<I>Salmonella Enteritidis</I>

1022

Visualization of Eggshell Membranes and Their Interaction with Salmonella enteritidis Using Confocal Scanning Laser Microscopy

J. W. WONG LIONG,l J. F. FRANK,l* and S. BAILEy2 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021

tCenter for Food Safety and Quality Enhancement, Department of Food Science and Technology, University of Georgia, Athens, GA 30602-2106; and 2Poultry Microbiological Safety Research, United States Department of Agriculture, Russell Research Center, Athens, GA 30605, USA

(MS# 96-255: Received 27 September 1996/Accepted 24 December 1996)

ABSTRACT systems, CSLM has been used to visualize in real time the formation of yogurt structure and the responsible microorgan- Confocal scanning laser microscopy (CSLM) was used to isms (11). Specimens for CSLM do not need to be dehy- visualize eggshell membrane and observe its interaction with drated or fixed like samples that are to be examined by Salmonella enteritidis. Two- and three-dimensional images of electron microscopy. Special fluorescent dyes can be used to fluorescein isothiocyanate (FITC)-stained egg membranes were visualize specific features of a specimen. CSLM's analytical obtained for observation of structure. Outer membrane fibers 1 to 7 power derives from its ability to exclude light not in the 11m in thickness could be seen emerging from the calcified layers of focal plane so as to produce images with higher resolution the eggshell. Inner membrane fibers 0.1 to 3 flm in thickness were interlaced with the outer membrane. The limiting membrane, when than normal light microscopy (6, 7). The data, in digital stained with FITC, appeared as particles that filled spaces between form, can be collected and used to produce two-dimensional, the inner membrane fibers several microns outward from the level three-dimensional, and volume renderings and temporal at which the inner membrane fibers first appeared. The outer image sequences (3). membrane layer, approximately 50 to 70 11m thick, and the inner Contamination of eggs by Salmonella spp. can occur by membrane, approximately 15 to 26 11m thick, consisted of several numerous routes. An understanding of the mechanisms of discontinuous layers that were discernible as shifts in fiber position contamination can provide a basis for developing and or orientation and changes in fiber size. Large egg membranes, introducing controls to prevent or reduce this contamination. which had been detached from the eggshell, were submerged in a Introduction of Salmonella spp. can occur by transovarian 109-CFU/ml suspension of S. enteritidis over a 24-hour period to infection of the egg or contact with a contaminated environ- observe cell penetration. Cells were able to penetrate 28 flm into the membrane after 24 hours. Under moist conditions, Salmonella cells ment (19). Little can be done to rid an egg of microorgan- did not appear to attach to the fibers but floated easily in between isms which have infected it during its formation, but by them. Under dry conditions, Salmonella cells adhered to the understanding the natural barriers that protect the egg from membrane fibers. CSLM could be a useful tool in examining the environmental contamination, preventive measures might be effects of current storage and handling practices on eggshells and developed. membranes. The egg has a number of protective mechanisms to prevent the successful invasion of microorganisms (14). Key words: Eggshell membrane structure, Salmonella enteritidis, There are chemical as well as physical barriers to prevent the confocal scanning laser microscopy entrance of microorganisms and reduce their viability (1). Salmonella cells are known to survive the chemical barriers Confocal scanning laser microscopy (CSLM) allows in whole egg (1). The physical barriers, including the cuticle, access to spatial and topographical information with mini- the shell, and the shell membranes, limit contamination from mal sample manipulation (3) and offers a rapid means of particulates and microorganisms (1,2,20). collecting data regarding the interaction of microorganisms The objective of this study was to characterize the with their environment. With samples stained using the membranes of a hen's egg with confocal scanning laser fluorescent dye pyronin Y, CSLM has been used to examine microscopy and visualize the interaction of S. enteritidis the attachment of Salmonella cells to chicken skin and with the membranes. obtain information about skin topography (12). In food MATERIALS AND METHODS

Cultures * Author for correspondence. Tel: 706-542-0994; Fax: 706-542-7472; S. enteritidis Rochester was obtained from the USDA, Russell E-mail: [email protected]. Research Center, Athens, GA. The culture was stored on slants of SALMONELLA IN EGGSHELL MEMBRANES 1023 trypticase soy agar (TSA, Difco, Detroit, MI) at 4°C and trans- matching shell were placed in a sterile Petri dish and submerged in ferred on a monthly basis. FITC stain. After staining for one hour, the shell pieces and membranes were rinsed for 30 seconds with sterile distilled water. Eggs Shell pieces and membrane were individually taped onto slides. Large size grade A brown-shelled eggs were used in all An MRC-600 confocal scanning laser microscope (Bio-Rad microstructure experiments. Eggs for S. enteritidis interaction Inc. Hemel Hempstead, England) with 60X oil immersion objec- experiments were obtained from the USDA, Russell Research tive (numerical aperture 1.4, Nikon, Japan) and an Ar/Kr laser Center, Athens, GA. These eggs originated from a 65-week-old operating at 488 nm was used to visualize the FITC-stained fibers. flock of Peterson roosters and Avian Farms hens. Images (768 by 512 pixels) were digitally captured as data sets of 256-level gray scale values. Stacks of 10 images taken at 0.5-l1m intervals were merged. Fiber diameters were measured at the Stain, buffer, and fixative solution preparations thickest portion of the fiber image to make sure that the entire cross A 40-g/liter stock solution of osmium tetroxide (Sigma section was accounted for. The diameters of three hundred fibers Chemical Company, St. Louis, MO) was prepared by dissolving 1 g each were measured in the inner and outer membrane layers. of osmium tetroxide crystals in 25 ml of distilled water. A stock solution of 0.2 M cacodylate buffer (Sigma Chemical Company, Limiting, inner, and outer membrane layer thicknesses were Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021 St. Louis, MO) was prepared by dissolving 10.7 g of sodium determined by stepping with the stage motor through the sample until the corresponding fibers changed in diameter or disappeared. cacodylate in 250 ml of distilled water. The pH was adjusted to 7.2 Interstitial spaces were measured between fibers in the inner and with 0.1 N HC!. The 10-g/liter OS04 fixative was prepared by mixing 25 ml of the 40-g/liter OS04 stock solution with 75 mlofthe outer membrane layers. 0.1 M cacodylate buffer. Stock solution of fluorescein isothiocyanate was prepared by Thermal differential contamination of shell and outer membranes dissolving 30 mg of fluorescein isothiocyanate (FITC) crystals by S. enteritidis (Sigma Chemical Company) in 100 ml of 0.5 M sodium bicarbon- Approximately, 0.1 ml of stock S. enteritidis was transferred ate. The working solution was made by diluting the stock solution to 10 ml of Trypticase soy broth (TSB, Difco Laboratories, Detroit, lOO-fold in 0.5 M sodium bicarbonate. MI) and incubated in 10 ml at 34°C for 24 h. The entire contents Stock solution of pyron in Y (Sigma Chemical Company) was were transferred to 300 ml of nutrient broth in a 500-ml beaker. The prepared by dissolving I g of pyronin Y crystals in 100 ml of sterile inoculated broth and a sterile control were incubated at 34°C for 24 distilled water. The working solution was made by diluting the h. The final cell concentration of the cell suspension was 105 stock solution IDO-fold in sterile distilled water. CPU/m!. The cell suspension (in broth) was cooled to a temperature of Preparation of egg membranes for scanning electron microscopy 21°C. Dry eggs equilibrated to 40°C were submerged in the cell Membranes were detached from the shell using flat Teflon- suspension for approximately one minute, removed, and stored in coated forceps. The membranes were placed on aluminum stubs, 100-ml beakers containing 10 ml of sterile distilled water. The and carefully anchored using an aluminum retaining ring. While beakers containing the eggs were incubated for 24 h at 34°C. The the stubs were being prepared, cork was used to hold the stubs eggs were then cracked open, the yolk and albumen removed, and upright in a petri dish containing sterile water to maintain a moist membranes detached from the shell pieces. The remaining outer environment. Membranes were vapor fixed with osmium tetroxide membrane on the shell and the cuticle on the outer surface of the to minimize actual manipulation of the membrane surfaces and to shell were stained by flooding with O.I-g/liter pyronin Y solution minimize dehydration. Once the samples were anchored, the water for 10 min. The pieces were then rinsed in 100 m1 distilled water for was drawn off and replaced with a lO-gIliter OS04 solution. The 30 s and mounted on a slide. dishes were covered, and the samples were allowed to vapor fix for The confocal scanning laser microscope with 60X oil immer- 30 minutes. The samples were removed from the Petri dishes sion objective and Ar/Kr laser operating at 568 nm was used to containing the OS04 and sequentially dehydrated by submersion in visualize the pyronin Y-stained bacteria. Scans were initially taken 30%, 50%, 70%, 80%, 85%, 90%, 95%, and 100% ethyl alcohol at the calcified layer and then at 1.0-l1m depth increments until (vol/vol) for 10 minutes at each concentration. organisms could be visualized. Singlex-y plane images (768 by 512 The fixed membranes were dried by critical point drying in pixels) were digitally captured as 8-bit data sets. liquid CO2 using a Samdri critical point dryer (Tousimis Research Corp., Rockville, MD). The stubs and membranes were coated with Preparation of S. enteritidis infected egg membranes for time 35 nm of gold-palladium using a Hummer X sputter coater sequence and penetration experiment (Anatech LTD, Alexandria, VA). The membranes were then Approximately, 0.1 ml of stock S. enteritidis culture was observed under a Philips 505 scanning electron microscope (Phil- transferred to 12 plastic centrifuge tubes containing 40 ml of ips, Eindhoven, Netherlands). A spot size of 20 nm and an energy nutrient broth and incubated at 34°C for 24 h. The cells were setting of 15 keV were used as these settings were found to produce centrifuged at 114 X g for 15 min. The pellets were then the highest quality images. Surfaces including the limiting mem- resuspended with 200 ml of 0.3 mM phosphate buffer (pH 7.2). The brane, inner membrane fibers, and outer membrane fibers were cell counts of the suspensions were obtained by making serial chosen at random and micrographed. dilutions on plate count agar (PCA, Difco), which were then One hundred individual inner membrane fibers were randomly incubated at 37°C for 24 h. The final cell concentration of the selected and their diameters measured, and a range was determined. suspension was 109 CFU/m!. The same was done for the outer membrane fibers. Membranes from six eggs were tom and placed into the cell suspension. The flask was shaken at 50 rpm. Control membranes Preparation of egg membrane sample for fiber characterization were suspended in 200 ml of sterile phosphate buffer. Membranes byCSLM were removed for analysis after 30 s, 1 min, 30 min, and 1, 4, 8, 12, Membranes were detached from the eggshells by using a pair and 24 h, immediately placed into 200 ml of phosphate buffer of flat Teflon-coated forceps. Pieces of membrane and their rinsing solution, and agitated on a plate shaker at 50 rpm for I min. 1024 WONG LIONG, FRANK, AND BAILEY

At the end of each time interval, three samples were removed 0.09 to 0.15 !lm. The CSLM range for the thickness of this and stained by flooding with O.I-glliter pyronin Y solution for 10 layer is 2.5 to 4.6!lm with the average at 3.6!lffi. Methods of min. The pieces were then rinsed in 100 ml distilled water for 30 s. specimen preparation in addition to differences in bird Sample membranes were viewed using one wet mount and two dry variety could account for the discrepancies. mounts. Wet mounts were created by placing a piece of the egg The CSLM measurements represent a small aggregate membrane and a drop of phosphate buffer onto a slide and sealing area, but they show that the limiting membrane is not of the cover slip edges. The wet mounts allowed movement of cells to uniform thickness and is not distinctly separate from the be visualized. Dry mounts were created by placing the membrane directly onto the slide and taping the edges down. Two dry mount inner membrane. The accuracy of thickness measurements is slides were made so the limiting membrane would be exposed on limited by the fact that the limiting membrane actually the first slide and the outer membrane on the second slide. intermeshes with the innermost region of the inner mem- Scans were initiated at the point where the membrane was first brane fibers rather than forming a separable and distinct visualized, and continued through the membrane until bacteria layer (Fig. 1). In electron micrographs the limiting mem- were no longer seen in four fields (768 by 512 pixels). The scans brane appears to be a solid sheet. Simons and Wiertz (13) did were achieved by stepping the stage in O.I-flm increments in the not obtain micrographs that revealed intermeshing between Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021 z-axis direction. the limiting membrane and inner membrane. Tan et al. (15) viewed the limiting membrane in a TEM cross section as a Three-dimensional visualization of membranes and S. enteritidis sheet with undulations or ripples that could be caused by the interaction fibers that run along the inner membrane side or artifacts Optical sections for volume renderings of infected membranes from manipulation. Tan et al. (15) also observed that the were collected by setting the image capture screen to 192 by 128 limiting membrane contacts inner membrane fibers. lines to adjust the amount of area scanned per pixel. The stage motor was set to step in increments equal to the length represented Close examination revealed that not all areas of the egg by one pixel. This allowed the data to be collected in pixel volumes are covered by a uniform limiting membrane. Inner mem- or voxels. A 50-flm-thick stack of optical sections with a spacing of brane fibers can be seen very near the albumen side of the 0.3 flm (resulting in data points with a volume of (0.027 flm3) were limiting membrane (Fig. 1), despite the 2.5-!lm minimum collected and transferred to a Silicon Graphics Personal IRIS thickness detected by CSLM. This thin region may have 4D/35G Workstation (Silicon Graphics, Fairfield, IA). some significance for bacterial infection of the albumen as it indicates a possible entry point. RESULTS AND DISCUSSION Characterization of inner and outer membrane layers The limiting membrane by CSLM Simons and Wiertz (13) described the limiting mem- Egg membranes, when viewed using CSLM, appeared brane as a thin continuous layer that covers the inner different from those visualized using scanning electron membranes and keeps the albumen in the avian egg. This is microscopy. Table 1 presents the fiber diameter ranges and confirmed by the CSLM data as the limiting membrane membrane thickness data. SEM data ranges were 0.3 to 2.0 appears to cover and fill voids in the inner membrane. Table !lm and 0.5 to 5.2!lm for the diameters of the inner and outer 1 compares CSLM data to data previously obtained from membrane fibers, respectively. The CSLM ranges were from electron micrographs. The protein component of this mem- less than 0.3 to 3 !lm and from 1 to 7 !lm for inner and outer brane appears as particles which range in diameter from less membrane fibers, respectively. Fiber types associated with than 0.1 to 2.0 !lm. These particles appear separated from the inner and outer membranes are easily distinguishable in each other, probably because only the protein component Figures 2a and 2b. CSLM does not have as high a resolution was stained (Fig. lA). Simons and Wiertz (13) reported that as SEM and TEM, which limits accuracy of the fiber the thickness of the limiting membrane was 2.7 !lm. Tan et measurements, but overall the upper limits of the measure- al. (15) estimated the layer to be much thinner, ranging from ments were higher with CSLM. The lower SEM values in

TABLE 1. Confocal scanning laser microscopy (CSLM) and scanning electron microscopy (SEM) measurements of eggshell membranes andfibers

Structure CSLMa SEM Literature (reference)

Limiting membrane Layer thickness 2.5-4.6 flm; avg. 3.6 f1ill 2.7 flm (13); 0.09-0.15 flm (15) Particle diameter <0.1-2.0 flm Inner membrane Layer thickness 15-26 flm; avg. 21 f1ill 15-25 flm (5) Fiber diameter <0.1-3 f1ill;avg. 1.76 f1ill 0.3-2 f1ill;avg. 0.82 0.08-1.11 f1ill(15) Outer membrane Layer thickness 50-70 flm; avg. 59 flm 50-70 f1ill(5) Fiber diameter 1-7 f1ill;avg. 5.1 flm 0.5-5.2 f1ill;avg. 3.9 f1ill 0.11-4.14 f1ill(15) a For each structure, n is approximately 300. b No value has been determined or reported. SALMONELLA IN EGGSHELL MEMBRANES 1025 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021

FIGURE 1. Confocal scanning laser microscope images of limiting membrane. (A) Limiting membrane as represented by FITC-stained protein particles. Scale bar = 15 Jim. (B) Limiting membrane protein particles encroaching among the inner membrane fibers. Scale bar = 15 Jim. FIGURE 2. Volume-rendered images of membrane fibers. Image (a) is a top view of a IO-Jim-thick inner membrane with remnants of this study may be due to shrinkage caused by dehydration the limiting membrane. Image (b) is a lIDO rotation of a I5-J1m- that occurs during the fixation and alcohol dehydration thick volume rendering of outer membrane fibers. The arrow in steps, even though SEM samples examined for this study image (b) indicates a void space where bacteria could penetrate. were fixed in a chamber containing distilled water to increase environmental moisture and minimize shrinkage. The ranges obtained with SEM in this study were also wider were obtained using both membrane displaced from the shell than those reported by Simons and Wiertz (13) and Tan et al. and fiber remaining on the shell. The two measurements (15) (Table 1). were added together to get an overall thickness estimate. Interfiber distances were not measured in the SEM Fiber measurements were obtained from fibers representing sections, but were determined in the CSLM specimens. The different regions within each membrane. distances ranged from less than 0.3 to 15 /JIll for the inner Simons and Wiertz (13) using TEM and SEM reported membrane layer and from less than 0.3 to 30 /lm for the outer that the inner membrane is subdivided into three layers and membrane layer. The arrow in Figure 2b indicates a passage the outer membrane may have as many as six subdivisions. large enough for a bacterial cell to pass. These divisions were not apparent with CSLM. Rather an The measured membrane thicknesses of the inner and interlacing of inner fibers with outer membranes was outer membrane layers taken from the equatorial region of observed, and determination of where one layer ended and the egg were highly variable. The thickness of the inner another layer began was not possible. CSLM can scan membrane ranged from 15 to 26 /JIll in with an average of 21 through several layers of fibers. As the scans are collected, /lm. The thickness of the outer membrane ranged from 50 to shifts in fiber positions and changes in fiber diameters are 70 /lm with an average of 59 /JIll. The varying thicknesses of observed. In the outer membrane a shift in fiber direction the membranes agree closely with the estimates of Burley could indicate presence of a subdivision, but interlacing of and Vadehra (5) provided in Table 1. In CSLM the working fibers obscured sublayer delineations. Simons and Wiertz distance was reduced by the shell thickness, which reduced (13), using TEM cross sections, were able to see distinct the accuracy of estimates of the outer membrane layer separations between these subdivisions, with each subdivi- thickness. Estimates of the outer membrane layer thickness sion being several fibers thick. The subdivisions described 1026 WONG LIaNG, FRANK, AND BAILEY by Simons and Wiertz (13) ranged from 1 to 9 fibers in thickness. Distinct separations were not observed using CSLM; rather the fibers were interlinked. This interlinking also occurred at the inner and outer membrane interface. These observations indicate that there is not a complete separation between the different fibrous layers. Interlinks or junctions on a common fiber mantle appear to act as sites for subsequent fibers to be laid down when the yolk and albumen pass through the oviduct. Fibers of different diameters can emanate from a single junction. Simons and Wiertz (13) observed different fiber cores occupying a common mantle. The fact that the fibers associate with each other and share a common mantle could explain how the fibers from the outer and inner layers interlace with each Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021 other. Fibers with the diameter of inner membrane fibers often appear a few microns into what was considered exclusively the outer membrane fiber layer. Tan et al. (15) used eggs from Brown Leghorns, and Simons and Wiertz (13) used eggs from White Leghorns and Rhode Island Reds. The breed of bird eggs used in this study is not known. This may be important as Simons and Wiertz (13) reported differing shell thicknesses from two different breeds of chicken hens. Some differences in shell thickness between eggs from the same species were also reported. Breed, age, environment, nutrition, and diseases all have an effect on the formation of the egg (15). There could be other unknown factors that may affect membrane microstructure. Aside from the differences in fiber dimensions, nodules associated with the inner and outer membrane fibers appar- FIGURE 3. Confocal scanning laser microscope images of egg- ent in SEM micrographs were not seen in CSLM images. shell and outer membrane fibers from eggs infected with Salmo- The nodules may be an artifact of sample preparation, or nella cells by the thermal differential technique. Salmonella cells there could be a carbohydrate or a lipid component of the are present around mammillary caps in image (a). Scale bar = 10 membranes that was not stained in CSLM sample prepara- jim. Outer membrane fibers are closely associated with Salmonella cells in image (b). Scale bar = 10 jim. tion. interaction ofS. enteritidis and eggshell membranes Fajardo et al. (8) demonstrated movement of Salmo- nella cells across the shell when eggs are cooled. In cases of thermal differential infection by submersion it was found that S. enteritidis cells readily move through the shell pores and penetrate into the outer layers of the membranes. Eggs infected by the temperature differential technique showed accumulations of S. enteritidis around the shell interior mammillary caps and outer membrane (Figure 3). Figure 3a is an image showing bacteria in pores around the mammillary caps. Approximately 3 J..Iminward from the mammillary caps, outer membrane fibers are covered with adhering bacteria (Figure 3b). Volume-rendered sections of a wet-mounted infected outer membrane showed free-floating bacteria. Figure 4 illustrates a 90° rotation of a three-dimensional volume rendering of outer membrane showing that the bacteria are both associated with fibers and free floating. The arrows indicate individual bacteria floating in void spaces. S. enteritidis cells appear to adhere to fibers when dry mounted. FIGURE 4. Volume-rendered images of membrane fibers and When examined for penetration over time, specimens Salmonella cells. Salmonella cells are evenly distributed through- showed that the bacteria can move quite easily across the out the IS-jim-thick outer membrane layer. The image shows both outer membrane as shown in Figure 5. While bacteria moved some bacteria floating free of the fibers and some associated with more than 28 J..Iffiinto the membrane during 24 h of the fibers in three-dimensional space. SALMONELLA IN EGGSHELL MEMBRANES 1027

35 but also prevents microbial invasion. If there are thin spots in the limiting membrane, the bacteria could penetrate the 30 whole membrane. CSLM visualization of bacteria and the membrane suggests that free liquid is important for penetration. The egg can become moist by various means. Immedi- I \ j j ately after being laid the egg is quite moist (5). Washing not only exposes eggs to a moist environment, but, if not done correctly, can produce a temperature differential (8). Com- mercially raised eggs are exposed to high humidity and an environment that may include fecal matter (4). Eggs in commercial distribution may sweat due to temperature fluctuations. Porosity and age of an egg can influence 0.02 0.5 1 4 12 24 bacterial penetration (10), so CSLM could be used to Time (h) characterize the effects that age, species, diet, and environ- Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021 FIGURE 5. Distance (J1m)to which Salmonella cells penetrate the ment have on egg formation and membrane characteristics. eggshell membrane over a 24-hour period. Manipulation of these factors might ensure that a consistent shell and membrane is formed (15, 16). However, environ- incubation, none were seen in the inner membrane layer. mental controls would still be necessary to reduce exposure Experiments were conducted in high-moisture environ- to contaminants. ments, which provides a good medium for movement of the organism. ACKNOWLEDGMENT The organisms could be seen to move when examined in a wet mount. This movement was probably Brownian This research was supported by state and Hatch funds allocated to the motion with possible movement in convection currents Georgia Agricultural Experiment Stations. The authors thank Dr. M. A. arising from the laser heat. Movement, however, was not Farmer, Director of the Center for Advanced Ultrastructure Research, University of Georgia, for valuable help. observed in the thermal differential infection experiment or in the dry mounts in the penetration versus time experiment when less moisture was present in the mounted samples. REFERENCES During the penetration experiment, the membrane was 1. Board, R. G., C. Clay, J. Lock, and J. Dolman. 1994. The egg: a suspended in buffer from the time of exposure (to the S. compartmentalized, aseptically packaged food, p. 43-62. In R. G. enteritidis) to the time of examination of the sample. Board and R. Fuller (ed.), Microbiology of the avian egg. Chapman & After 24 hours of penetration, clumps of bacteria Hall, London. appeared on the fibers. These could have been colonies, or 2. Board, R. G., and N. A. Halls. 1973. The cuticle: a barrier to liquid and particle penetration of the shell of the hen's egg. Br. Poult. Sci. areas of entrapment where bacteria had difficulty moving in 14:69-97. and out of the fibers. During active scanning of samples, 3. Brackenhoff, G. J., E. A. Van Spronsen, H. T. M. Van Der Voort, and bacteria could be seen moving in and out of the fibrous N. Nanninga. 1989. Three-dimensional confocal fluorescence micros- mesh. This movement could have been induced by microcon- copy. In D. L. Taylor and Y.-L. Wang (ed.), Methods in cell biology, vol. 30. Fluorescence microscopy of living cells in culture, pt. B. vection currents caused by heat from the laser. Quantitative fluorescence microscopy-imaging and spectroscopy. Aca- A membrane that was allowed to sit in a cell suspension demic Press, Inc., San Diego, CA. for more than 40 hours showed a large number of cells in the 4. Bruce, J., and E. M. Drysdale 1994. Trans-shell transmission, p. outer membrane fibers and several at the inner membrane 63-92. In R. G. Board and R. Fuller (ed.), Microbiology of the avian egg. Chapman & Hall, New York. layer. Longer exposure could have allowed cells to cross the 5. Burley, R. W., and D. V. Vadehra. 1989. The avian egg: chemistry and entire membrane. An egg that takes 20 days to hatch could biology. John Wiley & Sons, New York. become surface infected and with a humid environment 6. Caldwell, D. E., D. R. Korber, and J. R. Lawrence. 1992. Confocal allow bacteria to traverse the shell and its membranes. Such laser microscopy and digital image analysis in microbial ecology. In K. C. Marshall (ed.), Advances in microbial ecology, vol. 12. Plenum an environment has been shown by Grave and MacLaury (9) Press, New York. to aid in the transport of contaminants. 7. Caldwell, D. E., D. R. Korber, and J. R. Lawrence. 1992. Imaging of The membrane acts like a filter for the eggshell. The bacterial cells by fluorescence exclusion using scanning confocal laser model suggested by Simons and Wiertz (13), Tan et al. (15), microscopy. J. Microbiol. Methods 15:249-261. 8. Fajardo, T. A., R. C. Anantheswaran, V. M. Puri, and S. J. Knabel. and Tyler (18) appears valid. CSLM offers some new 1995. Penetration of Salmonella enteriditis in eggs subjected to rapid information not obtained by TEM or SEM methods. Mem- cooling. J. Food Prot. 58:473-477. brane fibers appear larger when observed with CSLM than 9. Graves, R. C., and D. W. MacLaury. 1962. The effect of temperature, with EM techniques. The membrane layers are seen as vapor pressure and absolute humidity on bacterial contamination of shell eggs. Poult. Sci. 41:1219-1225. intermeshing and not completely separated. This may have 10. Hartung, T. E., and W. J. Stadelman. 1963. Pseudomonasfiuorescens implications for our understanding of membrane synthesis. penetration as influenced by shell porosity, age of egg and degree of The presence of the interfiber spaces observable by CSLM bacterial challenge. Poult. Sci. 42:147-150. implies that bacteria of various sizes can work their way 11. Hassan, A. N., J. F. Frank, M. A. Farmer, K. A. Schmidt, and S. I. Shalabi. 1995. Formation of yogurt microstructure and 3-dimensional through them. Tung and Richard (17) suggested that the visualization as determined by confocal scanning laser microscopy. J. limiting membrane not only prevents the leakage of albumen Dairy Sci. 78:2629-2636. 1028 WONG LIaNG, FRANK, AND BAILEY

12. Kim, K. Y, J. F. Frank, and S. E. Craven. 1996. Attachment of G. Wells and C. G. Belyavin (ed.l, Egg quality--current problems and salmonella on modified poultry skin surface. J. Food Sci. 61 :442-448. recent advances. Butterworths, London. 13. Simons, P. C. M., and G. Wiertz. 1963. Notes on the structure of 17. Tung, M. A., and J. H. Richard. 1972. Ultrastructure of the hen's egg membranes and shell in the hen's egg: an electron microscopical shell membrane by electron microscopy. J. Food Sci. 37:277-281. study. Z. Zellforsch. 59:555-567. 18. Tyler, C. 1969. Avian egg shells: their structure and characteristics. 14. Solomon, S. E. 1991. Egg and eggshell quality. Wolfe Publishing Ltd, Int. Rev. Gen. Exp. Zoo1. 4:81-130. Hazell Books, Ltd, Aylesbury, England. 15. Tan, C. K., T. W. Chen, H. L. Chan, and L. S. Ng. 1992. A scanning 19. Vadehra, D. v., R. G. Baker, and N. B. Naylor. 1970. Infection routes and transmission electron microscopic study of the membranes of of bacteria into chicken eggs. J. Food. Sci. 35:5-6. chicken egg. Risto1. Histopatho1. 7:339-345. 20. Vadehra, D. v., R. G. Baker, and N. B. Naylor. 1970. Role of cuticle in 16. Tullett, S. G. 1987. Egg shell formation and quality, p. 123-146. In R. spoilage of chicken eggs. J. Food Sci. 35:61-62. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/60/9/1022/1665105/0362-028x-60_9_1022.pdf by guest on 02 October 2021

&lt;I&gt;Salmonella Enteritidis&lt;/I&gt;

<I>Salmonella Enteritidis</I>