International Journal of Food Microbiology 127 (2008) 121–128

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j f o o d m i c r o

The detection of food soils and cells on stainless steel using industrial methods: UV
illumination and ATP bioluminescence
Kathryn A. Whitehead ⁎, Lindsay A. Smith, Joanna Verran
School of Biology, Chemistry and Health Sciences, Manchester Metropolitan University, Chester St, Manchester M1 5GD, UK

A R T I C L E I N F O A B S T R A C T

Article history: Open food contact surfaces were subjected to organic soiling to provide a source for transfer of microbial
Received 2 April 2008 cells. Rapid industrial methods used for the detection of residual cells and soil e.g. ATP (adenosine
Received in revised form 23 June 2008 triphosphate) bioluminescence and an ultraviolet (UV) light detection method were assessed for their ability
Accepted 23 June 2008
to detect organic soils, or organic soil–cell mix on surfaces. A range of soils (complex [meat extract, fish
extract, cottage cheese extract]; oils [cholesterol, fish oil, mixed fatty acids]; proteins [bovine serum albumin,
Keywords:
Soil
fish peptones casein]; carbohydrates [glycogen, starch, lactose]); was used. Under UV, oily soils, mixed fatty
Food acids, cholesterol and casein were detected at low concentrations, with detection levels ranging from 1% to
Conditioning film 0.001% for different substances. Glycogen was the most difficult substance to detect at lower concentrations.
ATP bioluminescence Using UV wavelength bands (λ) of 330–380 nm, 510–560 nm and 590–650 nm, wavelength bands of 330–
UV light 380 nm, illuminated most of the soils well, whilst the wavelength band of 510–560 nm illuminated the fish
Surface extract, cholesterol and fatty acids; the 590–650 nm wavelength band illuminated the lactose. Soils at all
concentrations were detected by the ATP bioluminescence method; the complex soils gave the highest
readings. When complex soils were combined with Listeria monocytogenes Scott A or a non-pathogenic
Escherichia coli O157:H7, ATP measurements increased by 1–2 logs. For UV illumination, the L. monocytogenes
and cheese combination was the most intensely illuminated, with E. coli and meat the least.
UV illumination is a simple well established method for detecting food soil, with little change in findings
when microorganisms are included. Performance can be enhanced in certain circumstances by altering the
wavelength. ATP bioluminescence is a proven system for hygienic assessment being especially useful in the
presence of microorganisms rather than organic soil alone.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction materials in the presence or absence of microorganisms will be

Most equipment used for food preparation in the food industry is
constructed from stainless steel. Regular cleaning of the equipment is
required to prevent the build up of adsorbed organic material and
microorganisms. The material present on a given hygienic food
contact surface may contain organic material (food soil), inorganic
material (residue of cleaning agent), and microorganisms. The nature
of this mixture (biological and chemical) of viable and inert
components will vary depending on the environment (Verran and
Whitehead, 2006). In this paper, commercial ‘kits’ for soil detection
were compared.
The development of adsorbed layers, termed conditioning films, on a
surface is considered to be the first stage in biofilm formation
(Chamberlain, 1992). Open food contact surfaces do not normally
provide a solid–liquid interface for microbial attachment, thus a ‘true’
biofilm is unlikely to develop (Verran et al., 2002). However, organic
⁎ Corresponding author. Tel.: +44 161 247 1157; fax: +44 161 247 6365.
E-mail address: k.a.whitehead@mmu.ac.uk (K.A. Whitehead).
transferred onto the surfaces; a process known as soiling. Organic
soiling of a surface will affect cell–substratum interactions, and will
introduce additional cell–soil and soil–substratum interactions (Verran
and Whitehead, 2006). The influence of organic material on substratum
properties and cell attachment and retention will have an impact on
surface fouling and on cleaning regimes.
There are a number of methods available (Verran et al., 2002;
Verran and Whitehead, 2006) (in situ and in vitro) that may be used to
detect and/or quantify food soiling on surfaces, but to date the merits
of a number of these methods for their use in the food industry have
not been directly compared (Davidson et al., 1997; Verran et al., 2002).
Methods used to detect soils include the use of iodine for the detection
of starch, Nile blue for fat (Verran et al., 2002), a visual assessment of
margarine using beta carotene (Anderson et al., 1986) and milk and
casein soil via the Lowry method (Bohner et al., 1991). Rapid in situ
industrial methods include ATP bioluminescence or a UV light
detection method. Microscopy in vitro methods include Scanning
Electron Microscopy (SEM) (Gounga et al., 2007), and epifluorescent
microscopy (Whitehead et al., 2005). Methods used to detect changes
in the surface physicochemistry due to surface soiling include contact

0168-1605/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijfoodmicro.2008.06.019
122 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128

Fig. 1. UV illumination imaging of soiled surfaces. Some components fluoresced brightly (a) when illuminated with UV light whilst others were more difficult to detect (b and c).
a) Fish oil (oils), b) glycogen (proteins) and c) lactose (carbohydrate). Image size 20 mm × 20 mm.

angle (van der Mei et al., 2002), surface free energy, dispersive and houses where some subtypes may be able to persist for months or
polar measurements (Briandet et al., 2001), whilst chemical methods even years (Vogel et al., 2001; Wulff et al., 2006). The potential for
used to determine surface contamination include such methods as ready-to-eat foods to become cross-contaminated with L. monocyto-
Energy Dispersive X-ray (EDX) (Reid et al., 1994) and Fourier genes is well recognized either directly or via surfaces and equipment
Transform Infrared Spectroscopy (FTIR) (Koca et al., 2007; Pappas that have been in contact with raw materials (Mena et al., 2004).
et al., 2008). Clearly, some of these techniques are less suitable for use E. coli O157:H7 can be transmitted to humans through indirect or
on site in the food industry but supporting data generated may help direct contamination of foods (Bouvet et al., 2001). Undercooked
identify the most appropriate of the simpler methods. This, and ground beef and raw milk have been implicated in foodborne infection
current work in our laboratories compares a range of these methods to (Armstrong et al., 1996). Many strains of Shiga toxigenic E. coli are
determine their merits, shortfalls and relationships in terms of soil human pathogens causing illness ranging in severity from mild
detection. diarrhoea to severe renal complications that can result in death (Rivas
ATP bioluminescence has been widely used for the detection of et al., 2007). Cross-contamination during processing and subsequent
microbial contamination and food residues in the food industry handling and preparation of foods leads to the entry of these
(Griffith et al., 1994; Davidson et al., 1997), providing a real time pathogens into the food chain (Hood and Zottola, 1997; Kumar and
estimate of total surface cleanliness including the presence of organic Arand, 1998). With increasing concerns over biotransfer potential and
debris and microbial contamination (Davidson et al., 1999). It has been with the low minimum infectious doses for pathogens such as E. coli
successfully used for determining cell numbers in fish processing O157:H7, the detection of low levels of contamination is becoming
factories (Miettinen et al., 2001) and the dairy industry (Oulahal-Lagsir increasingly important (Davidson et al., 1999).
et al., 2000). The results from the sample are displayed in relative light The aim of this work was to compare two commercially available
units (RLU). When compared to contact agar slide method, results for methods, ATP bioluminescence and a UV light method to determine
ATP bioluminescence were shown to be poorly correlated to the total the nature and limit of detection when used to detect organic material
number of bacteria when used to evaluate the surface contamination in (food soils) and cell–soil fouling on stainless steel substrata.
fish processing factories (Miettinen et al., 2001). Further the ATP
method does not detect all components of a given soil (Lappalainen 2. Materials and methods
et al., 2000). Clearly if ATP is absent one would speculate that the
surface is clean, but some types of residual soil may remain undetected. In this paper, commercial ‘kits’ used for soil detection were
UV light (353 nm) can also be used for the detection of residual compared. The nature of the organic material soiling the surface will
cells and soiling on industrial surfaces (Pedersen, 2007) although (as be related to the food materials present. Thus a selection of complex
with ATP) no distinction is made between the two components. The soils (cheese, fish and meat) was used in this study. These complex
molecular configuration of organic material allows some organic
residues to fluoresce when illuminated by UV light (Adhikari and
Tappel, 1975). Thus, UV light may be used to detect residual soil when
Table 1
work surfaces are illuminated at appropriately 350 nm by highlighting The level of soil concentration (%) detected on stainless steel using UV illumination
areas in an industrial plant that need a more intensive cleaning. This (353 nm)
method is advantageous in that it does not require direct contact with
Organic material (soil) Level of soil detection (%)
the surface (which ATP bioluminescence and total viable counts
Meat 0.1
require). The ability of different wavelengths to detect separate Fish 0.1
components has not been previously explored. Cheese 0.01
Although the presence of soil alone provides an indication of the Cholesterol 0.001
effectiveness of cleaning and hygienic procedures, it is the micro- Fish oil 0.1
Fatty acids 0.001
organisms present that impact significantly on public health. Highly
BSA 0.05
publicized outbreaks of foodborne diseases caused by pathogens such Fish peptone 0.05
as Escherichia coli O157:H7 and Listeria monocytogenes Scott A have Casein 0.001
increased consumer concerns and interest in food safety (Sofos and Glycogen 1
Smith, 1993). L. monocytogenes is commonly isolated from food Starch 0.1
Lactose 0.05
production plants including fish slaughter houses and fish smoke-
 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128 123

Table 2 flocs were seen in the liquid. The liquid was filtered through doubled
UV wavelength bands at which individual organic material (soils) on surfaces were wood pulp (size 4) coffee filters (CO CO-OP, Manchester, UK). Then
most easily detected
0.1 M phosphate buffer was added (7.62 g KH2(PO)4 and 7.66 g K2H
Organic material (soil) Best detection wavelength (PO)4 L− 1). The pH was adjusted to 6.6. The optical density (OD)
Meat 330–380 nm (600 nm) was measured and adjusted to 0.21. The liquid was sterilised
Fish extract 510–560 nm by autoclaving at 100 °C for 30 min. The extract was cooled to 0 °C–
Cheese 330–380 nm
5 °C, which resulted in clearing of the liquid. The liquid was stored at
BSA 330–380 nm
Fish protein 330–380 nm 5 °C.
Casein 330–380 nm
Cholesterol 510–560 nm 2.1.3. Cheese extract
Fish oil 330–380 nm Cheese extract (Langley Farm low fat natural cottage cheese,
Fatty acids 510–560 nm
Glycogen 330–380 nm
Holmforth, W. Yorkshire) was made by draining off the liquid residue
Starch 330–380 nm from the cottage cheese. The remaining cottage cheese was placed
Lactose 590–650 nm into a stainless steel metal sieve with a 2 mm pore size and gently
pressed to remove all the liquid. The liquid was stored at −20 °C until
use when it was defrosted to room temperature.
organic soils are an ill-defined heterogeneous mix of molecules, thus
for these studies a range of chemically defined soil components (oils, 2.1.4. Chemicals
proteins and carbohydrates) were also included in the study. Cholesterol, BSA, glycogen, fish oil, starch, myristic, palmitic,
Three complex soils were used in this work (meat, fish and cheese) stearic, casein and lactose were all obtained from Sigma Aldrich
as were a range of chemically defined soils including proteins (bovine (Dorset, UK). Fish peptone was obtained from Fluka Biochemika
serum albumin (BSA), fish peptones, casein), oils (cholesterol, fish oil, (Switzerland). For the lactose, fish peptone, BSA, starch and glycogen
and a mixture of three fatty acids, myristic [21.79%] palmitic [58.29%] 1 ml of 10% solution was made in an Eppendorf tube using sterile
stearic [19.93%]), and carbohydrates (glycogen, starch and lactose). distilled water. Samples that were difficult to get into solution were
Each soil was made up in 10%, 5%, 1%, 0.5%, 0.1%, 0.05%, 0.01%, 0.005% vortexed for 5 min. The fish oil, mixed fatty acids and cholesterol were
and 0.001% solutions. Meat, cheese and myristic acid were stored at made separately into 10% solutions in Eppendorf tubes using chloro-
−20 °C and were defrosted to room temperature before use. Fish form. The fatty acids myristic (21.79%), palmitic (58.29%) and stearic
extract, BSA, fish oil were stored at 4 °C, whilst the cholesterol, (19.93%) were mixed into a percentage solution to represent the major
palmitic and stearic acids, fish peptone, casein and carbohydrates oils found in cheese (www.foodcomp.dk).
were stored at room temperature. Soils were made up from the
following components. 2.2. Microbiology

2.1. Soiling components
2.1.1. Extraction of meat exudates (method kindly provided by Brigitte
Carpentier AFSSA, France)
One kilogram of fresh rolled beef brisket (CO-OP, UK) was cut into
10 mm × 10 mm pieces. The meat pieces were put into a stainless steel
tray and covered in aluminum foil. The meat was covered by another
tray and weighed down with 8.4 kg of stainless steel sheets and frozen
at −20 °C for 24 h. The meat was defrosted at room temperature and
the meat exudates were poured off and the meat squeezed to recover
surplus exudates. The meat exudates were stored in 20 ml aliquots at
−20 °C until needed.
2.1.2. Fish extract (method kindly provided by Lone Gram, DIFRES,
Denmark)
Fish fillets (cod) were cut into cubes and 500 ml of tap water was
added kg− 1 fish. The fish was boiled for 5 min. The juice was pressed
from the fish flesh and the resulting solution was drained off. The
liquid was boiled for 5 min without stirring and left for 5 min until
The L. monocytogenes Scott A was a kind gift from Professor Lone
Gram (Danish Institute of Fisheries Research (DIFRES), Technical
University of Denmark). A non-pathogenic strain of E. coli O157:H7
was a kind gift from Dr Brigitte Carpentier (Agence française de
sécurité sanitaire des aliments (AFSSA), Maisons-Alfort, France).
Stock cultures were stored at −80 °C. To store at −80 °C, the stock
cultures were grown overnight in either tryptone soya broth (TSB)
(Oxoid, Hampshire, UK) at 30 °C for L. monocytogenes or for E. coli in
brain heat infusion broth (BHIB) (Lab M, Bury, UK) at 37 °C. Equal
amounts of culture and freezing mix were added together and were
incubated at the above temperatures for a least 30 min to 1 h. Samples
were dispensed into 1.5 ml sterile plastic screw capped tubes, and
were frozen. To resuscitate, the culture was defrosted at room
temperature and a loop of liquid was taken and streaked out onto
preferred media or broth and was incubated at the appropriate
temperature overnight. The remainder of the culture was returned to
the freezer for future use. The freezing mix was made of two solutions.
Solution A contained di-potassium hydrogen phosphate (K2HPO4)
12.6 g L− 1, potassium di-hydrogen phosphate (KH2PO4) 3.6 g L− 1 tri-

Fig. 2. Different wavelength bands of UV light on fish oil demonstrating that specific band wavelengths of UV light illuminate the fish oil more brightly. a) 330–380 nm, b) 510–560 nm
and c) 590–650 nm (n = 3).
124 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128

sodium citrate (Na3C6H5O7 × 2H2O) 0.9 g L− 1, ammonium sulphate
(NH4)2SO4) 1.8 g L− 1 and glycerol 300 g L− 1. This mixture was
autoclaved at 121 °C for 15 min. Solution B consisted of 1.8 g L− 1
magnesium sulphate (MgSO4 × 7H2O). The solution was filter sterilised
using a 10 ml Luer-Lok™ syringe (BDH, Poole, UK) and an Acrodisc®
filter (32 mm with 0.2 μm non-pyrogenic supar® membrane, Pall
Corporation, Cornwall, UK). Then 1 ml of solution B was added to
100 ml of solution A to produce the final freezer mix. All chemicals
used in the freezing mix were obtained from BDH (Poole, UK).
In preparation for retention assays, stock cultures of L. mono-
cytogenes were inoculated on tryptone soya agar (TSA) (Oxoid,
Hampshire, UK), and incubated at 30 °C overnight. Cultures were
stored at 4 °C. Ten milliliters of TSB was inoculated with a single colony
of L. monocytogenes and incubated at 30 °C overnight. One hundred
microliters of this culture was used to inoculate 100 ml TSB, which
was incubated at 30 °C for 18 h.
E. coli was inoculated onto brain heat infusion agar (BHIA, Lab M,
Bury, UK). The E. coli was grown at 37 °C for 24 h. Cultures were stored
at 4 °C. Ten milliliters of BHIB was inoculated with a single colony of E.
coli and incubated at 37 °C overnight. One hundred microliters of this
culture was used to inoculate 100 ml of BHIB, which was incubated at
37 °C for 18 h.
Following incubation, cells were harvested at 716 ×g for 10 min and
were washed once, by re-suspension in sterile distilled water,
vortexing for 1 min, and then centrifugation at 716 ×g for 10 min.
Cells were re-suspended to an OD of 1.0 at 540 nm in sterile distilled
water. Colony forming units ml− 1 (cfu ml− 1) were determined by serial
dilution and were 1.62 ± 0.92 × 108 cfu ml− 1 for L. monocytogenes and
1.38 ± 0.68 × 108 cfu ml− 1 for E. coli.
2.3. Surface preparation
2.3.1. Soiling of stainless steel
One hundred microliters of organic material was added to a
50 mm × 40 mm 304 2B finish 2 mm thick stainless steel plate. The
organic material was spread across the surface using a sterile plastic
spreader. Samples were dried at room temperature in a class 2 flow
hood. The top 10 mm of the plate was used for labeling and handling.
When cells were also added to the sample, 50 μl of cells and 50 μl of
soil were mixed together in an Eppendorf tube and were applied to
the surface as above. L. monocytogenes was mixed with the cheese or
the fish extract, whilst E. coli was mixed with the meat exudate. To
visualise cells alone on stainless steel, 100 μl of the prepared cell
suspensions was added to the surface and allowed to dry at room

Fig. 3. ATP results on conditioning films at concentrations of 10%, 5%, 1%, 0.5%, 0.1%. 0.05%, 0.01%, 0.005% and 0.001%. a) Complex materials (meat, fish, cheese), b) proteins (BSA, fish
protein, casein), c) oils (cholesterol, fish oil, fatty acids) and d) carbohydrates (glycogen, starch, lactose). — indicates level above (b29 RLU) in which the surface is not considered
hygienic.
 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128
temperature for 2 h in a class 2 safety cabinet. All samples were stored
at 4 °C.
2.3.2. Preparation of test pieces for SEM
The samples with soils ± cells were immersed in 4% v/v gluter-
aldehyde (Agar, Essex, UK) for 24 h at 4 °C. Samples were thoroughly
rinsed with 100 cm3 distilled H2O using a distilled water bottle at a 45°
angle, with a 3 mm nozzle. Samples were then dried in a class 2 flow
hood. Samples were stored at room temperature in a phosphorous
pentoxide (Sigma Aldrich, Dorset, UK) dessicator. The samples were
fixed to stubs for gold sputter coating, which was carried out using a
Polaron E5100 (Milton Keynes, UK) SEM sputter coater. Samples were
sputter coated at a vacuum of 0.0921 mbar, for 3 min, at 2500 V, in
argon gas at a power of 18–20 mA. Images of substrata were obtained
using a JEOL JSM 5600LV Scanning Electron Microscopy (Jeol Ltd.,
Herts, UK).
2.4. Detection methods
2.4.1. UV detection method
A UV lamp (Labino trac-pack) with a lamp range of 353 nm was
kindly provided by Eigil Appel Pedersen founder of the Bactoforce®
Company (Silkeborg, Denmark). When the Bactoforce light was
unavailable, a standard UV lamp (15 W bench lamp, 365 nm,
1680 μW cm− 2 at 305 mm, 115 VAC. 60 Hz, Cole-Parmer, London,
UK) with an optimum wavelength of 350 nm was used.
2.4.2. UV wavelengths
In order to determine the optimum wavelengths of the UV light to
illuminate the different soils, soiled samples were visualised using an
epifluorescence microscopy (Nikon Eclipse E600, Surrey, UK). Differ-
ent filters, 330–380 nm, 510–560 nm and 590–650 nm were used to
select specific wavelength bands of UV light. The microscope was
mounted with an F-View II black and white digital camera (Soft
Imaging System Ltd., Helperby, UK, supplied by Olympus, Hertford-
Fig. 4. SEM images demonstrating the distribution of cells when added to a fouled soil surface. a) E. coli on naked stainless steel, b) E. coli on stainless steel fouled with meat exudates,
c) L. monocytogenes on naked stainless steel, d) L. monocytogenes on stainless steel fouled with fish extract and e) L. monocytogenes on stainless steel fouled with cheese extract.
125
shire, UK). This system used a Cell F Image Analysis package (Olympus,
Hertfordshire, UK). Dried but unstained samples were visualised
under the differently filtered UV wavelength bands and imaged.
2.4.3. ATP bioluminescence
ATP bioluminescent (Hygiena, Herts, UK) measurements were
carried out as per manufacturer's instructions (Hygeina, 2007); however,
the manufacturer recommends swabbing a 100 mm × 100 mm area, but
in this study, a 40 mm× 40 mm area was swabbed for comparability with
the other experiments carried out in this work. The 40 mm× 40 mm area
of stainless steel was swabbed using the Ultrasnap ATP sampling device
(Herts, UK) and placed into the ATP sample device (Herts, UK) according
to the manufacturer's instructions.
3. Results
Some materials fluoresced more brightly than others (Fig. 1). At a
concentration of 10% w/v or v/v the fish oil (Fig. 1a), and fish protein
fluoresced brightly and whilst the cheese, fatty acids, casein, glycogen
(Fig. 1b), starch and lactose (Fig. 1c) were detected, they did not
fluoresce. The meat exudates, cholesterol, BSA and fish extract were
difficult to detect using this method. Detection levels varied with the
different components (Table 1). The detection levels ranged from 1% to
a low level of 0.001%. Cholesterol, mixed fatty acids and casein were
easily detected with the lowest amounts of organic soiling (0.001%)
although they were not the most fluorescent samples. Glycogen was
difficult to detect at lower concentrations (b1%). No one type of
material (i.e. complex, carbohydrates, proteins or oils) was more easily
detectable in comparison with other groups, although there was
variation between components within the groups.
A range of UV wavelengths was used to identify whether different
soil components could be better illuminated at particular wavelengths
in the UV spectrum. Filter blocks that selected for 330–380 nm, 510–
560 nm and 590–650 nm wavelength bands were used. In some
instances, the different wavelengths of light illuminated certain soils
126 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128
Fig. 5. Soiled stainless steel surfaces illuminated with UV light used to demonstrate the detection of cells mixed with soils. a) Meat extract and E. coli, b) fish extract and L.
monocytogenes and c) cottage cheese extract and L. monocytogenes (n = 3).
more brightly: oils and proteins (Table 2). For example, the
wavelength band of 330–380 nm illuminated the fish oil better than
the other bands of wavelengths tested (Fig. 2). This wavelength was
also best for proteins (Table 2). The 330–380 nm wavelength band
illuminated the majority of the soils tested and was especially good for
the detection of oils.
ATP bioluminescence was used to detect the presence of soiling
films on surfaces in the absence of cells. The ATP hygienic monitoring
system provided rapid hygienic analysis of surfaces contaminated
with biological matter as the monitoring system detected adenosine
triphosphate (ATP). The results from the soiled sample were displayed
in relative light units (RLU), which was the amount of light produced
based on the amount of contamination from the sample. The ratio was
approximately 1 1 RLU to every 1 1 fmol of ATP as determined by the
manufacturer (Hygiena, 2007). There are levels of detection where
b10 was considered a pass, 11–29 a caution and N30 was a fail. The 10%
meat soil was detected to a far greater extent than the other complex
materials, and as expected, the signal decreased with decreasing
concentration of soil. All the soils gave an ATP reading at all levels;
however, the complex materials gave the highest readings especially
at higher concentrations. Of the complex soils the meat and cheese
extracts were the only soiled surfaces to fail at concentrations of 10%
and 5% (Fig. 3).
When soils on the surfaces were combined with L. monocytogenes
Scott A or a non-pathogenic E. coli O157:H7, E. coli cells were
embedded in the soil (Fig. 4b) when compared to cells applied to
their surface on their own (Fig. 4a). L. monocytogenes cells applied to
the naked stainless steel surface (Fig. 4c) showed little difference in
Fig. 6. ATP bioluminescence readings for cell–soil mixes. Meat Ec = meat extract and E.
coli, b) fish Lm = fish extract and L. monocytogenes and c) cheese Lm = cheese extract and
L. monocytogenes.
cell distribution (Fig. 4d) when compared to cells mixed with the fish
extract. However L. monocytogenes cells were embedded in the cheese
soil (Fig. 4e); this soil was very thick and makes differentiation
between cells and soil difficult across the majority of the surface.
When cell–soil surfaces were observed using a regular UV lamp (λ
350 nm), the E. coli and the meat extract (Fig. 5a) were poorly
illuminated, the L. monocytogenes and the fish extract (Fig. 5b) were
visible, whilst the L. monocytogenes and the cheese was most clearly
apparent (Fig. 5c). ATP bioluminescence for the cell–soil mixtures
revealed that when taken as an average no difference was found
between results for different soil and samples (Fig. 6); however, the
range of values for each of the different samples was large. The
presence of microorganisms raised the readings by 1–2 logs but the
samples had been loaded with a high number of cells approximately
6.25 × 105 cfu cm− 2.
4. Discussion
The soils were selected to represent those found in meat, fish and
dairy industries. Three defined soil components for each ‘industry’,
lipids, proteins and carbohydrates were selected to evaluate their
detection by the two methods. Cholesterol is a lipid found in cell
membranes, mixed fatty acids represent those found in cheese and
animal oils, fish oil is self-explanatory. For proteins, BSA was used
since it is a serum albumin protein found in blood and therefore, meat,
fish peptone was used to represent fish proteins, and casein is the
most predominant phosphoprotein in cheese. The carbohydrates
selected were glycogen, found in animal cells, starch made up of
amylase and amylopectin, and lactose, which is found in milk.
Soil components were all detected at all concentrations, however
the defined soils and the fish extract gave a pass reading. These results
support the use of the ATP bioluminescence method but the hygienic
significance of increased soil that would pass in the absence of
microorganisms (and their ATP) requires some thought. In this
research, the ATP bioluminescence method detected soils at the
higher concentrations tested but the presence of a high microbial
contaminant (6.25 × 105 cfu cm− 2) enhanced performance. One
concentration of microorganisms was selected, since the focus of
this work was on the detection of soil and its impact on the detection
of microorganisms. However, since the concentration of cells present
on the surface is of importance, and is one which will invariably affect
the reliability of results, future work will assess the sensitivity of these
methods using different concentrations of microorganisms and
different proportions of microorganisms and soil. The concentrations
 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128
selected will be based on the levels of microorganisms detected on
food surfaces in situ.
UV detection provided a simple, non-invasive, non-quantifiable
method that could not discriminate between cell and soil fouling. Of
the different food components screened, oils were detected particu-
larly well. The UV wavelength band of 330–380 nm was appropriate
for the detection of residual soil in the meat and cheese industry, but a
UV wavelength band of 510–560 nm could be considered for improved
soil detection in the fish industry. Thus, performance of UV detection
could be optimised by using light sources with different wavelengths
in the UV spectra to highly illuminate target soils on food processing
equipment in different environments. It would be interesting to test
these UV wavelength bands of light on soils retained in vitro and in
situ on other surfaces found in the food industry such as rubber and
plastics in order to determine if certain UV wavelengths of light can
detect soiled surfaces whilst reducing the autofluorescence of the
surfaces onto which they are retained.
The need for rapid industrial detection methods such as these is
important in determining the efficiency of cleaning and disinfection
against organic soil and cell retention. Consideration might also be
paid to the nature of soil retention on the substratum. Key to the
selection of soil for this work was the retention of specific soils to a
defined substratum. Wildbrett and Sauerer (1989) showed that the
main fouling component for stainless steel was proteins, whilst fat
was the main fouling component of plastics. Thus, in terms of surface
hygiene, from our results, further work might be prudent in order to
determine whether it might be more appropriate to use stainless steel
rather than polymer materials in a food environment that is fat rich.
As each food processing environment is likely to have a particular
microbial ecology reflecting the food preservation conditions
(Bagge-Ravn et al., 2003), the mixture of microorganisms and soil
selected when testing cleaning and disinfection agents in vitro
should be appropriate to the intended environment of use. In ad-
dition, behavior of microorganisms such as L. monocytogenes is
strongly influenced by the food matrix (Gram et al., 2007). For
complex organic foulants, work by Gram et al. (2007) showed that
sanitizers were extremely efficient against L. monocytogenes attached
from a fish emulsion, whereas only a marginal effect was seen when
the same sanitizer was used against the organism attached from a
meat emulsion.
The detection of organic material is important since food soils
influence cell retention to a surface, and also inhibit cleaning and
disinfection processes. Much work has been carried out on the effect
of proteins on cell attachment, but less consideration has been given
to the effect of lipids and carbohydrates on cell retention. Our work
suggests that although a hygienic reading may be given for a surface
(for example the proteins), the surface may still retain a conditioning
layer of material that is only detectable by other methods. This will
obviously change substratum properties. Although it has been
suggested that generally adsorbed proteins reduce bacterial attach-
ment (Meadows, 1971; Fletcher, 1976; Orstavik, 1977; Fletcher and
Marshall, 1982; Helke et al., 1993; Hood and Zottola, 1997; Barnes
et al., 1999), increased attachment has also been observed (Meadows,
1971; Almakhlafi et al., 1994; Verran and Whitehead, 2006). Inhibition
of cell attachment to surfaces could be due to masking of the surface
chemical groups, which would stop the adhesive interactions with the
bacteria, or due to stearic stabilization (Rutter and Vincent, 1984).
Reduced attachment of cells to surfaces may also be due to proteins in
the bulk fluid phase competing for binding sites on the stainless steel
surface (Flint et al., 2007). However, it must be remembered that the
nature of the effect of the conditioning film appeared to vary with the
organism, substratum and protein under investigation (Barnes et al.,
1999). The increased attachment of cells to rubber and stainless steel
with either whey proteins or lactose has been demonstrated (Speers
and Gilmour, 1985). Lipopolysaccharides inhibited attachment of cells
to substrata if in the presence of cells or adsorbed to the substrata
127
before attachment, whereas dextrans have only demonstrated inhibi-
tion when in the presence of the cells (Pringle and Fletcher, 1986).
Both methods implemented in this study are used for detection only,
and do not indicate the nature of the material (or microorganisms)
present. Under UV, oily soils, mixed fatty acids, cholesterol and casein
were detected at low concentrations, with detection levels ranging from
1% to 0.001% for different substances. A wavelength band of 330–
380 nm, illuminated most of the soils. For combined cell–soil surfaces
the L. monocytogenes and cheese combination was the most intensely
illuminated. Soils at all concentrations were detected by the ATP
bioluminescence method; the complex soils gave the highest readings.
Shortfalls with these methods include detection of the fouled area once
the light has been removed or the area has been swabbed (ATP), there is
no indication as to where the fouling problem areas occurred. However,
these methods provide a very useful screen for cleaning procedures and
to highlight difficult to clean areas. The interactions between substrata,
soil components and detection methods warrant further work.
Acknowledgements
The authors would like to give their sincere thanks to Eigil Appel
Pedersen founder of Bactoforce® for his use of their specialist
equipment and to the partners of Work Package 11 (Hygienic
Processing Systems). This work has been part of and has been funded
by the European PathogenCombat consortium.
References
Adhikari, H., Tappel, A., 1975. Fluorescent probes from irradiated amino acids and
proteins. Radiation Research 61, 177–183.
Almakhlafi, H., McGuire, J., Daeschel, M., 1994. Influence of preadsorbed milk proteins
on adhesion of Listeria monocytogenes to hydrophobic and hydrophilic silica
surfaces. Applied and Environmental Microbiology 60, 3560–3565.
Anderson, M., Huff, H., Marshall, R., Naumann, H., 1986. A simple qualitative method for
detecting cleanliness of food contact surfaces. Journal of Food Protection 49,
342–346.
Armstrong, G., Hillingsworth, J., Morris, J., 1996. Emerging foodborne pathognes:
Escherichia coli O157:H7 as a model of entry of a new pathogen into the food supply
of the developed world. Epidemiology Review 18, 29–51.
Bagge-Ravn, D., Ng, Y., Hjelm, M., Christiansen, J.N., Johansen, C., Gram, L., 2003. The
microbial ecology of processing equipment in different fish industries — analysis of
the microflora during processing and following cleaning and disinfection.
International Journal of Food Microbiology 87, 239–250.
Barnes, L.M., Lo, M.F., Adams, M.R., Chamberlain, A.H.L., 1999. Effect of milk proteins on
adhesion of bacteria to stainless steel surfaces. Applied and Environmental
Microbiology 65, 4543–4548.
Bohner, B., Wilbrett, G., Gmeiner, M., 1991. Cleanability of slip resistant tiles. Tile Brick
International 7, 238–242.
Bouvet, J., Bavai, C., Rossel, R., Le Roux, A., Montet, M., Ray-gueniot, S., Mazuy, C.,
Arquilliere, C., Vernozy-Rozand, C., 2001. Prevalence of vertoxin producing Escher-
ichia coli and E. coli O157:H7 in pig carcasses from three French slaughterhouses.
International Journal of Food Microbiology 71, 249–255.
Briandet, R., Herry, J.M., Bellon-Fontaine, M.N., 2001. Determination of the van der
Waals, electron donor and electron acceptor surface tension components of static
Gram-positive microbial biofilms. Colloids and Surfaces B-Biointerfaces 21,
299–310.
Chamberlain, A.H.L., 1992. Biofilms and corrosion. Biofilms — Science and Technology
223, 207–217.
Davidson, C.A., Griffith, C.J., Peters, A.C., Fielding, L.M., 1997. Sensitivity and
reproducibility of ATP bioluminescence in determining microbial contamination.
Abstracts of the General Meeting of the American Society for Microbiology 97, 462.
Davidson, C., Griffith, C., Peters, A., Fielding, L., 1999. Evaluation of two methods for
monitoring surface cleanliness — ATP bioluminescence and traditional hygiene
swabbing. Luminescence 14, 33–38.
Fletcher, M., 1976. Effects of proteins on bacterial attachment to polystyrene. Journal of
General Microbiology 94, 400–404.
Fletcher, M., Marshall, K.C., 1982. Bubble contact-angle method for evaluating
substratum interfacial characteristics and its relevance to bacterial attachment.
Applied and Environmental Microbiology 44, 184–192.
Flint, S., Walker, K., Waters, B., Crawford, R., 2007. Description and validation of a rapid
(1 h) flow cytometry test for enumerating thermophilic bacteria in milk powders.
Journal of Applied Microbiology 102, 909–915.
Gounga, M.E., Xu, S.Y., Wang, Z., 2007. Whey protein isolate-based edible films as
affected by protein concentration, glycerol ratio and pullulan addition in film
formation. Journal of Food Engineering 83, 521–530.
Gram, L., Bagge-Ravn, D., Ng, Y.Y., Gymoese, P., Vogel, B.F., 2007. Influence of food soiling
matrix on cleaning and disinfection efficiency on surface attached Listeria
monocytogenes. Food Control 18, 1165–1171.
128 K.A. Whitehead et al. / International Journal of Food Microbiology 127 (2008) 121–128
Griffith, C.J., Mathias, K.A., Price, P.E., 1994. The mass media and food hygiene education.
British Food Journal 96, 16–21.
Helke, D.M., Sonners, E., Wong, A.C.L., 1993. Attachment of Listeria monocytogenes and
Salmonella typhimurium to stainless steel and Buna-N in the presence of milk and
individual milk components. Journal of Food Protection 56, 479–484.
Hood, S.K., Zottola, E.A., 1997. Growth media and surface conditioning influence the
adherence of Pseudomonas fragi, Salmonella typhimurium, and Listeria monocyto-
genes cells to stainless steel. Journal of Food Protection 60, 1034–1037.
Hygiena, 2007. Rapid ATP Hygiene Monitoring Guide. Herts, UK.
Koca, N., Rodriguez-Saona, L.E., Harper, W.J., Alvarez, V.B., 2007. Application of Fourier
Transform Infrared Spectroscopy for monitoring short-chain free fatty acids in
Swiss cheese. Journal of Dairy Science 90, 3596–3603.
Kumar, C.G., Arand, S.K., 1998. Significance of microbial biofilms in food industry: a
review. International Journal of Food Microbiology 42, 9–27.
Lappalainen, J., Loikkanen, S., Havana, M., Karp, M., Sjoberg, A.M., Wirtanen, G., 2000.
Microbial testing methods for detection of residual cleaning agents and
disinfectants — prevention of ATP bioluminescence measurement errors in the
food industry. Journal of Food Protection 63, 210–215.
Meadows, P.S., 1971. Attachment of bacteria to solid surfaces. Archiv Fur Mikrobiologie
75, 374.
Mena, C., Almeida, G., Carneiro, L., Teixeira, P., Hogg, T., Gibbs, P.A., 2004. Incidence of
Listeria monocytogenes in different food products commercialized in Portugal.
Food Microbiology 21, 213–216.
Miettinen, H., Aarnisalo, K., Salo, S., Sjoberg, A., 2001. Evaluation of surface
contamination and the presence of Listeria monocytogenes in fish processing
factories. Journal of Food Protection 64, 635–639.
Orstavik, D., 1977. Sorption of Streptococcus-faecium to glass. Acta Pathologica Et
Microbiologica Scandinavica Section B-Microbiology 85, 38–46.
Oulahal-Lagsir, N., Martial-Gros, A., Bonneau, M., Blum, L.J., 2000. Ultrasonic
methodology coupled to ATP bioluminescence for the non-invasive detection of
fouling in food processing equipment — validation and application to a dairy
factory. Journal of Applied Microbiology 89, 433–441.
Pappas, C.S., Tarantilis, P.A., Moschopoulou, E., Moatsou, G., Kandarakis, I., Polissiou, M.G.,
2008. Identification and differentiation of goat and sheep milk based on diffuse
reflectance infrared Fourier transform spectroscopy (DRIFTS) using cluster analysis.
Food Chemistry 106, 1271–1277.
Pedersen, E.A., (2007) Personal communication.
Pringle, J.H., Fletcher, M., 1986. Influence of substratum hydration and adsorbed
macromolecules on bacterial attachment to surfaces. Applied and Environmental
Microbiology 51, 1321–1325.
Reid, G., Davidson, R., Denstedt, J.D., 1994. XPS, SEM and EDX analysis of conditioning
film deposition onto ureteral stents. Surface and Interface Analysis 21, 581–586.
Rivas, L., Fegan, N., Dykes, G.A., 2007. Attachment of Shiga toxigenic Escherichia coli to
stainless steel. International Journal of Food Microbiology 115, 89–94.
Rutter, P.R., Vincent, B., 1984. Physicochemical interaction of the substratum,
microoganisms, and the fluid phase. In: Marshall, K.C. (Ed.), Microbial Adhesion
and Aggregation. Springer, Berlin, pp. 21–38.
Sofos, J.N., Smith, G.C., 1993. The headache of the United States meat industry: E. coli
O157:H7. Meat Focus International 2, 317–325.
Speers, J.G.S., Gilmour, A., 1985. The influence of milk and milk components on the
attachment of bacteria to farm dairy equipment surfaces. Journal of Applied
Bacteriology 59, 325–332.
van der Mei, H.C., White, D.J., Kamminga-Rasker, H.J., Knight, J., Baig, A.A., Smit, J.,
Busscher, H.J., 2002. Influence of dentifrices and dietary components in saliva on
wettability of pellicle-coated enamel in vitro and in vivo. European Journal of Oral
Sciences 110, 434–438.
Verran, J., Whitehead, K.A., 2006. Assessment of organic materials and microbial
components on hygienic surfaces. Food and Bioproducts Processing 84, 260–264.
Verran, J., Boyd, R.D., Hall, K.E., West, R., 2002. The detection of microorganisms and
organic material on stainless steel food contact surfaces. Biofouling 18, 167–176.
Vogel, B.F., Huss, H.H., Ojeniyi, B., Ahrens, P., Gram, L., 2001. Elucidation of Listeria
monocytogenes contamination routes in cold-smoked salmon processing plants
detected by DNA-based typing methods. Applied and Environmental Microbiology
67, 2586–2595.
Whitehead, K.A., Colligon, J., Verran, J., 2005. Retention of microbial cells in substratum
surface features of micrometer and sub-micrometer dimensions. Colloids and
Surfaces B-Biointerfaces 41, 129–138.
Wildbrett, G., Sauerer, V., 1989. Cleanability of PMMA and PP compared with stainless
steel. In: Kessler, H., Lund, D. (Eds.), Fouling and Cleaning in Food Processing.
Proceedings ICFC, pp. 163–177.
Wulff, G., Gram, L., Ahrens, P., Vogel, B.F., 2006. One group of genetically similar Listeria
monocytogenes strains frequently dominates and persists in several fish slaughter-
and smokehouses. Applied and Environmental Microbiology 72, 4313–4322.