Dear Editor,
I am writing to comment on the recent article "Eradication of Bacillus licheniformis Contamination while Maintaining Employee Safety in a Gnotobiotic Mouse Colony" which appeared in the December 2013 issue of LAS Pro by Julia Krout, et al.
I thank the authors for bringing attention to the endospore aka "spore" of Bacillus lichiniformis, because it is not only among the most resistant of spores to so-called chemical "sterilants," or more accurately, "sporicides," but it is also among the most resistant spores ever tested against "wet" heat (steam sterilization). Therefore this spore poses one of the greatest challenges in the field of gnotobiotics to both the chemical sporicide used for cold sterilization in the entry lock and to the wet steam sterilization of feed and water in the autoclave as well.
However, before addressing inadequate port sterilization, the main issue of the article, several facts concerning the sterilants tested and the test data provided should be noted.
Clidox-S® is not 3% chlorine dioxide as presented in the paper. It is 0.007%. The stoichiometry of the reaction of the Clidox-S® base (sodium chlorite) with the Clidox-S® activator (hydroacetic acid) shows that even if every single molecule of the 0.85% sodium chlorite in the Clidox-S® base were converted completely to chlorine dioxide, it would still not generate the 3% (30,000 ppm) of chlorine dioxide cited. The supplier of the Clidox-S®, Pharmacal, Inc., maintains there are ~70 parts per million (ppm) of chlorine dioxide when Clidox-S® is prepared as a sterilant at 1:3:1 (highest concentration and the one used in the study), which translates into 0.007% chlorine dioxide.
Although "4% sodium chlorite" is listed in the introduction as one of three sterilants currently in use within gnotobiotic facilities, sodium chlorite is not a sterilant. Therefore it cannot be used to maintain isolators in a gnotobiotic facility. The source of the 4% sodium chlorite was given as Exspor®. Exspor®, like Clidox-S®, comes as a two-part chlorine dioxide generating system. Neither the base (sodium chlorite) nor the activator (an organic acid) is a sporicide. Only when they are mixed together does the sodium chlorite undergo a chemical reaction turning it into chlorine dioxide gas, an entirely different molecule than sodium chlorite. The main difference between Exspor® and Clidox-S® is merely the type and concentration of surfactant (detergent) used in each. Both generate the same sterilant, chlorine dioxide. Chorine dioxide is not only a potent sporicide when used in aqueous solution, but has also been shown to be 1,075 times more sporicidal than ethylene oxide gas when used in the gaseous state1, a trait Dr. Trexler considers essential for any sporicide used in Gnotobiotics because air bubbles can become trapped below liquids2.
The greatest concern expressed to me was that the contact time was not given for the data presented in Table 1. Initially it was stated, "Procedures described by Majcher et al. (2008) were used to evaluate 7.35% hydrogen peroxide/0.23% PAA (etc listing the rest of the sporicides tested)", which would have meant a contact time of 20 minutes. However, it was then stated, "The protocol used for the testing was the Standard Guide for Assessment of Antimicrobial Activity Using a Time-Kill Procedure instead of Majcher et al. (2008)... ." This guide states, "For selection of contact times, a minimum time period should be selected based on the ability to reproducibly conduct the test sampling in this short time frame (for example, 15, 30, or 60 s). Other time points may be selected based on the intended use of the test material...." However, no contact time was indicated, so it is not known if it was just a matter of seconds, as suggested by the Guide, or whether it took minutes to kill the required minimum of one million spores.
The reader is therefore unable to evaluate the results, some of which make very little sense, e.g., 4.0% PAA gave 100% kill while a stronger solution of 4.5% PAA (with 22% hydrogen peroxide added as well) gave so little kill that it could not even be calculated, and was therefore recorded as TNTC (too numerous to count). The contact time for the entry port was also not given, which would be "...the intended use of the test material" also mentioned in the Guide. So the sporicidal test results could not be compared to the field condition under which the sporicide was to be utilized.
In my opinion, the role of biological burden should have been given more consideration. What is the probability that a million spores would fall from the ambient air into an entry port during the brief time it is open? With an exposure time of mere minutes, one would undoubtedly be challenged with no more than 10 to a 40 viable dust particles and about 1 to 4 of them would be sporeforming species, based on air samplings witnessed during my career. Then, even if all of those spores were Bacillus licheniformis, the sterilant would merely have to kill those few, not a million.
In stark contrast, pre-autoclaved feed, unlike ambient air, contains hundreds of thousands and often millions of bacteria/gram, a mixture of both sporeformers and nonsporeformers with vegetative cells and spores. In one study, spores of Bacillus licheniformis were the most heat resistant among 58 strains of 9 species of Bacillus tested, and, unlike the spores of the other species, could even withstand 135oC for a short period of time3. Moreover, many of the spores in feed pellets are buried deep within the pellets, which are dense and hard, and the spores are coated with proteins, carbohydrates, and fats. Together, these factors seriously challenge the ability of the steam to penetrate the center of each pellet, to bring it up to 121°C for the necessary period of time, and to completely destroy a potentially large number of spores. The number of “prevacs" or "pulses", and how much vacuum is drawn during each, is crucial for getting the steam to penetrate the center of the feed pellets.
The autoclave cycle for sterilizing water is also critical because it takes a long time to get so much mass up to 121°C. Spores of Bacillus licheniformis are ubiquitous and can be a challenge in water flasks as well as in feed and/or entry ports. During my career, an autoclave load of water flasks was given at least 90 minutes in order to sterilize the last flask in the center of the load, which should also contain a biological indicator suspended in the middle of the water of the middle flask in order to prove efficacy. A posting on the gnotobiotics listserv (gnotobiotics@listserv.uab.edu) by the senior author indicated they were using 60 minutes at 121oC for sterilizing their water.
Thus, in my opinion the most probable source of the Bacillus licheniformis contamination was either through the feed and/or water, rather than any inadequate sterilization of the port.
Lastly, I would like to address the use of an appropriate concentration of peracetic acid (PAA) for those who prefer it over chlorine dioxide. At an Annual Meeting of the Association for Gnotobiotics several decades ago, before the introduction of chlorine dioxide, a round table discussion was convened as one institution was experiencing a lot of personnel problems while using peracetic acid. It was discovered that everyone else was using 2% PAA except the facility having the personnel problems. That facility had gone from using 2% PAA to 4% with the rationale that if 2% is good, 4% is better.
Dr. Trexler has told the following story whenever this issue has come up. When he first started out, he used 0.1% PAA, but soon learned that very hard water would inactivate some of it, so he increased the concentration tenfold to 1.0%. Dr. Trexler has since insisted, "There is no reason to ever use peracetic acid above 1.0%." However, he acknowledged that few have followed his recommendation when he wrote, "A 2% aqueous solution of peracetic acid with wetting agent is used to sterilize isolators," and for introducing items via the entry lock, "Dilutions (1-2%) should be made in distilled or deionized water, with a small amount of wetting agent, preferably nonionic, added.”2
In the Majcher et al. study cited, just 0.3% PAA was used and they not only prepared it in hard water, but also added a "tripartite soil load" of tryptone, bovine serum albumin and mucin! Nevertheless, this solution of PAA was still able to qualify as a sterilant by killing a million spores of Bacillus licheniformis within 20 minutes. If just 0.3% PAA can kill over a million spores of Bacillus licheniformis within 20 minutes and qualify as a sterilant, why would one consider using over ten times that amount (4% PAA), thereby requiring personnel to wear expensive whole body suits and face gear, making isolator maintenance even more cumbersome, more labor intensive, and more expensive?
However, for those who chose to use PAA instead chlorine dioxide, they should also remember Dr. Trexler wrote that PAA needs to be neutralized with a sterile solution of sodium bicarbonate before the inside cap is opened (2). Otherwise the gnotobiotes will be exposed to PAA vapors that are highly irritating to eyes and damaging to nasal turbinates (Nota Bene: mice and rats can only breathe through their nares and not through their mouths), causing an enormous amount of stress to the animals. Technicians may be wearing respirators while exposed to PAA, but the gnotobiotes are not, and will be greatly stressed unless it is neutralized. Neutralizing PAA is yet another step requiring additional equipment and time, and is not necessary when using chlorine dioxide.
Allow me to reiterate that we owe these authors a note of thanks for bringing up the importance of Bacillus licheniformis in being one of the hardest contaminants to keep out of gnotobiotic isolators. All of one's sterilization methods need to be examined and proven to be effective with biological indicators.
I thank the authors for citing the chapter on Gnotobiotics by Dr. Trexler and myself in the AALAS publication, 50 Years of Laboratory Animal Science.
1.) Jeng DK, Woodworth AG. 1990. Chlorine dioxide gas sterilization under square-wave conditions. Appl Environ Micro 56:514-519.
2.) Trexler PC. 1983. Gnotobiotics. p 1-16 In: Foster HL, Small JD, Fox JG, editors. The Mouse in biomedical research, vol III. Burlington (MA) : Academic Press.
3.) Janstova B, Lukasova J. 2001. Heat resistance of bacillus spp. spores isolated from cow's milk and farm equipment. Acta Vet Brno 70:179-184.
Sincerely,
Roger P. Orcutt, PhD
Principal Consultant
