Cold Plasma Induced Modification of Stainless Steel Surfaces to Reduce Bacterial Biofilm Deposition

Development of bacterial biofilms on a variety of surfaces in food processing environments can lead to food contamination, transmission of diseases, and reduced shelf-life of food products. The objective of our research is to create surfaces that can inhibit bacterial attachment and biofilm formation.

In the biomedical field, surfaces treated with polyethylene oxide (PEO) and polyethylene glycol (PEG) have been shown to have antifouling properties. Classical methods for creating PEG/PEO-like coatings include: grafting, direct adsorption, covalently binding to a substrate, and crosslinking. These processes have been shown to increase wettability and render surfaces resistant to protein adsorption and cellular attachment.

Our research is aimed at antifouling properties on materials common to the food processing industry. Our approach is to use cold-plasma-enhanced processes to deposit coatings containing PEO/PEG-like structures onto various substrate surfaces. Plasma, the fourth state of the matter, is a partially ionized gas, containing charged and neutral particles, including electrons, positive and negative ions, atoms, and molecules. Plasmas can be broadly divided into cold plasmas and hot plasmas. Hot plasmas occur when the temperature of electrons and atomic and molecular species are extremely high. 99% of the matter in the universe is in the plasma state. Our Sun and the stars in the universe consist entirely of hot plasma, and the space within the stars of a galaxy is filled with plasma. Cold plasmas occur when the atomic and molecular species are at ambient temperature, whereas the electrons are at high temperatures. Cold plasma can be used to treat surfaces (e.g. oxidation, functionalization) or to deposit specific coatings onto organic and inorganic substrates.

Surfaces of stainless steel (type 304, #4 finish,) were plasma-modified using various gases and monomers. The newly developed surface chemistry was evaluated using electron spectroscopy for chemical analysis and attenuated total reflectance Fourier transform infrared spectroscopy. The presence of the coating and the surface topography was analyzed using atomic force microscopy. Water contact angle measurements were used to evaluate the wettability of the plasma-coated substrates. The contact angle concept is an indirect method used to estimate surface energy. It involves measuring the angle between the tangent to the surface of a liquid droplet made with the surface of the solid sample. Low water contact angle values indicate a hydrophilic (wettable) surface, whereas high water contact angles are indicative of a hydrophobic surface.

Samples were treated with 12-crown–4-ether and tri(ethylene glycol) dimethyl ether (triglyme) in a parallel-plate cold plasma reactor with heating facilities for the monomer and reaction chamber. High resolution electron spectroscopy for chemical analysis showed a high concentration of ethylene glycol-type functionality (C-O bonds) on the surfaces of the treated samples.

One of the first steps in biofilm formation is bacterial attachment. Both unmodified and plasma treated samples were subjected to 1-hour bacterial attachment and 1-day biofilm development by exposing them to a mixed culture of Salmonella typhimurium, Staphy lococcus epidermidis and Pseudomonas fluorescens. Bacterial biofilms were enumerated by plating on tryptic soy agar. Bacterial attachment to 12-crown–4-plasma-coated surfaces was decreased by 99% in comparison to unmodified samples. Also, biofilm development was 76% lower on the modified stainless steel. Triglyme-plasma coating reduced bacterial attachment by 74% and biofilm development by 83%. It was observed that biofilm bacteria are removed more easily from triglyme- and 12-crown–4-plasma coated samples in comparison to the unmodified stainless steel.

Cold plasma induced surface modification represents an attractive alternative for tailoring surface characteristics of materials. The process is dry and this greatly increases manufacturability over the wet chemistry approach, representing an important advantage for commercialization. Monomer costs are negligible because reactions occur at low pressure and the films are very thin. The coatings come from the reactor in a sterile form. In addition, the plasma treatment is limited to the outermost surface of a material and does not affect its mechanical properties. Thus, plasma treatment provides a powerful tool for altering performance characteristics of materials used in the food processing industry, and this technology shows promise as an approach to create surfaces that can help reduce biofilm formation.

• Subtyping of Vaginal Isolates of Group B Streptococci by Pulsed-field Gel Electrophoresis and Serotype
• Therapy with Botulinum Toxin
• Waterborne and Animal-to-animal Transmission of Escherichia coli O157:H7 and Subsequent Shedding Patterns in Holstein Bull Calves

Group B beta-hemolytic streptococci (GBS) recovered from the birth canal of pregnant women are the most common cause of neonatal sepsis in North America and Europe. Of the 25% of pregnant women that harbor GBS, disease of the newborn occurs in 1–2 of every 1000 deliveries, and infection has a case fatality rate of 6–20%. Once vaginal GBS infection during pregnancy is diagnosed, prophylactic therapy with an intrapartum antibiotic is practiced. Since the prophylactic treatment is unnecessary for 80% of patients, there is more to be learned about the epidemiology of these infections for working towards a better prevention method. To this end, we conducted pulsed-field gel electrophoresis (PFGE) and serotyping of GBS isolated from pregnant women.

Vaginal strains of GBS were recovered from 42 pregnant patients (35–37 weeks of gestation) from two hospitals in Madison, WI, over a seven month period in 1998. Isolates were molecularly characterized by pulsed-field gel electrophoresis (PFGE) at the Food Research Institute and serotyped at the Centers for Disease Control in Atlanta, Georgia. Analyses by PFGE revealed that the 42 strains could be further sub-divided into 24 genomic types following enzymatic digestion and 36 genomic types following digestion with a second enzyme. In contrast, the 42 strains could only be further sub-divided into 9 phenotypic groups via serotyping. These results substantiate that PFGE is a more discriminatory typing method than serotyping. Further analyses of these data revealed that strains of the same serotype typically belonged (i.e., clustered) to the same genomic group(s). We conclude that PFGE is an efficient, reproducible, and highly discriminating method for subtyping GBS, that along with serotyping, will be valuable for epidemiologic studies of GBS infection during pregnancy and for optimizing alternate treatment regimens.

The clostridia produce more protein toxins than any other bacterial genus and are a rich reservoir of toxins for research and medicinal uses. Research is under way to utilize clostridial toxins for drug delivery, prevention of food poisoning, and for the treatment of cancer and other diseases. The remarkable success of botulinum toxin as a therapeutic has created a new field of investigation in microbiology.

Prior to the start of the study, testing of fecal samples revealed that one calf (room 1) was already shedding E. coli O157:H7. Calves in adjacent pens and a pen across a walkway from the positive calf began to shed the same strain of E. coli O157:H7 during the 8 days following our initial screening. By the end of the 8-week study, 5 of 8 calves in room 1 had shed this strain of E. coli O157:H7. One of two negative calves began shedding after their water containers and water were switched with a calf that was shedding E. coli O157:H7.

In room 2, 4 of 8 calves received one liter of water containing ca. 10³–10⁴ CFU/ml of strain FRIK 1275. This O157:H7 strain was recovered from water and was the predominant strain recovered from a farm/herd in a previous study. Following inoculation, the calves were provided clean water for the remainder of the day, at which time water was pulled until they received a second liter of inoculum the following morning. The 4 inoculated calves shed strain FRIK 1275 24 h after administration. The length of shed in the naturally infected and inoculated calves ranged from 17–24 days at levels from 6.0 x 10¹ to 1.2 x 10⁶ CFU/g of feces. Removal and testing of lumen and intestinal-wall samples from O157:H7-positive calves did not indicate an association of the bacterium with a specific location in the bovine gastrointestinal tract. Results from this study demonstrate transmission of E. coli O157:H7 by water and animal-to-animal contact and suggest that this human pathogen is transient in cattle.

John M. Mansfield, Ph. D. Professor of Bacteriology	A new face around the FRI building is that of John Mansfield from the Department of Bacteriology. John recently moved his labs into the building after space was renovated for him as part of a College retention exercise. Although his research is not in the area of food microbiology, John has offered to dovetail some of his expertise with interests of the department. For example, his training and experience in the field of immunology may be useful for departmental investigators who are interested in learning how different bacterial toxins interact with T and B lymphocytes, and previously he has trained several students in structure-function relationships with respect to staphylococcal enterotoxins.  The current research carried out in the Mansfield lab is NIH-supported and addresses the complex biological relationship between host and parasite in African trypanosomiasis, a fatal protozoan disease of man and animals. Their research not only studies the immunogenetics of resistance to this disease, it also examines trypanosome gene expression that is important in regulating parasite virulence during infection. John’s research interest in parasitic diseases, and his role as the director of the UW Center for Research and Training in Parasitic Diseases, could be valuable for getting more campus investigators interested in food- and water-borne parasitic infections. Also, his immunology and eukaryotic molecular biology background may be useful in training departmental researchers how to set up certain types of experimental protocols with which they might not be familiar. And, his past experience with the NIH funding process as a Study Section Chair may come in handy for advising those who wish to pursue biomedical research funding from the NIH. John says that the department has been extremely friendly and interactive, and he is looking forward to a long and productive relationship with the group. He invites the faculty and trainees in the department to “drop in anytime” for a chat about science or related issues.  On the personal side, John is an avid watersports person. He is a competitive rower with the Mendota Rowing Club, and practices almost every day at 5:30 am. When the wind blows, he and his wife, Professor Donna Paulnock in the Medical Microbiology & Immunology Department, who is also a rower, can be seen windsurfing on Lake Mendota. They both enjoy the Madison lifestyle to its fullest!

• Salmonellae as Disease Fighters
• Listening to Corn...
• ...and Smelling Beef and Chicken

Since many bacteria causing intestinal infections act at the level of the intestinal mucosa, it seems logical to employ a bacterium which also interacts with the mucosa as a preventive agent. Several research groups have been working to develop avirulent strains of Salmonella spp. as vaccine carriers for antigens of pathogenic bacteria. Virulent salmonellae are attenuated by genetically inactivating two or more metabolic genes, for example those coding for enzymes involved in synthesis of aromatic amino acids. Such mutants persist in a mammalian host for only a short time and have essentially no potential for reversion to a virulent phenotype.

Vaccine vector strains of attenuated salmonellae have been constructed to contain antigenic determinants from tetanus toxoid, Helicobacter pylori urease, listeriolysin of L. monocytogenes, and from proteins of other foodborne bacteria. When these bacteria were fed to mice, they penetrated the intestinal mucosa, were taken up by white blood cells, and then elicited an antibody response to the foreign antigen. Subsequent challenge with live pathogenic bacteria or toxins revealed that the mice were resistant to infection. Immunization with these specific bacterial strains was not equally effective in all animal species, hence this system may need some fine tuning before it is useful in humans. This method of harnessing the infective ability of salmonellae to combat other enteric illness offers intriguing possibilities for controlling some serious foodborne illness.

Similarly, some Salmonella strains, weakened by mutations affecting metabolism of amino acids or purines, have been devised to target tumor cells in mice. Although these genetic changes prevent the salmonellae from growing well in normal cells, there are more nutrients available in tumor cells and therefore the bacteria thrive in tumor cells and concentrate there. Bacterial cell levels in tumor tissue can reach 10⁹ cfu/gram tissue to the detriment of the tumor. These bacterial strains can also be modified to carry genes affecting the survival of the tumor cells. Research is underway to make these strains carriers for anticancer drugs.

One problem caused by the injection of these cells into mice is the induction of septic shock by tumor necrosis factor. However, recently constructed mutant salmonellae, with alterations in lipid A, have been found to induce much lower levels of the tumor necrosis factor while, at the same time, they preferentially accumulate in tumor cells. In experiments with mice, these salmonellae significantly retarded growth of melanoma tumors without any evidence of septic shock. These experiments should also be considered preliminary in terms of their possible application to the treatment of human cancer. Never the less, they do provide a different perspective on our old nemesis, Salmonella.

References

• Fielder, M., and D. J. M. Lewis. Vaccination against bacterial gut infections. Curr. Opin. Infect. Dis. 11:591–596 (1998).
• Low, K.B., M. Ittensohn, T. Le, et al. Lipid A mutant Salmonella with suppressed virulence and TNF induction retain tumor-targeting in vivo. Nature Biotechnology 17:37–41 (1999).

In the midst of innumerable new “simple, rapid, and sensitive” visual methods developed for determining the safety of foods, two approaches based on other types of sensory information have emerged: photoacoustic spectroscopy and “electronic noses”.

Photoacoustic signals were actually discovered by Alexander Graham Bell when he observed that a beam of light focused on a sample produced an audible sound if the beam were turned on and off at an acoustic frequency. Strongly absorbing samples produce a louder sound than weakly absorbing samples. Apparently the pulses of light make an object heat up and cool down rapidly and thus generate sound waves.

This phenomenon is the basis of a technique called Fourier Transform Infrared Photoacoustic Spectroscopy (FTIR-PAS) which has recently been used by scientists to differentiate between moldy and clean corn kernels. When an infrared strobe light is shone on corn, the moldy kernels produce sounds at different wavelengths than those from uninfected kernels. The sounds are detected by a very sensitive microphone and the data is fed into a computer which can identify the contaminated corn. Although this technique is still in its developmental stages, it can identify moldy corn with a 96% accuracy compared to an 85% accuracy for a method based on the yellow-green fluorescence produced when UV light is shone on corn containing aflatoxins. The fluorescence test is used by the corn industry to assess aflatoxin but does not indicate other mycotoxins such as fumonisins. In contrast, corn infected by any species of mold will give a different FTIR-PAS signal than whole corn. Epidemiological data, such as plant stress, humidity, growing temperature and insect infestations, are combined with FTIR-PAS data to predict the most likely type of fungal infection.
 

...and Smelling Beef and Chicken

return to top

One familiar application of the so-called electronic nose is the carbon monoxide (CO) detector used in homes. This device detects the presence of CO and sounds an alarm. Similar types of instruments have been developed for detecting off-odors and monitoring quality in a variety of foods including meat, grains, cheese, fish, beer, and other beverages. In one recent study done in Sweden, an electronic nose was used to monitor the spoilage of vacuum packaged beef. Although there were some problems with the stability of some of the sensors, good correlation occurred between the degree of spoilage detected by the sensors and that determined by a trained sensory test panel. Furthermore, the e-nose was able to distinguish spoiled meat which originated in 4 different slaughterhouses. Presumably the spoilage bacteria contaminating each slaughterhouse were different, hence producing different proportions of volatile compounds.

Another recent USDA study demonstrated the ability of e-noses to differentiate between several bacteria, including Salmonella, Listeria, E. coli, and Pseudo monas isolated from processed poultry. Many of the same volatile compounds were produced by several different bacteria. However, the proportions of the gases released by these species were unique so that the e-nose could identify the bacteria by their smell. A research group in the UK has had similar success in distinguishing different spoilage fungi using an electronic nose. The volatile compounds produced by these organisms may allow early detection of pathogenic or spoilage organisms.

Many of the commercially available “noses” use metal oxide semiconductors which respond to either oxidizing or reducing compounds, such as ammonia or alcohols. Other types of sensors include conducting organic polymers and piezoelectric crystals. Earlier or more simple versions of such noses utilized an array of sensors, and the data from these sensors were analyzed statistically to determine patterns of volatile compounds. More recent devices incorporate artificial neural networks which are trained with known reference samples, thus can compare patterns of unknown samples to these references. In this way the electronic nose, like the human nose, recognizes odor patterns rather than specific compounds.

Although some technological problems have so far prevented electronic noses from reaching their full potential, with further development they are likely to be useful as an early warning system for food safety and food spoilage.

References

• Gordon, S. H., B. C. Wheeler, R. B. Schudy, et al. Neural network pattern recognition of photoacoustic FTIR spectra and knowledge-based techniques for detection of mycotoxigenic fungi in food grains. J. Food Prot. 61:221–230 (1998).
• Arnold, J. W., and S. D. Senter. Use of digital aroma technology and SPME GC-MS to compare volatile compounds produced by bacteria isolated from processed poultry. J. Sci. Food Agric. 78:343–348 (1998).

return to:
|  FRI Newsletter, Summer 1999 index  |  FRI Communications  |  FRI home page   |  Dept. Food Microbiology & Toxicology  |  UW–Madison  |