APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1976, p. 222-231Copyright ©D 1976 American Society for Microbiology

Vol. 32, No. 2Printed in U.S.A.

Volatiles Produced by Microorganisms Isolated fromRefrigerated Chicken at Spoilage

L. R. FREEMAN,' G. J. SILVERMAN,* P. ANGELINI, C. MERRIIT, JR., AND W. B. ESSELEN

Food Sciences Laboratory, U.S. Army Natick Research and Development Command, Natick, Massachusetts01 760,* and Department ofFood Science and Nutrition, University of Massachusetts, Amherst,

Massachusetts 01002

Received for publication 16 April 1976

Volatile components present at spoilage of refrigerated chicken breasts wereidentified using high-vacuum-low-temperature distillation techniques followedby analysis with combined temperature-programmed gas chromatography andmass spectrometry. A comparison was made of the compounds detected fromboth irradiated and non-irradiated muscle stored at 2 and 10°C under bothaerobic and anaerobic conditions. Isolates were randomly selected from thespoiled poultry, identified, and evaluated for their ability to produce volatilespoilage notes when grown on radiation-sterilized chicken. Several isolates thatproduced off-odors on sterile chicken breasts were examined. Twenty-two com-pounds were associated with spoilage. Some of the compounds found on bothirradiated and unirradiated samples were considered to play only a minor role inthe spoilage aroma or were present in low concentrations, since the aroma ofspoiled irradiated chicken lacked the harsh odor notes typical of spoiled unirra-diated chicken. Fifteen of the 22 compounds were considered to be unique tounirradiated, aerobically spoiled samples. Nine of these compounds, hydrogensulfide, methyl mercaptan, dimethyl sulfide, dimethyl disulfide, methyl acetate,ethyl acetate, heptadiene, methanol, and ethanol, were found on chicken spoiledat both 2 and 10°C. Xylene, benzaldehyde, and 2,3-dithiahexane were detectedonly in samples stored at 20C and methyl thiolacetate, 2-butanone, and ethylpropionate were associated with 10°C spoilage. Fifty-eight isolates randomlyselected from fresh, radiation-pasteurized, and unirradiated spoiled poultrywere classified taxonomically, and 10 of them, which produced spoilage odors on

sterilized chicken breasts, were selected for subsequent analysis of their vola-tiles. Isolates identified as Pseudomonas putrefaciens and Pseudomonas speciesthat were members of groups I and II of Shewan's classification, as well as

Flavobacterium and oxidative Moraxella, produced a number of the compoundsfound in the aroma of spoiled chicken. A total of 17 compounds were identified.Whereas no isolate produced all of the aroma compounds found in the aroma ofspoiled chicken, together they did produce the nine found in unirradiatedsamples spoiled at either 2 or 10°C, as well as methyl thiolacetate and xylene.Six compounds were present in the volatiles produced by the isolates but were

absent in the volatiles identified from spoiled chicken. These were hydrogencyanide, methyl isopropyl sulfide, 2-propane thiol, methyl propionate, ethylbenzene, and an unidentified compound.

The spoilage odor of meat, poultry, and fishat refrigeration temperatures is attributedmainly to microbial by-products (2, 10, 13, 15)and not to autolytic products from tissue (10,13, 17). Whereas spoilage odors have been sub-jectively described (1, 14, 15), only recentlyhave attempts been made to identify specificvolatile compounds present at spoilage and re-late them to the causative microorganism.

' Present address: Department of Food Science and Nu-trition, Brigham Young University, Provo, UT 84602.

These studies (10, 24-26) have been devotedmainly to fish, and, although meat and poultryhave been neglected, a considerable effort hasbeen expended on identifying their spoilage mi-croflora.

Despite the heterogeneous nature of the mi-croflora present on chicken immediately afterprocessing, the microflora during refrigeratedstorage undergoes a shift in the microbial spec-tra. At spoilage, the nonpigmented Pseudomo-nas organisms predominate, whereas thelarger initial percentage of Pseudomonas pu-

222

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOLATILES OF SPOILED CHICKEN 223

trefaciens, Achromobacter (Acinetobacter-Moraxella), and the fluorescent pseudomonadsform a much smaller proportion of the totalflora (1-4, 19). When poultry is subjected toagents capable of selectively altering the micro-flora, such as pasteurizing doses of radiation,the predominant organisms present at spoilageare the more radiation-resistant psychrotrophicorganisms, such as Acinetobacter-Moraxellaspecies (13, 14) or Microbacterium thermos-phactum (3). The refrigerated shelf life of theproduct is extended, since these organisms areless degradative toward the product and theirspoilage aroma is considerably less objectiona-ble.

The purpose of this study was to identifyvolatile compounds associated with chickenspoiled aerobically at refrigerated tempera-tures and relate them to the microorganismsresponsible for the production of these com-pounds. Poultry was also subjected to anaerobicspoilage and to pasteurization dosages of ioniz-ing radiation to study, for comparative pur-poses, the relationship of altered metabolismand microflora on the production of volatiles.

MATERIALS AND METHODSTreatment of muscle tissue. Fresh deboned

chicken breasts were obtained from a local distribu-tor. For aerobic storage studies, 350 g (- 10 g) offresh muscle was placed into a sterilized samplebottle having a closure suitable for drawing a vac-uum (23), and the neck was loosely covered with asterile 9-cm Pyrex petri dish. The bottle was thenenclosed in an oxygen-permeable, medium-density,polyethylene bag to aid in moisture control duringstorage at 2 and 10°C for 12 and 5 days, respectively.

Anaerobic conditions were established by vacuumevacuation of the container and sealing off the sys-tem during storage. Both of the above systems weremonitored by GasPak (BBL) disposable anaerobicindicators taped to the inside of the bottle.Samples irradiated at 0.25 Mrad (±0.025) were

heat sealed in laminated polyethylene pouches andirradiated at 0 to 2°C. After irradiation, the sampleswere removed from the pouch and placed into sam-ple containers for aerobic or anaerobic storage.

Chicken to be radiation sterilized for pure culturestudies was placed into laminated polyethylenepouches, the air was evacuated, flushed with N2,and reevacuated, and the package was heat sealed.The samples were then rapidly frozen, irradiated at-30°C, and stored at -40°C.

Microbiological sampling of surfaces. Aerobicplate counts were obtained from the chicken sam-ples upon arrival and during storage. The aerobicplate counts was obtained by moistening a cottonswab in 4% Trypticase soy broth (BBL) supple-mented with 0.5% yeast extract (Difco) (TSY broth)and swabbing a 6.45-cm2 surface area of the chickenfor 15 s. The swab end was then broken into a tubecontaining 9 ml of TSY broth. The contents werethen shaken, serial dilutions were made in addi-

tional tubes of TSY broth, and a 0. 1-ml sample wasuniformly distributed over the surface of prepouredagar plates containing 4.0% Trypticase soy agar(BBL) supplemented with 0.5% yeast extract (TSYagar). The plates were incubated at 2 and 20°C for 11and 2 days, respectively.

Isolation and classification of pure cultures. Iso-lates were randomly selected from aerobic platecount plates having 30 to 120 colonies, purified, andtaxonomically classified by use of the followingtests: Gram stain, motility, catalase, benzidine, cy-tochrome oxidase, oxidation-fermentation (glucose),pigment production, hippurate hydrolysis, litmusmilk, reducing compounds from gluconate, esculinhydrolysis, nitrate reduction, methyl red, Voges-Proskauer, arbutin hydrolysis, growth at 4, 20, and37°C, phosphatase, starch hydrolysis, skim milkagar proteolysis, Tween 80 hydrolysis, lecithinase,and growth in or on triple sugar iron agar, Sellersdifferential agar, Herellea agar, and blood agar. Ad-ditional tests were conducted on selected isolatesusing Enterotubes (Roche Diagnostics) and the APIMicrotube 50 research system (Analytab Products,Inc.). The isolates were broadly classified in aschema, illustrated in Fig. 1, which is partiallybased upon the study of Shewan (32). Tests weregenerally conducted at 20°C.With the exception of Pseudomonas fragi and P.

putrefaciens, which were obtained from R. Levin,University of Massachusetts, the known culturesused in this study were obtained from our culturecollection.

Evaluation of volatiles by isolates. Frozen, radia-tion-sterilized poultry breasts were thawed at 2°Cand inoculated by dipping into a 24-h TSY brothculture of a specific isolate grown at 20°C. The sam-ple was then drained, placed in a sterile petri dish,and incubated at 10°C for 4 days prior to odor evalu-ation by seven panelists. Isolates were consideredpositive when at least four of the seven panelistsassociated a distinct off-odor with a given sample.The volatiles of a number of cultures were alsoevaluated after growth on TSY agar.

For volatile analysis the inoculated samples wereincubated in the sample bottles aerobically at 10°Cfor 5 days. To eliminate the influence of chickenmuscle, selected isolates were inoculated onto TSYagar in a sample bottle enclosed in a polyethylenebag and incubated at 10°C for 5 days, and the headspace was analyzed for volatiles.

Collection and analysis of volatiles. The totalcondensate of each sample was collected at -196°Cby using a modification of the low-temperature-high-vacuum distillation technique of Merritt et al.(23). The "center fraction," or that fraction of thetotal condensate containing compounds exerting va-por pressure between the temperatures of -140 and-80°C, was separated from fractions designated ascarbon dioxide and water fractions. Its odor wasfound to be representative of spoilage aroma eventhough it represented less than 1% of the total con-densate. The center fraction was analyzed by a com-bined temperature-programmed gas chromatographand mass spectrometer instrument system (22).The components in the center fractions were sepa-

rated on a stainless-steel support, coated open tubu-

VOL. 32, 1976

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

224 FREEMAN ET AL.

lar (SCOT) column (50 feet [ca. 15.2 m] by 0.51 mm[ID]) with 1,2,3-tris(cyanoethoxy)propane as the sta-tionary phase and with a helium carrier gas flowrate of 30 cm/s (5 ml/min). The chromatograph usedwas a Bendix series 2200 laboratory gas chromato-graph, which was coupled directly to a Bendix modelMA-2 time-of-flight mass spectrometer. The temper-ature was programmed at a rate of 6°C/min from-65 to 125°C. This mass spectrometer was equippedwith a strip chart recorder to record the gas chroma-togram of the sample by monitoring the total ioncurrent of the mass spectrometer, an oscilloscope fordisplaying the mass spectra, and a recording oscillo-graph for recording the mass spectra of each gaschromatographic peak as it eluted from the column.

Electron microscopy. Representative motile andnonmotile microorganisms were examined in a Hi-tachi model HS-7 transmission electron microscopeby James V. Allen, Brigham Young University, forthe presence of flagella. All cultures were grown at20°C for 48 h in TSY broth before being prepared forexamination in the electron microscope.

RESULTSStorage temperature and microbial devel-

opment. The fresh poultry was consistentlyfound to have about 10' colony-forming units/cm2. After 12 days at 2°C or 5 days at 10°C,about 108 colony-forming units/cm2 were pres-ent, accompanied by a distinct spoilage odor.For poultry breasts irradiated with 0.25 Mradand subsequently stored at 2 and 10°C, themicrobial population increased to 107 to 108 col-ony-forming units/cm2 after 35 and 14 days,respectively.

Radiation-pasteurized and stored chickenbreasts did not display typical spoilage off-odorseven when the microbial count increased to 108

APPL. ENVIRON. MICROBIOL.

colony-forming units/cm2. Instead, a mild sweet-ish odor developed gradually, making it diffi-cult to precisely decide when the sample wasspoiled. This was, in agreement with otherstudies (12, 14), due to an alteration of the typesofmicroorganisms able to survive radiation andproliferate. Radiation and subsequent storagechanged the microflora, consisting mainly ofradiation-sensitive Pseudomonas, to one ofnonoxidative and oxidative Moraxella.General taxonomic classification of the iso-

lates. The isolates were placed in 11 of the 14alphabetic groups shown in Fig. 1 and theiridentity, based upon additional tests, is listedalphabetically in Table 1. No organisms werefound to belong to groups F, H, and P. Only twogram-positive rods were isolated. One, classi-fied in group A, was a catalase-positive, benzi-dine-negative rod identical to a culture of M.thermosphactum. The other isolate, in group B,was a Lactobacillus species, as it was both ben-zidine and catalase negative. The remainingorganisms were either gram-negative rods orcoccobacilli. All the gram-negative motile iso-lates were oxidase positive and could be sepa-rated into either groups C and D (group I ofShewan [32]; Shewan et al. [33]), group E(group II [32, 33]), or group G.The fluorescent isolates of group C grew at

4°C but not at 37°C, produced fluorescin, andcould be further differentiated, on the basis ofproteolytic and lecithinase activity (34), intoPseudomonas fluorescens and Pseudomonasputida.Group D contained a pigmented nonfluores-

cent organism that produced a brown diffusable

NONSPORE FORMING RODS

GRAM STAIN

(I)

CATALASE

FLCvPA GF

--IMOTILITY

-1I I

ROUP B OXIDA

Ii)O/F

I

| INERTOR

OXIDATIVE ALK

I IPIGMENT GROUP F

F_k NON-FLGOR I(

I I IsoRP C GROGP D GROGP E

SE

- l II I

GROUP H O/F

FERMENTATIVE ALK OXIDATIVE

I IIGROUP G OXIDASE GROUP L

IrLi

GGROUP J GROUP K

(-)PIGMSENT

I

I

+)+)-OXIDASE

I

I GROUP PO/F

rI IFERMENTATIVE INERT

I IGROUP M GROUP N

FIG. 1. Classification of isolates from unirradiated and irradiated, fresh and spoiled poultry. O/F, Reac-tion in Hugh-Leifson medium (11); ALK, alkaline; FLUOR, fluorescent.

-

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOLATILES OF SPOILED CHICKEN 225

TABLE 1. Incidence of isolates on irradiated andnon-irradiated poultry

No. of isolates atspoilage

Initial no. IraiGroupa Organism of atedisolates Unirra- (0.25

diate Mrad[±0.025])

A M. thermosphactum 1 (1)B Lactobacillus sp. 1 (1PC P. fluorescens and 4 (2)

P. putidaD Pseudomonas sp. 1 (1)E Pseudomonas sp. 8(4) 18 (12)G P. putrefaciens 1 2 (1)J Moraxella 1 (1) 2K Acinetobacter 1 1L Moraxella- oxidative 1 2 (1) 8M Flavobacterium 3 (3) 2 (1)N Flavobacterium 1

a These designations correspond to those listed in Fig. 1.bThe number in parentheses indicates those isolates pro-

ducing a pronounced aroma after growth on sterile chicken.

pigment when grown on Pseudomonas P agar.Group E consisted of nonpigmented organismsdesignated as Pseudomonas and contained thelargest number of isolates, 26 of the 58 studied.It was difficult to subdivide this group on thebasis of the tests performed, as no consistentpattern for differentiation was observed. One ofthe isolates, though, culture 37, was identifiedas P. fragi. The program of differentiation ofthese nonpigmented pseudomonads has beennoted by other investigators (1, 3, 8), and theseorganisms have been found to be the dominantorganisms found on spoiled, proteinaceousfoods (8, 19).The organisms of group G, although fermen-

tative, did not correspond to the fermentative,oxidase-positive, motile rods designated as Vi-brio and Aeromonas by Shewan (32). Aero-monas is described as being able to produce gasin addition to being fermentative and Vibrio asbeing nonpigmented and capable of hydrolyz-ing starch. None of these reactions could beascribed to the group G isolates ofthis study.Group G did contain salmon-to-brown, non-

diffusable, pigmented organisms correspondingto the published description of P. putrefaciens(18). These isolates reduced litmus milk rap-idly, were proteolytic on skim milk agar, re-duced nitrates, produced hydrogen sulfide,were capable of fermenting glucose in Hughand Leifson's medium (11), produced deoxyribo-nuclease, and were oxidase positive. Under an-aerobic conditions these isolates also reducedthe indicator in Baird-Parker's medium.Two of the three isolates in group G did not

decarboxylate lysine or ornithine and were cit-rate negative. The third culture was positivefor all three tests.McMeekin and Patterson (20) described iso-

lates of P. putrefaciens obtained from spoiledfoods that were in many respects dissimilar tothe isolates in group G. Their isolates were notfermentative on glucose and did not produceurease, whereas the strains in this study werefermentative on glucose and positive for ureaseactivity.Using the criterion of Baumann et al. (5, 6),

the oxidase-positive group J isolates were des-ignated Moraxella and the oxidase-negativegroup K isolates as Acinetobacter. Althoughthe group L isolates were oxidase-positive, non-motile organisms whose morphology variedfrom rod to coccobacillus and generallymatched the reactions of the group J isolates,they oxidatively attacked glucose. Thornley(36) described Moraxella species capable of oxi-dizing glucose aerobically, as well as Moraxellaspecies that did not. In this study group L hasbeen designated Moraxella-oxidative to distin-guish this characteristic towards glucose.Groups J, K, and L appear to correspond in

some respects to Thornley's (36) phenons 4i, 4ii,and 3, respectively. These groups, however,were not homogeneous and could be furtherdivided into subgroups (Table 2).Group M and group N were both classified as

Flavobacterium. Group M was subdivided intotwo subgroups on the basis ofnitrate reduction,litmus milk reactions, hydrolysis of starch, theproduction of hydrogen sulfide, the productionof indole, arbutin hydrolysis, and their reactionin Baird-Parker's medium. With the exception

TABLE 2. Reactionsa that subdivide groups J, K,and L

J K LTest----

Jl J2 Kl K2 Li L2 L3

Morphology CB CB CB CB R CB CBNitrate reduction + - + _ _ _ _Litmus milk Alk RedPhosphatase + -

Lactose Alk ADeoxyribonuclease - +Tween 80 + - _ + +Growth at 37C + -

Blood agar H -Lysine decarboxylase - +Ornithine decarboxylase - +Hippurate hydrolysis + - +Gluconate + - -

Egg yolk (lecithinase) - + +Sellers medium

Slant Alk Alk AButt NC NC Aa A, Acid; Alk, alkaline; NC, no change; CB, coccobacil-

lus; Red, reduction; R, rod; H, hemolysis.

VOL. 32, 1976

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

of motility, one subgroup of group M appearedto be identical to P. putrefaciens isolates ofgroup G, and this relationship will be the sub-ject of another publication. Examination in theelectron microscope demonstrated that the mo-tile group G isolates, designated P. putrefa-ciens, had a monotrichous polar flagellum andthe non-motile Flavobacterium sp. were devoidof flagellum.

Evaluation of the spoilage activity of iso-lates. The ability of isolates to produce a pro-nounced aroma on chicken is also presented inTable 1. Organisms from 9 of the 11 taxonomicgroups found on chicken were capable of pro-ducing a distinctive aroma. The aromas mosttypical of spoilage were associated with gram-negative rods of groups C through M, with themost pronounced belonging to groups Cthrough G. The aroma produced by M. ther-mosphactum and the Lactobacillus sp. isolates(groups A and B) had a predominant volatileacid character, whereas 20 of the 34 pseudo-monad isolates, 4 out of the 6 Flavobacterium,and 2 ofthe 16 Moraxella-Acinetobacter isolatespossessed pronounced aromas with typical chilltemperature spoilage notes. None of the 10Moraxella isolates (groups J and L) from irradi-ated chicken produced a distinctive aroma.

Volatile compounds of spoiled commercialchicken. Portions of chicken were allowed tospoil at either 2 or 10°C, and their volatileswere analyzed by gas chromatography-massspectrometry (Table 3). Fifteen of the 22 com-pounds identified were common to chickenspoiled at either storage temperature and, al-though the data is not presented, there was alarger concentration ofeach of these compoundsin the samples stored at 10°C than in thosestored at 2°C. Three additional compounds werefound only in the volatiles from samples storedat 10°C, and four additional compounds wereidentified only in those stored at 2°C. Acetonewas the sole compound found in the center frac-tion of fresh, unstored chicken breasts. Withthe exception of 2,3-dithiahexane, no additionalcompounds were identified when three centerfractions, derived from 1,050 g of chickenbreast, were pooled for analysis.Compounds identified after storage at 2 and

10°C under anaerobic conditions. Chickenstored anaerobically at 2°C for 21 days or at10°C for 8 days did not develop any typicalspoilage aroma. Only four compounds, acetone,ethanol, methanol, and toluene, were detected.

Volatile compounds identified from irradi-ated chicken breasts. Nine compounds wereconsistently present in the aroma of chickenanalyzed immediately after irradiation with apasteurizing (0.25 Mrad) or a sterilizing dose

TABLE 3. Volatile compounds associated withspoiled chicken breasts stored at 2a or JoboCStorage temp2 or 10°C

Only at 2°C

Only at 10°C

CompoundHydrogen sulfideMethyl mercaptanDimethyl sulfideDimethyl disulfiden-Heptane1-HepteneHeptadienen-OctaneAcetoneMethyl acetateEthyl acetateBenzeneMethanolEthanolTolueneBenzaldehydeXylenen-Pentane2,3-Dithiahexane'Methyl thiolacetate2-ButanoneEthyl propionate

Samples stored for 12 days.b Samples stored for 5 days.Only found when three center fractions were

combined.

(1.5 Mrad) (Table 4). Two additional com-pounds, 1-propane thiol and dimethyl trisul-fide, were only found occasionally. With theexception ofthe latter two compounds, all of thenine compounds initially present after irradia-tion persisted throughout anaerobic storage.Dimethyl disulfide disappeared from samplessubjected to both levels of irradiation andstored aerobically, whereas benzene was absentonly from those samples aerobically stored aftersubjection to 0.25 Mrad. A stronger irradiationodor persisted in samples receiving a dose of 1.5Mrad than in samples receiving a dose of 0.25Mrad.

Evaluation of the volatiles produced by iso-lates. Table 5 lists the aroma components oftwo known cultures, P. putrefaciens and P.fragi, and 10 isolates representing seven groupslisted in Table 1. A total of 16 known com-pounds, and an unidentified compound isolatedas a single gas chromatographic peak, wereproduced by the 12 cultures. This unknowncompound exhibited a mass spectrum whosemajor mass-to-charge ratios were 86, 70, 43, 42,41, and 27. It appeared to have a molecularweight of 102. Present but not listed in the bodyofthe table were eight compounds generated bythe irradiation used for sterilizing the chicken.The lowest number ofcompounds produced by aculture was two (cultures 20 and 44, groups A

226 FREEMAN ET AL.

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOLATILES OF SPOILED CHICKEN 227

n-Pentanen-Hexanen-Heptane1-Heptenen-OctaneAcetoneTolueneBenzeneDimethyl disulfide1-Propane thiolaDimethyl trisulfide"

n-Pentanen-Hexanen-Heptane1-Heptenen-OctaneAcetoneToluenen-Pentanen-Hexanen-Heptane1-Heptenen-OctaneAcetoneTolueneBenzeneDimethyl disulfiden-Pentanen-Hexanen-Heptane1-Heptenen-OctaneAcetoneTolueneBenzene

I Inconsistently found on samples irradiated at0.25 Mrad.

b Stored for 19 days at 2°C or 14 days at 10°C.'Stored at 2°C for 20 days or at 10°C for 5 days.

and M) and the highest number of compounds,six, was present in the head space ofP. fluores-cens (culture 60), a Pseudomonas (culture 48),and P. fragi (culture 37). These cultures pro-duced different combinations of their six com-pounds. Hydrogen cyanide (9) was identifiedfrom cultures ofP. fluorescens grown either onTSY agar or on chicken breasts. In fact, 12 ofthese compounds were produced in detectableamounts by cultures grown either on chicken oron TSY agar.Four compounds, benzaldehyde, 2,3-dithia-

hexane, 2-butanone, and ethyl propionate, were

isolated from the aroma of spoiled chicken butwere not found in the pure culture studies.

The volatile produced by the largest numberof cultures tested was dimethyl disulfide, fol-lowed by ethanol, methanol, and ethyl acetate.Of the 17 compounds, 11 were also found on

spoiled chicken. The remaining six compounds,methyl propionate, hydrogen cyanide, methylisopropyl sulfide, 2-propanethiol, ethyl ben-zene, and the unidentified compound, were as-sociated with five organisms, cultures 59, 60,18, 14, and 48.

DISCUSSIONA recent study of McMeekin (19) verified the

importance of the contribution of Pseudomo-nas, especially those of Shewan's groups I andII and, to a minor extent, IV (32), to the spoil-age aroma of chicken breast muscle. Of the 250cultures isolated by him from samples stored at2°C, 80% produced off-odors. Refrigerated stor-age favored the growth of organisms able togenerate spoilage aroma, and, at spoilage, thepredominate microflora was found to be orga-nisms belonging to Pseudomonas group II. Al-though the strains ofpseudomonads included ina comprehensive study by Stanier et al. (34)were isolated from a variety of environments,organisms corresponding to group II of Shewanwere omitted. Davidson et al. (8) has concludedthat the majority of commonly occurring pseu-domonads on meat and meat products do notconform with the species studied by Stanier etal. (34).

In Lhis study, involving a smaller number ofisolates, groups C and E, corresponding togroups I and II, respectively, were the predomi-nant organisms in unirradiated fresh andspoiled chicken. Group E, though, was the mostnumerous organism isolated at both test pe-riods. Half or more of the isolates from thesetwo groups were capable of producing obviousspoilage aromas. McMeekin did not recover P.putrefaciens in his initial study (19), althoughit was recovered in a subsequent study by theuse of a selective medium (20). In this study, P.putrefaciens as well as Moraxella and Acineto-bacter were isolated. Whereas all of the isolatesof P. putrefaciens studied by McMeekin andPatterson (20) produced off-odors, only one outof three cultures isolated in this study demon-strated this ability when grown on sterilechicken muscle.The following compounds are considered to

be responsible for or associated with the charac-teristic off-odor of chicken breasts undergoingnormal spoilage at refrigerated temperatures:hydrogen sulfide, methyl mercaptan, dimethylsulfide, heptadiene, methyl acetate, ethyl ace-tate, methanol, ethanol, dimethyl disulfide,benzaldehyde, xylene, 2,3-dithiahexane, 2-bu-

TABLE 4. Volatile compounds present duringaerobic and anaerobic storage of irradiated chicken

breastsStorage condition Compound

Immediately afterirradiation

0.25 MradbAerobic

Anaerobic

1.5 Mrad (aerobic)'

VOL. 32, 1976

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

TABLE 5. Volatilesa produced by cultures grown on irradiation-sterilized chicken for 5 days at 10°CKnown cul-

tures Isolates5~~~~~~PresentM. P. fluorescens Pseu- P pu Morax FIavoI onVolatile P.p- ther- Pseud- o-Flato- sonle~*P-P. o-~ .--. trPu- Mora-x- baumchpilkedtrefa- mos tida monas fragi monas trefa- ella-ox rium chicken

ciens fragi phac- (14, C) 59, C 60, C sp. (48, (37, E) sp. (15, (36enG) (18, L) p (4turn)E 3,G)(8 )M(20, A) D )M

Hydrogen cyanide +. +.Hydrogen sulfide +. + + +Methanol + + + + + + +Ethanol + + + + + + + +Heptadiene + +Methyl acetate + + + +Ethyl acetate + +C + + + +C +Methyl propionate +.

Methyl mercaptan +C + + + +Dimethyl sulfide + +. + + +Dimethyl disulfide +. +. + + + + + +Methyl isopropyl +

sulfideMethyl thiolacetate +C +2-Propanethiol +C

Xylene + +Ethyl benzene +Unidentified (mo- +C +C

lecular wt, 102) 1

a n-Pentane, n-hexane, n-heptane, n-octane, 1-heptene, acetone, benzene, and toluene were present in the head space ofradiation-sterilized chicken (1.5 Mrad) after incubation (see Table 4).

Numbers indicate culture number; letters indicate group.'*Also detected in the head space volatiles of cultures grown on TSY agar.

tanone, ethyl propionate, and methyl thiolace-tate. However, sensory evaluations of combina-tions of these compounds will be necessary toconfirm their relationship to the odor of typicalspoilage aroma.

It is also apparent that considerable researchremains to be accomplished in relating the com-ponents of spoilage aroma to their causativemicroorganisms. The fact that human sensoryevaluation can be considerably more sensitivefor the detection of many compounds than cur-rent analytical instruments makes correlationdifficult. Moreover, this study did not attemptto explore the effect of the growth of mixedcultures on subsequent end products.By comparing those compounds identified

after anaerobic storage of unirradiated musclewith those identified after aerobic storage ofunirradiated muscle, it is obvious that anaero-biosis prevented the development of a numberof compounds associated with the aerobicallydeveloped spoilage aroma. The observation alsoprovides a rationale in substantiating the ob-servation of Van den Berg et al. (37) that thestorage life of fresh poultry at refrigerated tem-peratures could be doubled under aseptic andanaerobic conditions.The hydrocarbons produced by irradiation

most likely originated from lipid compounds,benzene and toluene from phenylalanine, anddimethyl disulfide directly from cystine or me-thionine (22). Ethanol and methanol were notdetected in irradiated samples in this study,although Merritt (21) reported a small amountof ethanol present in chicken receiving a dose of0.5 Mrad. Ethanol and methanol were bothfound in aerobically and anerobically storedunirradiated samples, indicating the source ofthese two compounds was microorganisms thatwere eliminated by irradiation. The organismsmore sensitive to irradiation, Pseudomonasand Flavobacterium, produced these alcohols inpure-culture growth studies on chicken,whereas Moraxella, the more radioresistant or-ganism that predominated on irradiatedchicken after storage, did not.The compounds n-heptane, n-octane, 1-hep-

tene, acetone, benzene, and toluene were foundon samples spoiled aerobically at both 2 and10°C and on stored samples subjected to steriliz-ing and pasteurizing doses or irradiation. Theorigin of these compounds in irradiated sampleshas been discussed above, but in the unirra-diated samples it is possible that they wereproduced by aerobically growing microorga-nisms. Toluene was also found in unirradiated,

228 FREEMAN ET AL. APPL. ENVIRON. MICROBIOL.

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOL. 32, 1976

anaerobically stored samples, and acetone wasfound in all samples, including fresh chickenbreasts analyzed prior to storage.Dimethyl disulfide was present both in

spoiled unirradiated chicken and immediatelyafter irradiation. This leads one to speculate asto its contribution to the aroma of both theirradiation odor of chicken and the normalspoilage odor, since these aromas are dissimi-lar. The influence of dimethyl disulfide in botharomas may be dependent upon its concentra-tion. At lower concentrations it will not be asobjectionable, and its influence may be modi-fied by other compounds.

In pure culture studies of isolates from codand haddock inoculated onto sterile cod muscle,Herbert et al. (10) reported that the aroma frompseudomonad cultures produced either sweet-fruity or characteristic hydrogen sulfide-sul-fury odors and that where fruitiness was de-tected no sulfides were identified. Miller et al.(24) reported that a P. fragi, isolate 18, inocu-lated onto sterile fish muscle produced bothesters and sulfur compounds (dimethyl sulfide,ethyl acetate, and dimethyl disulfide) eventhough a fruity odor predominated. In the pres-ent study, culture 15 of group E, P. fluorescens(culture 60), Pseudomonas isolate 48, and Mor-axella-oxidative (isolate 18) produced severalsulfur compounds in addition to ester com-pounds. The known P. fragi culture and P.fragi culture 37 smelled predominantly fruity,and both produced at least one sulfur compoundin addition to methyl and ethyl acetate.Hydrogen sulfide and methyl mercaptan

were produced by the known P. putrefaciensculture and isolates designated as P. putrefa-ciens (isolate 36) and Pseudomonas sp. (isolate15). The production of hydrogen sulfide by P.putrefaciens strains has been well established(10, 16, 18, 20), and Lea et al. (17) reported thatP. putrefaciens when grown on chicken breastmuscle produced small quantities of hydrogensulfide at 1°C. Herbert et al. (10) found P. fragistrains to be hydrogen sulfide producers, as didNicol et al. (27) for Pseudomonas mephiticaisolated from beef.

Miller et al. (25) used peptone iron agar todetect hydrogen sulfide production by P. putre-faciens. In the current study, hydrogen sulfidewas detected for two of the three isolates byboth gas chromatographic-mass spectrometricanalysis and by triple sugar iron agar. How-ever, for culture 15, hydrogen sulfide was de-tected by gas chromatographic-mass spectro-metric analysis but not by triple sugar ironagar. This observation supports the position ofPadron and Dockstader (28) that triple sugar

VOLATILES OF SPOILED CHICKEN 229

iron agar does not reveal hydrogen sulfide pro-duction by organisms producing it in relativelysmall quantities.Methyl mercaptan may have originated from

the decomposition of methionine. Segal andStarkey (31) reported that a P. fluorescensstrain first deaminated methionine, followed bydemethiolation with the production of methylmercaptan. Further transformation would oc-cur upon oxidation to dimethyl disulfide. Milleret al. (25, 26) identified both methyl mercaptanand dimethyl disulfide when cultures of P. pu-trefaciens, Achromobacter, P. fluorescens, andP. perolens were grown in a sterile homogenateof sterile fish muscle. Miller et al. (24) alsoreported identification of dimethyl disulfidefrom the volatiles ofP. fragi when grown in fishhomogenate, but methyl mercaptan was ab-sent. In this study dimethyl disulfide accompa-nied methyl mercaptan in the head space offour cultures. In instances where dimethyl di-sulfide but not methyl mercaptan was present,it is likely that all of the methyl mercaptan wasoxidized to the former (16).Dimethyl sulfide was produced by group E,

cultures 37 (P. fragi) and 15, P. fluorescens(isolate 60), and the known P. fragi. Herbert etal. (10) reported dimethyl sulfide production byPseudomonas species from methionine, andMiller et al. (24) reported dimethyl sulfide pro-duction by P. fragi strain 18 when grown onsterile fish muscle. The disulfides are less vola-tile than the sulfides but have a more offensiveodor (16).

Herbert et al. (10) observed that Pseudomo-nas species of the P. fragi type produced fruityodors when grown on sterile cod muscle andthat this odor was probably due to the presenceof esters. In this study ethyl acetate was identi-fied with group E, cultures 37 (P. fragi) and 15,P. fluorescens (isolate 60), group D (culture 48),Moraxella-oxidative (isolate 18), and theknown P. fragi culture. Of these cultures onlyP. fragi and culture 37 were found to produceethanol in addition to the ester. These two orga-nisms were also the only cultures among the 12to produce methyl acetate together with metha-nol. However, P. fluorescens culture 60 did pro-duce the ester, methyl propionate, in additionto methyl thiolacetate. Reddy et al. (29, 30) andMiller (24) found that P. fragi isolated fromfruity-flavored cottage cheese and pasteurizedmilk, when grown in sterile homogenized milkat 21 and 7°C or in fish until a fruity aromadeveloped, produced ethyl acetate, ethyl butyr-ate, and ethyl hexanoate. The addition ofeither0.1 or 0.2% ethyl alcohol or butyric acid to thesubstrate markedly enhanced ester production.

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

Although both the known strain of P. fragiused in the present study and P. fragi culture37 produced methanol and methyl acetate inaddition to ethyl acetate and ethanol, no otheresters or alcohols were detected in samples con-taining these two isolates.

It cannot be assumed that isolates capable ofproducing pronounced spoilage odors on onesubstrate will do so on others. An isolate fromspoiled haddock, which was always accompa-nied by a strong spoilage aroma on TSY agar,failed to produce a strong off-odor on chickenbreast. Herbert et al. (10) reported that whenMoraxella-like strains were inoculated ontosterile scampi tails they produced strong spoil-age odors but that these strong odors were notpresent after growth on sterile cod muscle.

It appears that a relatively broad spectrum ofmicroorganisms isolated from commericallyprocessed chicken breasts at spoilage are capa-ble of producing compounds associated with thespoilage aroma of poultry stored at refrigera-tion temperatures. Whereas Pseudomonas spe-cies have traditionally been the organismsfound in the greatest abundance on unirra-diated chicken spoiled at chill temperaturesand appear to play a major role in the produc-tion of the typical spoilage odor, the data fromthis and other studies demonstrate that othertaxa contain organisms that can contribute tospoilage odor. It should be noted that approxi-mately 40% of the Pseudomonas isolates ingroups C and E did not produce strong offensiveodors on sterile chicken breasts. Herbert et al.(10) reported that spoilage odor-producing orga-nisms on spoiled cod and haddock never ac-counted for more than 10 to 20% of the totalflora. McMeekin (19) found that, whereas moststrains isolated from spoiled chicken were capa-ble of producing a spoilage odor, there werestains within each Pseudomonas group that didnot. The contribution that these non-odor-pro-ducing organisms make to the spoilage aromabouquet remains to be elucidated.

ACKNOWLEDGMENTSWe wish to express sincere thanks to M. L. Bazinet, B.

Y. K. Rao, and C. Dipietro for technical assistance.This study was supported by the U.S. Army Research

Office, contract no. DAAG17-72-G-0003.LITERATURE CITED

1. Ayres, J. C. 1960. The relationship of organisms of thegenus Pseudomonas to the spoilage of meat, poultryand eggs. J. Appl. Bacteriol. 23:471-486.

2. Ayres, J. C., W. S. Ogilvy, and C. F. Stewart. 1950. Postmortem changes in stored meats. I. Microorganismsassociated with development of slime on evisceratedcut-up podltry. Food Technol. 4:199-205.

3. Barnes, E. M., and C. S. Impey. 1968. Psychrophilicspoilage bacteria of poultry. J. Appl. Bacteriol. 31:97-107.

4. Barnes, E. M., and M. J. Thornley. 1966. The spoilageflora of eviscerated chickens stored at different tem-peratures. J. Food Technol. 1:113-119.

5. Baumann, P., M. Doudoroff, and R. Y. Stanier. 1968.Study of the Moraxelld group. I. Genus Moraxella andNeisseria catarrhalis group. J. Bacteriol. 95:8-73.

6. Baumann, P., M. Doudoroff, and R. Y. Stanier. 1968.Study of the Moraxella group. II. Oxidative-negativespecies (genus Acinetobacter). J. Bacteriol. 95:1520-1541.

7. Chen, T. C., H. W. Newar, and R. E. Levin. 1974.Identification of major high-boiling volatile com-pounds produced during refrigerated storage of had-dock fillets. Appl. Microbiol. 38:679-680.

8. Davidson, C. M., J. J. Dowdell, and R. G. Board. 1973.Properties of gram-negative aerobes isolated frommeats. J. Food Sci. 38:303-305.

9. Freeman, L. R., P. Angelini, G. J. Silverman, and C.Merritt, Jr. 1975. Production of hydrogen cyanide byPseudomonas fluorescens. Appl. Microbiol. 29:560-561.

10. Herbert, R. A., M. S. Hendrie, D. M. Gibson, and J. M.Shewan. 1971. Bacteria active in the spoilage of cer-tain sea foods. J. Appl. Bacteriol. 34:41-50.

11. Hugh, R., and E. Leifson. 1953. The taxonomic signifi-cance of fermentative versus oxidative metabolism ofcarbohydrates by various gram-negative bacteria. J.Bacteriol. 66:24-26.

12. Idziak, E. S., and K. Incze. 1968. Radiation treatmentof foods. I. Radurization of fresh eviscerated poultry.Appl. Microbiol. 16:1061-1066.

13. Ingram, M., and R. H. Dainty. 1971. Changes caused bymicrobes in spoilage of meats. J. Appl. Bacteriol.34:21-39.

14. Ingram, M., and M. J. Thornley. 1959. Changes inspoilage pattern of chicken meat as a result of irra-diation. Int. J. Appl. Radiat. Isot. 6:122-128.

15. Jay, J. M. 1972. Mechanism and detection of microbialspoilage in meats at low temperatures: a status re-port. J. Milk Food Technol. 35:467-471.

16. Kadota, H., and Y. Ishida. 1972. Production of volatilesulfur compounds by microorganisms. Annu. Rev.Microbiol. 26:127-138.

17. Lea, C. H., L. J. Parr, and H. F. Jackson. 1969. Chemi-cal and organoleptic changes in poultry meat result-ing from the growth of psychrophylic spoilage bacte-ria at 1 C. 3. Glutamine, glutathione, tyrosine, am-monia, lactic acid, creatine, carbohydrate, haem pig-ment and hydrogen sulfide. Br. Poult. Sci. 10:229-238.

18. Levin, R. E. 1968. Detection and incidence of specificspecies of spoilage bacteria on fish. Appl. Microbiol.16:1734-1737.

19. McMeekin, T. A. 1975. Spoilage association of chickenbreast muscle. Appl. Microbiol. 29:44-47.

20. McMeekin, T. A., and J. T. Patterson. 1975. Character-ization of hydrogen sulfide-producing bacteria iso-lated from meat and poultry plants. Appl. Microbiol.29:165-169.

21. Merritt, C., Jr. 1972. Qualitative and quantitative as-pects of true volatile components in irradiated foodsand food substances. Radiat. Res. Rev. 3:353-368.

22. Merritt, C., Jr., P. Angelini, M. L. Bazinet, and D. J.McAdoo. 1966. Irradiation damage in lipids, p. 225-240. In R. F. Gould (ed.), Advances in chemistryseries. American Chemical Society Publications,Washington, D.C.

23. Merritt, C., Jr., S. R. Bresnick, M. L. Bazinet, J. T.Walsh, and P. Angelini. 1959. Determination of vola-tile components of foodstuffs. Techniques and theirapplication to studies of irradiated beef. J. Agric.Food Chem. 1:784-787.

24. Miller, A., III, R. A. Scanlan,' J. S. Lee, and L. M.

230 FREEMAN ET AL.

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOLATILES OF SPOILED CHICKEN 231

Libbey. 1973. Identification of the volatile compoundsproduced in sterile fish muscle (Sebastes melanops)byPseudomonasfragi. Appl. Microbiol. 25:952-955.

25. Miller, A., III, R. A. Scanlan, J. S. Lee, and L. M.Libbey. 1973. Volatile compounds produced in sterilefish muscle (Sebastes melanops) by Pseudomonas pu-trefaciens, Pseudomonas fluorescens, and an Achro-mobacter species. Appl. Microbiol. 26:18-21.

26. Miller, A., III, R. A. Scanlan, J. S. Lee, L. M. Libbey,and M. E. Morgan. 1973. Volatile compounds pro-duced in sterile fish muscle (Sebastes melanops) byPseudomonas perolens. Appl. Microbiol. 25:257-261.

27. Nicol, D. L., M. K. Shaw, and D. A. Ledward. 1970.Hydrogen sulfide production by bacteria and sulfmy-oglobin formation in prepacked chilled beef. Appl.Microbiol. 19:937-939.

28. Padron, A. P., and W. B. Dockstader. 1972. Selectivemedium for hydrogen sulfide production by salmonel-lae. Appl. Microbiol. 23:1107-1112.

29. Reddy, M. C., D. D. Bills, and R. C. Lindsay. 1969.Ester production by Pseudomonas fragi. II. Factorsinfluencing ester levels in milk cultures. Appl. Micro-biol. 17:779-782.

30. Reddy, M. C., D. D. Bills, R. C. Lindsay, L. M. Libbey,A. Miller III, and M. E. Morgan. 1968. Ester produc-tion by Pseudomonas fragi. I. Identification andquantification of some esters produced in milk cul-tures. J. Dairy Sci. 51:656-659.

31. Segal, W., and R. L. Starkey. 1969. Microbial decompo-sition of methionine and identity of the resultingsulfur products. J. Bacteriol. 98:908-913.

32. Shewan, J. M. 1963. The differentiation of certain gen-era of Gram negative bacteria frequently encoun-tered in marine environments, p. 499-521. In C. H.Oppenheimer (ed.), Marine microbiology. Charles CThomas, Springfield, Ill.

33. Shewan, J. M., G. Hobbs, and W. Hodgkiss. 1960. Adeterminative scheme for the identification of certaingenera of Gram negative bacteria, with special refer-ence to the Pseudomonadacae. J. Appl. Bacteriol.23:379-390.

34. Stanier, R. Y., N. J. Palleroni, and M. Doudoroff. 1966.The aerobic pseudomonads: a taxonomic study. J.Gen. Microbiol. 43:159-271.

35. Thomas, C. P., P. S. Dimick, and J. H. MacNeil. 1971.Poultry product quality. 4. Levels of carbonyl com-pounds in fresh, uncooked skin from chickens andturkeys of different ages and sexes. J. Food Sci.36:537-541.

36. Thornley, M. J. 1967. A taxonomic study ofAcinetobac-ter and related genera. J. Gen. Microbiol. 49:211-257.

37. Van den Berg, L., C. P. Lentz, and A. W. Khan. 1964.Changes in quantity and water-holding and ion-bind-ing properties of chicken meat during above-freezingstorage under aseptic conditions. Food Technol. (Chi-cago) 18:171-173.

VOL. 32, 1976

on January 25, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from