive rod) and E. coli O157:H7 and S. Enteritidis (two gram- negative rods) made the same observation. Wuytack et al. (2002) investigated the resistance of five grampositive and six gram- negative bacteria to high-pressure homogenization. Without any exception, all the gram-positive bacteria were more resistant to high-pressure homogenization than the gram-negative bacteria. All these experiments provide convincing evidence to generalize the statement that gram-positive bacteria are more resistant to high- pressure homogenization than gram-negative bacteria. Kelemen and Sharpe (1979) concluded further that composition of the cell wall and more particularly peptidoglycan, determines the resistance to high-pressure homogenization. In continuation of these findings, Middelberg and O’ Neill (1993) developed their model for disruption off. coli by high-pressure homogenization in which one of the parameters, the mean effective strength, is correlated with the peptidoglycan cross-linkage, as discussed above (section 4 relations 10-11). By using transmission electron microscopy Vachon et al. (2002) observed damaged L. monocytogenes after the homogenization treatments. Cells surviving the treatment were intact while the injured cells had partially or completely lost their content due to mechanical damage. The same type of cell damage was also reported by Kheadr et al. (2002) for L. innocua. Based on these findings Vachon et al. (2002) concluded that high-pressure homogenization acts on bacterial cell walls. They further suggested that the higher resistance of L. monocytogenes in comparison with E. coll and S. Enteritidis was probably due to a larger number of peptidoglycan layers. The tight correspondence between peptidoglycan structure and high-pressure homogenization resistance suggests that high-pressure homogenization kills vegetative bacteria mainly through mechanical destruction of the cell integrity, caused by the spatial pressure and velocity gradients, turbulence (Doulah et al. 1975), impingement (Engler and Robinson 1981; Keshavarz-Moore et al. 1990) and/or cavitation (Save et al. 1994; Shirgaonkar et al. 1998), that occur in a liquid during high-pressure homogenization. Kelemen and Sharpe (1979) also postulated cell shape to be an additional factor determining bacterial resistance to high-pressure homogenization, rods predicted to be more easily disrupted than cocci. Results obtained by Wuytack et al. (2002) do not confirm this postulate. The coccus L. dextranicum was much more sensitive to highpressure homogenization than the rod L. plantarum. Of course, an experimental investigation of the effect of cell shape is complicated by the fact that rods and cocci from different bacterial species may not only differ in cell shape but also in amount or structure of peptidoglycan. Wuytack et al. (2002) further compared the resistance to high-pressure homogenization and to high hydrostatic pressure of a set of gram-positive and gram-negative bacteria in a similar pressure range (up to 300 MPa). It was found that, within each group (gram-positive or gram-negative bacteria) there were large differences in resistance to high hydrostatic pressure, but not to high-pressure homogenization. second, it was possible to completely discern gram-positive and gramnegative bacteria on the basis of their resistance to high-pressure homogenization, while based on resistance to high hydrostatic pressure both groups overlapped. Third, within the group of gram-negative bacteria there also existed another order in resistance to high-pressure homogenization than to high hydrostatic pressure. Based on these observations they suggested that the inactivation mechanisms of high-pressure homogenization differ from those of high hydrostatic pressure, at least over the pressure ranges studied. Presumably the high pressure developed during high-pressure homogenization is not to any major extent responsible for inactivation of microorganisms because the bacteria are only exposed to this high pressure for a very short time in the order of a second or less. For comparison, exposure time to high hydrostatic pressure treatment in this work was 15 min.

Yeasts. The basic structural components of the yeast cell wall are glucans, mannans, and proteins. The overall structure is thicker than in gram-positive bacteria, but the cells are also larger (Geciova et al. 2002). Yeast glucans are moderately branched molecules composed of glucose residues, primarily in β-(1-3) and β-(1-6) linkages. Yeast mannans are characterized by a backbone of mannose residues in a-(1-6) linkage having short oligosaccharide side chains composed of mannose units. Many of the proteins found in yeast cell wall are enzymes rather than structural components (Engler 1985). Glucan fibrils constitute the innermost part of the cell wall and give the cell its shape. These fibrils are covered by a layer of glycoproteins, beyond which there is a mesh of mannans covalently linked by 1,6-phosphodiester bonds.

The most investigated yeast cell walls are from Sacchammyces cerevisiae and Candida utilis, Brookman (1974) achieved 100% disintegration of S. cerevisiae (baker’s yeast) at 170 MPa with \one single pass through the homogenizer. It was shown by Engler and Robinson (1981) that C. utilis is more resistant to disruption by high-pressure homogenization than the more commonly studied 5. cerevisiae. Because of their larger size and different cell wall structure, disruption of yeasts is believed by most researchers to be easier than bacteria (Geciova et al. 2002). However, this is in contradiction with the findings of Kelemen and Sharp (1979), who found that 5. cerevisiae had the same resistance to high-pressure homogenization than the gram-positive coccus S. faecalis.

Bacterial spores. Certain species of bacteria produce endospores under unfavorable conditions. These are dormant differentiated cells that are very resistant to heat and other physical and chemical stresses. Structurally, an endospore consists of a core, surrounded by a cortex of peptidoglycan, a spore coat of protein and in some species a delicately thin layer called the exosporium. The core of a mature endospore differs greatly form the cytoplasm of the vegetative cell from which it is derived. It is rich in the spore- specific component calcium dipicolinate, contains only 10-30% of the water content of the vegetative cell, and thus the consistency of the core cytoplasm is that of a gel.

Spores from gram-positive Bacillus spp. and Clostridia spp. are greatly resistant to any treatment, including homogenization (Popper and Knorr 1990). Feijoo et al. (1997) investigated the inactivation by high-pressure homogenization of B. licheniformis spores in an ice cream mix. At an inlet temperature of 33C and process pressure ranging from 50 to 200 MPa, only an inactivation between 0 and 0.5 log-units was reached. Even this low level of inactivation may not represent direct disruption of dormant spores. The unavoidable increase in temperature of the treated ice cream mix due to shear and cavitation effects during the passage through the homogenizing valve may have stimulated germination of the spores in the ice cream mix (Church and Holverson 1956), rendering them sensitive to heat and mechanical forces.

Viruses. Little is known about the effect of high-pressure homogenization on viruses. Jean et al. (2001) evaluated the effectiveness of a high-pressure homogenization treatment for the inactivation of hepatitis A virus. They found that the hepatitis A virus was relatively resistant since 300 MPa and five passes through the homogenizing valve were needed to reach more than 1 log-unit inactivation.

Moroni et al. (2002) subjected bacteriophages-bacterial viruses- to a high-pressure homogenization treatment. They found, based on electron microscopy, that the inactivation of lactococcal bacteriophages c2, ski, and u136 probably resulted from breaking of the phage heads, from which genetic material was lost. They also observed that some phages lost their tail or a part of it, which would render them incapable of attachment. These physical effects on phage integrity do not appear to be specific to any part of the phage. Solomon et al. (1966) observed by electron microscopy that a high hydrostatic pressure treatment at 420 MPa produced empty heads in T4 bacteriophages. They suggested that this loss of genetic material was due to tail sheath contraction. Moroni et al. (2002) did not observe tail contraction but did observe non-specific damage explainable in term of mechanical phenomena including cavitation, shearing, turbulence and impact on the stationary surface in the device (Lanciotti et al. 1994).

5.2.2. Influence of Cell Concentration

Most studies report that cell concentration has no discernible influence on cell disruption efficiency over a wide range of cell concentrations and operating pressures (Hetherington et al. 1971; Agerkvist and Enfors 1990; Harrison et al. 1991; Moroni et al. 2002). A few studies however contradict this (Sauer et al. 1989; Middelberg et al. 1991; Kleinig et al. 1995). Vachon et al. (2002) investigated the influence of the initial bacterial load on the inactivation by high-pressure homogenization at 200 MPa and 250C. Firstly they inoculated 10 mM phosphate buffered saline with L. monocytogenes, E. coli O157:H7 and 5. Enteritidis. The cell concentration varied from 104 to 109 CFU/ml. The highest degree of inactivation was obtained with the lowest initial load. In a second step, they also investigated the effect of cell concentration in milk using the same microorganisms. Contrary to the buffer results, the initial bacterial concentration of the milk samples had no impact on the effectiveness of high-pressure homogenization. Agerkvist and Enfors (1990), Harrison et al. (1991) and Kleinig et al. (1995) explained the decrease in disruption with increased cell concentration by the increased viscosity of the homogenate. However, experiments done by Diels et al. (2005a) with E. coli demonstrated that the viscosity of cell suspensions containing 10^sup 5^-10^sup 8^ cfu/ml was not notably different, and neither was the efficiency of high-pressure homogenization inactivation on these suspensions.

5.2.3. Influence of Growth Phase of Micro-Organisms

Another important parameter that influences the microbial inactivation by high-pressure homogenization is the growth phase or growth rate of cells. The growth rate of exponentially growing cells is higher than that of stationary phase cells, because growth in stationary phase is limited due to the exhaustion of essential nutrients in the growth medium and due to the accumulation of toxic waste products produced by the cells. Experiments with Alcaligenes eutrophus done by Harrison et al. (1991) showed that 2 to 3 passes at pressures exceeding 60 MPa were required for maximum disruption of stationary phase cells, while actively growing A. eutrophus cells were disrupted in a single pass. They concluded that the increased sensitivity of the exponentially growing cells to mechanical cell disruption resulted from their rapid growth. In order to extend the cell wall during cellular growth, it is locally cleaved by autolysins, and this results in weakened areas. Engler (1990) reached the same conclusion based on his experiments with baker’s yeast. Considerable other changes occur in peptidoglycan structure during the transition from exponential phase to stationary phase (Pisabarron et al. 1985). In particular, the degree of crosslinkage increases significantly (Middelberg et al. 1992). The formation of a novel cross bridge between two adjacent meso-di-aminopimelic acid residues (A^sub 2^pm-A^sub 2^pm) in the peptidoglycan can also be responsible for this increase, as shown in E. coli (Glauner et al. 1983 and 1988). The amount of A2pm-A2pm cross-links increased drastically under growth conditions where the cell was unable to synthesize the normal amount of alanine-A2pm acid crosslinks (Glauner et al. 1988; Seltmann and Hoist 2002). Glauner et al. (1988) concluded that the ability to form peptide bridges via the A2pm-A2pm can be considered as a backup system which guarantees an extent of cross-linkage sufficient to ensure the mechanical integrity of the sacculus.

5.3. Characteristics of the Fluid

5.3.1. Viscosity

Fluid viscosity has an effect on some of the proposed mechanisms of cell disruption by high-pressure homogenization, which include turbulence (Doulah et al. 1975), cavitation (Save et al. 1994), impact with solid surfaces (Engler and Robinson 1981; Keshavarz- Moore et al. 1990) and extensional stress (Shamlou et al. 1995) (see section 1.3 and 1.4). Kleinig et al. (1995) proposed that the increased levels of high-pressure homogenization inactivation upon dilution of E. coli cell suspensions could be explained by the lower viscosity of diluted cell suspensions as observed by Harrison et al. (1991). They did not measure the viscosity in their own study however. As discussed in section 4.2.2 and opposed to Harrison et al. (1991), Diels et al. (2005a) found that the viscosity of E. coli cell suspensions containing 105-108 cfu/ml was not different and that the bactericidal efficiency of high-pressure homogenization was independent of the cell concentration. Diels et al. (2005a) further demonstrated that bacterial inactivation by high-pressure homogenization was inverselvyrelated to the initial fluid viscosity.

While fluid viscosity can influence microbial inactivation by high-pressure homogenization, the homogenization treatment itself can also change fluid viscosity. Some studies have shown degradation of high molecular weight compounds subjected to strong shearing forces in a viscous flow (Morawetz 1983). Floury et al. (2002) investigated the degradation of methylcellulose, a polymer that is extensively used in food, cosmetic, and pharmaceutical preparations, during high-pressure homogenization. They found that the very strong flow at the entrance of the homogenizing valve, and the resulting frictional forces encountered by the fluids during high-pressure homogenization induced irreversible degradation of long molecules. They also postulated that the breakdown of the covalent bonds trough the homogenizing valve is a random phenomenon. Experiments have shown that repeatedly passing the solution though the homogenizer continues to reduce the viscosity until a limiting value is achieved (Pandolf, 1998). It also appears that the end viscosity reached is proportional to the product of the homogenization pressure and the number of passes. Two passes at a certain pressure (P^sub 1^) is equivalent to one pass at two times this pressure (P^sub 2^ = 2P^sub 1^) (Pandolf 1998).

5.3.2. Buffer versus Real Food

The fluid in which microorganisms are suspended can also have an influence on microbial inactivation. This is well documented for inactivation processes such as heat treatment and high hydrostatic pressure treatment (Juneja and Elben 2000; Juneja et al. 2001 ; Erkmen and Dogan 2004; Dogan and Erkmen 2004; Van Opstal et al. 2005). Because high-pressure homogeniza\tion is already used in the preparation and/or stabilization of emulsions and suspensions it is important to know the differences in microbial inactivation in different media.

Vachon et al. (2002) studied the inactivation of L. monocytogenes and E. coli O157:H7 in 10 mM phosphate buffered saline and milk. After three successive high-pressure homogenization treatments at 200 MPa, the inactivation of E. coli was 7.2 log-units in buffer, compared to only 5.3 log-units in milk. For L. monocytogenes the achieved reductions were respectively 3.6 and 1.7 log-units. Jean et al. (2001) evaluated the inactivation of hepatitis A virus by high- pressure homogenization suspended in wastewater, milk and apple juice. At 300 MPa and after five passes through the homogenizing valve the inactivation level in milk was only 84.4%, while in wastewater and apple juice it was respectively 95 and 99.97%. The high sensitivity of hepatitis A virus in apple juice was, according to the authors, due to an enhancement of viral susceptibility by low pH. In a subsequent experiment, they investigated the inactivation of hepatitis A virus in milk samples with different fat content. Using 200 MPa and five passes, no viral inactivation was observed in milk containing 2 and 3.25% fat, while 4.3% of the added hepatitis A viruses were inactivated in milk containing 0 or 1 % fat. Kheadr et al. (2002) also detected less inactivation of L. innocua in full fat milk compared to skim milk. All these authors ascribe the high resistance in milk to the supposed protective effect of milk fat, by analogy to the well-established protective effect of fat on inactivation by heat (MacDonald and Sutherland 1993) and hydrostatic pressure (Garcia-Graells et al. 1999). However, the results obtained by Diels et al. (2005a) plead against a specific protective role of fat. They demonstrated that the inactivation of E. coli MG1655 in three dairy drinks of strongly different viscosity and composition, corresponded to the inactivation achieved in polyethylene glycol containing 10 mM potassium phosphate buffers of the same viscosity, suggesting that product composition is not a major parameter affecting high-pressure homogenization inactivation. The difference in bacterial inactivation between buffer and milks with different fat contents can be explained based only on their differences in viscosity.

5.3.3. Additives

The fact that high-pressure homogenization does not cause metabolic injury, as described by Wuytack et al. (2002 and 2003), may limit the applications of this technique in hurdle technology. If the cells surviving high-pressure homogenizing treatment are not injured, it can indeed be anticipated that they will not be sensitized to other treatments or to unfavorable conditions such as low pH, high NaCl concentration, or antimicrobials. On the other hand, because it is suggested that the cell wall is the main vital target, lytic enzymes, or other treatments that weaken the cell wall, may increase microbial sensitivity to high-pressure homogenization (Baldwin and Robinson 1990; Harrison et al. 1991; Vogels and Kula 1992; Middelberg 2000; Wuytack et al. 2002). Treatment of B. cereus with 0.5 mg/g cellosyl (i.e., an extracellular lytic enzyme produced by Streptomyces coelicolor) before homogenization resulted in 98% cell disrupture after a single pass at 70 MPa, while less than 40% of the not enzymatically treated cells were broken under the same experimental conditions (Vogels and Kula 1992). The addition of SDS, EDTA, lysozyme, and monovalent cations before homogenization was reported by Harrison et al. (1991) to increase the sensitivity of Alcaligenes eutrophus. Popper and Knorr (1990) investigated the effect of adding lysozyme (0-3 g/ml) and chitosan (0-100 g/ml) to cell suspensions of Streptococcus lactis, B. subtilis, and E. coli before homogenization to weaken the cell wall. No synergistic effect of chitosan/homogenization and lysozyme/homogenization were observed. The authors assumed that under the used experimental conditions the integrity of cells surviving the pre-treatment was only little affected and that a synergistic effect could possibly only be observed at higher pressures than those applied in their study (i.e.,

Yeasts, such as S. cerevisiae and Candida utilis, may be effectively weakened by pre-treatment with zymolyase (a lytic enzyme isolated from Oerskovia xanthineolytica) (Engler and Robinson 1981; Baldwin and Robinson 1994). The disruption of Candida utilis obtained using a zymolyase pre-treatment approached 95% with four passes at a pressure of 95 MPa, as compared with only 65% disruption without pre-treatment. In general, however, the cost of these enzyme preparations can be quite high and recovery and recycle is difficult and costly to implement. Significantly enhanced disruption is therefore required to justify this cost (Baldwin and Robinson 1994; Middelberg 2000). Often, a simpler and more practical strategy is simply to increase the number of homogenizer passes (Middelberg 2000).

A relevant question in this context is if the homogenization treatment on itself has an effect on the activity of antimicrobial enzymes or peptides used. Zapico et al. (1999) investigated the effect of homogenization of milk on nisin activity and reported losses up to 64% in whole milk and 62 % in fat-in-water emulsion. The homogenization of skim milk containing 100 IU/ml nisin resulted in a loss of activity of 20%. The loss of nisin activity produced by the homogenization could, according to Walstra and Jenness (1984), be due to increased adsorption to the milk fat globules, whose exposed surface will be increased by homogenization as a consequence of disruption. Vannini et al. (2004) reported that the addition of lysozyme and the lactoperoxidase system before homogenization treatment enhanced the instantaneous pressure efficacy on almost all the tested bacteria. For example, L. monocytogenes was only slightly sensitive to high-pressure homogenization, but the addition of lysozyme (3.3 mg/100 ml) or the lactoperoxidase system (lactoperoxidase: 30 g/ml; H^sub 2^O^sub 2^: 26.5 mM and KSCN: 19.0 mM) before high-pressure homogenization induced a significant viability loss within three hours after treatment and an extension of the lag phase of the survivors in skim milk during incubation at 37C. In the same study, the inactivation of E. coli by high- pressure homogenization was also claimed to be synergistically enhanced in the presence of the lactoperoxidase enzyme system. However, the dosage of the lactoperoxidase system used in this work was so high that it caused rapid inactivation of E. coli and L. monocytogenes even without high-pressure homogenization treatment, and under such conditions it is difficult to observe a synergistic action. According to the authors, the synergistic effects of high- pressure homogenization and the antimicrobial enzymes can be attributed to three factors: (1) a direct effect of the pressure on the integrity of walls or outer membranes of the microorganisms; (2) a subsequent increased penetration of the enzymes through the damaged walls and membranes; and (3) an indirect stimulatory effect of the process on the enzymes, caused by small structural changes affecting the active sites. The last hypothesis was supported by the finding that the antimicrobial activity of a lysozyme solution on L. plantarum increased after homogenization. However, this is in contradiction with the findings of Diels et al. (2005b), who found that high-pressure homogenization up till 300 MPa did not affect the antimicrobial activity of nisin or lysozyme.

6. CONCLUSION

Although high-pressure homogenization has been known for long to disrupt microbial cells, the interest to use this process as a unit operation to reduce the microbial burden of foods or pharmaceutical, cosmetic or other biodeteriorable products has only emerged recently. The finding by several researchers that gram-negative bacteria are invariably more sensitive to high-pressure homogenization than gram-positive bacteria, has led to the widespread view that high-pressure homogenization kills bacteri\a by an “all or nothing” physical disruption of the cell wall. This view is further supported by the apparent inability of high-pressure homogenization to induce sublethal injury in bacteria. High- pressure homogenization differs substantially from high hydrostatic pressure treatment in this respect: while the former acts by a single hit mechanism based on cell disruption, the latter is believed to kill by a multiple hit mechanism involving protein denaturation and membrane damage. Insights in the mechanisms of bacterial inactivation by these processes can support the intelligent design of combined treatments which act synergistically on bacterial inactivation, by combining a process with another process or with antimicrobial components with a specific mode of action. Several studies demonstrated that bacterial inactivation by high-pressure homogenization can be accurately predicted by mathematical models. The limited number of process and product parameters that apparently determines bacterial inactivation by high- pressure homogenization is certainly an advantage for the development of industrial applications for this process.

REFERENCES

Agerkvist, I., and Enfors, S. 1990. Characterization of E. coli cell disintegrates from a bead mill and high-pressure homogenizers. Biotechnol. Bioeng. 36:1083-1089.

Baldwin, C.V., and Robinson, C.W. 1990. Disruption of Sacchammyces cerevisiae using enzymatic lysis combined with high- pressure homogenization. Biotechnol. Technol. 4:329-334.

Baldwin, C.V., and Robinson, C.W. 1994. Enhanced disruption of Candida utilis using enzymatic pretreatment and high-pressure homogenization. Biotechnol. Bioeng. 43:46-56.

Beltaos. S., and Rajaratnam. N. 1974. Impinging circular turbulent jets. ASCE J. Hydraulic Eng. 100:1313-1328.

Botha, C.J.A. 1993. A study of hydrodynamic cavitation as a method of water treatment, M.Sc. Thesis, University of Natal, Durban, September 1993.

Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.

Brookman. J.S.G. 1974. Mechanism of cell disintegration in a high- pressure homogenizer. Biotechnol. Bioeng. 16:371-383.

Brookman, J.S.G. 1975. Further studies on the mechanism of cell disruption by extreme pressure extrusion. Biotechnol. Bioeng. 17:465- 479.

Brucato, A., and Micale, G. 1998. Numerical prediction of flow fields in baffled stirred vessels:comparison of alternative modeling approaches. Chem. Eng. Sci. 53:3653-3684.

Burgaud, I., Dickinson, E., and Nelson, RV. 1990. An improved high-pressure homogenizer for making fine emulsions on a small scale. Int. J. Food Sd. Technol. 25:39-46.

Church, B.D., and Holverson, H.O. 1956. Effect of heating and aging on germination and glucose oxidation of spores of aerobic bacilli. In J.L. Roberts (ed.), Bacteriological Proc.: Pmc. 56th General Mtg. Soc. Am. Bacterol, p. 45.

De Pedro, M.A., Quintela, J.C., Hflltje, J.V., and Schwarz, H. 1997. Murein segregation in Escherichia coli. J. Bacterial. 179:2823- 2834.

Deshimaru, J. 1994. A study of cavitation in oils: Effect of base oils and polymers. Jpn. J. Tribohgy 36:531-542.

Dickinson, E., and Stainby, G. 1988. Emulsion Stability. In E. Dickinson and G. Stainby, (eds.), Advances in Food Emulsions and Foams, Elsevier Applied Sciences, London, Chapter 1.

Diels, A.M.J., Wuytack, E.Y., and Michiels, C.W. 2003. Modeling inactivation of Staphylococcus aureus and Yersinia enterocolitica by high-pressure homogenization at different temperatures. Int. J. Food Microbiol. 87:55-62.

Diels, A.M.J.. Callewaert, L., Wuytack E.Y., Masschalck, B., and Michiels, C.W. 2004. Moderate temperatures affect Escherichia coli inactivation by high-pressure homogenization only through fluid viscosity. Biotechnol. Prog. 20:1512-1517.

Diels, A.M.J., Callewaert, L., Wuytack, E.Y, Masschalck, B., and Michiels, C.W. 2005a. Inactivation off. coli by high-pressure homogenization is influenced by fluid viscosity but not by water activity and product composition. Int. J. Food Microbiol. 101:281- 291.

Diels, A.M.J., De Taye, J., and Michiels, C.W. 2005b. Sensitization of Escherichia coli to antibacterial peptides and enzymes by high-pressure homogenization. Int. J. Food Microbiol. 105:165-175.

Dogan, C., and Erkmen, A. 2004. High-pressure inactivation kinetics of Listeria monocytogenes inactivation in broth, milk, and peach and orange juices. J. Food Eng. 62:47-52.

Doulah, M.S., Hammond, T.H., and Brookman, J.S.G. 1975. A hydrodynamic mechanism for the disintegration of Sacchammyces cerevisiae in an industrial homogenizer. Biotechnol. Bioeng. 17:845- 858.

Engler, C.R., and Robinson, C.W. 1981. Disruption of Candida utilis cells in high pressure flow devices. Biotechnol. Bioeng. 23:765-780.

Engler, C.R. 1985. Disruption of microbial cells. In Comprehensive Biotechnology, ed. M. Moo-Young, 305-324, Pergamon Press, Oxford.

Engler, C.R. 1990. Cell disruption by homogenizer. In J.A. Asenjo (ed.). Separation processes in biotechnology. Marcel Dekker, New York, 95-105.

Erkmen, O., and Dogan, C. 2004. Kinetic analysis of Escherichia coli inactivation by high hydrostatic pressure in broth and foods. Food Micmbiol. 21:181-185.

Feijoo, S.C., Hayes, W.W., Watson, C.E., and Martin, J.H. 1997. Effect of microfluidizer technology on Bacillus licheniformis spores in ice cream mix. J. Dairy Sci. 80:2184-2187.

Floury, J., Desrumaux, A., Axelos, M.A.V., and Legrand, J. 2002. Degradation of methylcellulose during ultra-high pressure homogenization. Food Hydrocolloid. 16:47-53.

Floury, J., Bellettre, J., Legrand, J., and Desrumaux. A. 2004. Analysis of a new type of high pressure homogenizer. A study of the flow pattern. Chem. Eng. Sci. 59:843-853.

Garcia-Graells, C., Masschalck, B., and Michiels, C.W. 1999. Inactivation of Escherichia coli in milk by high-hydrostatic pressure treatment in combination with antimicrobial peptides. J. Food Prot. 62:1248-1254.

Geciova, J., Bury, D., and Jelen, P. 2002. Methods for disruption of microbial cells for potential use in the dairy industry-a review. Int. Dairy J. 12:541-553.

Glauner, B., and Schwarz, U. 1983. The analysis of murein composition with high-pressure liquid chromatography. In K. Hackenback, J.-V. Holtje, and H. Labischinski (eds.), The Target of Penicillin, W. de Gruyter & Co., Berlin, pp. 29-14. Berlin.

Glauner, B., Holtje, J.-V., and Schwarz, U. 1988. The composition of the murein of Escherichia coli. J. Biol. Chem. 263:10088-10095.

Gray, P.P., Dunnill, P., and Lilly, M.D. 1972. The continuous- flow isolation of enzymes. In G. Terui (ed.), Proc. IV IS: Fermentation Technology Today, Society of Fermentation Technology, Kyoto, pp. 347-352.

Guerzoni, M.E., Vannini, L., Chaves Lopez, C., Lanciotti, R., Suzzi, G., and Gianotti, A. 1999. Effect of high-pressure homogenization on microbial and chemico-physical characteristics of goat cheeses. J. Dairy Sci. 82:851-862.

Harrison, S.T.L., Chase, H.A., and Dennis, J.S. 1991. The disruption of Alcaligenes eutrophus by high pressure homogenization:key factors involved in the process. Bioseparalion 2:155-166.

Harrison, S.T.L., and Pandit, A.B. 1992. The disruption of microbial cells by hydrodynamic cavitation. 9lhlnternational Biotechnology Symposium, Washington, DC.

Hauben, K.J.A., Wuytack, E.Y., Soontjens, C.C.F., and Michiels, C.W. 1996. High-pressure transient sensitization of Escherichia coli to lysozyme and nisin by disruption of outer-membrane permeability. J. Food Prot. 59:945-950.

Hayashi, R. 1989. Application of high pressure to engineering food processing and preservation. Philosophy and development. In W.E. Spiess and M. Schulet (eds.), Engineering of Food, Elsevier Applied Science, London, pp. 815-826.

Henning, U. 1975. Determination of Cell Shape in Bacteria. Annu. Rev. Micmbiol. 29:45-60.

Hetherington, P.J., Follows, M., Dunnill P., and Lilly, M.D. 1971. Release of protein from bakers’ yeast (Saccharomyces cerevisiae) by disruption in an industrial homogenizer. Chemical engineering Research and Design: Trans. I .ChemE. 49:142-148.

Hoover, D.G., Metrick, C., Pepinean, A.M., Forkas, D.F., and Knorr, D. 1989. Biological effects of high hydrostatic pressure on food microorganisms. Food Technol. 3:99-107.

Hull, A.S., and Middelberg, A.P.J. 1993. Evidence that septated E. coli disrupt preferentially during high-pressure homogenization. Trans. I. ChemE. 71:264-266.

Jean, J., Biais, B., Darveau, A., and Fliss, 1.2001. Detection of Hepatitis A Virus by the Nucleic Acid Sequence-Based Amplification Technique and Comparison with Reverse Transcription-PCR. Appl. Environ. Micmbiol. 67:5593-5600.

Juneja, V.K., and Eblen, B.S. 2000. Heat inactivation of Salmonella Typhimurium DT104 in beef as affected by fat content. Lett. Appl. Microbiol. 30:461-467.

Juneja, V.K., Eblen, B.S., and Ransom, G. M. 2001. Thermal Inactivation of Salmonella spp. in Chicken Broth, Beef, Pork, Turkey and Chicken: Determination of D- and Z-values. J. Food Sci. 66:146- 152.

Kelemen, M.V., and Sharpe, J.E.E. 1979. Controlled cell disruption: a comparison of the forces required to disrupt different microorganisms. J. Cell Sci. 35:431-141.

Keller, A.P. 2001. Cavitation Scale effects: empirically found relations and the correlation of cavitation number and hydrodynamic coefficients. CAV2001: lecture.001: Fourth International Symposium on Cavitation, 1-18.

Keshavarz-Moore, E., Hoare, M., and Dunnill, P. 1990. Disruption of baker’s yeast in a high-pressure homogenizer: New evidence on mechanism. Enzyme Microbial Technol. 12:764-770.

Kheadr, E.E, Vachon, J.F., Paquin, P., and Fliss, I. 2002. Effect of dynamic high pressure on microbiological, Theological and microstructural quality of Cheddar cheese. Int. Dairy J. 12:435- 446.

Kleinig, A.R., Mansell, C.J., Nguyen, Q.D., Badalyan, A., and Middelberg, A.P.J. 1995. Influence of broth dilution of the disruption of Escherichia coll. Biolechnol. Technol. 9:759-762.

Kleinig,A.R., O’Neill, B.K., and Middelberg, A.P.J. 1996. The effect of homogenizer impact distance on the disruption of Escherichia coli. Biotechnol. Tech. 10:199-204.

Kleinig, A.R., and Middelberg, A.P.J. 1996. The correlation of cell disruption with homogenizer valve pressure gradient determined by computational fluid dynamics. Chem. Eng. Sci. 51:5103-5110.

Kleinig, A.R., and Middelberg, A.P.J. 1997. Numerical and experimental study of a homogenizer impinging jet. Am. I. Chem. Eng. J. 43:1100-1107.

Koch, A.L. 1998. Orientation of the peptidoglycan chains in the sacculus of Escherichia coli. Res. Microbiol. 149:689-701.

Kumar, P.S., and Pandit, A.B. 1999. Modeling hydrodynamic cavitation. Chem. Eng. Technol. 22:1017-1027.

Labischinski, H., Goodell, E.W., Goodell, A., and Hochberg, M.L. 1991. Direct proof of a “more-than-single-layered” peptidoglycan architecture of Escherichia coli W7, a neutron small-angle scattering study. J. Bacterial. 173:751-756.

Lanciotti, R., Sinigaglia, M., Angelini, P., and Guerzoni, M.E. 1994. Effects of homogenization pressure on the survival and growth of some food spoilage and pathogenic microorganisms. Lett. Appl. Microbiol. 18:319-322.

Lanciotti, R., Gardini, F., Sinigaglia, M., and Guerzoni, M.E. 1996. Effects of growth conditions on the resistance of some pathogenic and spoilage species to high pressure homogenization. Lett. Appl. Microbiol. 22:165-168.

Lander, R., Manger, W., Scouloudis, M., Ku, A., Davis, C., and Lee, A. 2000. Gaulin Homogenization: a mechanistic study. Biotechnol. Prog. 16:80-85.

Leduc, M., Frehel, C., Siegel, E., and van Heijenoort, J. 1989. Multilayered distribution of peptidoglycan in the periplasmic space of Escherichia coli. J. Gen. Microbiol. 135:1243-1254.

Lengeler, W.D., Drews, G., and Schlegel, H.G. 1999. Biology of the Prokaryotes. Georg Thieme, Stuttgart.

Lowry, O.H., Rosenbrough, NJ., Fair, A.L., and Randall, R.J. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275.

MacDonald, F., and Sutherland. A.D. 1993. Effect of heat treatment on Listeria monocytogenes and gram-negative bacteria in sheep, cow and goat milks. J. Appl. Bacterial. 75:336-343.

Madigan, M.T., Martinko, J.M., and Parker, J. 2000. Brock biology of microorganisms. Prentice-Hall Inc., Upper Saddle River, NJ.

Masucci, S.F. 1985. Improving the efficiency of high-pressure homogenizers for cell disruption applications, B.S. thesis. Massachusetts Institute of Technology.

Matias, V. R. F., Al-Amoudi, A., Dubochet, J., and Beveridge, T. J. 2003. Cryo-Transmission Electron Microscopy of Frozen-Hydrated Sections of Escherichia coli and Pseudomonas aeruginosa. J. Bacterial. 185:6112-6118.

Middelberg, A.P.J., O’Neill, B.K., and Bogle, I.D.L. 1991. A novel technique for the measurement of disruption in high-pressure homogenization: Study on E. CO/I containing recombinant inclusion bodies. Biotechnol. Bioeng. 38:363-370.

Middelberg, A.P., O’Neill, B.K., and Bogle, I.D.L. 1992a. A new model for the disruption of Escherichia coli by high-pressure homogenization Part 1. Model development and verification. Trans. I. ChemE. Pan C, 70:205-212.

Middelberg, A.P., O’Neill, B.K., Bogle, I.D.L., Gully, N.J. Rogers, and A.H., Thomas, C.J. 1992b. A new model for the disruption of Escherichia coli by high-pressure homogenization Part II. A correlation for the effective cell strength. Trans. I. ChemE. Part C, 70:213-218.

Middelberg, A.P.J. 1993. Extension of the wall-strength model for high-pressure homogenization to multiple passes. Trans. 1. ChemE. Part C, 71:215-219.

Middelberg, A.P.J., and O’Neill, B.K. 1993. A correlation for the effective strength of Escherichia coli during homogenization. Biotechnol. Prog. 9:109-112.

Middelberg, A.P.J., and O’Neill, B.K. 1994. A simplified model for the disruption of Escherichia coli: The effect of cell septation. Biotechnol. Prog. 10:109-113.

Middelberg, A.P.J. 1995. Process-scale disruption of microorganisms. Biotechnol. Adv. 13:491-551.

Middelberg, A.P.J. 2000. Microbial cell disruption by high- pressure homogenization. In Methods in Biotechnology 9: Downstream processing of proteins: Methods and protocols, ed. M.A. Desai, 11- 21. Humana Press Inc., Totowa, NJ.

Miller, J., Rogowski, M., and Kelly, W. 2002. Using a CFD model to understand the fluid dynamics promoting E. coli breakage in a high-pressure homogenizer. Biotechnol. Prog. 18:1060-1067.

Morawetz, H. 1983. Macromolecules in solution, 2nd ed. Krieger Publishing. Reprint of 1975 John Wiley edition; High Polymers Series, Vol. 21 S. Near Fine.

Moroni, O., Jean, J., Autret, J., and Fliss, I. 2002. Inactivation of lactococcal bacteriophages in liquid media using dynamic high pressure. Int. Dairy J. 12:907-913.

Pagan, R., and Mackey, B. 2000. Relationship between membrane damage and cell death in pressure-treated Escherichia coli cells: Differences between exponential- and stationary-phase cells and variation among strains. Appl. Environ. Microbiol. 66:2829-2834.

Pandolf, W.D. 1998. High-pressure homogenization: Latest technology expands performance and product possibilities. Chem. Process 61:39-43.

Paquin, P. 1999. Technological properties of high-pressure homogenizers: The effect of fat globules, milk proteins and polysaccharides. Int. Dairy J. 9:329-335.

Pisabarron A.G., de Pedro, M.A., and Vazquez, D. 1985. Structural modifications in the peptidoglycan of Escherichia coli associated with changes in the state of growth of the culture. J. Bacterial. 161:238-242.

Popper, L., and Knorr, D. 1990. Applications of high-pressure homogenization for food preservation. Food Technol. 44:84-89.

Rosina E. Y. and Tistruga O.N. 2001. The special features of the sound capillary effect in the dielectric liquids with high viscosity. Russian Acoustical Society, XI, 113-116.

Rovinsky, L.A. 1994. The analys