research article sampling and homogenization...

Research ArticleSampling and Homogenization Strategies Significantly Influencethe Detection of Foodborne Pathogens in Meat

Alexander Rohde12 Jens Andre Hammerl1 Bernd Appel1

Ralf Dieckmann1 and Sascha Al Dahouk1

1Federal Institute for Risk Assessment Diedersdorfer Weg 1 12277 Berlin Germany2Department of Biology Chemistry and Pharmacy Free University Berlin Takustraszlige 3 14195 Berlin Germany

Correspondence should be addressed to Alexander Rohde alexanderrohdebfrbundde

Received 12 March 2015 Accepted 1 May 2015

Academic Editor Pierre Colin

Copyright copy 2015 Alexander Rohde et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Efficient preparation of food samples comprising sampling and homogenization for microbiological testing is an essential yetlargely neglected component of foodstuff control Salmonella enterica spiked chicken breasts were used as a surface contaminationmodel whereas salami and meat paste acted as models of inner-matrix contamination A systematic comparison of differenthomogenization approaches namely stomaching sonication and milling by FastPrep-24 or SpeedMill revealed that for surfacecontamination a broad range of sample pretreatment steps is applicable and loss of culturability due to the homogenizationprocedure is marginal In contrast for inner-matrix contamination long treatments up to 8 min are required and only FastPrep-24 as a large-volume milling device produced consistently good recovery rates In addition sampling of different regions of thespiked sausages showed that pathogens are not necessarily homogenously distributed throughout the entire matrix Instead inmeat paste the core region contained considerably more pathogens compared to the rim whereas in the salamis the distributionwas more even with an increased concentration within the intermediate region of the sausages Our results indicate that samplingand homogenization as integral parts of food microbiology and monitoring deserve more attention to further improve food safety

1 Introduction

Despite the rise of novel molecular and high-throughputdetection methods the recovery isolation and enumerationof bacterial pathogens in food are still primarily basedon culture techniques the current gold standard in foodmicrobiology [1 2] The continuing dominance of traditionalmicrobiological detection methods in foodstuff control isattributed to the goal to prove the absence or presenceof living pathogenic bacteria which is indispensable toassess the actual health hazard for consumers The merepresence of bacterial DNA which represents the target formany rapid techniques like PCR cannot predict the risk ofinfection However the formation of visible colonies requiresthe successful recovery of the target bacteria out of a foodmatrix in a viable and replication-competent state Thussample preparation is critical for the successful subsequent

microbiological detection and has to be adapted to therespective food matrix [3 4]

The initial extraction of the pathogen is usually per-formed by applying mechanical forces of varying magnitudeto homogenize the food matrix [5 6] In addition to simpleprocedures such as vortexing or manual release varioustechnical solutions are commercially available Peristalticblenders like the Stomacher or alternatively the Pulsifierare probably the most prominent ones for microbiologicaldetection [7ndash9] The widely used Stomacher consists of twoquickly movable paddles which disperse the input foodsamplebuffer mix for enrichment via cultivation Otherhomogenization methods using beads to mill food or theapplication of ultrasound might be also employed for thispurpose Such applications may be also used for the dis-ruption of cells to release proteins or DNA for moleculardetection techniques or further downstream purification

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 145437 8 pageshttpdxdoiorg1011552015145437

2 BioMed Research International

processes [10 11] Many novel systems (eg FastPrep-24)also offer the possibility of adapting the homogenizationdevice to the food matrix by adding further components likequartz sand or beads of variable sizes The effects of thesediverging approaches of sample pretreatment on cell viabilityand test sensitivity have been insufficiently investigated sofar Furthermore the instructions in the highly standardizedand widely accepted ISO standards for the identification ofmicrobes in food are usually rather vague and unspecific withrespect to the sampling process as well as the sample pretreat-ment and homogenization (in contrast to the downstreamdetection procedures) Thus the current use mainly dependson the availability of the devices mentioned above as wellas on personal preferences and rarely considers the physicalproperties of the foodmatrix Due to limited resources somelaboratories may entirely rely on manual homogenization orsimple vortexing

Bacterial pathogens such as Salmonella enterica mightbe located on the surface of a food product due to crosscontamination during slaughtering in case of meat or duringharvest and subsequent transport In contrast processedproducts like sausages or cheese can get contaminated insidethe food product during the production process [12ndash14] Littleis known about the nature of the microbial burden whetherit is evenly distributed throughout the entire product orwhether a microbial gradient towards the surface is presentIn the latter case an arbitrarily taken sample might causefalse-negative results Likewise the use of different homog-enization approaches for the extraction of inner microbialcontamination and the microbial survival rate after theimposed shear forces have not been systematically comparedand neither has the physical detachment of bacteria fromthe surface been evaluated [15 16] In case of Salmonellacontaminated fresh produce as well as pathogenic bacteria onfish studies indicate the significance of sample preparation[17ndash19]

In this study the applicability of four differentmechanicalhomogenizing devices (stomaching by Bagmixer 400millingby FastPrep-24 or SpeedMill and sonication by the BransonSonifier Table 1) for pathogen isolation and conventionaldetection by cultivation for processed and unprocessed meatproducts was evaluated As a proof of principle usingSalmonella enterica surface contamination was establishedon chicken breasts and inner-matrix contamination wasestablished in pork sausages of soft and hard consistenceThe microbial survival rate of this Gram-negative pathogenand the recovery success of each method were assessedAdditionally the influence of sample taking on the testoutcome was investigated

2 Materials and Methods

21 Bacterial Strains and Growth Conditions Exemplarily forGram-negative Enterobacteriaceae the Salmonella entericaserovar Typhimurium (S enterica) reference strain DSM11320 (DSMZ Braunschweig Germany) was used for allexperiments S enterica was cultivated under aerobic con-ditions at 37∘C in lysogeny broth (LB) medium For spikingexperiments aliquots of overnight cultures were transferred

into fresh LB and cultivated under rotational shaking (GFLshaking incubator 3033 180ndash200 rpm)until an optical density(OD588

) of 05 was reached These cultures were seriallydiluted using 1 buffered peptone water (wvol) Bacterialtiters were enumerated after plating 100120583L of the dilutionsteps on selective media (Xylose Lysine Deoxycholate (XLD)agar Oxoid Wesel Germany) and incubation for 22ndash26 h at37∘C

22 Spiking of Food Products Two types of meat contamina-tion were simulated surface contamination and inner-matrixcontamination in which the pathogen can be distributedthroughout the entire food matrix In all food samples theabsence of Salmonella prior to spiking was confirmed accord-ing to DIN EN ISO 65792002 (microbiology of food andanimal feeding stuffs-horizontal method for the detection ofSalmonella spp) Only sporadic occurrence of other bacteriasuch as Serratia Hafnia or Citrobacter was found

For artificial surface contamination chicken breasts pur-chased at local supermarkets (Berlin Germany) in Jan-uary 2014 were used One or two cubical-shaped pieces ofchicken breast were cut using sterile equipment and razorsto obtain meat samples with a weight of 4 g (BagMixer 400Interscience Germany) 3 g (FastPrep-24 (MP BiomedicalsFrance) and sonication) and 015 g (SpeedMill Analytik JenaAG Germany) On the surface of the samples peptone watercontaining Salmonella (volume addition of 360 120583L for stom-aching 270120583L for FastPrep-24 and sonication and 135 120583Lfor SpeedMill) was evenly applied to obtain spiked sampleswith a final bacterial load of approximately 3 times 105 CFUgcorresponding to 3 times 104 CFUmL in the homogenate Thesesamples were then incubated for 1 h at 4∘C Meat sampleswithout artificial contamination were used as negative con-trols

For the simulation of inner-matrix contaminationcoarsely ground smoked or air-dried salami with a highdegree of hardness and German Mettwurst (meat paste) afinely ground spreadable sausage of soft consistency wasproduced in the technology facilities of the Federal Institutefor Risk Assessment These sausages made of lean porkand bacon were spiked under BSL-2 conditions withinthe production process with Salmonella in six differentconcentrations ranging from 1CFUg to 108 CFUg Onepreparation without Salmonella was used as negative controlBriefly 37 kg of meat (24 kg lean pork 13 kg bacon) wasmixed with 50mL of Salmonella solution containing therespective pathogen concentration while being mincedand flavoured (90 g nitrite salting mix 12 g paprika and8 g black pepper) in an automated meat cutter (HFMFleischereimaschinen Germany) Smoked salamis werecured for five days at 18∘C in the Bastra MC 500 GermanMettwurst was cured for 2 h and afterwards allowed to ripenfor six days at 18∘C before being stored at minus20∘C until use

23 Homogenization Procedure and Sausage Sampling Toevaluate the efficacy of various methods to homogenizemeat and meat products four devices with diverging tech-nical approaches were systematically compared (Table 1)

BioMed Research International 3

Table 1 Properties of the chosen homogenization devices

Method Stomaching(Bagmixer 400) FastPrep-24 SpeedMill Branson Sonifier

450

Principle Blending bymovable paddles Bead-mediated milling Bead-mediated milling Sonication

Handling + + + +minusPortability and on-site usage minus minus + minus

Adaptability to different matrices +minus ++lowast

++lowast +minus

Current usage for detection by cultivation ++ minus minus minus

Parallel sample preparation minus+

(2ndash48)lowastlowast+

(2ndash20) minus

Suitability for high volumeslowastlowastlowast ++(lt400mL)

+(lt50mL)

minus

(lt2mL)+

(lt50mL)Available volume rangelowastlowastlowast +minus + +minus +Avoidance of heat generation + +minus +minus +minusPerformance in this study

Surface contamination + + + +Inner-matrix contamination Variable + +minus minus

++ excellent + good +minus ambiguous and minus poorlowastVarious matrix-specific kits and beads for sample preparation are commercially availablelowastlowastThe parallel preparation of 48 samples is only possible for volumes smaller than 2mL Two samples can be homogenized simultaneously for the highestvolume inputlowastlowastlowastExact volumes depend on the sizes of the used bags BD Falcon tubes and lysis tubes

The SpeedMill is a small portable milling apparatus suitablefor on-site sample processing and easily adaptable to differentsample types Likewise FastPrep-24 is a large-volumemillingtool (adaptable also to small-volume samples) which alsooffers a broad range of modular adaptations by adding quartzsand or beads Stomaching was chosen because it is one of themost frequently usedways to prepare food samples Stomach-ing is considered as a gentle homogenization method sinceheat development is marginal and the peristaltic movementsof the paddles distribute the input energy on a large areareducing potential peaks in shear forces Finally sonication isa rather old but simple method which exerts a distinct kindof mechanical force compared to stomaching or milling

All food samples were diluted 1 10 in buffered peptonewater (thus 36mL for stomaching 27mL for sonication andFastPrep-24 135mL for homogenization in the SpeedMill)For stomaching samples were placed in sterile stomacherbags (BagPage 400mL Interscience) Homogenization wasperformed at the highest intensity (paddle distance 7mm)at indicated time intervals Milling was performed with theFastPrep-24 system at a velocity of 5ms (500Watts) andwiththe SpeedMill (150 Watts) In case of the FastPrep-24 deviceBD Falcon 50mL tubes (BD Biosciences Germany) werefilled with three ceramic beads with a size of 635mm (1410158401015840Ceramic Sphere MP) For the SpeedMill 2mL innuSpeedLysis tubes E with 24ndash28mm ceramic spheres (AnalytikJena) were employed Sonication was performed under con-tinuous cooling with circulating chilled water at the highestintensity (output power 400Watts) in BD Falcon 50mL tubesplaced in the water-filled cup horn of a Branson Sonifier 450(Branson Ultrasonics) An initial aliquot was taken beforehomogenization after vortexing for 20 s Further aliquots

were taken after 30 s and 1 2 4 and 8min of homogenizationAs the FastPrep-24 system requires a cooldown period of5min after 1min of homogenization food samples werecooled on ice for this time period after each minute ofhomogenization All aliquots were diluted appropriately inbuffered peptone water and plated twice in suitable dilutionsteps on selectivemedium After an incubation period of 24 hat 37∘C colonies on XLD were enumerated To monitor theaccompanying flora of the food products 100 120583L of sampleswas also plated on LB agar and cultivated under aerobic andmicroaerobic conditions Ambiguous colony morphologiesand sporadically observed putatively accompanying flora onXLD agar plates were analyzed by the MALDI Biotyper(Bruker Germany) following the instructions of the manu-facturer All homogenization experiments were performed inthree independent tests

To compare the different homogenization devices forthe inner-matrix contamination whole cross sections of thesausages were used covering all regions of the sausages Todetermine the spatial distribution of Salmonella within thesausages four different regions were investigated the innercore (A) the outer rim (B) whole cross sections covering theentire area of the sausage (C) and the intermediate region(D) without outer rim and inner core For the inner coresausage slices were cut and circular center pieces with adiameter of approximately 7mmwere taken For the rim theouter 3mm of the sausages was used A total of 3 g out ofeach region was diluted 1 10 in buffered peptone water andhomogenized via 8min of FastPrep-24 treatment as describedabove All distribution experiments were again performed inthree independent tests


24 Statistical Analyses Two-tailed unpaired Studentrsquos 119905-testwas used to evaluate the significance of the results assumingunequal variance between the two compared groups119875 valuesbelow 005 were considered as significant All data are givenas means with standard deviations

3 Results and Discussion

31 Surface Contamination To compare and evaluate thefour different homogenization approaches (Table 1) we firstdetermined the release of bacteria from chicken breastsurfaces artificially spiked with S enterica Immediatelyafter 20 s of vortexing considerable numbers of Salmonellawere already found in the medium without homogenizationHowever 8min of homogenization resulted in an increaseof the number of bacteria released after stomaching andFastPrep-24 treatment whereas sonication and SpeedMilltreatment led to a decrease in pathogen recovery compared tothe CFU counts obtained after an initial rinse and vortexingstep (Figure 1) Independent of the applied method therewas no significant difference between the recovery rates ofbacteria from homogenized and nonhomogenized samples(119875 gt 005) However the intermethod comparison revealeda rather good performance of both FastPrep-24 and stom-aching on the one hand and CFU losses by sonication andSpeedMill on the other hand Notably the different methodsdiffered substantially in the degree of sample disintegrationas well as in the amount of generated foam which hampersaccurate pipetting SpeedMill and to a lesser extent FastPrep-24 produced the greatest amounts of debris and foam whilesonication causes only a slow increase in liquid turbidity andonly limited fragmentation of the food samples

To rule out that the homogenization procedure by itselfexhibits a negative effect on the viability of S entericathereby explaining the inferior results of SpeedMill andsonication we investigated the effects on pure Salmonellacultures diluted in buffered peptone water No CFU loss wasfound for SpeedMill even after 8min of homogenization incontrast sonication of pure cultures for 8min diminishedthe CFU count by roughly one-fourth in accordance withthe reduction seen for the surface contamination of chickenbreast This might indicate that the inferior performance ofsonication is indeed a result of slow bacterial killing whileSpeedMill treatment does not affect bacterial viability in purecultures but seems to be unable to separate bacteria efficientlyfrom the food surface The results obtained for S entericaare probably valid for other disease-causing members of theEnterobacteriaceae like Escherichia coli or Yersinia spp Inaddition it is likely that smaller bacteria like Campylobacteror Gram-positive pathogens like Listeria monocytogenes arealso not affected in their viability after extensive homoge-nization because mechanical shear forces in general are moreharmful for larger objects than for smaller ones Howeversmaller beads and sphere materials other than ceramic mightexhibit much more unfavorable shear forces In accordancewith this assumption the manufacturers of FastPrep-24 andSpeedMill MP Biomedicals and Analytik Jena recommendthe use of considerably smaller beads to lyse bacteria forsubsequent molecular detection methods

00

05

10

15

20

25

CFU counts withouthomogenization

FastP

rep-

24

Spee

dMill

Stom

achi

ng

Soni

catio

n

CFU

8min

CFU

0min

Figure 1 Changes in pathogen detection of chicken surface con-tamination after homogenization The indicated bars express thenormalized pathogen concentrations released from spiked chickenbreast samples after 8min of homogenization in relation to the CFUcount after 20 s of sole vortexing

In contrast to inner-matrix contamination for whichonly scarce information is available so far it has beenreported that for surface contamination extended periodsof stomaching and other homogenization methods did notsignificantly enhance the detection of the pathogen comparedto simple rinsing procedures [11 15 20] It was suggested bySharpe and others that this might be the result of a ldquomassactionrdquo effect which prevents a higher pathogen release afterequilibrium between the liquid phase and the food surfacehas been reached Although this hypothesis has not yet beenproven it illustrates that longer homogenization periods andharsh homogenization methods are not necessary for surfacecontamination instead soaking hand-massaging or the useof swabs for those food products might be equally applicableor even better procedures [6 18 21 22]

32 Inner-Matrix Contamination To elucidate the efficiencyof mechanical disruption for the detection of inner-matrixcontamination two sausages with an artificial Salmonellacontaminationwere producedDuring the production ripen-ing and storage of the spiked sausages the number ofpathogens in the sample dropped by 3 to 4 log units Inthe salamis slightly higher concentrations of S entericacompared to the meat paste were found despite being spikedin equal amounts which might be a result of the shortercuring and ripening period (in total five days for the salamisand six days for the meat paste) For the homogenizationexperiments sausages primarily spiked with 108 CFUg werechosen to yield sufficient CFU counts for an adequateinterpretation In contrast to surface contamination thetwo sausage types showed substantial differences in termsof pathogen release (Figure 2) For the soft finely mincedmeat paste FastPrep-24 the large-volume milling deviceshowed a superior performance extracting seven timesmore pathogens after 8min than stomaching (Figure 2 leftcolumn) The low-volume SpeedMill enabled intermediateextraction success whereas sonication which was unableto substantially break up the sausage matrix yielded noCFUs at all (data not shown) Interestingly independent


Meat paste8000

4000

00 1 2 4 8

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

FastP

rep-

24

8000

4000

00 1 2 4 8

Stom

achi

ng

15000

10000

5000

00 1 2 4 8

Stom

achi

ng15000

10000

5000

00 1 2 4 8

Spee

dMill

Salami15000

10000

5000

00 1 2 4 8

FastP

rep-

248000

4000

00 1 2 4 8

Spee

dMill

Homogenization t (min)


Homogenization t (min) Homogenization t (min)



Figure 2 Homogenization of inner-matrix contamination Release of Salmonella from whole cross sections of internally contaminatedsausages after pretreatment by FastPrep-24 stomaching and SpeedMill for 0 30 s and 1 2 4 and 8min was monitored

of the system the longer the homogenization period waschosen the higher the number of detectable CFUs wasalthough a burst of pathogen release was already foundafter 30 s of homogenization Simple vortexing and rinsingprocedures (before homogenization 119905 = 0) yielded noCFUs demonstrating that this procedure is only suitablefor surface contamination For the hard coarsely groundsalami FastPrep-24 treatment and stomaching showed acomparable release of S enterica (Figure 2 right column) Incontrast to themeat paste no initial burst of pathogen releasewas detected Instead after a short lag phase a continuousnearly linear release rate of bacteria was measured Aftershorter homogenization periods remarkably less bacteriawere extracted than after longer treatment for example thenumber of extracted bacteria was 10-fold less after 1mincompared to 8min Similar to the meat paste SpeedMilltreatment of the salamis was rather ineffective and sonicationwas not able to homogenize the matrix effectively In bothsausages the amount of accompanying flora was marginalindicating that the influence of other bacteria did not playa major role A third sausage type smoked salami showedresults similar to the air-dried salami However due to thevery low Salmonella concentration after the smoking process

(data not shown) a statistically reliable comparison was notpossible

For routine examinations of food products homogeniza-tion is usually performed for a rather short duration (eg 1-2min) to save time and by using soft mechanical treatment(stomaching or blending at low intensities) to avoid loss ofbacterial viability However the conventional mild samplepreparation is not necessarily preferable because the results inFigure 2 show that longer treatments and harsher conditionsare beneficial to determine inner-matrix contamination anddo not affect the bacteria Interestingly stomaching yieldedsignificantly different results for the salami and the meatpasteThe inferior performance of stomaching for Salmonellaextraction out of the meat paste may result from the veryhigh reduction ratio and thus a very small meat particle sizeof the meat paste The meat was spiked during the mincingprocess and Salmonella therefore sticks to the surface ofthese particles In the coarsely ground salami the meatparticle size during the spiking procedure was much biggerand as a consequence in comparison to the meat paste theaccessible surface area was much smaller To transfer thepathogen to the dilution medium it is necessary to bring thepathogen in contact with the liquid phase by homogenizing


Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate

Figure 3 Pathogen distribution within meat paste and salami The schematic drawings on top of the bars indicate the examined region ofthe sausages (black) Pathogen concentrations were determined after 8min of FastPrep-24 treatment and are given in relation to the totalconcentration in whole cross sections

the solid matrix Consequently harsh mechanical treatmentslike the FastPrep-24 system which enable the rapid exchangebetween the particulate solid phase and the liquid phasemight bemuchmore effective for finelyminced sausages (egmeat paste) with a high particulate surface area whereasin case of more granular matrices with small overall surfaceareas FastPrep-24 and stomaching perform similarly In linewith these considerations a recent study with tomatoesinternally contaminated with S enterica showed that ashort but harsh blending step was more effective than alonger stomaching treatment [19] It would be interesting toevaluate whether suggested alternatives to stomaching likethe Pulsifier which was reported to perform equally forsurface contamination while generating less debris [8 9 23]provide different results

33 Analysis of Inner-Matrix Pathogen Distribution Notonly homogenization but also the region of sampling mightinfluence the detection of a pathogen Although the pathogenhas been evenly distributed throughout the entire meat massduring the production process it is possible that this evendispersal might change during the sausage ripening There-fore we divided the sausages into different regions coveringthe outer rim the core and whole cross sections of the foodproducts and determined the concentration of S enterica ineach region (Figure 3) For homogenization the FastPrep-24treatment was chosen because this system had shown a goodperformance in both sausage typesThe core of themeat pastecontained up to fourfold higher concentrations of Salmonellathan the outer rim demonstrating a sharp gradient towardsthe center of the sausage matrix In contrast the distributionof S enterica in the salami was relatively homogenous andno bacterial enrichment in the core was identified Howeverthe concentration of Salmonellawas slightly increased withinthe intermediate region of this sausage type compared to rimand core Further investigations are necessary to determine

whether the pathogen distribution within the meat paste isbased on the production process or on specific parameterslike rigidity water content ingredients or grain size of themeat

Standard operation procedures are rather vague in termsof sample taking in general they require ldquorepresentativerdquosamples which can be interpreted as whole cross sectionsin case of sausages However our study shows that theassumption of a homogenous pathogen distribution withinthe matrix is not necessarily realistic For the meat pastewe identified surprisingly high differences in pathogen loadpresumably resulting from a prolonged survival in thesausage core which might be a result of a different wateractivity or pH both major factors for pathogen inactivation[24 25] Different types of sausages or in general any foodproduct with a suspected inner-matrix contamination (thusincluding not only meat products but also fresh produce ordairy products) might have a characteristic distribution (andsurvival) pattern for a particular bacterial species includingthe formation of bacterial aggregates [26ndash29] Thus it mightbe worthwhile to elucidate the pathogen distribution in otherfood products since a more sophisticated and risk-basedsampling of food regions with higher bacterial loads mightenable better detection limits

4 Conclusions

The results presented in this work demonstrate the piv-otal role of sampling and homogenization for the reliabledetection of pathogens in specific food products It becomesevident that the general lack of precise advices regardingsample pretreatment might be responsible for considerableinterlaboratory differences in pathogen detection This notonly is important for microbiological investigations but alsomight be suitable as a general reference point for otherwhole-cell detection methods for example fluorescence


in situ hybridization [30] However choosing the appropriatehomogenization device should consider not only the effi-ciency of detection but also the ease of handling costsand high-throughput capabilities Summing up to enableimproved pathogen detection methods standardized andharmonized sample preparation protocols are needed fordifferent food matrices

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The work of Alexander Rohde and Jens Andre Hammerl wassupported by a grant of the Federal Ministry of Educationand Research and was executed within the framework ofthe project ZooGloW (FKZ 13N12697) The authors thankDirk Meyer and Niels Bandick for technical support in theproduction of the artificially contaminated sausages

References

[1] T A McMeekin Ed Detecting Pathogens in Food WoodheadPublishing Cambridge UK 1st edition 2003

[2] B Ge and J Meng ldquoAdvanced technologies for pathogen andtoxin detection in foods current applications and future direc-tionsrdquo Journal of the Association for Laboratory Automation vol14 no 4 pp 235ndash241 2009

[3] T S Hammack I E Valentin-Bon A P Jacobson and W HAndrews ldquoRelative effectiveness of the Bacteriological Analyti-cal Manual method for the recovery of Salmonella from wholecantaloupes and cantaloupe rinses with selected preenrichmentmedia and rapid methodsrdquo Journal of Food Protection vol 67no 5 pp 870ndash877 2004

[4] W H Andrews and T S Hammack ldquoFood sampling andpreparation of sample homogenaterdquo in Bacteriological Ana-lytical Manual US Food and Drug Administration AOACInternational Rockville Md USA 2001

[5] M Kanki J Sakata M Taguchi et al ldquoEffect of samplepreparation and bacterial concentration on Salmonella entericadetection in poultry meat using culture methods and PCRassaying of preenrichment brothsrdquo Food Microbiology vol 26no 1 pp 1ndash3 2009

[6] K H Seo R E Brackett I E Valentın-Bon and P SHolt ldquoComparison of homogenization methods for recoveringSalmonella Enteritidis from eggsrdquo Journal of Food Protectionvol 66 no 9 pp 1666ndash1669 2003

[7] A N Sharpe and A K Jackson ldquoStomaching a new concept inbacteriological sample preparationrdquo Applied Microbiology vol24 no 2 pp 175ndash178 1972

[8] D-H Kang R H Dougherty and D Y C Fung ldquoComparisonof Pulsifier119879119872 and Stomacher119879119872 to detach microorganismsfrom lean meat tissuesrdquo Journal of Rapid Methods and Automa-tion in Microbiology vol 9 no 1 pp 27ndash32 2001

[9] V C H Wu P Jitareerat and D Y C Fung ldquoComparison ofthe pulsifier and the stomacher for recovering microorganismsin vegetablesrdquo Journal of Rapid Methods amp Automation inMicrobiology vol 11 no 2 pp 145ndash152 2003

[10] A N Sharpe and D C Kilsby ldquoUltrasound and vortex stirringas bacteriological sampling methods for foodsrdquo Journal ofApplied Bacteriology vol 33 no 2 pp 351ndash357 1970

[11] A N Sharpe ldquoSeparation and concentration of samplesrdquo inDetecting Pathogens in Food T A McMeekin Ed pp 52ndash68Woodhead Publishing Cambridge UK 1st edition 2003

[12] A S Gounadaki P N Skandamis E H Drosinos and G-J E Nychas ldquoMicrobial ecology of food contact surfacesand products of small-scale facilities producing traditionalsausagesrdquo Food Microbiology vol 25 no 2 pp 313ndash323 2008

[13] M Mataragas P N Skandamis and E H Drosinos ldquoRiskprofiles of pork and poultry meat and risk ratings of variouspathogenproduct combinationsrdquo International Journal of FoodMicrobiology vol 126 no 1-2 pp 1ndash12 2008

[14] W Rivera-Perez E Barquero-Calvo and R Zamora-SanabrialdquoSalmonella contamination risk points in broiler carcassesduring slaughter line processingrdquo Journal of Food Protectionvol 77 no 12 pp 2031ndash2034 2014

[15] J F Hannah J A Cason J R Richardson et al ldquoEffect ofstomaching on numbers of bacteria recovered from chickenskinrdquo Poultry Science vol 90 no 2 pp 491ndash493 2011

[16] J M Jay and S Margitic ldquoComparison of homogenizingshaking and blending on the recovery of microorganisms andendotoxins from fresh and frozen ground beef as assessed byplate counts and the Limulus amoebocyte lysate testrdquo Appliedand Environmental Microbiology vol 38 no 5 pp 879ndash8841979

[17] A P Jacobson V S Gill K A Irvin H Wang and T SHammack ldquoEvaluation of methods to prepare samples of leafygreen vegetables for preenrichment with the bacteriologicalanalytical manual Salmonella culture methodrdquo Journal of FoodProtection vol 75 no 2 pp 400ndash404 2012

[18] P C Nedoluh S Owens E Russek-Cohen and D C WesthoffldquoEffect of sampling method on the representative recoveryof microorganisms from the surfaces of aquacultured finfishrdquoJournal of Food Protection vol 64 no 10 pp 1515ndash1520 2001

[19] H Wang V S Gill K A Irvin et al ldquoRecovery of Salmonellafrom internally and externally contaminated whole tomatoesusing several different sample preparation proceduresrdquo Journalof AOAC International vol 95 no 5 pp 1452ndash1456 2012

[20] A N Sharpe ldquoFood sample preparation and enrichment forrapid detectionrdquo in Rapid Detection Assays for Food and WaterS A Clark K CThompson CWKeevil andM S Smith Edspp 129ndash137 Royal Society of Chemistry London UK 2001

[21] A N Sharpe C I B Kingombe P Watney L J Parrington IDudas and M P Diotte ldquoEfficient nondestructive sampler forcarcasses and other surfacesrdquo Journal of Food Protection vol 59no 7 pp 757ndash763 1996

[22] A P Jacobson T S Hammack andWH Andrews ldquoEvaluationof sample preparation methods for the isolation of salmonellafrom alfalfa and mung bean seeds with the bacteriologicalanalytical manualrsquos Salmonella culture methodrdquo Journal ofAOAC International vol 91 no 5 pp 1083ndash1089 2008

[23] D Y C Fung A N Sharpe B C Hart and Y Liu ldquoThepulsifier a new instrument for preparing food suspensionsfor microbiological analysisrdquo Journal of Rapid Methods andAutomation in Microbiology vol 6 no 1 pp 43ndash49 1998

[24] R Lindqvist and M Lindblad ldquoInactivation of Escherichia coliListeria monocytogenes and Yersinia enterocolitica in fermentedsausages during maturationstoragerdquo International Journal ofFood Microbiology vol 129 no 1 pp 59ndash67 2009


[25] M Mataragas K Rantsiou V Alessandria and L CocolinldquoEstimating the non-thermal inactivation of Listeria monocy-togenes in fermented sausages relative to temperature pH andwater activityrdquoMeat Science vol 100 pp 171ndash178 2015

[26] H Korkeala and P Makela ldquoCharacterization of lactic acidbacteria isolated from vacuum-packed cooked ring sausagesrdquoInternational Journal of Food Microbiology vol 9 no 1 pp 33ndash43 1989

[27] H U Liepe ldquoKeimverteilung in einer Rohwurstrdquo Fleischwirt-schaft vol 67 no 1 pp 1266ndash1267 1987

[28] U Paulat Einfluss der Probenentnahme auf das Ergebnis derbakteriologischen Analyse am Beispiel des FleischerzeugnissesRohwurst University of Veterinary Medicine Hannover Foun-dation Hannover Germany 1999

[29] JWWimpenny L Leistner L VThomas A JMitchell K Kat-saras and P Peetz ldquoSubmerged bacterial colonies within foodandmodel systems their growth distribution and interactionsrdquoInternational Journal of Food Microbiology vol 28 no 2 pp299ndash315 1995

[30] A Rohde J A Hammerl B Appel R Dieckmann andS Al Dahouk ldquoFISHing for bacteria in foodmdasha promisingtool for the reliable detection of pathogenic bacteriardquo FoodMicrobiology vol 46 no 1 pp 395ndash407 2015

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


processes [10 11] Many novel systems (eg FastPrep-24)also offer the possibility of adapting the homogenizationdevice to the food matrix by adding further components likequartz sand or beads of variable sizes The effects of thesediverging approaches of sample pretreatment on cell viabilityand test sensitivity have been insufficiently investigated sofar Furthermore the instructions in the highly standardizedand widely accepted ISO standards for the identification ofmicrobes in food are usually rather vague and unspecific withrespect to the sampling process as well as the sample pretreat-ment and homogenization (in contrast to the downstreamdetection procedures) Thus the current use mainly dependson the availability of the devices mentioned above as wellas on personal preferences and rarely considers the physicalproperties of the foodmatrix Due to limited resources somelaboratories may entirely rely on manual homogenization orsimple vortexing

Bacterial pathogens such as Salmonella enterica mightbe located on the surface of a food product due to crosscontamination during slaughtering in case of meat or duringharvest and subsequent transport In contrast processedproducts like sausages or cheese can get contaminated insidethe food product during the production process [12ndash14] Littleis known about the nature of the microbial burden whetherit is evenly distributed throughout the entire product orwhether a microbial gradient towards the surface is presentIn the latter case an arbitrarily taken sample might causefalse-negative results Likewise the use of different homog-enization approaches for the extraction of inner microbialcontamination and the microbial survival rate after theimposed shear forces have not been systematically comparedand neither has the physical detachment of bacteria fromthe surface been evaluated [15 16] In case of Salmonellacontaminated fresh produce as well as pathogenic bacteria onfish studies indicate the significance of sample preparation[17ndash19]

In this study the applicability of four differentmechanicalhomogenizing devices (stomaching by Bagmixer 400millingby FastPrep-24 or SpeedMill and sonication by the BransonSonifier Table 1) for pathogen isolation and conventionaldetection by cultivation for processed and unprocessed meatproducts was evaluated As a proof of principle usingSalmonella enterica surface contamination was establishedon chicken breasts and inner-matrix contamination wasestablished in pork sausages of soft and hard consistenceThe microbial survival rate of this Gram-negative pathogenand the recovery success of each method were assessedAdditionally the influence of sample taking on the testoutcome was investigated

2 Materials and Methods

21 Bacterial Strains and Growth Conditions Exemplarily forGram-negative Enterobacteriaceae the Salmonella entericaserovar Typhimurium (S enterica) reference strain DSM11320 (DSMZ Braunschweig Germany) was used for allexperiments S enterica was cultivated under aerobic con-ditions at 37∘C in lysogeny broth (LB) medium For spikingexperiments aliquots of overnight cultures were transferred

into fresh LB and cultivated under rotational shaking (GFLshaking incubator 3033 180ndash200 rpm)until an optical density(OD588

) of 05 was reached These cultures were seriallydiluted using 1 buffered peptone water (wvol) Bacterialtiters were enumerated after plating 100120583L of the dilutionsteps on selective media (Xylose Lysine Deoxycholate (XLD)agar Oxoid Wesel Germany) and incubation for 22ndash26 h at37∘C

22 Spiking of Food Products Two types of meat contamina-tion were simulated surface contamination and inner-matrixcontamination in which the pathogen can be distributedthroughout the entire food matrix In all food samples theabsence of Salmonella prior to spiking was confirmed accord-ing to DIN EN ISO 65792002 (microbiology of food andanimal feeding stuffs-horizontal method for the detection ofSalmonella spp) Only sporadic occurrence of other bacteriasuch as Serratia Hafnia or Citrobacter was found

For artificial surface contamination chicken breasts pur-chased at local supermarkets (Berlin Germany) in Jan-uary 2014 were used One or two cubical-shaped pieces ofchicken breast were cut using sterile equipment and razorsto obtain meat samples with a weight of 4 g (BagMixer 400Interscience Germany) 3 g (FastPrep-24 (MP BiomedicalsFrance) and sonication) and 015 g (SpeedMill Analytik JenaAG Germany) On the surface of the samples peptone watercontaining Salmonella (volume addition of 360 120583L for stom-aching 270120583L for FastPrep-24 and sonication and 135 120583Lfor SpeedMill) was evenly applied to obtain spiked sampleswith a final bacterial load of approximately 3 times 105 CFUgcorresponding to 3 times 104 CFUmL in the homogenate Thesesamples were then incubated for 1 h at 4∘C Meat sampleswithout artificial contamination were used as negative con-trols

For the simulation of inner-matrix contaminationcoarsely ground smoked or air-dried salami with a highdegree of hardness and German Mettwurst (meat paste) afinely ground spreadable sausage of soft consistency wasproduced in the technology facilities of the Federal Institutefor Risk Assessment These sausages made of lean porkand bacon were spiked under BSL-2 conditions withinthe production process with Salmonella in six differentconcentrations ranging from 1CFUg to 108 CFUg Onepreparation without Salmonella was used as negative controlBriefly 37 kg of meat (24 kg lean pork 13 kg bacon) wasmixed with 50mL of Salmonella solution containing therespective pathogen concentration while being mincedand flavoured (90 g nitrite salting mix 12 g paprika and8 g black pepper) in an automated meat cutter (HFMFleischereimaschinen Germany) Smoked salamis werecured for five days at 18∘C in the Bastra MC 500 GermanMettwurst was cured for 2 h and afterwards allowed to ripenfor six days at 18∘C before being stored at minus20∘C until use

23 Homogenization Procedure and Sausage Sampling Toevaluate the efficacy of various methods to homogenizemeat and meat products four devices with diverging tech-nical approaches were systematically compared (Table 1)




450




++lowast +minus




(2ndash20) minus


+(lt50mL)

minus

(lt2mL)+













00

05

10

15

20

25


FastP

rep-

24

Spee

dMill

Stom

achi

ng

Soni

catio

n

CFU

8min

CFU

0min





Meat paste8000

4000

00 1 2 4 8

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

FastP

rep-

24

8000

4000

00 1 2 4 8

Stom

achi

ng

15000

10000

5000

00 1 2 4 8

Stom

achi

ng15000

10000

5000

00 1 2 4 8

Spee

dMill

Salami15000

10000

5000

00 1 2 4 8

FastP

rep-

248000

4000

00 1 2 4 8

Spee

dMill











Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate






4 Conclusions






Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




450




++lowast +minus




(2ndash20) minus


+(lt50mL)

minus

(lt2mL)+













00

05

10

15

20

25


FastP

rep-

24

Spee

dMill

Stom

achi

ng

Soni

catio

n

CFU

8min

CFU

0min





Meat paste8000

4000

00 1 2 4 8

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

FastP

rep-

24

8000

4000

00 1 2 4 8

Stom

achi

ng

15000

10000

5000

00 1 2 4 8

Stom

achi

ng15000

10000

5000

00 1 2 4 8

Spee

dMill

Salami15000

10000

5000

00 1 2 4 8

FastP

rep-

248000

4000

00 1 2 4 8

Spee

dMill











Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate






4 Conclusions






Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology






00

05

10

15

20

25


FastP

rep-

24

Spee

dMill

Stom

achi

ng

Soni

catio

n

CFU

8min

CFU

0min





Meat paste8000

4000

00 1 2 4 8

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

FastP

rep-

24

8000

4000

00 1 2 4 8

Stom

achi

ng

15000

10000

5000

00 1 2 4 8

Stom

achi

ng15000

10000

5000

00 1 2 4 8

Spee

dMill

Salami15000

10000

5000

00 1 2 4 8

FastP

rep-

248000

4000

00 1 2 4 8

Spee

dMill











Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate






4 Conclusions






Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Meat paste8000

4000

00 1 2 4 8

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

(CFU

g)

FastP

rep-

24

8000

4000

00 1 2 4 8

Stom

achi

ng

15000

10000

5000

00 1 2 4 8

Stom

achi

ng15000

10000

5000

00 1 2 4 8

Spee

dMill

Salami15000

10000

5000

00 1 2 4 8

FastP

rep-

248000

4000

00 1 2 4 8

Spee

dMill











Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate






4 Conclusions






Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Meat paste

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Salami

00

05

10

15

20

25

Nor

mal

ized

CFU

coun

t

Rim

Cor

e

Cros

s sec

tion

Inte

rmed

iate






4 Conclusions






Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology





Acknowledgments


References







































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

research article sampling and homogenization...

Documents