A Novel Method for Sampling and Long-Term Monitoring of Microbes That Uses Stickers of Plain Paper

As a ubiquitous bacterium, Listeria monocytogenes has a propensity to enter food production areas inadvertently via fomites such as door handles and switches. While the bacterium might not be in direct contact with the food products, knowing the microbial status of the surroundings is essential for risk assessment. Our investigation into a novel quantitative PCR (qPCR)-based sampling system with the highest sensitivity and ability to monitor over long periods of time, yet based on paper, proved to be cost-effective and reasonably convenient to handle.

ABSTRACT

Detection of pathogens is crucial in food production areas. While it is well established, swabbing as a state-of-the-art sampling method offers several drawbacks with respect to yield, standardization, overall handling, and long-term monitoring. This led us to develop and evaluate a method that is easier to use at a lower cost and that should be at least as sensitive. After evaluating sundry promising materials, we tested text-marking paper stickers for their suitability to take up and release Listeria monocytogenes with their nonsticky paper side over a 14-day time period using quantitative PCR. The recovery rate was similar to that in previous studies using conventional swabs, and we also confirmed the feasibility of pooling besides resilience to cleansing and disinfection. In a proof-of-concept experiment that sampled several locations, such as door handles, the occurrences of L. monocytogenes and Escherichia coli were determined. The results suggest that the presented sticker system might offer a promising cost-effective alternative sampling system with improved handling characteristics.

IMPORTANCE As a ubiquitous bacterium, Listeria monocytogenes has a propensity to enter food production areas inadvertently via fomites such as door handles and switches. While the bacterium might not be in direct contact with the food products, knowing the microbial status of the surroundings is essential for risk assessment. Our investigation into a novel quantitative PCR (qPCR)-based sampling system with the highest sensitivity and ability to monitor over long periods of time, yet based on paper, proved to be cost-effective and reasonably convenient to handle.

INTRODUCTION

Foodborne pathogens can cause serious diseases and death. Further, recalls of potentially contaminated goods result in significant economic damage. A prominent example is Listeria monocytogenes, a Gram-positive, facultative anaerobic, and ubiquitous human pathogen which has the ability to adhere to surfaces commonly encountered in food processing environments (1). Outbreaks of listeriosis continue to occur sporadically, and the mortality rate is as high as 20% (2). Since L. monocytogenes is able to grow under refrigeration conditions (4°C) (3), and given that clinical symptoms are frequently compiled only after delays (4), listeriosis is an especially serious diagnosis. For these reasons, national and international standards and regulations for cleansing, disinfection, and monitoring (5) of foodborne pathogens have implemented zero tolerance for L. monocytogenes in ready-to-eat foods (6–8). However, only the most reliable methods with the lowest detection limits will achieve this goal. Common assays for monitoring pathogens comprise conventional microbiological methods based on growth and are thus time-consuming to perform (9). In order to reduce processing time and costs, faster and more reliable alternative detection methods are being investigated (10).

However, a prerequisite for accurate microbiological data is effective sampling. In the food production facility, conventional swabbing as a standard method can only expose a momentary snapshot. For example, it is not possible to reconstruct information about yesterday’s status after cleansing has been performed. In addition, when moistened swabs or contact-plate sampling methods are used, they bring with them growth medium into a supposedly clean environment, making subsequent disinfection necessary. To obtain comparable results, swabbing should be performed within a defined area with reference to ISO 18593:2004 (11) or the FSIS directive 10,300.1 (12). This is not very practical with complex surfaces such as door handles, light switches, and other typical fomites (13). The method itself intrinsically has a low ability to take up bacteria from dry surfaces (14) and is associated with highly varied recovery rates (15) averaging 20% (16).

A promising departure from growth-based methods is quantitative PCR (qPCR), which provides faster results. Theoretically, the detection limit of this method approaches one copy of the target gene (17). Practically, it is an especially sensitive tool that has undergone enormous developmental progress over the past several years (18). Although qPCR cannot distinguish between living and dead cells, the results reflect the occurrence of (present or past) contamination, including the presence of viable but nonculturable cells (VBNCs) (19). Consequently, advantages are to be anticipated when qPCR can be combined with an improved but cost-effective sampling method.

Sticking a thin collecting device to the surface of interest for a longer period of time than the swabbing process may overcome some of these problems. Ideally, the material used should be bacteriostatic and widely insensitive to moisture or abrasion. Additionally, a porous structure should provide adhesive force and inter alia capillary action and incorporate adhesive mechanisms for cell membranes (20). It should inhibit neither DNA extraction nor the performance of qPCR.

Consequently, the purpose of this study was to determine the suitability of plain sterile paper to collect bacterial cells and to recover them for quantitative analysis. The fomite potential of paper has already been widely discussed (21, 22), and in the food industry, it is regarded as a noncritical material (23). To establish the suitability of this method, deliberate contamination (spiked) experiments were performed. Finally, both Listeria monocytogenes (24) and Escherichia coli (25) were detected in a proof-of-concept study using qPCR.

RESULTS

Regular sampling is mandatory for proper monitoring of microbes in sensitive environments. This procedure is commonly performed using swabs. Besides a rather small recovery (26, 27), detection results obtained only represent a momentary snapshot and subsequent disinfection after sampling is advised, which makes the sampling process laborious. Therefore, in this study, we examined an alternative surface suitable for trapping bacteria. Attached to the surface of interest as a sticker, it can remain in place for long time periods while collecting contaminants. In this manner, we investigated the appropriateness of text-marking stickers, comprising plain paper surfaces, as an alternative sampling system.

DNA recovery from stickers is sufficient and constant over time.

The suitability of text-marking stickers as sampling alternatives for swabs was initially assessed quantitatively and qualitatively using both molecular and microbiological methods. L. monocytogenes ΔprfA mutant and E. coli strains were applied to stickers in concentrations ranging over four logarithmic units. After 24 h, the stickers were either quantitatively analyzed using plate count methods or qualitatively analyzed with enrichment in tryptic soy broth (TSB) or Half Fraser medium for microbiological analysis (Table 1). The enrichment method resulted in random detection of E. coli at any concentration, while L. monocytogenes was detected at 100 CFU. Poor results were obtained using the plate count method, where almost no growth was achieved at any tested concentration. In contrast, qPCR provided quantitative results for both bacteria at all concentrations, although recovery of L. monocytogenes was lower than for E. coli (Fig. 1). Thus, further experiments focused on qPCR analysis as the detection method of choice.

TABLE 1.

Growth of L. monocytogenes ΔprfA mutant and E. coli after drying on stickers

Input (CFU)	No. of positive findings fora :
	L. monocytogenes ΔprfA in TSB	L. monocytogenes ΔprfA Half Fraser medium	E. coli in TSB
10	1	0	1
102	8	5	1
103	9	8	1
104	9	9	1

^aNumber of positive findings (growth) out of 9 after stickers incubated in TSB (L. monocytogenes and E. coli) or Half Fraser medium (L. monocytogenes) from three independent experiments performed in triplicate.

FIG 1.

Quantification of L. monocytogenes and E. coli from artificially contaminated stickers over a broad dynamic range. DNA from stickers artificially contaminated with four 10-fold logarithmic dilutions (starting at 80 CFU for E. coli and 10 CFU for L. monocytogenes) was extracted and quantified using qPCR (y axis). Control DNA (input, applied on stickers) was simultaneously extracted and analyzed as a reference (x axis). Symbols and error bars denote standardized mean differences and standard deviations, respectively (n = 3 independent experiments with three repetitions each). For the sake of clarity, only positive y-error bars of are displayed (negative y-error bar values are identical to the positive values).

To determine stability over time, five sets comprising six sterile stickers were contaminated artificially with different counts of the L. monocytogenes ΔprfA mutant. Two stickers each were contaminated with 5 CFU, 50 CFU, or 500 CFU and were stored for up to 14 days before DNA extraction. For comparative purposes, the same inoculum was plated on tryptic soy agar (TSA) plates, or DNA was extracted directly after dilution (schematically presented in Fig. 2). Recovery from the stickers was rather varied, at around 30%, but did not distinctly decrease after 14 days of storage (Fig. 3). This suggests the possibility of sampling over 2 weeks as well. The results are similar to those obtained from recovery from sponge-sticks in a previous study (16).

FIG 2.

Schematic representation of the artificially contaminated sticker setup. UV-treated stickers were artificially contaminated by the addition of diluted bacteria suspensions at desired concentrations. After respective incubation times, DNA from stickers and controls was extracted and analyzed using qPCR. In parallel, cells were plated on TSA plates as controls (described in more detail in Materials and Methods). O/N, overnight; neg., negative.

FIG 3.

Stability of recovery over time. Stickers were artificially contaminated with L. monocytogenes ΔprfA mutant and DNA extracted and analyzed with qPCR after 0, 1, 3, 7, and 14 days. Bars and errors bars represent grand means of recovery (outcome [qPCR]/input [qPCR]) and standard errors of four (days 0 and 3) or three (1, 7, and 14 days) independent experiments, including two different bacterial concentrations in duplicate.

Cleansing and disinfection have minor impacts on bacterial detection with stickers.

Surfaces in food processing plants are anticipated to be cleansed regularly. Therefore, to test whether the paper stickers convey advantages or disadvantages compared with conventional sampling, artificially contaminated (L. monocytogenes ΔprfA mutant) stickers were treated with water, soapy water, and a disinfection agent to simulate routine cleansing practices. As a control, ceramic tiles were artificially contaminated, treated the same way as the stickers, and sampled using sponge-stick swabs. The results summarized in Fig. 4 reveal that after cleansing and/or disinfection, distinctly more bacteria could be detected using stickers.

FIG 4.

Recovery after cleansing and disinfection. Surfaces or stickers applied to surfaces were artificially contaminated with 10³ to 10⁴ CFU of L. monocytogenes ΔprfA mutant. After drying, surfaces were washed and sampled, and DNA was extracted and analyzed with qPCR. Bars represent the grand mean of recovery (outcome [qPCR]/input [qPCR]) with the standard error of five independent experiments performed in duplicate.

Pooling of up to six stickers.

The surface of one sticker measured only 50 mm², which is much smaller than commonly suggested for swabbing. In order to optimize sampling density versus effort (labor power and materials), we decided to process more than one sticker per DNA extraction sample rather than using larger stickers (28). Although the number of stickers to be processed at once is restricted by the volume of prelysis buffer used in the first step of the NucleoSpin kit, we found that six stickers per sample proved to be a good compromise when retaining the original protocol. To test whether pooling six stickers leads to a loss of information due to possible dilution by empty stickers or increased quantities of insoluble material, two pools were tested against reference samples (Fig. 5a). The two pools contained the same bacterial count. A single sticker carrying the same or 1/6 of the contamination level as both pools served as a control. As demonstrated in Fig. 5b, pooling appeared to be adequate, whereby similar results were obtained independent of the number of (empty or contaminated) stickers. Despite relatively high variation, only the total amount of DNA in the sample appears to be relevant. Thus, this pooling approach might help reduce sample numbers. Statistically, a higher number of applied stickers increases the chance of detecting minor contaminants.

FIG 5.

Pooling of stickers. (a) Schematic representation of the pooling approach demonstrates the different samples, as follows: a single contaminated sticker, a single sticker with 1/6 contamination level, a pool containing six contaminated stickers, and a pool containing one contaminated and five empty stickers. (b) Results show that pooling of six stickers does not lead to a great loss of information. Bacterial cell equivalents (BCE) were determined using qPCR. Bars represent the standardized mean difference with the standard deviation of the results from four independent experiments.

Sampling on-site and proof of concept for successful detection of bacterial contamination using stickers.

After investigating the suitability of the new method with artificial contamination experiments, the stickers were utilized in a proof-of-concept experiment to establish whether bacteria can be captured with this system. For this purpose, stickers were applied at several locations that underwent frequent hand contact, such as door handles or light switches of toilets, for 1 to 7 days. In the first setup, for detecting L. monocytogenes, sampling using sponge-sticks was performed in parallel. In the second setup, qPCR for E. coli was additionally performed to supplement the occurrence of positive results and to monitor another species. Swabbing was omitted in this setup, but three time periods were included. Since it was demonstrated that the prfA assay can detect and quantify even down to a single molecule (17), each positive signal in qPCR was rated as a positive result.

The results summarized in Tables 2 and 3 show that stickers detected both bacterial species repeatedly from several locations, suggesting suitability as an appropriate on-site sampling/detection system. Further, in the first setup, the stickers detected similar or even higher occurrences of L. monocytogenes than with the conventional swab system (Table 2). Finally, an analysis of stickers that were applied for periods of 1 to 7 days indicated their suitability for sampling and detection, essentially independently of the date of contamination (Table 3).

TABLE 2.

Detection of L. monocytogenes on-site using stickers and swabs

Sticker or swab	No. of positive findings by sitea
	1	2	3	4	5	6	7	8	9	10
Sticker	2	2	3	7	2	5	3	4	0	2
Swab	1	0	2	1	1	1	0	1	0	2

^aNumber of L. monocytogenes-positive findings out of 7 in qPCR of seven independent trials (setup 1).

TABLE 3.

Detection of L. monocytogenes and E. coli on-site with stickers

Day	No. of positive findings/total no. of testings by pass fora :
	E. coli			L. monocytogenes
	1st	2nd	3rd	1st	2nd	3rd
1	4/5	2/5	0/5	1/5	1/5	0/5
3	4/4b	2/5	0/5	1/4b	0/5	0/5
7	3/4b	1/5	1/5	0/4b	1/5	0/5

^aNumber of positive findings in qPCR on five door handles tested three times from three time periods (setup 2).

^bOne sticker was unfortunately lost.

Accumulation of free DNA on frequently used door handles.

Besides the monitoring of microbes in our proof-of-concept study, the prfA internal amplification control (IAC) assay (29) was examined in parallel because the lyophilized internal amplification control was accidently distributed in one room more than 10 years ago. Startlingly, this synthetic oligonucleotide of 100 bp could still be detected on the door handle of this room and even accumulated over time on the stickers, demonstrating the stability of DNA and the ability of the new sticker system to detect it effectively (Fig. 6).

FIG 6.

Accumulation of synthetic IAC on stickers over time. qPCR (IAC assay) of DNA extracted from stickers applied to a door handle demonstrates an accumulation over time of synthetic DNA on stickers that was distributed in this room. Results are representative of three independent experiments. dR, baseline-corrected raw fluorescence.

DISCUSSION

The aim of this study was to investigate the ability of plain paper to trap bacterial pathogens and related DNA. Although there was variation in recovery from artificially contaminated stickers, the stickers provided results similar to those previously obtained with swabs (16). Preservation of DNA on the sticker also showed the method to be very reliable over time. The presented detection is qPCR based since evaluation using microbiological growth methods demonstrated poor results, and, furthermore, this molecular method was faster. No inhibitory factors impairing DNA extraction or qPCR were encountered. However, as positive findings are a statement for the presence of DNA, this does not inevitably originate from living cells. Therefore, the method thus detects living, dead, and viable but nonculturable cells (VBNCs). While detection of nongrowing cells in the past was often discussed as a disadvantage (1), increasing interest and awareness of VBNCs today highlight that nongrowing cells are also a potential threat (30). Further, the detection of nongrowing cells advantageously attests to badly cleansed areas, as most chemical disinfectants alone cannot remove DNA.

As shown in the proof-of-concept experiment, synthetic DNA is effectively captured. Thus, used stickers can also be used to detect “flying” DNA that can also produce severe problems (31). Contamination of this nature might occur more often than anticipated in establishments and has also been demonstrated with peptides (32). Although contamination with artificial DNA is unlikely to occur in the food industry, where it is rarely used, there is still a risk of contamination with PCR products. Guidelines for laboratory practice (33) must be followed strictly to minimize risks, e.g., neither to open nor to autoclave PCR tubes prior to disposal. All essential controls must be included in the monitoring setup.

Since regular cleansing and disinfection of surfaces are obligatory in food processing environments, the applicability of stickers after washing was tested and compared to swabbing results from surfaces. Swabs showed similar results prior to cleansing, but the yield from the stickers tended to be higher. This might be secondary to the adherent nature of the stickers themselves, which have a good affinity for attaching bacteria. Nevertheless, it must be acknowledged that despite advantages, the sticker itself might have the potential to distribute contaminants, even while studies have shown higher microbial transfer rates from nonporous surfaces (34) and no outbreaks from paper as the source of contamination are documented (23). Reproducible regrowth of L. monocytogenes and E. coli attracted to stickers was observed only at high concentrations (Table 1). However, to circumvent this possible hazard in the future, stickers supplemented with bacteriostatic components might be beneficial. Despite offering promising data, further tests on-site are necessary to complete data sets. Initial experiments on-site did show inconsistencies in sticker compound stability. In some cases, the stickers became dog-eared and detached spontaneously from the adhesive tape. A slight improvement was obtained by prior preparation of the compound on a surface of similar geometry to the door handles to which they were subsequently applied. Yet, as many stickers remained fast without any inconsistencies, this problem is not insurmountable.

Despite the small surface area of stickers, measuring only 0.5 cm², even more positive samples were obtained from them than with swabs. The swabbed area in comparison was at least 10 times greater, demonstrating the capabilities of the sticker system. While the size of the areas to be sampled is suggested to be between 10 and 100 cm² (6), areas greater than 100 cm² (35) or as large as possible clearly carry advantages (7). However, the practical benefit of standardized areas is of less relevance to irregularly shaped fomites. The use of multiple stickers spread widely over tested areas in an array formation should increase the probability of obtaining positive findings in comparison with a single swabbed area.

A newly developed paper sticker system to sample surfaces for microbial contamination and that is suitable for molecular detection methods, has demonstrated several advantages over the sponge-stick swabbing system, despite comparable losses and variation in recoveries. A major advantage of stickers is in handling, as they are easy to distribute and to collect, and no further processing steps, such as centrifugation, are necessary for subsequent DNA extraction. The results also indicate that cleansing and disinfection only slightly impair the results obtained from the stickers, suggesting that prolonged interval sampling should be possible. Additionally, it is not necessary to disinfect monitored surfaces after usage, as should be the case when using sponge-stick swabs. The presented detection system appears to be a promising alternative for effective sampling of bacterial contaminations.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Listeria monocytogenes EGDe and a ΔprfA mutant (36) (part of the collection at the Institute of Food Safety, Food Technology and Veterinary Public Health, Department for Farm Animal and Public Health in Veterinary Medicine, University of Veterinary Medicine, Vienna, Austria) and Escherichia coli TOP 10F′ (Invitrogen, Carlsbad, CA, USA) were grown overnight in tryptone soya broth with 0.6% (wt/vol) yeast (TSB-Y; Oxoid, Hampshire, UK) at 37°C; the optical density at 610 nm (OD₆₁₀) was measured with an HP 8452 spectrophotometer (Hewlett-Packard, Waldbronn, Germany; an OD₆₁₀ of 0.6 approximates 10⁸ CFU/ml).

DNA standard.

As a DNA standard for qPCR quantification, one milliliter of an L. monocytogenes (strain EGDe or ΔprfA mutant) or E. coli overnight culture was used for DNA isolation with the NucleoSpin tissue kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany), following the protocol instructions for Gram-positive bacteria. The DNA was eluted twice with 50 μl double-distilled water (ddH₂O) at 70°C. The DNA concentration was measured using the Qubit double-stranded DNA (dsDNA) broad-range kit (Fisher Scientific, Vienna, Austria). The copy number of the single-copy gene (EGDe, E. coli) or single-integrated internal amplification control was calculated using the DNA molecular weight of L. monocytogenes (1 ng of DNA equals 3.1 × 10⁵ copies of the genome) or E. coli (1 ng of DNA equals 1.8 × 10⁵ copies of the genome).

qPCR.

The prfA qPCR assay for detecting L. monocytogenes was modified after that by Rossmanith et al. (24). One qPCR reaction mixture of 25 μl final volume contained 1× reaction buffer (Fisher Scientific, Vienna, Austria), 3.5 mM MgCl₂, 0.5 μM each primer (Table 4), 0.25 μM each probe (Table 4), 200 μM each dATP, dTTP, dGTP, and dCTP, 1.5 U of Platinum Taq (Fisher Scientific), and 12 μl of template DNA.

TABLE 4.

Primers, probes, and thermal program of qPCR assays

Assay	Primer or probea	Primer or probe sequence (5′–3′)b	Length of product (bp)	PCR program	Reference
prfA assay (L. monocytogenes)	Lip1	GAT ACA GAA ACA TCG GTT GGC	274	94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	37
	Lip2	GTG TAA TCT TGA TGC CAT CAG G		94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	37
	LIP Probe2	FAM-CAG GAT TAA AAG TTG ACC GCA-BHQ1		94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	24
IAC assay (L. monocytogenes ΔprfA)	Lip1	GAT ACA GAA ACA TCG GTT GGC	100	94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	24
	Lip2	GTG TAA TCT TGA TGC CAT CAG G		94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	24
	p-lucLm5	HEX-TTC GAA ATG TCC GTT CGG TTG GC-BHQ1		94°C for 2 min, 45 cycles of 94°C for 15 s and 64°C for 60 s	24
sfmD assay (E. coli)	Ert2F	ACT GGA ATA CTT CGG ATT CAG ATA CGT	106	94°C for 2 min, 50 cycles of 94°C for 15 s and 60°C for 60 s	25
	Ert2R	ATC CCT ACA GAT TCA TTC CAC GAA A		94°C for 2 min, 50 cycles of 94°C for 15 s and 60°C for 60 s	25
	Ert2P	FAM-CAG CAG CTG GGT TGG CAT CAG TTA TTC G-BHQ1		94°C for 2 min, 50 cycles of 94°C for 15 s and 60°C for 60 s	25

^aAll primers and probes were obtained from Eurofins (Ebersberg, Germany).

^bFAM, 6-carboxyfluorescein; BHQ1, black hole quencher 1.

The sfmD qPCR assay for detection of E. coli was modified after that by Kaclíková et al. (25). One qPCR reaction of 25-μl final volume contained 1× reaction buffer, 3.5 mM MgCl₂, 0.3 μM each primer (Table 4), 0.2 μM probe (Table 4), 200 μM each dATP, dTTP, dGTP, and dCTP, 1 U of Platinum Taq (Fisher Scientific, Vienna, Austria), and 12 μl of template DNA.

The qPCR was performed as previously published in an Mx3000p real-time PCR thermocycler (Stratagene, La Jolla, CA, USA) using the thermal programs listed in Table 4, and the analysis was performed with the MxPro software (adaptive baseline settings).

Stickers.

Commercially available text-marking stickers (Markierungspunkte Ø 8 mm, permanent, no. 3013 yellow, 3175 white, 3179 green; Avery of CCL Industries, Inc., Toronto, Canada) were applied to a strip of adhesive tape (tesafilm transparent, 15 mm no. 57370-02; tesa SE, Norderstedt, Germany) using sterile tweezers, followed by sterilization with UV-C radiation for 15 min (Sylvania G30W T8, 10-cm distance; Feilo Sylvania, Erlangen, Germany). The sticker compound was then attached to the surface of interest.

Artificial contamination of stickers.

For artificial contamination of stickers (Fig. 2), bacteria were washed and log-diluted in 1× phosphate-buffered saline (PBS). A 5-μl droplet of the respective bacterial suspension was applied to each sticker to achieve approximately 10, 100, 1,000, or 10,000 CFU per sample. Bacterial suspensions were dried for at least 1 h or until any visible moisture had evaporated, or they were kept at room temperature for 1, 3, 7, or 14 days. After the respective storage times, stickers were transferred into a 1.5-ml Eppendorf tube using sterile tweezers for subsequent DNA extraction. As references, an equivolume inoculum was transferred directly to DNA extraction and to TSA with yeast (TSA-Y) plates to obtain reference values. In parallel, artificially contaminated stickers were incubated for 1 h in 500 μl 1× PBS and vortexed, and the entire supernatant was plated to TSA-Y. Alternatively, stickers were transferred to Half Fraser broth (Biokar Diagnostics, Beauvais, France) or TSB medium and the bacterial growth assessed after 24 h at 30°C or 37°C, respectively.

Experiments were performed at room temperature (22°C to 25°C), and relative humidity levels were between 40 and 60%.

DNA recovery and isolation from stickers.

Stickers detached with sterile tweezers and transferred into 1.5-ml Eppendorf tubes were used directly for DNA extraction with the NucleoSpin tissue DNA extraction kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany) by adding the prelysis buffer on top of the stickers. The original protocol for Gram-positive bacteria was followed with the modification of DNA elution twice with 24-μl ddH₂O (70°C) in order to reduce the volume, yielding a 48-μl elution volume instead of 100 μl.

Sampling with sponge-sticks.

The performance of stickers was compared with sponge-stick swabbing (Sponge-Stick with buffered peptone water broth; 3M, St. Paul, MN, USA). After surface sampling, sponges were soaked with 10 ml 1× PBS and stomached for 2 min. The liquid was centrifuged for 5 min at 8,000 × g, and the obtained pellet subsequently was used for DNA extraction with the NucleoSpin kit (elution twice with 48 μl H₂O at 70°C).

Cleansing and disinfection.

Soapy water was prepared by diluting Exact AC (E. Mayr, Vösendorf, Austria) in water to concentrations commonly using for cleansing surfaces. For disinfection of surfaces, mikrozid AF liquid (Schülke & Mayr, Norderstedt, Germany) was applied by wiping. Two-minute exposure times were also tested for comparison.

Materials.

Line art was drawn with Noodler’s Ink no. 14005 and the manuscript typed on an IBM 1391403.