Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

1992 Classification of enterococci and their roles in spoilage of pork products and as sanitary indicators in pork processing Linda Marie Knudtson Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Agriculture Commons, Food Microbiology Commons, and the Microbiology Commons

Recommended Citation Knudtson, Linda Marie, "Classification of enterococci and their roles in spoilage of pork products and as sanitary indicators in pork processing " (1992). Retrospective Theses and Dissertations. 10124. https://lib.dr.iastate.edu/rtd/10124

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, begiiming at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor, Ivll 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9311504

Classification of enterococci and their roles in spoilage of pork products and as sanitary indicators in pork processing

Knudtson, Linda Marie, Ph.D.

Iowa State University, 1992

UMI 300N.ZeebRd. Ann Arbor,MI 48106

Classification of enterococci and their roles in spoilage

of pork products and as sanitary indicators in pork processing

by

Linda Marie Knudtson

A Dissertation Submitted to the

Graduate Faculty in Partial Fulfillment of the

Requirements for the Degree of

DOCTOR OF PHILOSOPHY

Department: Microbiology, Immunology and Preventive Medicine Major: Microbiology

Approved:

Signature was redacted for privacy. In Charge of Major Work Signature was redacted for privacy. For the Major

Signature was redacted for privacy. Major Department Signature was redacted for privacy. For the Graduate College

Iowa State University Ames, Iowa

1992 il

TABLE OF CONTENTS

GENERAL INTRODUCTION 1

Dissertation Format 2

LITERATURE REVIEW 4

History of the Enterococci 4

Origin of the species of the genus Enterococcus 5

Importance of the Enterococci 12 Habitat of the enterococci 12 Enterococci as fecal indicators in food and water 14

Enterococci in Meats 18 Enterococci as a cause of food spoilage 18 Hazard analysis critical control points 19 Meat sampling techniques 21

Enterococci in Food Poisoning 23

Clinical Importance of the Enterococci 24

Isolation and Identification Methods for the Enterococci 30 Selective media 30 Rapid methods for identification 36 Molecular approaches to identification 42

PAPER 1. ROUTINE PROCEDURES FOR ISOLATION AND IDENTIFICATION OF ENTEROCOCCI AND FECAL STREPTOCOCCI 45

ABSTRACT 47

INTRODUCTION 48

MATERIALS AND METHODS 50

RESULTS AND DISCUSSION 52

ACKNOWLEDGMENTS 57

LITERATURE CITED 58 iii

PAPER 2. ENTEROCOCCI IN PORK PROCESSING 76

ABSTRACT 78

INTRODUCTION 79

MATERIALS AND METHODS 81

RESULTS 84

DISCUSSION 88

ACKNOWLEDGMENTS 91

LITERATURE CITED 92

PAPER 3. COMPARISON OF fGTC AND KF AGARS TO ENUMERATE ENTEROCOCCI AND FECAL STREPTOCOCCI IN MEATS 105

ABSTRACT 107

INTRODUCTION 108

MATERIALS AND METHODS 110

RESULTS AND DISCUSSION 112

ACKNOWLEDGMENT 116

LITERATURE CITED 117

PAPER 4. ANTIBIOTIC RESISTANCE AMONG ENTEROCOCCAL ISOLATES FROM CLINICAL AND ENVIRONMENTAL SOURCES 124

ABSTRACT 126

INTRODUCTION 127

MATERIALS AND METHODS 129

RESULTS AND DISCUSSION 130

ACKNOWLEDGMENT 133 iv

LITERATURE CITED 134

PAPER 5. COMPARISON OF FOUR LATEX AGGLUTINATION KITS TO RAPIDLY IDENTIFY LANCEFIELD GROUP D ENTEROCOCCI AND FECAL STREPTOCOCCI 141

ABSTRACT 142

LITERATURE CITED 148

GENERAL SUMMARY AND DISCUSSION 150

LITERATURE CITED 157

ACKNOWLEDGMENTS 171

APPENDIX A; MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM PORK CARCASSES SAMPLED IN THREE PLANTS 172

APPENDIX B: MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM TWO PORK PROCESSING PLANTS 176

APPENDIX C; MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM RETAIL SAMPLES 179

APPENDIX D: MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM WATER SAMPLES 181

APPENDIX E; SOURCES AND IDENTIFICATIONS OF CLINICAL ISOLATES FROM MERCY HOSPITAL MEDICAL CENTER, DES MOINES, IOWA 183 1

GENERAL INTRODUCTION The classification of the enterococci has undergone significant changes in recent years. In 1984, Schleifer and Kilpper-Balz (132) used DNA-DNA and DNA-rRNA hybridization studies to show that the enterococcus group of the streptococci was a separate genus, Enterococcus. Currently, 19 species belong to this genus; E. avium (26), E. casseliflavus (26), £ cecorum (152), E. columbae (39), E. dispar (27), E. durans (26), E. faecalis (132), E. faecium

(132), E. flavescens (121), E. gallinarum (26), E. hirae (58), E. malodoratus (26), E. mundtii (25), E. pseudoavium (24), E. raffinosus (24), E. saccharolyticus, E. seriolicida (89), E. solitarius, (24) and E. sulfureus (102).

Generally, the habitat of the enterococci is the intestinal tracts of warm blooded animals (111), including domestic (126) and wild animals (110).

Enterococci also have been isolated from the feces of reptiles (110) and birds

(37) as well as from insects (113). Some enterococci can establish an epiphytic relationship with plants (113) and they have been isolated from the soil, where they are considered contaminants (25, 26).

Despite their ubiquity in nature, the natural habitat of the enterococci is the intestinal tracts of man and animals and their occurrence in food or water implies either indirect or direct fecal contamination (62). The relative resistance of these organisms to adverse conditions, such as tolerance to extremes in temperature, pH, and salinity is advantageous in the examination of sea water, soft drinks and dried, frozen and processed foods where conforms might not have survived (108). Although the enterococci are not entirely host specific, they do show some degree of host specificity (70). The proportions of enterococcal species differ in various animal and human feces. 2

Therefore, the identification of enterococci rather than their enumeration was proposed to pinpoint the origin of fecal contamination (124). The enterococci also cause spoilage of pork and other meat products (64), and they have been implicated as a cause of diarrhea in suckling rats (48) and possibly humans

(71). Besides possible food poisoning, enterococci cause a number of other human infections, such as urinary tract infections, bacteremia, and endocarditis (114).

In this study, a schema for differentiating 13 species of the genus

Enterococcus and S. bovis and S. equinus by using rapid test kits and a minimum of supplemental tests was proposed. Using this schema, identifications of isolates from pork slaughtering plants and processed pork were made; counts also were compared. Two media selective for enterococci, fGTC and KF agars, were compared for recovery of enterococci from pork carcasses, and processed pork, beef, and poultry. Identities of isolates on both media also were compared. Antibiotic resistances of enterococci isolated from pork, water, and clinical infections were examined. Species comparisons of antibiotic resistance were made for 13 enterococcal species and S. bows and S. equinus. Finally, four Lancefield grouping kits were compared for their ability to accurately group the enterococcal isolates, most of which are Lancefield group D.

Dissertation Format

This dissertation includes five papers that are manuscripts. Paper 1 appeared in the September 1992 issue (Volume 9) of Applied and

Environmental Microbiology. Paper 2 has been accepted for publication in the Journal of Food Protection. Paper 3 has been submitted to Applied and 3

Environmental Microbiology as a research note. Paper 4 has been submitted to the Journal of Food Protection. Paper 5 has been submitted to the Journal of Rapid Methods and Automation in Microbiology as a research note. All papers follow the style of the journal to which they were submitted. A general literature review precedes the first manuscript, and a general summary and discussion follows the fifth manuscript. There are separate lists of references for each of the manuscripts as well as a separate list for the Introduction, Literature Review, and Summary and Discussion.

The doctoral candidate, Linda M. Knudtson, was the principal investigator in these studies, but worked in conjunction with several other investigators in the collection of the pork carcass samples. 4

LITERATURE REVIEW

History of the Enterococcl

The term "Enterococcus" was first used by Theircelin and Jouhaud in

1903 as a generic name to describe a Gram-positive diplococcus of intestinal origin. The type species was called Enterococcus proteiformis. In 1906,

Andrewes and Horder renamed this type species Streptococcus faecalis (3).

They believed that it belonged in the genus Streptococcus because it formed chains under some cultural conditions. In later classifications, the term

"enterococcus" was used by many to designate Streptococcus species of fecal origin which had in common some of the outstanding characteristics of

Streptococcus faecalis. In 1937, Sherman (138) proposed a classification scheme which separated the streptococci into four divisions. One of these divisions, the enterococcus division, was comprised of organisms that grew at 10 and 45°C, in 6.5% NaCI, and at pH 9.6, and which survived 60°C for 30 min. The ability to hydrolize esculin was also noted. Although the classification has changed, some of these characteristics are still used today to help identify enterococcl. In 1970, Kalina (82) recommended that current nomenclatural and taxonomic decisions should be reconsidered because chain formation was observed to be the result of unfavorable environmental conditions and the main type of cell arrangement was the diplococccus form.

The pleomorphic nature of the enterococcl also set them apart from the rest of the streptococci. In 1984, Schleifer and Kilpper-Balz (132,133) used DNA-

DNA and DNA-rRNA hybridization to show that Streptococcus faecalis and

Streptococcus faecium were so distantly related to other streptococci. 5

including Streptococcus bovis, that they should be transferred to another genus. They proposed that the genus Streptococcus should be divided into three genera. The first genus, Streptococcus sensu stricto, comprised the majority of the known species, particularly the pyogenic and oral streptococci.

The second genus, Lactococcus, encompassed all lactic Lancefield group N streptococci. Finally, the genus Enterococcus, as previously proposed by

Kalina, contained the enterococcal streptococci. It should be noted that the fecal species S. bovis and Streptococcus equinus, though they reacted with group D antisera and shared several characteristics with the enterococci, were not considered members of the new genus Enterococcus, as they were more closely related to the streptococci by 16S rRNA studies.

The genus Enterococcus consists of gram-positive, facultatively anaerobic organisms that are ovoid in shape and may appear in a gram stain in short chains, in pairs, or as single cells. Like streptococci, they are catalase negative, although some strains do produce pseudocatalase (52,132). Most

(but not all) react with group D antisera and some react with group Q. Hydrolysis of L-pyrrolidonyl-p-naphthylamide (PYR) is a characteristic feature, although some species are PYR-negative (39,128,152). Most strains of the newly defined genus Enterococcus possess the characteristics summarized by Sherman (138), although there are species of enterococci that grow poorly at 10 or 45°C. Another key test is the ability to hydrolyze esculin in the presence of bile (49). However, some enterococci require up to 48 hours for the correct reaction to occur. 6

Origin of tlie species of the genus Enterococcus

Enterococcus faecalis was among the first Enterococcus species recognized. Previously Streptococcus faecalis, this group of organisms was transferred to the genus Enterococcus in 1984 by Schliefer and Kilpper-Balz

(132) as proposed by Kaiina (82). This species was first described by

Thiercelin and Jouhaud in 1903, and named Enterococcus proteiformis. In

1906, Andrewes and Horder (3) grouped this species with the streptococci and renamed it Streptococcus faecalis. In 1937, Sherman (138) recognized that several species of streptococci showed only slight variations to S. faecalis isolates and suggested that these species be considered varieties of 5. faecalis becoming S. faecalis var. faecalis, S. faecalis var. liquifaciens, and S. faecalis var. zymogenes. These varieties were differentiated by their ability to be proteolytic and hemolytic (35). Since that time several investigators have questioned these varieties (16, 34, 59, 80,111) because hemolysis and proteolysis were inconsistent within these subspecies. The abilities to liquify gelatin and hemolyze red cells, upon which these varieties were based, were later found to be controlled by plasmids (78, 119) which explains the earlier inconsistencies. Therefore, these subspecies are no longer recognized.

The change from S. faecium to Enterococcus faecium was also proposed by Schleifer and Kilpper-Balz (132). Streptococcus faecium isolates were initially considered to be the same species as S.faecalis. In

1919, Orla-Jensen (cited in (34)) proposed dividing the two species on the basis of fermentation characteristics, tolerance to heat and sodium chloride, and temperature limits of growth. Many researchers supported this division on the basis of many different criteria, and eventually it was generally accepted. 7

Shattock (137) supported the distinction between S. faecalis and S. faecium by reexamining the physiological characteristics of 350 fecal streptococci.

She claimed that S. faecium actually did comprise a legitimate species.

Barnes (5) added another important criterion for differentiation by showing that

S. faecium does not have the ability to reduce tetrazolium to formazan in a glucose medium at pH 6, whereas S. faecalis does reduce tetrazolium.

Whittenbury (151) stated that the two species could be separated on the basis of tellurite tolerance and reducing ability. Many subspecies of S. faecium also existed, but these have all been transferred to other species.

Enterococcus durans is an example of a species that was initially considered a subspecies of S. faecium. It was isolated from dairy products and named Streptococcus durans by Sherman and Wing (139). Later, Deibel,

Lake and Niven (35) noted its similarity to S. faecium and suggested a varietal status for the organism, as did Barnes (6). In 1984, Collins et al. (26) proposed that S. durans (or S. faecium subsp. durans) be named Enterococcus durans.

Isolates of another subspecies of S. faecium, S. faecium subsp. mobilis, were shown to be members of the species Streptococcus casseiifiavus. This subspecies was first described as the motile strains of S. faecium isolated from plant material (93). S. casseiifiavus also began as a subspecies of S.

faecium, S. faecium subsp. casseiifiavus. This subspecies was proposed by

Mundt and Graham (112) to describe a group of motile, yellow pigmented

streptococci isolated from vegetation. These strains were elevated to species level by Vaughan et al. in 1979 (148). It was proposed in 1984 that all strains of S. casseiifiavus, on the basis of biochemical, chemical and genetic data 8

(59, 80,148), should be transferred to the genus Enterococcus as Enterococcus casseliflavus (26).

Pette (59) isolated some streptococcal strains from Gouda cheese in

1955 which he named S. faecalis var. malodoratus. A numerical phenetic

study by Jones et al. (81) indicated that these strains were more closely

related to S. faecium than S. faecalis. In 1984, Collins et al. and others (26,

59) determined that S. faecalis var. malodoratus was sufficiently distinct to be considered a separate species. It was renamed Enterococcus malodoratus.

The species Streptococcus avium was proposed by Nowlan and Deibel (118) to accommodate streptococci resembling S. faeca/Zs and S. faeciumoi serological group Q. The numerical phenetic study of Jones et al. (81) in 1972 indicated that some similarity existed between S. avium and S. faecium but the similarities were not sufficient for S. avium strains to be included in the species S. faecium. In 1984, Collins et al. (26) also proposed transferring S. avium to the new genus as Enterococcus avium.

Barnes et al. (8) isolated from the intestines of young chickens a number of streptococci of serological group D which were apparently distinct from S. faecalis, S. faecium and S. durans. The DNA homology data performed by Farrow et al. (59) indicated that the streptococci isolated by

Barnes represented a distinct species. This finding was supported by Bridge and Sneath (15), who named this group of isolates Streptococcus gallinarum.

This species was transferred by Collins et al. (26) in 1984 to the genus

Enterococcus and renamed Enterococcus gallinarum. This species also included a variety of motile strains isolated from clinical infections which had 9

been called E. faecalis or £ faecium (although neither is considered motile) (25).

Schleifer and Kilpper-Balz (132) reported that 3 strains designated E.

faecium showed very little DNA sequence homology (<40%) with the type

strain of E. faecium. These three strains were ancestrally related to a common

strain (NCDO 1258) which also was found to be distinct from E. faecium and

E. durans. They suggested that these strains may represent a new species

(132). Other strains, intermediate between E. durans and E. faecium, which

caused growth depression in young chickens (63) were believed to belong to a new species. Farrow and Collins (58) studied these "atypical" E. faecium strains, as well as some unclassified group D streptococci from Sharpe and

Fewins (136). Their results (58) indicated that these strains represented a new species of enterococcus, for which the name Enterococcus hirae was proposed.

Until recently a number of "enterococcus-like" strains remained unclassified. Some of these were yellow-pigmented isolates from plants (148) and soil (144). Primary studies indicated that they were not related to the yellow-pigmented E. casselifiavus (148). Other isolates from cows (69) had previously been shown by Farrow et al. (59) to be distinct from recognized

Enterococcus species. Studies on these isolates by Collins et al. (25) found them to belong to a new species and the name Enterococcus mundtii was proposed. E. mundtii has also been isolated from normally sterile body sites in two septic patients (84); this was the first detailed description of E. mundtii

as a cause of human infection. 10

Unclassified enterococci from clinical sources were studied by Collins et al. (24). Nucleic acid studies and penicillin-binding protein patterns demonstrated that 6 human strains, a single clinical isolate and a strain from bovine mastitis were all genetically distinct from each other and from all previously described Enterococcus species. For the 6 human strains, the name Enterococcus raffinosus was proposed due to their ability to metabolize raffinose. For the strain from bovine mastitis, the name Enterococcus pseudoavium was proposed. For the single clinical isolate (from ear exudate), the name Enterococcus solitarius was proposed.

De Vriese et al. (40) described a species. Streptococcus cecorum, as carboxyphilic strains isolated from the caeca of chickens. It resembles the enterococci in many respects, but it fails to grow at 10°C and in 6.5 % NaCI, characteristics generally attributed to enterococci. It also fails to react with group D antiserum, unlike all other enterococcal species to date. Sequencing of 16S rRNA by reverse transcriptase was performed by Williams et al. (152) to clarify its taxonomic position. Streptococcus cecorum was clustered with the enterococcal species and was only distantly related to the other streptococci, suggesting that S. cecorum is phylogenetically a member of the genus Enterococcus. Thus, the name Enterococcus cecorum was proposed.

Further studies on £ cecorum carried out by De Vriese et al. (38) showed that this species was not limited to growth in chickens. It was isolated from the intestines of pigs, cattle, horses, a mallard duck, and canaries (where it appears to predominate).

Reverse transcriptase sequencing of 16S rRNA was performed on another Streptococcus species to clarify its phylogenetic position. 11

Streptococcus saccharolyticus, first described by Farrow et al. (1984), included atypical S, bows-like strains isolated from cows and straw bedding.

Although distinct from S. bovis, S. saccharolyticus failed to react with group D antisera and was placed in the group Streptococcus sensu stricto. It did, however, more closely resemble members of the genus Enterococcus.

Research by Rodrigues and Collins (128), using reverse transcriptase sequencing, revealed that S. saccharolyticus was only distantly related to members of the genus Streptococcus sensu stricto and formed a distinct group with E. faecalis. Therefore, it was proposed that S. saccharolyticus should be reclassified as Enterococcus saccharolyticus comb. nov.

A new species of enterococcus, closely related to E. cecorum and E. avium, was proposed by De Vriese et al. (39), This species, Enterococcus columbae, dominates in the intestinal flora of domestic pigeons. Another new species of Enterococcus, Enterococcus dispar, was proposed by Collins et al. (27). This species was discovered during a survey of atypical enterococci from human sources. They were isolated from synovial fluid and stool. It is most closely related to E. hirae in biochemical characteristics.

Streptococcicosis, a disease of fish, is the most important disease of yellowtail fish in Japan. The disease agent, previously identified as a streptococcus, did not fit into any of the species yet described. A study by Kusada et al. (89) determined that the causative agent of the disease is a new species of Enterococcus, Enterococcus seriolicida.

In 1991, Martinez-Murcia and Collins proposed yet another new species of enterococcus (102). Enterococcus sulfureus was proposed as the name for 3 unknown non-motile, yellow-pigmented enterococci originating 12

from plants. The isolates were closely related to E. casseliflavus and E. mundtii, both yellow-pigmented enterococcal species.

The newest members of the genus Enterococcus, Enterococcus flavescens, are yellow pigment-producing strains of clinical origin. Three

yellow-pigmented strains of enterococci (£ casseliflavus, E. mundtii, and E.

sulfureus) had already been described (25, 26,102). Almost all of these had been isolated mainly from environmental materials (plants, soil etc.) and rarely from clinical material (130). Pompei et al. (121) studied four atypical yellow- pigmented group D isolates from severe human infections (122). By using

DNA-DNA hybridization tests and fatty acid content determinations, they found these 4 isolates comprised a new species and proposed the name Enterococcus flavescens.

Importance of the Enterococci Habitat of the enterococci

Enterococci are found in the feces of most healthy adults. When enterococci from feces have been identified to species level, many studies report that E. faecalis is most common and is found in higher numbers than E.

faecium (114). Besides the alimentary tract of humans , enterococci also reside in many other animals. Enterococcal species have been isolated from buffalo, cattle, sheep, camels, pigs, horses, mules, donkey, rabbits, chickens and geese (126). Mundt (109,110) surveyed the incidence of enterococci in wild animals, reptiles and birds. He noted that many wild mammals and reptiles excreted enterococci (34). Although Mundt did not associate enterococci with birds, new species of enterococci have been defined. E. 13

cecorum was isolated from canaries and a mallard duck (37), E. hirae and E. avium from chickens (26, 58), and E. columbae from domestic pigeons (39).

These species had not been defined at the time of Mundt's survey. Although enterococci are natural residents of the intestinal tracts of man and other warm- and cold-blooded animals, their resistance and tolerance, together with their low minimum and high maximum temperature limits of growth, fit them for survival and even growth under diverse conditions in nature (138). In humans, they are infrequently found in vaginal and oral specimens (such as dental plaques) (114). Mundt (110) studied the distribution of enterococci in insects and concluded that in the insect digestive tract, enterococci are transient, and their occurrence on the insect exterior was due most probably to mechanical transfer. Extensive studies verified the common occurrence of enterococci on plants, particularly on domestic plants. It was believed that an epiphytic relationship occurred since the bacteria could establish a cycle in plants with transmission in the plant seed (113). Contamination of plants probably occurred through insects and wind. Several species of enterococci have been isolated from plants and soil. E. casseliflavus, E. mundtii, and E. sulfureus are all yellow-pigmented enterococci and all have been isolated from plants and soil (25, 26,102). There is general agreement that enterococci are not native to soil, and their presence in soil samples represents contamination from either plant or animal sources. In this environment, the enterococci are disseminated most probably by wind, rain, and insects. 14

Enterococci as fecal indicators in water and food

An ideal indicator organism should 1) be associated in high numbers

with feces and/or intestinal pathogens, 2) not multiply in the environment, 3)

die off less rapidly than potential pathogens, and 4) withstand the processing

conditions undergone by the substrate (62). Questions had arisen regarding the complete reliance on the use of Escherichia coli as the sole indicator

organism. Its relatively rapid death in water, in comparison with pathogenic

organisms (including enteric viruses), had led some public health officials to question its utility as the sole index microorganism for fecal contamination (34). In the routine procedures employed , E. coli may be confused with physiologically similar bacteria. Many workers have emphasized the importance of enterococci as indicators of fecal pollution (7,108,126,151). In the past, the acceptance of enterococci as pollution indicators also had been questioned. E. coli was easier to detect in water, whereas there was a lack of quantitative recovery methods for enterococci. The absence of information regarding the sources of enterococci, questions concerning their classification, and their significance in the water supply also were concerns (34). Questions regarding their classification were answered with a new classification system (132). Many studies have been conducted on enumeration of enterococci and many selective and differential media are available (72). Although enterococci are widely distributed in nature, the major natural habitat of these organisms is the intestinal tract of man and animals (62). Thus, the occurrence of these bacteria in water implies either direct or indirect fecal contamination. Characteristically, enterococci are not 15

detected in waters free from fecal contamination even though they are found on plants and insects that could contaminate water.

The ratio fecal coliforms/fecal streptococci was proposed by Geldreich and Kenner in 1969 (124). It was used to identify the origin of pollution of surface waters. A ratio above 4 suggests human contamination and a ratio below 0.7 contamination by animals. The persistence of both indicators after discharge of feces in water, however, is unclear. According to Bartley and

Slanetz (9), after initial contamination both groups may exhibit a slight increase in numbers (presumably due to the organic matter in the water and the temperature) followed by a pronounced decrease. An early study of the coWiorm-Streptococcus ratio in various water sources indicated a more rapid decrease in conforms. Recent studies also have shown that E. coli is more sensitive than enterococci to natural water (124). It is important to note that S. bovis also dies off rapidly in water. As £ coli numbers decrease the ratio approaches unity and the ratio is no longer accurate for identification of fecal origin. It is estimated that the ratio would only be valid for the first 24 hours

(124). The choice of selective media adds variability to the ratio approach.

Many investigators have reported a lack of correlation between

Enterococcus and E. coli counts, and the unreliability of enterococcal counts as a reflection of fecal contamination is established rather well. Because the proportions of the various enterococcal species are not the same in different animal and human feces, the identification of enterococcal species rather than their enumeration was proposed as a possible solution (124). None of the enterococci can be considered absolutely host specific, but some species evidence a degree of host specificity (70). There are reports indicating that 16

while S. faecalis (£ faecalis) predominates in the human intestines, S. bovis

and S. faecium (E. faecium) are the predominant species of animals, including

swine (126). E. faecalis dominated the enterococcal flora of human origin and birds feces, whereas it was absent or in low numbers in cow, pig, sheep, or

rabbit feces. E. faecium, on the other hand, showed a wider range (124),

Studies on the flora of British pigs revealed that S. faecium (E. faecium) is

commonly present in the pig intestine, while £ faecalis is rare (7). Kjellander

(87) in 1960, also observed that S. faecalis (E. faecalis) was more numerous in humans, whereas S. faecium (E. faecium) was of animal origin. The presence of S. bovis in humans would be associated with disease (12, 97). According to De Vriese et al. (41), the enterococcal flora of poultry comprises a relatively small number of frequently occurring species. E. faecium, E.

cecorum, E. faecalis, E. hirae and E. durans were regularly isolated, while E.

mundtii, E. casseliflavus, E. gallinarum and E. avium were only rarely isolated.

Another study by De Vriese et al. (43) revealed that no E. avium were isolated from chickens. This finding is odd since the original strains of E. avium were isolated from the feces of chickens (26). A study of enterococci isolated from calves, young cattle, and dairy cows (42) revealed that enterococcal species are rare in ruminating cattle, but E. faecalis was isolated from nearly half of the pre-ruminating calves. Variation in the enterococcal flora of animals has been associated with many factors. Geographical location, diet, age, species of the

animal, and even seasonal effects have been observed to contribute to this

variation (34). Despite this potential for variation, significant differences in the distribution of the enterococci and fecal streptococci have been observed in 17

host animals. It is believed that identification of enterococci and fecal

streptococci provides valuable information on the origin of pollution (124).

Enterococci are important as fecal indicator organisms in foods

because they are readily isolated in large numbers from human and animal

feces. The relative resistance shown by their cells against adverse conditions, such as tolerance to extremes in temperature (64), pH, and salinity, proves

advantageous in the bacteriological examination of sea water, soft drinks and

dried, frozen and processed foods (108) where coliforms might not have

survived. The use of enterococci as a fecal indicator in frozen foods was discussed by Deibel (34). Most, but not all, of these foods undergo some

thermal processing or precooking prior to freezing, so bacteria found in these

products represents recontamination after thermal processing prior to

freezing. Since E. coli is particularly susceptible to freezing and enterococci

are quite resistant, enterococci provide a more reliable index to the sanitary

history of frozen food. Enterococci may occur in comminuted, cured meat

products, either as the result of survival of thermal processing or from

postprocessing contamination. These products are heated to kill most vegetative bacteria; however, processors assume that raw sausage mix does

not contain excessive bacterial numbers. The high heat and salt resistance of the enterococci as well as high initial population in the sausage mix are

factors contributing to their survival in marginally processed products. Even

under ideal conditions the products could be recontaminated during

subsequent slicing and prepackaging. In these instances, the occurrence of

enterococci does not necessarily suggest direct fecal contamination (34).

Quite often enterococci become established in food plants and grow in areas 18

far removed from the original source of fecal contamination. Therefore,

caution and discretion must be exercised in attributing significance to the

numbers and species of enterococci and fecal streptococci present in foods

(70).

Enterococci in meats

Enterococci as a cause of food spoilage

Enterococci present in foods are not only important as indicators of

fecal pollution, they are also a serious cause of food spoilage, especially in meats. There are many reasons for spoilage in meat and meat products. The

following factors can be important: meat from sick or stressed animals; poor

slaughtering hygiene; inadequate chilling of meat during transport, cutting or

storage; poor hygiene amongst personnel; insufficient cleaning and

disinfection of equipment and food-handling surfaces; processing of heavily

contaminated meat and other ingredients; inadequate heat treatment for refrigerated products or for canned products that can be stored unrefrigerated; and inadequate refrigeration of products after purchase (73). Whether the

origin of the enterococci lies with fecal pollution or contamination of the food plant, the presence of food spoilage organisms greatly decreases the shelf-life of meat products and other foods. A number of interrelated factors influence the shelf-life of meat, specifically holding temperature, atmospheric oxygen, indigenous enzymes, light, and microorganisms. Although some deterioration of meat will occur in the absence of microorganisms, microbial growth is by far the most important factor in relation to keeping the fresh quality of meat (91 ).

The initial bacterial load of meat has a major influence on its shelf-life. At

lower cooking temperatures, the shelf life and microbiological safety of the 19

product are reduced because a greater number of organisms survive the process. The enterococci are the major group of heat-resistant organisms that survive the processing of cured meat products (64). However, excessively high numbers of thermoduric enterococci in cured meats indicate inadequate processing. Both E. faecalis and £ faecium have been implicated in the spoilage of these products. Enterococci have also been shown to cause food spoilage in canned hams (116,136). The shelf life of sliced, prepackaged ham (and sometimes other similarly prepared cured meats) also may be dictated by the initial numbers of contaminating enterococci (70). These bacteria produce sour flavors, discoloration, gas, slime, and milky exudates.

Dykes et al. (46) performed quantitation of microbial populations associated with the manufacture of vacuum-packaged, smoked Vienna sausages. They concluded that the lactic acid bacteria contaminated sausage surfaces as a result of manufacturing and handling processes. Complete avoidance of enterococci in these products is difficult, and control of numbers rather than avoidance of occurrence must be practiced. Hazard Analvsis Critical Control Points

Good manufacturing practice (GMP) during slaughter, as described in the Recommended International Code of Hygienic Practice for Fresh Meat, includes all necessary measures to produce meat with the lowest possible microbial contamination. This is attainable only if the whole process is strictly controlled. The hazard analysis critical control point (HACCP) concept, as reviewed by Pierson and Corlett, Jr. (120) and Tompkin (146), provides a systematic approach to achieve this end. If the goal of reduced foodborne illness is to be achieved, it is necessary to identify the errors which are 20

involved in food preparation. HACCP plans must be customized to the unique conditions existing within each establishment, but it is possible to generalize on similar procedures in the production of meat and poultry products.

The first step toward establishing a HACCP plan in a meat or poultry plant is to conduct a hazard analysis. Hazard has been defined as the unacceptable contamination, growth or survival by microorganisms of concern to safety or spoilage: and/or the unacceptable production or persistence in foods of microbial metabolic products such as toxins, enzymes or amines.

The purpose of hazard analysis is to identify potential problems which could occur in an operation. Examples of potential hazards in a meat or poultry plant include; raw materials with a history of causing microbiological problems; sites of contamination in the process; and the potential for microorganisms to survive or multiply during production, storage, distribution, or use.

After hazards have been identified, procedures must be established for their control. The definition for critical control point (CCP) is important to the meat and poultry processor and regulator because it defines the limits of what should be achieved when a HACCP program is established. The current

ICMSF definition for CCP is "a location, practice, procedure or process at which control can be exercised over one or more factors which, if controlled, could minimize or prevent a hazard" (146). This definition states the value of a

CCP, while limiting it to steps in a process where some degree of control is possible. The definition allows for and encourages the adoption of CCPs to minimize contamination with enteropathogens during the slaughtering processes and subsequent handling of raw meat and poultry. There are two 21

levels of CCP, depending on the confidence that the hazard can be controlled.

A CCP1 will assure control (ie. cooking to 63°C assures destruction of salmonellae in raw meat), while a CCP2 can only minimize the hazard (proper care in eviscerating reduces incidence of salmonellae on fresh meat). Control means managing the conditions of an operation to maintain compliance with established criteria. If the criteria for each CCP are met, the hazards will be reduced or eliminated. A key element of the HACCP system is the use of measurements to monitor and verify whether established criteria are being met. All of the monitoring and verification measurements that are taken must be recorded. Included must be action taken when the established criteria have been exceeded.

The concept of HACCP is applicable to a wide variety of problems. It is a common sense approach to preventing problems. It is the best system currently available for improving the microbiological safety of food. In the

Federal Republic of Germany a new slaughter hygiene monitoring system (CR monitoring) was tested. It is the German equivalent to HACCP. According to Schutz et al. (135), during a 7-month experimental period using CR monitoring, it was found to be an effective means of control and an efficient monitoring system. If carried through consistently, it causes a noticeable reduction in visible carcass dirt. Meat samplino techniques

Sampling of carcass surfaces for microbiological examination should be used to identify critical control points in slaughterlines. The data should then be used to monitor the process to attain a good end product. Sampling techniques have to be accurate and precise to make valid microbiological 22

comparisons of different steps in slaughtering plants. Various techniques have been suggested for determining bacterial colony counts on carcasses.

Those commonly used are the agar contact technique, the swab technique, and the excision technique. Snijers et al. (143) compared four sampling techniques: the excision, the double swab, agar contact and modified agar contact. In the agar contact technique (called agar sausage technique by other authors) an agar surface is pressed onto the test area and incubated. The modified technique involves homogenization of impressed agar slices in peptone saline and pour plating of the samples. The swab technique relies on rubbing one swab (or two swabs) on a test surface, transferring it (them) to a dilution bottle, mixing to release the bacteria from the swab, diluting, and plating on appropriate media. In the excision technique, pieces of tissue are removed and homogenized in a solution and plated. Snijers et al. (143) showed that the agar contact technique cannot be used for determining contamination of carcasses because plates were overcrowded. The double swab and modified agar contact techniques yielded significantly higher standard deviations than did the excision technique. The excision technique was the most suitable method in view of its high accuracy and precision.

Several other researchers obtained similar results (2, 60). Nortje et al. (117) also reported that the excision technique was the most reliable. Morgan et al.

(107) stated that the method used to collect microbial samples from carcasses should be simple, non-destructive, reproducible and economical. When pork carcasses were sampled, the excision technique was admittedly the most reliable. It is used as the standard against which all other methods were evaluated. This technique, however, is not practical in a commercial 23

environment due to carcass mutilation; swabbing, despite its drawbacks, is still the most universally used method. Lasta and Fonrouge (94) used the swab technique to ascertain whether small sampling areas (10 to 100 cm^) from bovine carcasses were characteristic of the hygiene level in the slaughtering plant. They concluded that the most heavily contaminated areas of the carcass are generally small; the microorganisms were not evenly distributed. Contamination is random and comes from touching or rubbing the carcass with hands, clothes, tools, equipment, hides, and spattering with different materials (feces, water, pus, etc.). Based on their results, they concluded that the bacterial count from relatively small areas of the carcass (less than 100cm2) was not an adequate indicator of hygiene.

Homogenization by blending and stomaching was compared by Dickson for the recovery of Listeria monocytogenes from inoculated beef tissue (44). No difference between these two methods was detected. He did find, however, that phosphate buffer was slightly inferior to buffered peptone as a diluent.

Sampling methods for poultry carcasses are somewhat different.

Lillard (96) compared the whole carcass rinse with the stomaching or blending of excised skin for sampling broilers. She demonstrated that these three sampling methods most commonly used resulted in the isolation of equivalent levels of aerobic bacteria and Enterobacteriaceae, but only a small portion of the total bacterial load was recovered by any sampling method.

Enterococci in Food Poisoning

Enterococci also have been implicated as a cause of food poisoning.

Murray (114) stated that this was a misconception which probably arose because of the occurrence of enterococci in the intestinal tract and, therefore. 24

on fecally contaminated food. In a recent study, a strain of Enterococcus hirae

from a stool of a human with diarrhea was implicated as the cause of diarrhea

in suckling rats (48). In 1924, enterococci were suggested as the cause of

food borne illness in an outbreak ascribed to milk. The most convincing

evidence that enterococci cause food poisoning was obtained in feeding tests

on human volunteers by Carey et al. In 1931 (19). These volunteers became

ill when they ingested whole-cell enterococcal cultures. But these results could not be consistently repeated by other investigators. A collection of

strains implicated in various food-poisoning outbreaks indicated an

approximate equal distribution of the S. faecalis and S. faecium species (34).

Hartman et al. in 1965 (71) suggested that for enterococcal food poisoning to

occur, the conditions must be "right". The question then, is what are these

conditions and how do we avoid them when processing foods implicated in

enterococcal food poisoning? Variations in temperature, pH, freshness of cultures, and synergism all have been suggested as possible contributing

factors in enterococcal food poisoning (71).

Clinical Importance of the Enterococci

Urinary tract infections (UTI) are commonly caused by enterococci. The rate of urinary colonization by enterococci rises in patients who have been instrumented, received antibiotics (especially cephalosporins), have structural

abnormalities, and/or recurrent UTIs (95, 114). In healthy patients enterococci cause less than 5% of UTIs.

Enterococci also cause an estimated 5-15% of the cases of bacterial

endocarditis (20). As with other human enterococcal infections, most isolates 25

are identified as E. faecalis. However, otfier species also cause this disease.

E. avium, E. casseliflavus, E. durans, E. gallinarum, and E. raffinosus as well

as E. faecium also have been isolated (53). Although enterococcal endocarditis usually affects older adults (56-59 years of age), it occasionally occurs in children (20,114). Fifty percent of men with enterococcal

endocarditis have a history of previous enterococcal urinary tract infection

(UTI) or genitourinary tract instrumentation. Forty-three percent of women have a history of childbirth or abortion in the preceding three months.

Patients with underlying valvular heart disease and IV drug users also are at risk,

Enterococcal bacteremia is much more common than enterococcal endocarditis. The incidence of enterococcal bacteremia appears to be increasing (20, 95,114). There has been a steady increase in patients with enterococcal bacteremia since 1975, even though there was no increase in admissions. This increase was entirely due to an increase in nosocomial bacteremias (100). Possible sources of enteroccoccal bacteremia include infection or colonization of the genitourinary, gastrointestinal, and hepatobiliary tracts. The most common source in several studies had been the urinary tract. Enterococci also have been reported in secondary bacteremia in patients with postoperative wound infections, pyelonephritis, and many gynecological infections. Intravascular catheters are also a major source of enterococcal bacteremia (100). Mortality of enterococcal bacteremia has generally been high because of underlying complicating factors. These factors include association with burns, hospital-acquired infections, and serious underlying illness. 26

Enterococci have been implicated in several other human infections. In intra-abdominal infections, the role of enterococci is controversial. It has been suggested that enterococci are pathogenic in these infections only synergistically with anaerobes (20). Although group B streptococci are the most common cause of neonatal infections, it has been well documented that enterococci also can cause infections in neonates. An outbreak of E. faecium occurred in premature infants with severe underlying disease, nasogastric tubes, and multiple intravascular devices. Positive blood cultures and CSF samples were recovered. It was concluded that all the E. faecium isolates were the same strain, and the outbreak was spread by hospital personnel. In addition to neonatal meningitis, enterococci can cause central nervous system

(CNS) infections in older children and adults. Most cases are related to an underlying disorder such as a long-term primary illness, invasive CNS procedures, prior antibiotic therapy or all three (114).

Enterococcal infections are most often nosocomial. These nosocomial infections account for as much as 10% of all hospital infections (95). The numbers are increasing. In 1984, they were the third leading cause of urinary tract infections and the sixth leading cause of bacteremia (77). To account for the increased infection rate, attention has been drawn to the growing use of broad-spectrum antibiotics, cephalosporins in particular. Treatment with these agents, which lack appreciable activity against the enterococci, would be expected to provide this organism with a selective growth advantage, leaving patients vulnerable to superinfection (77). It is also possible that the increase may be caused in part by factors other than exposure to antibiotic therapy.

Greater reliance on indwelling urinary catheters and intravascular devices in 27

the care of the hospitalized patient may be another possible cause. Patient- to-patient transmission, and even interhospital spread of the organism can occur. The epidemiology of nosocomial infection caused by enterococci is similar to that caused by methicillin-resistant staphylococci. Therefore, control measures such as those taken for methicillin-resistant staphylococci should be taken to prevent nosocomial spread of enterococci. The most likely way these resistant bacteria are spread is from an infected patient by transient carriage on the hands of personnel. Nosocomial enterococci often show high- level resistance to aminoglycosides (95).

Enterococci do not possess the potent virulence factors associated with certain other bacterial species. However, enterococci possess a number of characteristics which make them particularly capable of surviving and causing disease in this antibiotic era. They are intrinsically resistant to a number of antimicrobial agents, including p-lactam antibiotics and other agents that inhibit cell wall synthesis, the polymixins, and lincosamides (clindamycin and lincomycin) (47). Two mechanisms are responsible for resistance to p-lactam antibiotics, low affinity of the penicillin-binding-proteins and production of p- lactamase (61). They also are intrinsically resistant to clinically low levels of aminoglycosides. However, cell wall active agents plus aminoglycosides have been effective in the therapy of serious infections caused by enterococci.

To make matters worse, enterococci have recently acquired resistance to a number of clinically important antimicrobial agents, including high-levels of aminoglycosides. In one animoglycoside, gentamicin, high level resistance is due to the presence of a 2'-phosphorylating enzyme which inactivates the drug. This enzyme also has 6'-acetylating activity and is mediated by a 28

transmissible genetic determinant comprised of two fused genes. IVIost isolates studied also have streptomycin adenylating activity and are resistant to streptomycin (47). One of the major reasons for the rapid dissemination of antibiotic resistance determinants in enterococci is the ability of these organisms to exchange genetic elements among one another and with other bacteria, particularly other gram-positive cocci including staphylococci (22,

104). Aminoglycoside-modifying enzymes responsible for high-level aminoglycoside resistance to gentamicin in E. faecalis infections have been shown to be very similar to those isolated from cultures of aminoglycoside resistant Staphylococcus aureus. Schaberg and Zervos (131) were able to identify gentamicin resistance genes in strains of gentamicin resistant E.

faecalis when a staphylococcal plasmid encoding a similar resistance determinant was used as a probe, but not when a deletion mutant of the identical plasmid was used. Many of the important resistance determinants have been shown to reside on transmissible plasmids. Resistance to chloramphenicol, for example, is mediated by a plasmid carrying chloramphenicol acetyltransferase (114). In addition, enterococci are able to transfer resistance on transposons, without transmission of plasmids. The intestinal location of these organisms and their abundance of plasmids and transposons suggests that they may serve as a significant reservoir of genetic information available to other gram-positive bacteria residing in the gut (22). The recent acquisition of plasmid-borne resistance to gentamicin and p- lactamase-associated penicillin resistance, has left vancomycin as one of the few remaining drugs to treat serious enterococcal infections. Very recently, however, plasmid-borne vancomycin resistance was detected in a few clinical 29

isolates (140). Vancomycin resistance was inducible and transferable by conjugation or mobilization between species of enterococci. A membrane protein which may block the access of the antibiotic to its peptidoglycan target is responsible for the resistance (140).

With the increased emergence of antibiotic resistance, rapid and reliable methods of detecting both low-level resistance and high-level resistance to a number of antibiotics has become increasingly important. Routine susceptibility testing (especially for aminoglycosides and |3-lactam antibiotics) of clinical enterococcal isolates is of primary importance in the care of patients suffering from enterococcal infections (66). Knowledge of the prevalence of these resistant strains cultured from a hospital's patient population can be used as a guide for appropriate antimicrobial therapy (75).

The association of antibiotic resistance and species identification of isolates has also received considerable attention. With the recent changes in classification, and the knowledge that the rapid methods used in hospitals are not adequate to accurately identify the enterococci, many researchers are reclassifying previously isolated enterococci and performing antibiotic susceptibilities. They are finding differences in patterns of resistance between different species of enterococci. Moellering et al. (105) stated that

Enterococcus faecium displays greater resistance than E. faecalis to penicillin and certain synergistic drugs. Louie et al. (98) reported that E. faecalis and £ faecium are the most predominant species of enterococci encountered in human infections. In general, E. faecium strains also are less susceptible to p- lactams and aminoglycosides than are E. faecalis strains. E. faecium strains also are often more refractory to the synergistic effect of the antibiotic 30

combination. The possible existence of similar susceptibility differences among the more recently described species of Enterococcus suggests an asssessment of possible correlations between a certain species and a given susceptibility pattern could provide valuable information (129). Gray et al. (65) found several differences in antibiotic resistance patterns between different species. More than half of E. faecium isolates were resistant to ampicillin, but all E faecalis strains were sensitive to ampicillin. High-level gentamicin resistance was seen in E faecalis, but no other enterococcal species. Two studies on E raffinosus from clinical infections were performed. In one study,

E. raffinosus isolates showed higher levels of resistance to penicillin and kanamycin than E avium isolates (66). The second study revealed that high- level ampicillin resistance existed among isolates of E raffinosus, but there were no significant differences between patients with ampicillin-resistant E raffinosus and those with ampicillin-sensitive E. raffinosus (21).

Isolation and Identification Methods for the Enterococci

Selective media

As the importance of the enterococci as fecal indicators, in food spoilage, in clinical infections, and antibiotic resistance emerged, the desire to isolate and grow these organisms increased. Many media have been developed to select for enterococcal species and differentiate them from other types of bacteria. Applications of media parallel to a great extent the development of taxonomy of the streptococci (72). Initially, most media were designed to isolate streptococci associated with human and animal infections. Many of these early media may have been selective for most species of enterococci, but the nomenclature of the group was not defined well enough 31

to recognize the importance of the discovery. Improvements on these media were made to adapt them to the isolation of the enterococci.

Krumwiede and Pratt (88), in 1914, found that some streptococci were more resistant than were other bacteria to gentian violet. This led to the addition of small amounts of crystal violet to glucose broth to produce a selective medium. Edwards devised an agar medium containing crystal violet and esculin. This was to become the predecessor of other esculin-containing media. It is important to note that the concentrations of these inhibitors must be carefully chosen. The concentration must be high enough to inhibit other bacteria, but not so high as to inhibit the enterococci.

Media containing dyes alone were not selective enough, so other inhibitors were studied, Hartmann (cited in (72)) examined the relative inhibition of E. coli and streptococci by a variety of compounds, and found that azide was the only compound that inhibited £ coli more than the streptococci at a relatively wide range of concentration. Azide exerts its primary function by inhibiting metalloporphyrin enzyme systems, such as catalases and cytochrome c oxidases. Electron transport is interrupted. Azide penetrates some cells only as the disassociated acid, so the pH of the medium can have a great effect on the selective properties of the medium. The addition of

0.02% sodium azide allowed Mallmann (101) to estimate selectively the numbers of streptococci in samples of sewage. McKenzie (103) preferred thallous-crystal violet medium. The use of a single selective ingredient, azide, leaves much to be desired when grossly contaminated samples are tested.

Thus, the use of azide alone in media is restricted to preenrichment prior to confirmation in a more selective medium. One exception to this is M- 32

enterococcus agar devised by Slanetz and Bartley (141). This medium is used for recovery of enterococci from water samples. Lachica and Hartman (90) modified this medium by adding Tween 80, KH2PO4 and NaHCOa. They called this modified medium Tween-carbonate medium and used it to recover enterococci from frozen foods. Higher counts were obtained by using the new Tween-carbonate medium than the original M-enterococcus medium of

Slanetz and Bartley, According to Hartman et al. (72) the use of Tween and carbonate may reverse in part the inhibitory power of the azide, and the carbonate itself may be contributing to the selectivity of this medium.

In 1961, Kenner et al. (86) described new solid and liquid media for enumerating enterococci. These media contained, among other ingredients, sodium azide, bromcresol purple, and 2,3,5-triphenyltetrazolium chloride

(TTC). These media were called KF streptococcal broth and agar. Kenner's data (86) indicate that typical strains of S. bovis, S. equinus, S. mitis, S. salivarius, and the enterococcal group and its biotypes produce growth in or on KF media. Growth on KF agar was counted as all colonies on the plate with a red or pink color visible with 15X magnification after 48 hours of incubation. Negative growth was observed with S. cremoris, S. lactis, S. pyogenes, S. thermophilus, and S. uteris as well as some non-streptococcal species.

Raibaud et al. (125), also in 1961, devised yet another medium to enumerate and identify dominant streptococci, this time in pigs. This medium

(AGAT) contained sodium azide and a mixture of sodium glutamate and acridine orange as inhibitors, plus triphenyl tetrazolium chloride as a redox indicator. This medium was supposed to allow selective enumeration of 33

Streptococci in the presence of large numbers of lactobacilli. Azide-containing

media possess disadvantages for certain applications because of the

instability of azide (72) and and the failure of some streptococci, such as S.

bovis to initiate growth on azide-containing media (45).

Besides azide, other inhibitory substances have been studied. One of these was thallium salts. After reports that certain gram-positive cocci were

more resistant than other bacteria to thallium salts, McKenzie examined the

selectivity of thallium salts in detail (103). He developed a thallium acetate (TA)-crystal violet broth. The selectivity of TA is not affected by minor changes in pH, although the selectivity of sodium azide is. Barnes (5) incorporated

thallous acetate into a tetrazolium agar medium. The medium was used to

determine numbers and types of fecal streptococci in bacon factories (4). The thallous acetate suppressed most other organisms and the tetrazolium salt differentiated Streptococcus faecalis and its variants from other Lancefield

group 0 organisms. S. faecalis reduced the tetrazolium to the insoluble red

formazan so colonies had dark red centers; S. faecium, S. durans, and S. bovis did not reduce it and formed white to very pale pink colonies.

Thallium salts also were used in conjunction with other ingredients to increase the selectivity for certain organisms. Thallium acetate was incorporated into media with 0.5% NaCl, in spite of the fact that McKenzie

(103) mentioned that TA, in concentrations of greater than 0.1%, reacted with

NaCl to yield insoluble and nonselective thallium chloride. Barnes (4),

however, could not find evidence that concentrations of NaCl up to 0.5%

reduced the inhibitory properties of TA. The use of esculin in media containing thallium salts was also investigated. However, the variability in the 34

ability of an organism to ferment esculin was noted years ago (72); some organisms were weakened and give a negative result, yet their activity could be regained on subculture. Lachica and Hartman (90) described the use of citrate in combination with thallous acetate. Citrate functioned as the primary energy source, as well as a selective agent. It was observed that citrate

(1.0%) was inhibitory to some lactobacilli but stimulatory to enterococci (18).

Gentamicin had been used by Black and Van Buskirk (14) for isolating P-hemolytic streptococci; growth of staphylococci and nearly all gram-negative bacilli was inhibited. Based on these favorable results, Donnelly and Hartman

(45) developed a gentamicin-based medium containing thallous acetate for the selective isolation of all group D streptococci. Gentamicin and TA were the major selective agents. NaHCOa, Tween 80, and KH2PO4 were added as specified by Lachica and Hartman (90) to stimulate the growth of group D streptococci. Esculin was included because group D streptococci hydrolyzed it (50), in the presence of ferric citrate, forming dark halos of ferric salts. This medium was superior to TA, KF, and PSE agars for the enumeration of fecal streptococci in fecal and surface-water samples (45). This medium also was tested for its ability to enumerate fecal streptococci in frozen foods (145). In overall efficiency, the GTC agar recovered significantly greater numbers of presumptive fecal streptococci. Both Donnelly and Hartman (45) and Thian and Hartman (145) believed that it was important to incorporate more differential abilities to this medium, to eliminate false positives and distinguish between species. This was accomplished by Littel and Hartman (97) in 1983. GTC medium was modified by incorporating a colorimetric starch substrate and a fluorogenic substrate, allowing differentiation of colonies on the agar 35

surface. The incorporation of amylose azure and 4-methylumbelliferyl-a-D-

gaiactoside (a-D-MUGAL) allowed differentiation of the fecal streptococci into

three phenotypic groups: starch hydrolysis and fluorescence; no starch

hydrolysis but fluorescence; and no starch hydrolysis or fluorescence (97).

Cooper and Ramadan (30) discovered that a broth containing 0.2%

potassium tellurite was suitable for isolating streptococci from the feces of various animals. The use of tellurite agar also was proposed. In 1940, the

complimentary action of penicillin and tellurite was devised. In the same year,

it was shown that enterococci were resistant to penicillin and tellurite. All

enterococci tested were resistant to 0.1% tellurite, while resistance of other

streptococci varied (72). The mechanism of growth inhibition by tellurite is not fully understood. Of the enterococci, E. faecalis is resistant to 0.5% tellurite,

which it reduces to pure tellurium metal (72). This resistance to higher

concentrations of tellurite was used to differentiate the species S. faecalis from

S. faeciunr, S. faecalis could grow in the presence of 0.4% tellurite while S,

faecium could not (35). This information was critical in supporting the belief that these two were distinct species.

Other inhibiting agents have been incorporated into media to select for enterococci. Sodium taurocholate, bile salts, phenylethyl alcohol, selenite, and tetrathionate have all been studied for their usefulness as inhibiting agents. More information on these agents is available in a review by Hartman et al. (72). Rapid methods for identification

In 1933, Rebecca Lancefield detected proteins on the cell walls of

streptococci that were group specific (92). The soluble substances upon 36

which these groups were based were first noted by Hitchcock in 1924 (76).

He called them 'residue antigens' and believed that they were common to almost all hemolytic streptococci. This view held until the specific nature of these antigens was discovered by Lancefield. Based on the presence of these antigens, the streptococci were divided into groups A through V. The group D antigen, produced by most enterococci, contains glycerol teichoic acids. Most of these teichoic acids of group D antigen are buried in the cell wall. The intracellular location of these teichoic acids have made serologic identification of organisms bearing this type of group-specific antigen difficult.

Nevertheless, the Lancefield grouping scheme became the primary method of grouping streptococcal isolates, including enterococci. Today, the Lancefield groups are still widely used to identify groups of streptococci. However, the original Lancefield grouping procedures have been replaced by more rapid methods, in which latex beads coated with antibodies to each specific group antigen are used. Numerous commercial kits are available. Their applications have been reviewed (127, 149).

Despite the heavy reliance placed by some investigators on serological grouping, streptococci had been investigated by other techniques, and much information has been accumulated from a variety of different kinds of study.

Once it was discovered that organisms could be grouped by using simple biochemical tests, many workers devised classification schemes using physiological tests. These include the use of improved or additional biochemical tests, including detection of enzymes (111), resistance to chemical compounds and antibiotics (36), studies of the peptidoglycan cell 37

wall (133), and studies of biochemical pathways and electron transport systems (80).

Numerical taxonomy, performed by Bridge and Sneath (16), involved the use of many biochemical tests, colony morphology, tolerance to temperature and inhibitory compounds, and antibiotic resistance. Data were collected on different streptococcal strains and a complicated mathematical approach was used to determine the degree of relatedness among strains of streptococci.

Facklam and co-workers (49, 50) presented the results of extensive studies on identifying the enterococci by using biochemical tests. They used the bile-esculin test, growth at 45°C and at pH 9.6, reduction of tetrazolium and resistance to tellurite as well as acid production from a number of carbohydrates to separate the enterococci and fecal streptococci into 3 divisions. An abbreviated battery of five tests was recommended to presumptively identify pathogenic streptococci on a routine basis (54). The bile-esculin reaction and salt tolerance were major determinants to differentiate enterococci from other pathogenic streptococci. Using these criteria, 97% of the enterococci were correctly identified. Gross et al. (67) used Facklam's battery of tests, plus pyruvate and arginine tests, for presumptive speciation of the group D streptococci. Confirmatory carbohydrate tests were also recommended. Gross et al. (67) found that certain differences in fermentation abilities within the species may be dependent on the source of isolation. In 1985, Facklam et al. (56) described a test based on the hydrolysis of pyrrolidonly-b-naphthylamide (the PYR test) that could replace the 6.5% NaCI tolerance test to presumptively identify 38

group D enterococci. They found that the failure of some enterococci to grow on salt agar led to misidentifications. The PYR test was rapid, convenient and specific for the group D streptococci. The PYR test will be described in more detail later.

As the classification of the enterococci changed, schemes had to be updated to reflect these changes. In 1989, Facklam et al. (53, 57) published a scheme to identify all of the accepted new species of the genus Enterococcus.

A battery of conventional tube tests was used to distinguish between species.

The type strain of each species was tested and then unknown strains of previously isolated enterococci were identified by comparing them to the expected results obtained by using the type strain. Conventional tests and commercially available systems were used by Ruoff et al. (130) to determine the species identities of clinical isolates of enterococci. Strict adherence to the scheme developed by Facklam et al. (53, 57) resulted in misidentification of lactose-negative E. faecalis as E. solitarius. This problem was overcome by performing additional tests.

As the number of tests increased, the difficulty of handling conventional tube tests became obvious. New commercial rapid test systems, utilizing panels of specific tests, emerged. The commercial systems used to identify enterococci have been reviewed by several researchers. Colman and Ball

(29) reviewed the APl-20 Strep system (Analytab Products, Plainview, New

York). In this system, dehydrated substrates are incorporated into microcupules attached to stiff cardboard strips. The inoculum rehydrates the substrates in the cupules. Colman and Ball (29) examined 965 streptococci by using the APl-20 Strep (API-20S) and established methods, and found that 39

with supplemental tests the API-20 Strep could identify the streptococci, Facklam et al. (51) compared the API-20S system with the Automicrobic

Gram-Positive Identification system (GPI; Vitek Systems, Hazelwood, Miss.).

Very few differences in identification were detected between the two systems.

The 20S system was more convenient, but it required more supplemental testing. The Vitek system required less supplemental testing, but it required the use of expensive equipment. Another evaluation of the GPI system (17) revealed two minor difficulties when compared with the conventional scheme of Facklam and Collins (53). First, three Enterococcus casseliflavus isolates that were arginine negative, sorbitol negative, sorbose negative and mannitol positive could not be identified using Facklam's scheme. Secondly, some strains of E. faecalis were misidentified as E. solitarius. Bryce et al. (17) recommended that tests for tellurite reduction and ribose fermentation should be added to Facklam's scheme to avoid these misidentifications. The GPI was also compared with the API Rapid Strep system (79). The API Rapid Strep system is very similar to the API-20 Strep system, except that a few tests differ, and the Rapid Strep system has a 4 hour (like the 20S) or 24 hour incubation.

A comparison of identification of isolates from bovine mammary glands showed that both systems were accurate in the identification of Streptococcus species of bovine origin. Facklam et al. (55) also evaluated the API Rapid

Strep System to identify the streptococci. They concluded that the Rapid

Strep system was more rapid and efficient than conventional identification systems, but improvement in the data base was needed (55). Tritz et al. (147) evaluated a different panel, the MicroScan panel (Baxter Healthcare, West

Sacramento, Calif.), to identify Enterococcus species. They compared results 40

obtained by using MicroScan panel with those obtained by using conventional media. Incubation times for conventional media were as long as

96 hours, whereas the MicroScan panel yielded results within 18-24 hours.

The results of tests recommended by Facklam et al. (53) agreed with the

MicroScan results, except for six isolates. Tritz et al. (147) concluded that

MicroScan panels were reliable for the identification of E. faecalis and E.

faeciunr, however, modification of the data base would be necessary to identify other Enterococcus species.

The traditional technique of identifying an isolate as an enterococcus was by using bile-esculin and 6.5 % NaCI tolerance tests. This combination of tests was accurate, but required an incubation period of up to 48 hours (49).

Most enterococci also reduce litmus milk with a species-specific enzyme, but litmus milk is costly and difficult to prepare. New rapid techniques to identify group D streptococci have emerged. One such method utilizes the hydrolysis of L-pyrolidonyl-p-naphthylamide (PYR) by a specific aminopeptidase that is produced by these organisms when isolated on selective media (32).

Facklam et al. (56) incorporated the substrate into agar. Daly et al. (32) evaluated a technique in which a synthetic substrate is incorporated onto a filter paper strip. Hydrolysis of the substrate can usually be detected visually within 1 minute by adding a coupling dye. A major drawback of this test is that

Streptococcus pyogenes is also PYR-positive. To differentiate them from the enterococci, the PYR test was combined with a rapid chromogenic test for beta-glucosidase. A test strip was devised in which a novel beta-glucosidase test using indoxyl glucoside as one substrate was combined with the PYR substrate. Only enterococci were positive for both enzymes (33)(85). 41

Study of the murein (peptidoglycan) type was also used to differentiate enterococcal species. The amino acid sequence of the murein, in particular the nature of the diamine acid and the sequence of the interpeptide bridge connecting the stem peptides, define the murein type (133). Most enterococci possess the murein type Lys-D-Asp, containing D-isoasparagine as a cross- bridge. The presence of the murein type Lys-Alag-s in E faecalis facilitates a separation of this species from all other enterococci (83). This test is not generally used to identify enterococci on a routine basis.

A novel approach to enterococcal speciation is analysis of bacteriolytic activity patterns (123). By varying media, substrate, pH, and additives, seven major groups (lyogroups) of bacteriolytic activity against Micrococcus lysodeikticus and Micrococcus luteus were defined. One to four species fit into each lyogroup. This method is new and has not yet become readily available. Cellular fatty-acid analysis has also been used as an aid in the classification of streptococci and enterococci (1,150). This method requires cumbersome technology such as gas chromatography and possibly mass spectrometry. However, it does have potential in laboratories that process large numbers of samples because it yields results within hours once pure cultures are available for testing An immunological method was proposed

(68) in which a monoclonal antibody against E. faecalis allowed differentiation within minutes of £ faecalis from the non-enterococcal fecal streptococci, S. bovis and S. equinus. This method has not yet been commercialized.

Tests employing fluorogenic substrates are being introduced for identification of the streptococci. Beighton et al. (10) utilized 4- methylumbelliferyl-linked fluorogenic substrates to test for the production of a 42

range of glycosidase activities. A combination of conventional fermentation and liydrolytic tests was used to devise a sclieme for differentiation of oral and vividans streptococci. The fluorescent tests required only three hours. Tests employing three fluorogenic 4-methylumbelliferyl-conjugated substrates and the lectin of Dolichos biflorus were developed for the identification of p- hemolytic streptococcal colonies from throat cultures (142). It is based of the generation of fluorescence when free 4-methylumbelliferone is released by enzyme hydrolysis of the substrate. This identification system has not been widely accepted. This non-serological method was unique in that it permitted identification of groups C, F, and G as well as A and B. It is rapid, simple and specific.

Molecular approaches to identification

Several new methods have been employed within the last decade to study the relationships between the streptococci and the enterococci. The application of nucleic acid hybridization and sequencing techniques provided significant insights into the natural relationships among the streptococci.

Along with DNA-DNA hybridization studies, oligonucleotide cataloguing of

16S rRNA of several species of streptococci showed, as stated previously, that the streptococci should be divided into three genera (11, 99,133). The first group, those bacteria which conformed to the enterococcus division of

Sherman (138), was renamed Enterococcus (132). DNA-DNA and DNA-RNA hybridization studies were used to determine into which species new strains fell. Collins et al. (28) used reverse transcriptase sequencing of 16S rRNA to determine the interrelationships of Pediococcus and Aerococcus species which resemble the enterococci physiologically. The small-subunit rRNA is 43

highly conserved and is recognized as a powerful molecular chronometer (154). Different degrees of sequence conservation allows comparison of closely related species as well as those of great geneological distances.

Later the same techniques were used (153) to clarify the intragenic relationships between species of enterococcl. Several species groups were revealed. Enterococcus avium, E. malodoratus, E. pseudoavium, and E. raffinosus formed a distinct group as did E. durans, E. faecium, E. hirae, E. mundtii, E. casseliflavus, and E. gallinarum. E. cecorum, E. columbae, E. faecalis, and E. saccharolyticus formed another group. Enterococcus solitarius displayed a closer affinity with Tetragenococcus fialophiius than with other enterococci. Therefore, the precise phylogenetic placement of E. solitarius remains unclear, although its affinity with Tetragenococcus halophilus merits further investigation.

In 1990, Murray (115) compared genomic DNAfrom different enterococcal isolates by using restriction endonucleases with infrequent recognition sites in attempts to aid in species identification and to possibly identify strains within a species. This "genomic fingerprinting" was also used on strains of Streptococcus suis (106). The method provides a means of studying the epidemiology of both swine and human infections.

DNA probes are receiving much attention for their use in identifying bacteria. Construction of DNA probes for the specific identification of species of streptococci allowed rapid and confident identification of S. oralis (134),

Acridium ester-labeled, chemiluminescent DNA probe tests also have been developed for Enterococcus species (31). The probe is a DNA oligomer complimentary to a rRNA target, and the DNA-RNA hybrids are measured by 44

using a luminometer. Lactococci and enterococci were also identified by

colony hybridization with 23S rRNA-targeted oligonucleotide probes (13).

Specific sequences of the 23S rRNA of one species of Lactococcus and three

species of Enterococcus were identified, and complimentary oligonucleotide probes were synthesized. The probes were specific when used in a dot blot

assay. The Lactococcus probe was also successfully used for the specific

enumeration of lactococci as CFU in a mixed population in fermented milk. However, only one strain of each species was used. 45

PAPER 1; ROUTINE PROCEDURES FOR ISOLATION AND IDENTIFICATION OF ENTEROCOCCI AND FECAL STREPTOCOCCI 46

Routine Procedures for the Isolation and Identification of Enterococci and Fecal Streptococci

LINDA M. KNUDTSON AND PAUL A. HARTMAN*

Department of Microbiology, Immunology and Preventive Medicine, Iowa

State University, Ames, Iowa 50011 Telephone: (515)294-8824 FAX: (515)294-6019

^Corresponding author 47

ABSTRACT

Over the past six years, a revised classification of the streptococci and enterococci, based primarily on molecular techniques such as 16S rRNA sequencing and DNA-DNA hybridization, emerged. However, little attention was placed on routine physiological tests that could be used in food and clinical laboratories to differentiate between species of a new genus

Enterococcus, and fecal Streptococcus spp. The purpose of this study was to devise a convenient and reliable system to identify the enterococci and fecal streptococci by using conventional procedures. Fifty-nine strains of 13 Enterococcus spp., including the type strains and many strains used by previous investigators, were characterized by using conventional tube tests, the API Rapid Strep system, and MicroScan Pos ID panels. Results were compared with each other and with previously published results. A comparison of conventional tube tests versus published tube test results yielded 17 discrepancies. Although not all tests were done with each of the three systems, 28 discrepancies between results obtained with the API system and those obtained with conventional tube tests were found. There were 24 discrepancies between results obtained with the MicroScan Pos ID panel and those obtained with conventional tube tests. There were 12 discrepancies between the results with the API Rapid Strep system and those with the

MicroScan Pos ID panels. We devised flow charts of key tests that might be used to identify cultures without resorting to nucleic acid analysis and other labor- and equipment-intensive analyses. 48

INTRODUCTION

Enterococcus spp. inhabit the intestinal tracts of warm-blooded animals

and of Insects (16). Therefore, these bacteria have been useful as indicators

of fecal contamination in water and in foods (5). Because of the prevalence of

certain species of enterococci in the intestinal tracts of swine, enterococci are implicated in the spoilage of pork products (20). Enterococci and fecal

streptococci are also receiving increased attention because of their role in serious human infections, such as endocarditis and bacteremia (17) and diarrheal diseases in neonates (7).

In 1984, the genus Streptococcus was divided into three genera: Enterococcus, Lactococcus, and Streptococcus. (19). The genus

Enterococcus now contains 18 species, differentiated primarily by the results of 16S rRNA and DNA-DNA hybridization studies. In addition to the 13 species described in Table 1, five new species have been proposed: E.

sulfureus (15), which includes yellow-pigmented strains isolated from plants;

£ columbae (6), a species that dominates the intestinal flora of domestic pigeons: E. dispar{A), composed of two strains of human origin; E. seriolicida

(14), a fish pathogen; and E. saccharolyticus (18), previously named Streptor vous saccharolyticus.

Initially, these species were phenotypically characterized by using API

50CH and 20S systems in conjunction with conventional tube tests (1-4, 6,12,

14,15,18). However, the API data base includes only 6 of the 13 tested enterococcal species, and the data base for another system, the MicroScan

system, contains only 4 of the species. An identification system with conventional tube tests exclusively was introduced in 1989 by Facklam and 49

Collins (9). However, conventional tube tests are cumbersome, costly, and labor-intensive, and the results are often difficult to accurately reproduce in different laboratories.

In this study, we tested 59 strains of 13 species of enterococci and fecal streptococci. These included selected strains of each enterococcal species used in the original 16S rRNA studies as well as each type strain. Two different rapid systems, as well as conventional tube tests were compared. A scheme was developed to identify 13 species of enterococci, and

Streptococcus bovis and S, equinus. The test scheme uses either the API

Rapid Strep or MicroScan system, plus a minimum of supplemental tests. 50

MATERIALS AND METHODS

Strains. A total of 59 strains of 13 enterococci and S. bovis and S. equinus were collected (Table 1). Type strains used were obtained from the

American Type Culture Collection (Rockville, Md.) or Dr. Richard Facklam

(Centers for Disease Control, Atlanta, Ga.). Cultures were also obtained from the National Collection of Food Bacteria (Shinfield, Reading, England),

MicroScan (Baxter Diagnostics, Deerfield, III.), Marcia Etheridge (Baltimore, Md.) (7) and the culture collection of Paul A. Hartman (Iowa State University, Ames).

Stock cultures were maintained on Brain Heart Infusion (BHI) slants at

5 to 10°C and transferred monthly. Stock cultures were also frozen in 10% glycerol and stored at -70 to -lOO'C.

Tests, All cultures were gram stained as they were obtained to verify that they were gram-positive cocci. Each culture was also tested for the presence of catalase before being inoculated to the test panels. The orf/70-nitrophenyl-p-D-galactopyranoside (ONPG) test (21) was used to determine the presence of p-galactosidase. All other conventional tube tests were performed as described by Facklam and coworkers (8, 9,10) and Gross et al (13). Motility was determined by using modified Difco motility medium (Difco Laboratories, Detroit, Mich.) or wet mounts of cultures grown in

BHI broth at SO'C and 37°C (21). Cultures were monitored for yellow pigmentation on a cotton swab used to pick up growth from BHI agar plates incubated overnight. 51

Serogrouping for groups A, B, C, D, F, and G was accomplished by using the Streptex grouping kit (Wellcome Diagnostics, Research Triangle Park, N.C.). ^-Hemolysis was determined by observation of 24-h growth on a plate of tryptone soy agar-5% sheep blood (BBL Microbiology Systems,

Cockeyville, Md.).

Inoculum preparation for the API Rapid Strep system (Analytab

Products, Inc., Plainview, N.Y.) was carried out on blood agar plates as indicated in the manufacturer's directions but without anaerobic incubation.

The test strips were inoculated, overlayed with mineral oil where specified, incubated and read as indicated in the manufacturer's directions. After 4 h incubation, Zyme A and B reagents (Analytab) were added to the enzyme tests, ninhydrin solution was added to the hippurate test, and reagents A and

B were added to the Vogues-Proskauer (VP) test; the results of these tests were recorded. The results of the remaining tests were determined after incubation of the panel for 18-24 h.

Inoculum preparation for the MicroScan Pos ID panels (Baxter

Diagnostics, Deerfield, III.) was carried out according to the log phase technique specified by the manufacturer. The panels were inoculated, covered, and incubated at 37°C. After 18-24 h of incubation, appropriate reagents were added and the tests were interpreted as indicated in the manufacturer's instructions. 52

RESULTS AND DISCUSSION

Table 2 shows the results of conventional tube tests. A comparison of our results with published results (9) revealed 17 discrepancies among 87 instances in which comparisons could be made. Every attempt was made to perform the tests as specified in the literature, but the discrepancies appeared repeatedly. These discrepancies are probably due to the difficulty in accurately reproducing tube test results in different laboratories.

There was also a poor correlation between results obtained with the

API and MicroScan systems and conventional tube tests (Table 2). There were 28 discrepancies between API (see Table 4) and conventional tube test results (Table 2). There were 24 discrepancies between MicroScan results

(Table 3) and conventional tube test results (Table 2).

Twelve discrepancies between the API Rapid Strep results and

MicroScan Pos ID system results were observed (Tables 2 and 3). VP test results were more variable when the API Rapid Strep strips were used than when MicroScan panels were used. Enterococcus cecorum and S. bovis test results varied unpredictably on the API panels, but they were always positive on the MicroScan Pos ID panels. On the other hand, S. equinusvias VP positive when tested with API strips, but VP negative on MicroScan panels. Because the same a-naphthol and KOH solutions were used for the VP tests on both panels, these discrepancies are caused by something other than the detection reagents. p-Galactosidase results for Enterococcus faecalis and E. pseudoavium were negative on API Rapid Strep panels, whereas positive results were 53

obtained for botti species with MicroScan Pos ID panels. Alkaline phosphatase test results also differed. E. faecalis was alkaline phosphatase negative when run on the API panel but positive when tested with the

MicroScan Pos ID panel. These discrepancies probably are the results of differences between the methodologies of the API Rapid Strep and MicroScan system p-galactosidase and alkaline phosphatase tests. The API Rapid Strep panel utilizes a naphthol-linked substrate, and color development is detected after the addition of Zyme A and B reagents furnished by API. The MicroScan Pos ID p-galactosidase and alkaline phosphatase tests use para-nitrophenyl- p-D-galactopyranoside (PNPG) and para-nitrophenyl phosphate, respectively, as substrates. Both enzyme reactions release para-nitrophenol which generates a yellow color. p-Galactosidase tests were also conducted in test tubes (21) with ONPG. ONPG is very susceptible to cleavage by p- galactosidase and, as in the MicroScan test, releases a yellow product. The product in this case, however, is ort/70-nitrophenol. The only difference between the two tests is the orientation of the nitrogen on the phenyl group, para or ortho. The ONPG tube test results, except for those with S. bovis and

S. equinus, correlated well with the MicroScan results(Table 2). It seems that the API Rapid Strep methodology, at least for the p-galactosidase test, is not as sensitive as either the PNPG methodology of the MicroScan system or the

ONPG tube methodology.

Discrepancies in carbohydrate fermentation tests of the API and

MicroScan systems were also observed. Three discrepancies in the sorbitol test were observed: Enterococcus malodoratus showed a positive API result but was variable on MicroScan panels; £ mundtii was also positive for 54

sorbitol wlien run on the API panel, but all four strains were negative on the

MicroScan panel; and £ solitarius was sorbitol negative in the API tests and sorbitol positive in the MicroScan tests. When a conventional tube method for sorbitol fermentation was used (9), E. malodoratus, E. mundtii, and E. solitarius produced variable, positive, and variable results, respectively (Table

2). Thus, manual test results failed to support the validity of the results of either panel as superior to those of the other. Two discrepancies were observed in the inulin test (Tables 3 and 4). E. malodoratus was uniformly inulin negative on API panels; variable results were obtained when

MicroScan panels were used. Only one discrepancy existed in the raffinose test: variable results were obtained with Enterococcus hirae on API Rapid

Strep test strips, whereas negative results were obtained on MicroScan Pos

ID panels.

The six discrepancies seen in the carbohydrate fermentation tests for sorbitol, inulin, and raffinose are highly variable. There was no consistent pattern wherein one panel was positive while the other was negative. Both systems use phenol red as the pH indicator, but there is a slight difference in methodology. The API Rapid Strep system requires a sterile mineral oil overlay on all the carbohydrate tests to reduce oxidative metabolism of the carbohydrates. The MicroScan system does not utilize a mineral oil overlay.

Therefore, the discrepancies could be caused by the ability of some bacteria to oxidatively metabolize the carbohydrates in the MicroScan system but not in the API system. If this was the only cause of the discrepancies, one would expect more negative results from the API system and more positive results 55

from the MicroScan system. As was pointed out above, however, this did not occur.

Also Included In Tables 3 and 4 are two tests that were not performed on either panel. Pigmentation is important for differentiating pigmented Enterococcus casseliflavus and E. mundtii from nonpigmented enterococci and fecal streptococci. Motility tests are used to distinguish motile Enterococcus gallinarum and E. casseliflavus from nonmotile species.

Figures 1 through 3 depict selected tests that can be used to identify all 13 species of enterococci as well as S. bovis and S. equinus. The tests shown in Tables 3 and 4 were used to construct the identification schémas, with the exception of the sucrose test (Table 2), which is needed to distinguish

E. hirae (positive) from E. durans (negative). Also included in Figures 1 through 3 are the Lancefield group D reactions. They were included to aid in the differentiation of E. cecorum and E. pseudoavium (group D negative) from enterococci and fecal streptococci that produce the group D antigen.

Not all of the tests that could be used to differentiate between species were included in these schémas. Each strain was tested several times, and some tests yielded inconsistent results (i.e., a positive result the one time and a negative result the next). To obtain the most consistent and reliable

Identification, only tests that produced the most consistent and reproducible results were Included in the identification schémas. These flow charts can be used with either the API Rapid Strep system or the MicroScan system. Some tests depicted are only available on the API Rapid Strep strips and are noted with an asterisk (*). Figure 2 shows that E. mundtii and E. casseliflavus are differentiated by the sorbitol test. As stated above, E. mundtii is only sorbitol 56

positive when tested with the API Rapid Strep kit; when E, mundtiils tested with the MicroScan system, the result is negative. Thus, when the MicroScan system is used, differentiation between these two species must be determined by using the motility, inulin, and raffinose results only. Figure 2 also shows that E faecalis is p-galactosidase negative. As discussed above, this is true only when E faecalis is tested on the API Rapid Strep panel; when tested with the MicroScan panel or ONPG tube test, the result is positive. The enterococci can be distinguished from other gram-positive catalase-negative cocci. The pyrrolidonylase (pyrrolidonyl arylamidase or pyrrolidonyl peptidase) test differentiates Enterococcus spp. from

Leuconostoc, Lactococcus and Pediococcus spp. Another test, which is included in the API Rapid Strep panels or can be conducted as a tube test, is for leucine aminopeptidase activity. This test differentiates Enterococcus spp. from all other non-streptococcal isolates (11).

In conclusion, we have described key tests that can be used to differentiate between species of enterococci and fecal streptococci. These tests can be performed with either the API Rapid Strep panel or MicroScan system panel. Additional tests include motility, pigmentation, and sucrose tests. The Lancefield group D determination is optional. This test scheme is a rapid and reproducible way to differentiate the enterococci and fecal streptococci. 57

ACKNOWLEDGEMENTS

This work was supported by a U. S. Department of Agriculture grant to the Food Safety Consortium. Journal paper No. J-14918 of the Iowa

Agriculture and Home Economics Experiment Station, Ames, project 2991.

We thank R. R. Facklam and M. E. Etheridge for providing cultures. 58

REFERENCES CITED

1. Collins, M. D., R. R. Facklam, J. A. E. Farrow, and R.

Williamson. 1989. Enterococcus raffinosus sp. nov., Enterococcus

solitarius sp. nov. and Enterococcus pseudoavium sp. nov. FEMS

Microbiol. Lett. 57:283-288.

2. Collins, M. D., J. A. E. Farrow, and D. Jones. 1986. Enterococcus mundtii sp. nov. Int. J. Syst. Microbiol. 36:8-12.

3. Collins, M. D., D. Jones, J. A. E. Farrow, R. Kilpper-Bâiz, and

K. H. Schlelfer. 1984, Enterococcus avium nom. rev., comb, nov.; E.

casseliflavus nom. rev., comb, nov.; £ durans nom. rev., comb, nov.; £

gallinarum comb, nov.; and £ malodoratus sp. nov. Int. J. Syst. Microbiol.

34:220-223.

4. Collins, M. D., U. M. Rodrigues, N. E. Pigott, and R. R.

Facklam. 1991. Enterococcus dispar sp. nov. a new Enterococcus

species from human sources. Lett. Appl. Microbiol. 12:95-98.

5. Deibel, R. H. 1964. The group D streptococci. Bacteriol. Rev. 28:330- 366. 59

6. Devriese, L. A., K. Ceyssens, U. M. Rodrigues, and M. D. Collins. 1990. Enterococcus columbae, a species from pigeon intestines. FEMS Microbiol. Lett. 71:247-251.

7. Etheridge, M. E., R. H. Yolken, and S. L. Vonderfecht. 1988.

Enterococcus hirae implicated as a cause of diarrhea in suckling rats. J.

Clin. Microbiol. 26:1741-1744.

6. Facklam, R. R. 1972. Recognition of group D streptococcal species of

human origin by biochemical and physiological tests. J. Clin. Microbiol. 23:1131-1139.

9. Facklam, R. R., and M. D. Collins. 1989. Identification of

enterococcus species isolated from human infections by a conventional

test scheme. J. Clin. Microbiol. 27:731-734.

10. Facklam, R. R., L. G. Thacker, B. Fox, and L. Eriquez. 1982.

Presumptive identification of streptococci with a new test system, J. Clin.

Microbiol. 15:987-990. 60

11. Facklam, R. R., and J. A. Washington II. 1991. Streptococcus and

related catalase-negative gram-positive cocci, p. 238-257. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy

(ed.). Manual of clinical microbiology, 5th ed. American Society for

Microbiology, Washington, D. C.

12. Farrow, J. A. E., and Wl. D. Collins. 1985. Enterococcus hirae, a new

species that includes amino acid assay strain NCDO 1258 and strains causing growth depression in young chickens. Int. J. Syst. Microbiol. 35:73-75.

13. Gross, K. C., M. P. Houghton, and L. B, Senterfit. 1975.

Presumptive speciation of Streptococcus bovis and other group D

streptococci from human sources by using arginine and pyruvate tests. J. Clin. Microbiol. 1:54-60.

14. Kusuda, R., K. Kawai, F. Salati, 0. R. Banner, and J. L. Fryer. 1991. Enterococcus seriolicida sp. nov., a fish pathogen. Int. J. Syst. Bacteriol. 41:406-409.

15. Martinez-Murcia, A. J., and M. D. Collins. 1991. Enterococcus

sulfureus, a new yellow-pigmented Enterococcus species. FEMS

Microbiol. Lett. 80:69-74. 61

16. Mundt, J. O. 1986. Enterococci and lactic acid streptococci, p. 1063- 1066. In P. H. A. Sneath, N. S. Mair, M. E. Sharpe, and J. G. Holt (ed.),

Sergey's manual of systematic bacteriology. The Williams and Wilkins

Co., Baltimore.

17. Murray, B. E. 1990. The life and times of the Enterococcus. Clin.

Microbiol. Rev. 3:46-65.

18. Rodrigues, U., and M. D. Collins. 1990. Phylogenetic analysis of Streptococcus saccharolyticus based on 16S rRNA sequencing. FEMS Microbiol. Lett. 71:231-234.

19. Schlelfer, K. H., and R. Kilpper-Bâiz. 1984. Transfer of Streptococcus faecalis and Streptococcus faecium to the genus

Enterococcus nom. rev. as Enterococcus faecalis comb. nov. and

Enterococcus faecium comb. nov. Int. J. Syst. Bacterid. 34:31-34.

20. Sharpe, M. E., and B. G. Fewins. 1960. Serological typing of strains

of Streptococcus faecium and unclassified group D streptococci isolated from canned hams and pig intestines. J. Gen. Microbiol. 23:621-630. 62

21. Smibert, R. M., and N. R. Krieg. 1981. General characterization, p.

409-443. In P. Gerliardt, R. G. E. Murray, R. S. Costilow, E. W. Nester, W.

A. Wood, N. R. Krieg, and G. B. Phillips (éd.). Manual of methods for

general bacteriology. American Society for Microbiology, Washington, D.C. 63

Table 1; List of strains used ^

Species Strains (s) E. avium ATCC 14025, 49462, 49463, 49465; NCDO 2366; SS -1^ ; A" ; B"

E. casseliflavus ATCC 25788; NCDO 2376; 7765F''; 443»; A8''

E. cecorum ATCC 43198; NCDO 2674

E. durans ATCC 19432; NCDO 498; PAH 940»; 15-20"

E. faecalis ATCC 4082,19433, 35038, 49477, 49478; NCDO 581; R1873"; 334-2"

E. faecium ATCC 349,19434; NCDO 502; R169a»; R281a''; 2100"; 2124"

E. gallinarum ATCC 49573; NCDO 2311, 2315, 2704

E. hirae ATCC 9790; NCDO 1631,1648, 2683; ME''

E. malodoratus ATCC 43197; NCDO 847

E. mundtii ATCC 43186; NCDO 582, 2374, 2377

E. pseudoavium SS-1277''; NCDO 2138

E. raffinosus SS-1278''

E. solitarius SS-1279''

S. bovis ATCC 9809, 33317; H-12" ; H-24"

S. equinus ATCC 9812

= See tlie text for sources not listed in footnotes b through d. " From culture collection of Paul A. Hartman. ® From M. E. Etheridge. '' From R. R. Facklam. !

Table 2: Conventional tube test results Test result®

species (no. Of strains) llllillilllœ œ Q. CO s X _l CO

E. avium (8) M " + (+)* + + b,c4 E. casseliflavus (5) + (+) - - + + d E. ceœrum (2) + + b,cjd E. durans (4) +(-)--- - v

E. faecalis (8) + v'' +" + d d E. faecium (7) +(-)-- v + (-)+- + +

£. gallinarum (4) +-• ++ + + W + E. hirae (6) + cM b,c4 E. malodoratus (2) + V - + +

£. mundtii (4) + v''

+ E. pseudoavium (2) (+)'' -

£. raffinosus (1) + + - + + + + + +

E. solitarius (1) + +W + . - + V v'' + - +W S. bovis (4)

S.equinus (1) - - - + + +^''^ - + 9+, Positive reaction (100%); -, negative reaction (100%); (+), 75% or greater siiow positive reaction; (-), 75%

or greater show negative reaction; v, variable (some strains positive, some strains negative). b Discrepancy between Table 3 MicroScan results and Table 2 results.

c Discrepancy between published tube test results and Table 2 results. d Discrepancy between Table 4 API results and Table 2 results. (/) rn !Ti m m rn M m m m fn m m m m c/> §• S 2 (D §• 01 i I ! t CO 1 o5 I 1^' ! S I 00 I 1 1 I I s ! 13 I 13 S oi CO S I + + c < c < < < < < < c c + + Crystal Violet 0 + + + + + + + + + + + + + + Micrococcus Screen % Nitrate 1

' + + + + + + + + + + + + + + Voges-Proskauer 8 + + + + + + + + + + + + + + + Optochin (/) + • + • • Phosphatase + + + + + + + + + + + + + + + 40% Bile Escuiin

• • + + + + + + + + + + • + + L-Pyndidcnylase

• • + • • + + + + + + • + • Arginine

• • • • + + + + + + + + + + < B-Galactoacbse < • • • • • • ' • ' • • • • • Urea

• + • + • • + • + • • • + + ' Raffinose • + • + + + + + + + 2: + + + + Lactose + < + + + + + + + + + + + + Trehatose + + + + + + + + + + + + + + + Mannose • • + + + + + + + + + + • + < 6.5% NaCi

• ' • < + + + • • • • + Sorbitol

• • • + • + • + + • + + Arabinose

• • • + + + + + + + + + + + + Ribose < + • • • • c + • • + + • Inulin < • + + + + + + + + • + + Mannitol + + + + + + + + + + + + + + + Bacitracin

• • ' • • • • • • Pyruvate

• ' • • • • • • Hemolysis

• • • ' • + + • Pigment

+ • • • + • Motility a+, Positive reaction (100%); -, negative reaction (100%); (+), 75% or greater show positive reaction; (-), 75% or greater show negative reaction; v, variable (some strains positive, some strains negative). (oçcrnmmrnrnrnrnMm m m m m m m 2 S" (D iI I i I 88 i I1 S' 1 1 if 5 5 # O) 1 10) |.w 1 s CO ï i i 3 S 3 J3 S iCO 3 M M I s CL. + < + + + + + + + + + < + + o- » Voges-PtDSkauer I < + ' -i • ' ' < < < < < ' ' Hippurate Hydrolysis + + + + + + + + + + + + + + Esculin $ + + + + + • + + + + + ' + + Pyirolidonylase c « < + + < < î + 3 + • 3 + + • a-Galactosidase

' • ' ' • • • ' • • • + • • Q-Glucuronidase

' • < ' » + + + + + o- + + + 13-Galactosidase

' ' ' ' • • • • • • + • ' o- Alkaline Phosphatase + + + + + + + + + + + + + + Leucine Arylamidase

• + ' ' + • + + + + + ' + ' Arginine Dehydrolase

' ' + + + + + + + + + + + + Ribose

• ' + • + ' • + + • • + + L-Arabinose i

< + + + + + ' + + + ' ' + + Mannitol $

' + + + ' • • + • ' • + o- » Sorbitol + ' + + + + + + + + + + + Lactose < + + + + + + + + + 3 + + + Trehalose

3 • • • • a- • + X • • + + • Inulin

+ ' + ' • + •«! + ' ' + + ' » Raffinose

+ ' < ' < < + + + + < + + < Starch

+ ' ' ' • • ' < • • • • • ' Glycogen

' ' ' ' ' • ' ' ' ' • ' • • Hemolysis

' ' ' ' + ' ' • ' ' • • + ' Pigment

• ' ' ' ' ' ' + ' ' ' ' + ' Motility a+, Positive reaction (100%); -, negative reaction (100%); (+), 75% or greater show positive reaction; (-), 75%

or greater show negative reaction; v, variable (some strains positive, some strains negative). ^ Indicates where discrepancies exist between the API Rapid Strep and the MicroScan system. Fig. 1: Flow chart for differentiating the pyrrolidonylase-negative enterococci and fecal streptococci. The data

shown are condensed from Tables 3 and 4 . The asterisk (*) indicates that a test is available with the API Rapid Strep but not with the MicroScan system. GRAM-POSITIVE COCCI CATALASE- ESCULIN+ PYRROLIDONYLASE-

GROUP D + GROUP D + GROUP D- P-GALACTOSIDASE - P-GALACTOSIDASE - P-GALACTOSIDASE + RIBOSE - RIBOSE - RIBOSE + LACTOSE - LACTOSE + LACTOSE + RAFFINOSE- RAFFINOSE + RAFFINOSE + STARCH -* STARCH +* STARCH +* GLYCOGEN -* GLYCOGEN +* GLYCOGEN +*

S. equinus S. bovis E. cecorum Fig. 2: Flow chart for differentiating the pyrrolidonylase- and arabinose-positive enterococci. The data shown

are condensed from Tables 3 and 4. The asterisk (*) indicates that a test is available with the API

Rapid Strep but not with the MicroScan system. A superscript a indicates a test that has a positive result with the API Rapid Strep and a negative result with the MicroScan system. A superscript b indicates a

test that has a negative result with the API Rapid Strep and a positive result with the MicroScan system. GRAM-POSITIVE COCCI CATALASE- ESCULIN+ PYRROLIPONYLASE+ ARABINOSE + PIGMENT + PIGMENT I MOTILITY - MOTILITY + SORBITOL SORBITOL - INULIN - INULIN + RAFFINOSE - RAFFINOSE + ARG NINE- ARG NINE E mundtii E casseliflavus

RAFFINOSE + RAFFINOSE - a-GALACTOSIDASE +* a-GALACTOSIDASE E raffinosus E avium

INULIN - INULIN + INULIN - MOTILITY - MOTILITY + MOTILITY - SORBITOL - SORBITOL - SORBITOL + RAFFINOSE - RAFFINOSE + RAFFINOSE - P-GALACTOSIDASE + P-GALACTOSIDASE + P-GALACTOSIDASE E faecium E gallinarum E faecalis Fig. 3: Flow chart for differentiating the pyrrolidonylase-positive and arabinose-negatlve enterococci. The data shown are condensed from Tables 2, 3, and 4. The asterisk (*) indicates that a test is available with

the API Rapid Strep but not with the MicroScan system. GRAM-POSITIVE COCCI CATALASE- ESCULIN+ PYRROLIDONYLASE+

ARABINOSE -

MANNITOL + MANNITOL -

SUCROSE + SUCROSE - ^ oi E. hirae E. durans

LACTOSE - LACTOSE+ LACTOSE + LACTOSE+ RAFFINOSE RAFFINOSE - RAFFINOSE + RAFFINOSE ARGININE + ARGININE - ARGININE +/- ARGININE + STARCH STARCH -* STARCH +/-* STARCH +*

E. solitarius E. pseudoavium E. malodoratus f^^oalis 76

PAPER 2: ENTEROCOCCI IN PORK PROCESSING 77

Enterococci in Pork Processing

LINDA M. KNUDTSON AND PAUL A. HARTMAN*

Department of Microbiology, Immunology and Preventive Medicine,

Iowa State University, Ames, Iowa 50011-3211

Telephone: (515) 294-8824

FAX: (515)294-6019

^Corresponding author 78

ABSTRACT

The purpose of this study was to determine the numbers and species of

enterococci encountered on pork carcasses during different stages in the

slaughter process as well as on pork products. Three hog slaughtering plants

were surveyed, each 3 times at four processing points. Each hog was

swabbed at two sites on the carcass. Specimens were plated on two different

enterococcal media, KF Streptococcal agar and fluorescent gentamicin-

thallous-carbonate agar. Retail and spoiled pork sausage products also were examined. Isolates were speciated by using the API Rapid Strep and Baxter

MicroScan Pes ID panels. Contamination levels varied between plants as the

carcasses progressed down the processing line; the highest counts were

obtained directly before packaging in plants A and C. The highest count for plant B occurred at the first stage of sampling. More Enterococcus faecalis than Enterococcus faecium were isolated from the pork carcasses. Pork

sausage results also are presented. Enterococci are useful as an indicator of pork sanitation and to detect critical control points during processing. In some instances, high levels of enterococci are associated with spoilage of pork sausage. 79

INTRODUCTION

Although enterococcl are widely distributed in nature (6), the natural habitat of these organisms is the intestinal tract of man and animals (4).

Enterococcl also have been isolated from plants and insects (14), where they

are believed to establish an epiphytic relationship (3). The importance of

enterococci as indicators of fecal pollution has been documented by several

workers (1,13). These bacteria are sometimes used as fecal indicator organisms because they are readily isolated in large numbers from human and animal feces. Their relative resistance to adverse conditions, such as tolerance to extremes in temperature (5), pH, and salinity, is advantageous when determining the sanitary history of moderately heated, frozen, salted, or other foods and drinks in which conforms might not have survived (19).

However, because of their ability to grow in environments far removed from the original source of fecal contamination, caution and discretion must be exercised in attributing significance to the numbers and types of enterococci and fecal streptococci present in foods. Once introduced into a food- processing plant, enterococci can become established, and the subsequent contamination of a food product does not necessarily indicate fecal pollution.

Many researchers have suggested the possibility of using enterococci as indicators of fecal pollution by using differences in species distribution in different hosts as a means to pinpoint the source of contamination (1, 8,15,

16). None of the enterococci can be considered absolutely "host specific," but some species show a degree of host specificity (6). Variation in the enterococcal flora of animals has been associated with many factors; such as 80

geographical location, diet, age, species of the animal, and even seasonal effects (3). Despite this potential for variation, significant differences in the distribution of enterococci and fecal streptococci have been observed in host animals, and predominating species have been identified in some common hosts.

In this study, numbers and species of enterococci present on pork carcasses during fabrication and subsequent processing were examined.

Levels of enterococcal contamination were determined, and possible sources of contamination were considered. The importance of enterococci as spoilage organisms was investigated, as well as their importance and usefulness as fecal indicators to determine critical control points in meat processing. 81

MATERIALS AND METHODS Sampling location

Three pork slaughtering plants located in the Midwest were chosen for sampling. All three plants were typical modern pork slaughtering plants of similar design and line speeds. Each plant was visited at random on three different occasions. All testing was performed on Tuesdays and Wednesdays, from June through September, to reduce daily variation. On Tuesday, three pork carcasses were selected at random at two areas of the slaughter, immediately after singeing and polishing and after the final rinse. On

Wednesday, three carcasses from the previous day's kill were swabbed after 18-24 h in the carcass cooler. Each carcass was swabbed at two locations, the midpoint of the loin and the outside of the ham. The side of the carcass that was swabbed was randomly chosen. Also sampled on Wednesday were six boneless loins. Each loin was swabbed on the lean side immediately before packaging.

Sampling procedure The pork carcasses were sampled on the slaughter line, so as to disrupt normal production as little as possible. The six loins were sampled on nearby tables in the packaging area to avoid interference with the line.

Personnel involved wore sterile gloves to avoid contamination. Sampling was performed by using a moistened swab technique. Sterile cotton swabs were moistened in 0.1% phosphate buffered saline at pH 7.0-7.2. The moistened swabs were uniformly stroked 12 to 15 times across the surface of the carcass inside a sterile 100-cm2 template. Swabs were then rotated and stroked 12 to 15 times perpendicular to the original swabbing direction. Swabs were 82

placed in 10 ml 0.1% phosphate buffered saline and stored on ice (approx.

4°C) for transport to the laboratory. Samples were plated within 4 to 7 h.

Tubes containing swabs were agitated on a vortex mixer for 20-30 seconds before samples were diluted in 0.1% peptone water and plated.

Fresh, expired, or spoiled pork sausage samples also were examined. These samples consisted of vacuum-packaged pork sausage chubs or uncased sausage links furnished by processing plants or purchased at a retail store. Five grams of pork sausage sample were aseptically transferred to a sterile Stomacher bag containing 45 ml of 0.1% peptone water diluent. Each sample was mixed by stomaching for 2 min in a Colworth Stomacher 400.

The samples were then decimally diluted in 9-ml 0.1% peptone water blanks and plated on two different media selective for enterococci and fecal streptococci.

Microbial enumeration

KF streptococcal agar (KF, Difco Laboratories, Detroit, Mich.; 7) was prepared according to the manufacturers' instructions (Difco manual, 10th edition, 1984). One-ml portions of appropriate dilutions were used for KF agar pour plates. Duplicate plates were incubated at 37°C for 48 h. Colonies exhibiting a red or pink color were counted as streptococci (7). The second medium, fluorescent gentamicin-thallous-carbonate (fGTC) agar (12), was prepared as specified. Appropriate dilutions of 0.1 ml were plated in duplicate, and the plates were incubated at 37°C for 24 h. All except pinpoint colonies were counted as enterococci. Positive fluorescence and starch hydrolysis (zone of clearing) were also recorded. Total counts were performed only on the processed pork sausage samples. Tryptone Glucose Extract 83

(TGE) agar (Difco Laboratories) was prepared as specified, and 1-ml portions

of the appropriate dilutions were plated in duplicate by using the pour plate technique. The plates were incubated at 37°C for 48 h.

Statistical analysis

The method of analysis by Cochran and Cox (2) for combining

experiments was used where plants are considered to be experiments, visits

within plants to be replications within experiments, and type-stage

combinations to be treatments. Treatment by plant and treatment by visit within plant sources of variation were pooled for treatment error. Variation

among samples for any given plant, visit, and treatment was combined for a pooled error. An unweighted means analysis was used whenever the number of samples differed among treatments.

Species identification

After the KF and fGTC plates had been counted, 3 colonies of each

colony type on each medium were streaked for isolation on Brain Heart

Infusion agar (Difco Laboratories). After 24 h of incubation, a gram stain and catalase test were performed to verify gram-positive, catalase-positive colonies. The cultures also were tested on Bile Esculin (BE) agar (Difco

Laboratories) for the ability to grow in bile and hydrolyze esculin. Cultures positive for these tests were then identified to species by using the classification schema developed by Knudtson and Hartman (10). This schema was developed by using the API Rapid Strep and MicroScan Pos ID panels. 84

RESULTS

A comparison of mean log counts of enterococci at each stage In the

slaughter process at each of three plants is shown in Fig, 1. Each point of the

figure represents 36 samples. The differences between plants and stages

were significant (P < 0.01). At stage 1, immediately after polishing, high mean enterococcal counts were obtained from plants A and B. These results are similar to those of Snijers et al. (18), who reported that a significant increase in contamination on carcass surfaces occurred during polishing in a back- scraping machine. Counts for plant C, however, were significantly lower than for plants A and B. Different sanitation practices or newer equipment could explain this difference, but the actual reason for these differences is not known. At stage 2, after the final rinse, mean enterococcal counts for plants

A and B decreased significantly, approaching those of plant C. After 18-24 h in the carcass coolers (stage 3), the mean enterococcal count from plant A increased significantly, whereas counts at plant B actually decreased; there was only a slight increase at plant C. Counts of samples taken immediately before packaging (stage 4) at plants A and C were similar. At plant B, however, counts were higher. It is evident that, although all three plants have similar operations, significant differences in contamination existed. This indicates that stage 3 is an important critical control point in pork processing plant A.

Overall differences among plants are shown in Fig. 2. When mean enterococcal counts were calculated for all samples taken at each plant, there were significant differences between the plants (P = 0.0001). Total mean 85

counts were highest at plant A, then plant B, and finally, plant C. Further studies are needed to determine the cause of these differences.

Fluorescent gentamicin-thallous-carbonate agar counts were used in

Fig. 1 and 2 because all enterococci and fecal streptococci grow equally well on fGTC agar media, whereas the growth of E faecalis is favored on KF medium. KF agar counts were performed, and these counts were lower in almost every instance; data not shown (9).

Identifications were performed on colonies picked from both fGTC and

KF agar plates. From all samples collected at the 3 plants, 175 enterococci were isolated. Fig. 3 represents the proportions of enterococcal species identified: 79% of the enterococci were E faecalis, 11% E faecium, 3% E pseudoavium, 2% each E. malodoratus and E solitarius, and 1% each E. casseliflavus and E durans. One percent of the enterococci isolated were not identified (2 cultures) and may represent one or two of the 5 new species of enterococci not included in the classification schema used to identify the majority of isolates (10). Proportions of enterococci isolated separately from each plant and from the two different isolation media resembled closely those represented in Fig. 3 (9).

The processed pork sausage data (Fig. 4) includes all 3 media used, fGTC, KF, and TGE. The differences between the fresh pork sausage counts at plants A and B could be due to sample differences. Plant A samples consisted of retail pork sausage chub samples, whereas plant B samples included chub samples and uncased link samples. The uncased link samples undergo more processing steps and had higher counts than the less processed chubs. The lower counts found on KF medium vs. fGTC medium 86

were consistent throughout the study; almost invariably, counts were higher on the less inhibitory fGTC medium than on KF agar. The spoiled pork sausage samples from plant A showed signs of spoilage before reaching the expiration date (gas was produced, bulging the packages). The expired pork sausage from plant B showed no signs of spoilage, although the mean counts were just as high as for the spoiled samples. Although mean enterococcal counts (fGTC) of spoiled and expired pork sausage were similar, identities of organisms from these 2 sources differed. Samples from the spoiled pork sausage yielded an almost pure culture of enterococci; very few other organisms were obtained (but not every colony was tested). On the other hand, the expired samples contained mostly gram-positive bacilli, resembling lactobacilli, which grew on both KF and fGTC media. Enterococcus species were isolated, but counts were much lower than they appear in Fig. 4, About one of every 10 colonies tested was an Enterococcus. These results indicate that enterococci are not good shelf-life indicators.

Fig. 5 represents the proportions of enterococcal species recovered from the pork sausage samples. (A) Represents proportions of enterococcal species recovered from fresh pork sausage samples obtained from plants A and B. The individual proportions from each plant were comparable, so data were combined. As with the pork carcass results, E. faecalis was the predominant Enterococcus species in fresh pork sausage; 83% of the isolates were identified as E. faecalis, 11% of isolates were E. malodoratus, and 6% were E. hirae. E. faecium was not isolated. In the spoiled pork sausage from plant A (B, Fig. 5), E. faecium vias predominant (84%); only 14% were E. faecalis and 2% £ durans. (C) represents proportions of enterococci from 87

expired pork sausage. In expired pork sausages, 44% of the enterococci were £ faecalis, 23% £ pseudoavium, and 11% each were £ malodoratus, £ rafiinosus, and £ durans. 88

DISCUSSION

The pork slaughtering plant data show that, although plants may be of

sinnilar overall design and line speed, bacterial counts on carcasses at

different points during processing can differ significantly. Undoubtedly, differences in equipment (such as conveyer belts) and in operating protocols

(such as personnel training and hygiene, equipment sanitation and

maintenance, and plant sanitation) exist among plants.

Airborne contamination within pork slaughter and processing

establishments warrants some attention. Kotula and Emswiler-Rose (11) reported that airborne contamination in one pork facility was 10 times higher than in previously studied dairy facilities. During slaughter and processing, bacteria on the surface of animals, employees, and equipment could become

airborne and cause contamination. The present work has defined some potential critical control points. Further studies are needed to elucidate factors

contributing to high contamination levels so that appropriate remedial action can be taken.

Several researchers have described the predominance of E faecium in

the intestinal tract of swine and other animals (1, 8,15,16) whereas E.

faecalis is believed to reside predominantly in human intestines (1, 8, 15, 16).

Therefore, the predominance of E faecalis ii. the hog carcass samples is puzzling. Because enterococci are associated with the intestinal tracts of pigs, enterococci isolated from a hog carcass in a slaughtering plant presumably

would arise from fecal contamination of hog carcasses by hog fecal matter; therefore, E faecium should predominate. The overall predominance of E. 89

faecalis isolated from all 3 plants indicates that the hogs slaughtered in these plants had an intestinal flora in which £ faecalis predominated, or that the enterococcal contamination arose from other sources. Because E. faecalis is the predominant species in the human intestine, human contamination of the hog carcasses should be considered as a possible source of the carcass contamination. The hygienic practices of employees would be of critical importance in controlling contamination of human origin. Another explanation would be the establishment of E. faecalis in the plant. Contamination of the carcasses would occur when they came in contact with equipment or other contaminated materials. The enterococcal contamination could occur at several points in the processing line, allowing recontamination at any of several stages in the slaughtering operation. This would explain the wide variation in counts at different stages among plants.

From the pork slaughtering plants, the pork enters the pork processing plant and is made into fresh pork sausage. Fresh sausage samples harbor roughly the same species of enterococci found on the pork carcasses (Fig.

5A). Spoiled pork sausages (Fig. 5B), on the other hand, contained predominantly E. faecium, with a smaller percentage of E. faecalis. Barnes and Ingram (1) reported similar findings; S. faecium (E. faecium) caused spoilage in canned hams. Enterococci predominating in hog intestines also were identified as S. faecium (E. faecium) and unclassified group D strains

(probably new Enterococcus spp.). S. faecalis (E. faecalis) was not isolated.

When testing was performed in bacon factories, the predominant enterococcus was S. faecalis (E. faecalis). Sharpe and Fewins (17) serologically typed the S. faecium strains isolated from pig intestines and 90

canned hams. Two serological types were present in both the samples from pig intestines and canned hams. But these serotypes are widespread and have been isolated from samples from several different countries. It is possible, however, that E. faecium arising from hog fecal contamination causes spoilage of pork sausages if present in high numbers. Even though fecal contamination with £ faecium was not discovered on the hog carcasses that we examined, it is possible that this contamination occurs infrequently.

When it does occur, contamination of a single hog carcass may be sufficient to instigate spoilage of sausage or hams made from the carcass. More carcasses must be assayed to test this hypothesis.

In conclusion, because of the ability of enterococci to grow in environments outside their original source, enterococcal counts alone are not always accurate indicators of fecal pollution. But, in conjunction with species identification, they can indicate possible sources of contamination and help to define critical control points in need of further attention. 91

ACKNOWLEGMENTS

Supported by a USDA grant to the Food Safety Consortium. Journal

Paper No. J-14979 of the Iowa Agriculture and Home Economics Experiment

Station, Ames; project 2991. We thank C. Lynn Knipe for allowing us to be part of this study, Anita McVey for assistance with the statistical analyses, and Christopher Guyer for technical assistance. 92

LITERATURE CITED

1. Barnes, E. M., and M. Ingram. 1955. The identity and origin of faecal

streptococci in canned hams. Ann. Inst. Pasteur Lille. 7:115-119.

2. Cochran, W. G., and G. M. Cox. 1957. Analysis of the results of a

series of experiments, pp. 545-568. In Experimental designs, 2nd ed.

John Wiley and Sons, Inc., New York.

3. Deibel, R. H. 1964. The group D streptococci. Bacterid. Rev. 28:330- 366.

4. Garg, S. K., and B. K. Mital. 1991. Enterococci in milk and milk products. Crit. Rev. Microbiol. 18:15-45.

5. Gordon, C. L. A. 1991. Thermal susceptibility of Streptococcus faecium strains isolated from frankfurters. Can. J. Microbiol. 37:609-612.

6. Hartman, P. A., R. H. Deibel, and L. M. Sleverding. 1992.

Enterococci, pp. 523-531. In C. Vanderzant, D. F. Splittstoesser (eds.).

Compendium of methods for the microbiological examination of foods, 3rd ed. American Public Health Association, Washington, D. C. 93

7. Kenner, B. A., H. F. Clark, and P. W. Kabler. 1961. Fecal

streptococci. I. Cultivation and enumeration of streptococci in surface waters. Appl. Microbiol. 9:15-20.

8. Kjellander, J. 1960. Enteric streptococci as indicators of fecal

contamination of water. Acta Pathol. Microbiol. Scand., Suppl. 136,. 48:9- 124.

9. Knudtson, L M. 1992. Classification of enterococci and their roles in

spoilage of pork products ans as sanitary indicators in pork processing.

Ph.D. thesis. Iowa State University, Ames.

10. Knudtson, L. M., and P. A. Hartman. 1992. Routine procedures for

isolation and identification of enterococci and fecal streptococci, Appl. Environ. Microbiol. 58; control #509-92.

11. Kotula, A. W., and B. S. Emswiler-Rose. 1988. Airborne microorganisms in a pork processing establishment. J. Food Prot. 51:935-

937.

12. Littel, K. J., and P. A. Hartman. 1983. Fluorogenic selective and

differential medium for isolation of fecal streptococci. Appl. Environ.

Microbiol. 45:622-627. 94

13. Moussa, R. S. 1965. Species differentiation of faecal and nonfaecai enterococci. J. Appl. Bacteriol. 28:466-472.

14. Mundt, J. O., A. H. Johnson, and R. Khatchikian. 1958. Incidence

and nature of enterococci on plant materials. Food Res. 23:186-193.

15. Fourcher, A. M., L. A. Devriese, J. F. Hernandez, and J. M.

Delattre. 1991. Enumeration by a miniaturized method of Escherichia

coii, Streptococcus bovis and enterococci as indicators of the origin of faecal pollution of waters. J. Appl. Bacteriol. 70:525-530.

16. Ramadan, F. M., and M. S. Sabir. 1963. Differentiation studies of

fecal streptococci from farm animals. Can. J. Microbiol. 9:443-450.

17. Sharps, M. E., and B. G. Fewins. 1960. Serological typing of strains

of Streptococcus faecium and unclassified group D streptococci isolated

from canned hams and pig intestines. J. Gen. Microbiol. 23:621 -630.

18. Snijers, J. M. A., M. H. W. Janssen, G. E. Gerats, and G. P.

Cortiaensen. 1984. A comparative study of sampling techniques for

monitoring carcass contamination. Int. J. Food Microbiol. 1:229-236.

19. Thian, T. S., and P. A. Hartman. 1981. Gentamicin-thallous-

carbonate medium for isolation of fecal streptococci from foods. Appl. Environ. Microbiol. 41:724-728. 1. Mean log counts of enterococci on fGTC agar at each stage in the slaughter process at each plant.

Stage 1 samples were taken immediately after singeing and polishing, stage 2 after the final rinse, and stage 3 after a 24-h chill. Stage 4 samples were from loins immediately before packaging. CM- 8 - k 4^ o c 7 - il Plant A (0 Plant B 0) o>CO 6 " A"" Plant C D) o 5 -

1 I I r Stage 1 Stage 2 Stage 3 Stage 4 Fig. 2. Comparison of mean log counts for all stages at each plant. The comparison shows the overall

enterococcal counts at each plant. Mean enterococcal count

log CFU/100 cm^

o ro O) CO

"O D) 3 >

"O Q) 3 w

"O 03 3 o

86 Fig. 3. Proportional representation of 175 enterococcal species isolated from all samples taken from pork

slaughtering plants. Complete circle represents 100% of enterococci isolated. 2%

3%

11% E. faecalis E. faecium E. pseudoavium I E. malodoratus I E. solitarius I E. casseliflavus ] E. durans I Ent. no ID Fig. 4. Comparison of mean log counts of pork sausage samples from pork processing plants. Each sample

was plated on 3 different media: fluorescent gentamycin-thallous-carbonate (fGTC) agar, KF

Streptococcal (KF) agar, and Tryptone Glucose Extract (TGE) agar. Fresh A represents fresh pork sausage produced by processing plant A. Fresh B represents fresh pork sausage from plant B. Spoiled A indicates pork sausage samples that showed signs of spoilage before the expiration date; these were

supplied by plant A. Expired B samples are those whose expiration date was reached; these were supplied by plant B. 9

8

7

6

5

4

3

2

1 Fresh A Fresh B Spoiled A Expired B Processed pork sausage Fig. 5. Proportions of enterococci isolated from processed pork sausage

samples. (A) Enterococcal spp. isolated from fresh pork sausage from plants A and B. (B) Enterococcal spp. isolated from spoiled pork

sausage from plant A. (C) Enterococcal spp. isolated from expired

pork sausage from plant B. Complete circles represent 100% of enterococci isolated from the respective samples. 104

11%

83%

B 14%

E. faecalis E. pseudoaviun E. durans E. hirae E. faecium E. malodoratus 84% E. raffinosus

11% 105

PAPER 3; COMPARISON OF FLUORESCENT GENTAMICIN-THALLOUS- CARBONATEAND KF STREPTOCOCCAL AGARS TO ENUMERATE ENTEROCOCCI AND FECAL STREPTOCOCCI IN MEATS 106

Comparison of Fluorescent Gentamicin-Thallous-Carbonate and KF Streptococcal agars to Enumerate Enterococci and Fecal Streptococci in Meats

LINDA M. KNUDTSON AND PAUL A. HARTMAN*

Department of Microbiology, Immunology and Preventive Medicine Iowa State University, Ames, Iowa 50011 Telephone: (515)294-8824 FAX: (515)294-6019

^Corresponding author 107

ABSTRACT

Two selective and differential media were compared for their abilities to enumerate enterococci and fecal streptococci in pork, beef, and poultry products. Counts obtained on KF Streptococcal (KF) agar were compared with counts obtained by using Fluorescent Gentamicin-Thallous-Carbonate (fGTC) agar. Reactions of 13 known enterococcal species also were observed. All 13 species of enterococci as well as S. bovis and S. equinus grew equally well on fGTC agar. KF Streptococcal medium allowed growth of most species of enterococci, but not S. bovis and S. equinus. Comparisons between the two media showed that counts on fGTC agar were consistently and significantly higher than counts on KF agar for all sample sources. This trend was not observed, however, when pure cultures of most known species of enterococci were quantitated by using both media. 108

INTRODUCTION

EnterococcI have been considered an alternative potential indicator of fecal pollution of foods because they have an advantage over coliforms, the primary fecal indicator, in that they are more resistant than coliforms to most environmental insults. This trait is shared by many potential pathogens; therefore, enterococci can help determine the sanitary history of moderately heated, frozen, dried or salted foods, and other foods where coliforms might not have survived (13). Enterococci are also important in human infections such as endocarditis and bacteremia. They may be resistant to clinically important antibiotics (9).

Many media have been developed to isolate and enumerate enterococci (11). In 1961, Kenner et al. (3) described new solid and liquid media (KF streptococcal broth and agar) for enumerating enterococci. These media contained, among other ingredients, sodium azide, bromcresol purple, and 2,3,5-triphenyltetrazolium chloride (TTC). A new medium, gentamicin- thallous-carbonate (GTC) agar, was reported in 1978 (1). It was superior to

KF agar for the enumeration of fecal streptococci in fecal and surface-water samples. GTC agar was modified (fGTC agar) to make it more differential by

Littel and Hartman (8) in 1983 by incorporating a colorimetric starch substrate plus a fluorogenic substrate, allowing differentiation of colonies on the agar surface. Gentamicin and thallous acetate were the major selective agents. NaHCOs, Tween 80, and KH2PO4 were added as specified earlier by

Lachica and Hartman (7) to stimulate the growth of group D streptococci. The incorporation of amylose azure and 4-methylumbelliferyl-a-D-galactoside 109

allowed differentiation of the fecal streptococci into three phenotypic groups: starch hydrolysis and fluorescence, no starch hydrolysis but fluorescence, and no starch hydrolysis or fluorescence (8).

With the recent changes in classification of this group of organisms, questions have arisen on the selectivity of the media to new members of the genus Enterococcus. The abilities of these media to recover all species of enterococci and fecal streptococci, as well as the differentiating characteristics of each species on both media, are the subject of this study. 110

MATERIALS AND METHODS

Cultures

Fifty-nine strains of 13 species of enterococci and Streptococcus bovis and Streptococcus equinus were collected from various sources (5). Cultures also were isolated from pork carcasses during slaughter, fresh and spoiled pork sausage, poultry, and beef products. All cultures were isolated according to Knudtson and Hartman (6). Media

KF Streptococcal agar (KF; Difco Laboratories, Detroit, Mich.) was prepared according to the manufacturers' instructions. One-ml portions of appropriate dilutions were used for KF agar pour plates. Duplicate plates were incubated at 37°C for 48 h. Colonies exhibiting a red or pink color were counted as streptococci (3). The second medium, fGTC agar (8) was prepared as specified by the authors. One-tenth ml portions of appropriate dilutions were plated in duplicate, and the plates were incubated at 37°C for

24 h. All except pinpoint colonies were counted as enterococci. Positive fluorescence and starch hydrolysis (zones of clearing) also were recorded. Species identification

After the KF and fGTC plates had been counted, three colonies of each colony type on each medium were streaked for isolation on Brain Heart

Infusion agar (Difco). After 24 h of incubation, a gram stain and catalase test were performed to verify gram-positive, catalase-positive colonies. The cultures also were tested on Bile Esculin (BE) agar (Difco Laboratories) for their abilities to grow in bile and hydrolyze esculin. Cultures positive for these 111

tests were identified to species by using tlie classification schema developed by Knudtson and Hartman (5). This schema was developed by using API

Rapid Strep and MicroScan Pos ID panels. The identities of the isolates were compared with results of known strains on both media. 112

RESULTS AND DISCUSSION

In every instance, enterococcus counts made on pork, beef, and poultry samples by using fGTC agar were significantly higher (P < 0.01) than counts on KF agar (Figure 1). One explanation is that a larger proportion of injured enterococci grew on fGTC agar than on KF agar. In 1961, Kenner et al. (3) stated that KF agar allowed the growth of S. bovis, S. equinus and the enterococcal group consisting of S. faecalis, and its varieties S. faecalis var. liquifaciens and zymogenes, and related S. faecalis biotypes. Significant classification changes have occurred since that time, and many of the newly classified species of enterococci may not have been tested for growth on KF agar. fGTC agar allows grov^h of S. faecalis, S. faecium, S. bovis, S. equinus, and S. avium (8), This medium was developed before the classification changes that began in 1984 (12). Newly classified species of enterococci had not been tested on this medium either.

When known species of enterococci were plated on each medium (Table 1), fGTC agar allowed growth of all enterococci and fecal streptococci tested. Strains on fGTC agar were differentiated into three groups according to the ability to fluoresce and hydrolyze starch. E. faecalis, E. pseudoavium, E. solitarius and S. equinus comprised a group that was both fluorescence- and starch-negative. S. bovis was positive for both fluorescence and starch hydrolysis. Most of the rest of the enterococcal species tested were fluorescence-positive and starch-negative. When E. avium and £ faecium were tested, only 38% and 71% of isolates produced fluorescence, respectively. Littel and Hartman (8) reported similar findings; when 87 strains of S. faecium were plated on fGTC agar, 80 (92%) showed fluorescence, and 113

7 (8%) did not. KF agar did not allow growth of known strains of E cecorum,

S. bovis or S. equinus (Tablel), other enterococci grew on this medium. This medium contains sodium azide as a selective agent. Some streptococci, such as S. bovis, cannot initiate growth on media that contain azide (1). Most species reduced tetrazolium to some degree, and E. faecalis reduced it the most strongly (Table 1). Differences in the intensity of tetrazolium reduction were difficult to visualize and could be determined accurately only if plates containing different species were compared side-by-side. When counts obtained from plating diluted samples of known strains were compared (data not shown), there were no significant differences in numbers on the two media

(except for the three species that did not grow on KF agar).

To determine if differences in counts obtained from meat samples plated on the two media were a result of the recovery of species on fGTC agar that did not grow on KF agar, the identities of 175 isolates collected from pork carcasses were tabulated (Table 2). The distribution of most species of enterococci was similar on both media. No E. cecorum, S. bovis, or S. equinus was isolated from the pork carcasses by using either medium. Only

E. durans and E, casseliflavus were isolated on fGTC agar and not on KF agar. These accounted for only very small percentages of isolates and could not be sufficient to account for the differences in counts.

Another possible explanation for higher recoveries on fGTC agar than on KF agar might be that organisms other than enterococci and fecal streptococci grew on fGTC agar, giving false-positive results. However, more than 95% of isolates from the colony types counted as enterococci on fGTC agar (pinpoint colonies were not counted, although some were identified to 114

assure they were not enterococci), were identified as enterococci or fecal streptococci (4). These results were similar to those reported by Littel and

Hartman (8), who found that 90% of isolates from sewage and fecal samples were enterococci or fecal streptococci. Some care must be exercised, however, when using fGTC agar. Knudtson and Hartman (6) found that when samples (such as expired pork sausage) highly contaminated with lactobacilli were plated on fGTC agar, enterococcal colonies could not be discerned among a heavy background of Lactobacillus colonies.

Hartman et al. (2) stated that many, if not a large majority, of the microorganisms in certain foods and natural environments may be in a different physiological state than similar strains cultivated in the laboratory.

With laboratory stock cultures as test material, continuation of bacterial growth is the concern, whereas with bacteria from the natural environment the problem seems to be not only growth, but also growth initiation (10). The overall effect is that a medium will usually be more inhibitory to bacteria from natural products than to rapidly growing laboratory cultures. If this is the case, then it is possible that the differences between counts made on fGTC and KF agars could be a result of the inhibitory properties of KF agar (which contains sodium azide) to enterococci and fecal streptococci present in natural products.

In conclusion, fGTC agar is a selective and differential medium for enterococci and fecal streptococci. It is as good as, if not better than, KF agar to enumerate enterococci and fecal streptococci in foods. fGTC agar also enables colony differentiation that might be useful in some circumstances. It 115

should be noted that fGTC agar can yield falsely high counts when excessive numbers of lactobacilli are present. 116

ACKNOWLEDGMENT

Supported by a U.S.D.A. grant to the Food Safety Consortium. Journal

Paper No. J-XXXXX of the Iowa Agriculture and Home Economics Experiment Station, Ames, project 2991. 117

REFERENCES CITED

1. Donnelly, L. S., and P. A. Hartman. 1978. Gentamicin-based medium

for the isolation of group D streptococci and application of the medium to

water analysis. Appl. Environ. Microbiol. 35:576-581.

2. Hartman, P. A., G. W. Reinbold, and D. S. Saraswat. 1966. Media

and methods for isolation and enumeration of the enterococci. Adv. Appl.

Microbiol. 8:253-289.

3. Kenner, B. A., H. F. Clark, and P. W. Kabler. 1961. Fecal

streptococci. I. Cultivation and enumeration of streptococci in surface waters. Appl. Microbiol. 9:15-20.

4. Knudtson, L. M. 1992, Classification of enterococci and their roles in

spoilage of pork products and as sanitary indicators in pork processing. Ph.D. thesis. Iowa State University, Ames.

5. Knudtson, L. M., and P. A. Hartman. 1992. Routine procedures for

isolation and identification of enterococci and fecal streptococci. Appl.

Environ. Microbiol. 58:3027-3031.

6. Knudtson, L. M., and P. A. Hartman. 1992. Enterococci in pork

processing. J. Food Prot. in press control # JFP-92-119. 118

7. Lachica, R. V. F., and P. A. Hartman. 1968. Two improved media for isolating and enumerating enterococci in certain frozen foods. J. Appl.

Bacterid. 31:151-156.

8. Littel, K. J., and P. A. Hartman. 1983. Fluorogenic selective and differential medium for isolation of fecal streptococci. Appl. Environ.

Microbiol. 45:622-627.

9. Murray, B. E. 1990. The life and times of the Enterococcus. Clin. Microbiol. Rev. 3:46-65.

10. Ray, B. (éd.). 1989. Injured index and pathogenic bacteria: Occurrence and detection in foods, water and feeds. CRC Press, Inc, Boca Raton, Florida.

11. Reuter, G. 1985. Selective media for group D streptococci. Int. J. Food Microbiol. 2:103-114.

12. Schleifer, K. H., and R. Kiipper-Bâiz. 1984. Transfer of

Streptococcus faecalis and Streptococcus faecium to the genus

Enterococcus nom. rev. as Enterococcus faecalis comb. nov. and

Enterococcus faecium comb. nov. Int. J. Syst. Bacteriol. 34:31-34. 119

13. Thian, T. S., and P. A. Hartman. 1981. Gentamicin-thallous-

carbonate medium for isolation of fecal streptococci from foods. Appl. Environ. Microbiol. 41:724-728. 120

Table 1; Growth characteristics of known species of enterococci and fecal streptococci on fGTC and KF agars

fGTC agar KF agar No. of SPECIES Strains Growth F s Growth Color

E avium 8 100% 38% 0 100% PINK

E. casseliflavus 5 100% 100% 0 100% PINK

E. cecorum 2 100% 100% 0 0 E. durans 4 100% 100% 0 100% PINK E. faecalis 8 100% 0 0 100% DK RED

E. faecium 7 100% 71% 0 100% PINK

E. gallinarum 4 100% 100% 0 100% RED

E. hirae 6 100% 100% 0 100% PINK

E. malodoratus 2 100% 100% 0 100% RED

E. mundtii 4 100% 100% 0 100% RED E. pseudoavium 2 100% 0 0 100% RED E. raffinosus 1 100% 100% 0 100% RED

E. solitarius 1 100% 0 0 100% RED

S. bows 4 100% 100% 100% 0 S. equinus 1 100% 0 0 0

% Indicates the percentages that showed a positive result, whether it be growth, (F) fluorescence, or (S) starch hydrolysis. Color indicates the color of colonies on KF agar. 121

Table 2: Species isolated from pork carcasses on fGTC and KF agars

fGTC agar KF agar Species identified # isolated % # isolated % E. faecalis 49 69 87 84

E. faedum 15 19 7 6

E. durans 2 3 0 0

E. casseliflavus 1 2 0 0

E. malodoratus 1 2 2 2

E. solitarius 1 2 2 2

E. pseudoavium 1 2 4 4

Enterococcus NO ID 1 2 2 2

TOTAL 71 100 104 100

# Isolated indicates the number of each species isolated on each type of medium. % Indicates the percentage each species that was isolated. Fig. 1. Comparison of mean log counts on fGTC and KF agars of four sample sources (pork

carcasses=/100cm2; other products =/gram). All of the differences between the two media significant (P < 0.01). o o ro c CO (0 0) E O) o

Pork carcasses Pork products Beef products Poultry products SAMPLE SOURCE 124

PAPER 4: ANTIBIOTIC RESISTANCE AMONG ENTEROCOCCAL ISOLATES FROM CLINICAL AND ENVIRONMENTAL SOURCES 125

Antibiotic Resistance among Enterococcal Isolates from Clinical and Environmental Sources

LINDA M. KNUDTSON AND PAUL A. HARTMAN*

Department of Microbiology, Immunology and Preventive Medicine Iowa State University, Ames, Iowa 50011 Telephone: (515)294-8824 FAX: (515)294-6019

^Corresponding author 126

ABSTRACT

Antibiotic resistance among enterococci and fecal streptococci was examined by testing 149 isolates from pork, water, and clinical material, as well as 50 known species, for resistance to 27 different antimicrobial agents. Tests were performed by using MicroScan Pos MIC type 6 panels. Pork isolates exhibited less resistance than either water or clinical isolates to most antibiotics, although a larger proportion of pork isolates than others was resistant to tetracycline. Comparisons of antimicrobial resistance patterns between enterococcal species revealed that Enterococcus faecium was most resistant to p-lactam antimicrobials, especially ampicillin, whereas

Enterococcus faecalis appeared to be the most resistant to the synergistic effects of antimicrobial combinations. Vancomycin resistance was observed in one Enterococcus hirae isolate from water. Enterococcal isolates from any of the sources tested did not show multiple resistance to antibiotics (such as gentamicin, ampicillin, streptomycin, and vancomycin) used to treat serious infections caused by gram-positive cocci. 127

INTRODUCTION Recent attention has focused on enterococci because of their remarkable and increasing resistance to antimicrobial agents (12). This resistance allows them to survive in environments in which antimicrobial agents are heavily used. Antimicrobials are given to food animals, such as swine, to improve their growth rate and feed conversion (3). However, many people believe that the use of antimicrobials in humans or animals is often followed by appearance of resistant microorganisms (13). Studies by Cohen and Tauxe (2) suggested that the antimicrobial drugs to which food animals were exposed provided selective pressure that lead to the appearance and persistence of drug resistant strains. Specifically, they associated the occurrence of certain drug-resistant Salmonella sp. to antimicrobial use in food animals. If antimicrobial use in food animals is linked to antimicrobial resistance in Salmonella, Enterococcus species in the intestinal tract might also be expected to develop similarly elevated resistance patterns. Antimicrobial resistance in enterococci can be divided into two general types, intrinsic and acquired. Intrinsic resistance is present in all or most strains of those species, and the genes appear to reside on the chromosome.

Intrinsic resistance includes resistance to semisynthetic, penicillinase- resistant penicillins, cephalosporins, and low levels of aminoglycosides and clindamycin. Acquired resistance results from a mutation in cellular DNA or acquisition of new DNA. Examples of acquired resistance include resistance to chlorampenicol, erythromycin, tetracycline, high levels of aminoglycosides and clindamycin, penicillin by means of penicillinase, fluoroquinolones, and vancomycin (12). Several researchers have shown that differences in 128

antimicrobial susceptibility exist between Enterococcus faecium and

Enterococcus faecalis (4,10,11). The possible existence of similar susceptibility differences among the more recently described species of

Enterococcus needs to be determined. An assessment of possible correlations between a certain species and a given susceptibility pattern could provide valuable information (14). Few data are available on antimicrobial resistance patterns of isolates from widely divergent environmental sources, conducted in a single laboratory by using a common procedure.

In this study, enterococci and fecal streptococci were collected from three sources: pork, including pork carcasses and processed pork products; water. Including samples from rivers, lakes, and well water; and clinical material from several sources. These samples, as well as known strains, were tested for resistance to 24 antimicrobials and 3 synergy screens with

MicroScan Pos MIC type 6 panels. The results were examined for differences in resistance patterns between enterococci from the three different sources

(pork, water, clinical). Differences in resistance patterns between different species of enterococci and fecal streptococci also were examined. 129

MATERIALS AND METHODS Cultures

Fifty known strains of 13 species of enterococci and Streptococcus bovis and S. equinus were collected from various sources and their identities confirmed (7). Cultures also were isolated from pork carcasses during slaughter and from fresh and spoiled pork products. All pork sample cultures were isolated according to Knudtson and Hartman (8). Enterococci from water samples were isolated by the membrane filter method (1). Clinical isolates were collected from patients at one hospital over a two-year period (6).

Primary isolation on blood agar plates revealed colonies that, upon gram- staining, were gram-positive cocci. Catalase, bile esculin, and Lancefield grouping tests confirmed that the isolates were group D enterococci. Species identifications were carried out as indicated by Knudtson and Hartman (7).

Antibiotic susceptibility testing

Inoculum preparation for the MicroScan Pos MIC type 6 panels (Baxter Diagnostics, Deerfield, III.) was carried out by the log-phase technique specified by the manufacturer. Panels were inoculated, covered, and incubated at 37''C. After 18 to 24 hours of incubation, results were interpreted as indicated in the manufacturer's instructions. 130

RESULTS AND DISCUSSION

The percentages of water, pork and clinical isolates that were resistant to 27 different antimicrobial agents are shown in Table 1. Generally, smaller proportions of the pork isolates were resistant than either the water or clinical isolates. Only with cefazolin, imipenem, and tetracycline were larger proportions of the pork isolates resistant than the clinical isolates. Only in three instances, with penicillin, erythromycin, and tetracycline did higher proportions of the pork isolates exhibit resistance than the water isolates

(Table 1). Water isolates were more resistant than either the pork or clinical isolates to all cephalosporins, amikacin, gentamicin, imipenem and rifampin.

Only one isolate, from water, was resistant to vancomycin.

It has been suggested that animal husbandry practices have contributed to the dissemination of antibiotic resistance among intestinal isolates (2,3,13). Langlois et al. (9) stated that the selection for resistant bacteria brought about by the use of antimicrobials would not be easily reversed by partial restriction of antibiotics used in veterinary practice because the resistant bacteria are passed on from one animal generation to another. Resistant fecal coliforms were present in swine at the time of slaughter, regardless of recent administration of antimicrobial agents (9). If this were true for enterococci, isolates from pork carcasses and processed pork should show antimicrobial resistance patterns similar to the clinical isolates. However, the incidence of acquired resistance generally was lower in isolates from pork than from water or clinical material. The major exception was resistance to tetracycline, which is a common additive to swine feeds. 131

When resistance patterns of known strains and environmental isolates were compared (Table 2), several trends could be seen. All enterococci possessed a degree of resistance to the aminoglycosides and cephalosporins, as would be predicted by their intrinsic resistance. Intrinsic resistance to the semisynthetic penicillins, ticarcillin and oxacillin, also was prevalent throughout most species of the enterococci (Table 2). Enterococcus cecorum, however, was not resistant to any of these antimicrobials for which it should carry intrinsic resistance. It does, however, show resistance to clindamycin along with the other enterococci (Table 2). As stated in the literature (4,10), Enterococcus faecium appears to carry the highest amount of resistance to p-lactam antimicrobials, especially ampicillin. On the other hand, it was also stated that E. faecium strains are often more refractory to the synergistic effects of antibiotic combinations. Our results show that E. faecalis was more often resistant to synergistic combinations of antibiotics. Grayson et al. (5) stated that Enterococcus raffinosus showed higher levels of resistance than E. avium to penicillin; this was supported by our findings. Resistance to vancomycin was seen only In one strain of E. hirae, isolated from a sample of creek water.

In conclusion, pork and pork products did not harbor enterococci with levels of antibiotic resistance that were substantially higher than those possessed by enterococci obtained from water or from clinical material with the exception of resistance to tetracycline. No more resistance to antibiotics used clinically to treat gram-positive infections in humans was observed in any isolate from pork than was seen in water or clinical isolates. Therefore, antibiotic resistant enterococci and fecal streptococci of pork origin do not 132

present an exceptional public health hazard. Clinical and water isolates carry more varied antibiotic resistance patterns than isolates from pork. Also, antibiotic-resistance patterns differ among enterococcal species, even those isolated from environmental sources. 133

ACKNOWLEDGEMENT

Supported by a U.S.D.A. grant to the Food Safety Consortium. Journal

Paper No. J-XXXXX of the Iowa Agriculture and Home Economics Experiment Station, Ames, project 2991. 134

LITERATURE CITED

1. American Public Healtii Association. 1989. Standard methods for the

examination of water and wastewater, 17th ed., American Public Health Association, Washington, D.C.

2. Cohen, IW. L., and R. V. Tauxe. 1986. Drug-resistant Salmonella in the United States: An epidemiological perspective. Science 234:964-969.

3. DuPont, H. L., and J. H. Steele. 1987. Use of antimicrobial agents in

animal feeds: Implications for human health. Rev. Infect. Dis. 9:447-460.

4. Gray, J. W., D. Stewart, and S. J. Pedler. 1991. Species

identification and antibiotic susceptibility testing of enterococci isolated

from hospitalized patients. Antimicrob. Agents Chemother. 35:1943-1945.

5. Grayson, M. L., G. M. Eliopoulos, C. B. Wennersten, K. L. Rouff, K. Klimm, F. L. Sapico, A. S. Bayer, and R. C.

Moellering, Jr. 1991. Comparison of Enterococcus raffinosus m\\\

Enterococcus avium on the basis of penicillin susceptibility, penicillin- binding protein analysis, and high-level aminoglycoside resistance.

Antimicrob. Agents Chemother. 35:1408-1412.

6. Knudtson, L. M. 1992. Classification of enterococci and their roles in

spoilage of pork products and as sanitary indicators in pork processing.

Ph.D. thesis. Iowa State University, Ames. 135

7. Knudtson, L. M., and P. A. Hartman. 1992. Routine procedures for

isolation and identification of enterococci and fecal streptococci. Appl.

Environ. Microbiol. 58:3027-3031.

8. Knudtson, L. M., and P. A. Hartman. 1992. Enterococci in pork processing. J. Food Prot. in press control # JFP-92-119.

9. Langlois, B. E., K. A. Dawson, I. Leak, and D. K. Aaron. 1988. Antimicrobial resistance of fecal colifoms from pigs in a herd not exposed to antimicrobial agents for 126 months. Vet. Microbiol. 18:147-153.

10. Louie, M., A. E. Simor, S. Szeto, M. Patel, B. Kreiswirth, and

D. E. Low. 1992. Susceptibility testing of clinical isolates of Enterococcus faecium and Enterococcus faecalis. J. Clin. Microbiol. 30:41-45.

11. Moeilering, R. C., Jr., O. M. Korzeniowski, M. A. Sande, and

C. B. Wennersten. 1979. Species-specific resistance to antimicrobial

synergism in Streptococcus faecium and Streptococcus faecalis. J. Infect. Dis. 140:203-208.

12. Murray, B. E. 1990. The life and times of the Enterococcus. Clin.

Microbiol. Rev. 3:46-65. 136

13. Neu, H. C. 1992. The crisis in antibiotic resistance. Science 257:1064- 1072.

14. Rouff, K. L. 1990. Recent taxonomic changes in the genus

Enterococcus. Eur. J. Clin. Microbiol. Infect. Dis. 9:75-79. 137

Table 1: Percentages of water, pork and clinical isolates resistant to 27 different antimicrobial agents

Water Pork Clinical . ... ^ . %resistant %resistant %resistant Antibiotics tested (49)3 (50) (50) Aminoglycosides Amikacin 96 48 90 Gentamicin 88 36 74 ..b Gentamicin synergy screen — 36 Streptomycin synergy screen 6 — 32 CeohalosDorins Cephalothin 59 32 38 Cefazolin 73 30 26 Cefuroxime 94 84 84 Cefotaxime 94 60 84 Ceftriaxone 92 62 76 Penicillins

Amoxicillin/K clavulanate " " 2 Ampicillin — - 4 Ampicillin/Sulbactam - — 4 Oxacillin 96 84 96 Penicillin — 2 4 Ticarcillin/K clavulanate 94 68 98 Other 6-lactams Imipenem 10 6 4 Miscellaneous Chloramphenicol — — — Ciprofloxacin 29 6 62 Clindamycin 92 86 92 Erythromycin 35 38 62 Nitrofurantoin ~ — — Norfloxacin 31 2 46 Rifampin 49 20 26 Sulfamethoxazole 100 98 100 Tetracycline 37 88 56 Trimethoprim/Sulfamethoxazole 22 - 36 Vancomycin 2 — -

= Numbers in parentheses indicate number of isolates tested. All isolates were susceptible. Table 2: Percentages of resistance among enterococci and fecal streptococci.

3 Indicates number of strains tested.

^Numbers indicate the percentage of strains resistant. PAH isolates were susceptible. <^/S= trimethoprim/sulfamethoxazole. & ^ Q 0 to 5 S ^ _ o I i f I i ! I I f I Ï I <538-§4§«g,SgëâS8SS-| UjUjUjtjLjUjUjUjUjUjUjUjUjUjCoC/juj Aminoglycosides Amikacin 78 50 56 88 59 71 92 45 ÏÔÔ ÏÔ ÏÔÔ 50 - - ^ Gentamicin 60 67 - 56 70 65 57 75 45 100 10 100 88 06ntârnicin synsrQy 60 "• •* *" 24 — — — — — -- -- — — — — Streptomycin synergy 22 6 14 50 6

Cephalothin 20 22 — 56 42 62 71 50 18 50 20 50 — — 100 59 Cefazolin 20 22 — 56 69 68 86 58 36 83 40 50 50 -- 100 73 Cefuroxime 100 78 50 89 87 94 100 67 82 100 60 100 50 — 100 94 Cefotaxime 20 56 — 89 86 79 100 67 36 83 30 100 50 —- — 94 Ceftriaxone 20 44 — 78 82 82 100 67 45 85 20 100 50 —— — 92 Penicillins Amoxicililn/K clavulanate — -- — — — 3 — — — — -- -- " — -- — Ampicillin — — — — — 6 —- — — -- — — — -- — — Ampicillin/Sulbactam - - ~ — — 6 14 — — — — — — — — — Oxacillin 100 100 — 89 93 91 100 67 100 100 80 100 100 — 100 96 Penicillin — — — — — 9 — — — — 50 — —- -- — Ticarcillin/K clavulanate 100 100 — 67 90 79 100 67 82 100 20 100 100 — ICQ 94 Other Q-lactams Imipenem — 11 — — 1 21 14 17 9 " — -- " — - 10 Miscellaneous

Chloramphenicol — — — — 2 — — — — — — — — " — — Ciprofloxacin 80 56 — 22 40 35 29 8 27 33 — — — 75 " 29 Clindamycin 80 100 100 78 93 88 86 92 73 100 60 100 100 -- 100 92 Erythromycin 40 33 100 44 46 53 14 8 36 33 40 50 — — -- 35 Nitrofurantoin — — — — — — " — — -- — —— -- —— -- —— Norfloxacin 60 33 — 22 33 21 29 17 27 67 -- — — 100 —» 31 Rifampin " 22 — 11 28 53 29 8 9 " -- 50 50 — 100 49 Sulfamethoxazole 100 100 100 100 95 100 100 100 100 100 70 100 100 100 100 100 Tetracycline 60 33 — 67 64 35 71 8 73 — 60 50 50 — -- 37 T/S" 40 22 50 22 24 24 100 8 -- 17 — —— 50 50 — 22 Vancomycin — — — — — — — 8 140

PAPER 5; COMPARISON OF FOUR LATEX AGGLUTINATION KITS TO RAPIDLY IDENTIFY LANCEFIELD GROUP D ENTEROCOCCI AND FECAL STREPTOCOCCI 141

Comparison of Four Latex Agglutination Kits to Rapidly Identify Lancefield Group D Enterococci and Fecal Streptococci

LINDA M. KNUDTSON AND PAUL A. HARTMAN*

Department of Microbiology, Immunology and Preventive Medicine Iowa State University, Ames, Iowa 50011 Telephone: (515)294-8824 FAX: (515)294-6019

•Corresponding author 142

ABSTRACT

Enterococci isolated from water (50 strains), clinical material (50 strains), pork products (25 strains), and other foods (25 strains) as well as 50 known strains were used to compare the abilities of four latex streptococcal grouping kits to correctly identify group D isolates. The Streptex kit (Wellcome

Diagnostics) was 98.5% accurate and easiest to interpret. The Slidex Strepto kit (Vitek Systems) and Strepslide kit (NCS Diagnostics) also were acceptable; they were 96.5% and 96.0% accurate, respectively. When the

Bacto Strep Grouping kit (Difco Laboratories) was used, 99% of the group D isolates were positive for both group D and group B, including enterococcal strains that are group D-negative. 143

Most streptococci, enterococci, and lactococci possess group-specific antigens, wliich are usually carbohydrate structural components of the cell wall. Group D antigens, however, are glycerol teichoic acids. The teichoic acids are buried in the cell wall, which makes them difficult to detect (3).

Lancefield (5) showed that these antigens could be extracted in soluble form and identified by precipitin reactions with appropriate antisera. The

Lancefield precipitin procedure is tedious and labor intensive and has essentially been replaced by convenient latex agglutination and coagglutination methods. In these methods, latex beads are coated with antisera to group-specific antigens. These latex particles agglutinate strongly in the presence of the homologous antigen and remain in smooth suspension if the specific antigen is absent.

Enterococci and fecal streptococci belong to Lancefield group D. Four latex kits for the grouping of streptococci were compared for their abilities to accurately identify 200 Lancefield group D bacteria. The four kits were the Streptex kit (Wellcome Diagnostics, Research Triangle Park, N. C.), Slidex

Strepto-Kit (Vitek Systems, Hazelwood, Mo.), Strepslide (NCS Diagnostics,

Buffalo, N. Y.), and Bacto Strep Grouping kit (Difco, Laboratories, Detroit,

Mich.). The cultures examined included 50 enterococci and fecal streptococci isolates from culture collections, 50 isolates from water, 50 from clinical material, 25 from pork products, and 25 from other foods (4).

Cultures from the five sources were retrieved from frozen storage

(-70°C in 10% glycerol) and grown on Brain Heart Infusion agar (Difco) plates to ensure culture purity. Colonies from these plates were used to perform tests on all four kits according to the manufacturer's instructions for each kit. 144

The Streptex kit (Wellcome) employs an enzyme extraction procedure using a proteolytic fraction from Streptomyces griseus. The extraction procedure included a 10-minute incubation at 37°C that was simple to perform. Of the 200 enterococci and fecal streptococci tested, only three isolates identified as enterococci were group D-negative (1 E. faecium from water, 1 E. faecalis from pork, and 1 E. faecium from other foods). There were

10 instances in which enterococci were positive for Lancefileld group G in addition to group D (3 E. malodoratus an6 7 E faecalis), and 9 of these strains were also group G-positive when tested with other kits. Some strains of group

D enterococci also possess group G antigen (1); this point is discussed by the manufacturers.

The Slidex Strepto-Kit (Vitek) also employs an enzyme extraction, but mutanolysin from Streptomyces globisporus is used in this kit. The extraction procedure was essentially the same as for the Streptex procedure. Of the same 200 enterococci and fecal streptococci tested, 193 (96.5%) were group

D-positive. Nine isolates produced strong group D agglutination and weak agglutination with all other antisera. The manufacturer stated that strong agglutination of one of the latex suspensions indicates the group of the test organism and that weak reactions that occur with other latex suspensions should be ignored. Thus, the weak false-positive reactions were an annoyance but did not result in false identification if care was taken in interpreting the results. Seven isolates were group D-negative with the Slidex

Strepto kit (4 E. faecium, 1 E. durans, 1 unidentified enterococcus from water, and 1 E. durans from food). No isolates positive for group D and group G only were observed. 145

The Strepslide (NCS Diagnostics) employs an extraction enzyme of unspecified origin. A 10-minute incubation period at 37°C is used as with the other two kits. When the 200 isolates were tested with this kit, 192 were group

D-positive, and 8 were group D-negative (3 E. faecium, 2 E. faecalis and 1 E.c/urans from water, and 1 E. faecium and 1 E. Wrae from food). There were

25 instances of agglutination with other group specific antisera (2 with group B, 2 with group C, 7 with group F, and 1 with all groups). Most (but not all) reactions were weaker than the agglutination reaction with group D. All these isolates were accurately grouped by one of the other kits. There were 12 isolates that reacted with both group D and group G antisera (2 E. malodoratus, 7 E. faecalis, 2 E. faecium, and 1 E. gallinarum)] the E. malodoratus and E. faecalis were the same as those seen with the Streptex kit.

The Bacto Strep Grouping Kit (Difco) employs an acid-dependent extraction procedure. Antigen Extractant 1 contains a pH indicator that changes color with the addition of Extractants 2 and 3. When combined and allowed to react with bacteria, these extractants release antigenic material from the bacterial cell walls. There is a 5-minute room-temperature incubation with this kit. For suspected group B or group D isolates, the antigen extraction procedure is replaced by a specific procedure for groups B and D. This procedure entails direct testing of colonies from an agar plate by mixing a loop of inoculum directly into a drop of latex suspension. When the 200 isolates were tested by using both methods (the extraction method for groups A, C, F, and G and the direct method for groups B and D) only two were group D- negative (1 E. malodoratus from pork, and 1 S. bovis from other foods). 146

Agglutination was seen in 7 instances with group F, and in 13 instances with the group G antisera (5 E. hirae, 3 E. faecalis, 2 E. malodoratus, 1 £ faecium, 1 E. casseliflavus, and 1 unidentified enterococcus). Only 3 of the group G- positive isolates (2 E. faecalis and 1 E. malodoratus) were positive on the

Streptex or Strepslide kits. Some form of agglutination or clumping was observed in almost every test with the group B and D latex. The agglutination patterns were uneven and very difficult to interpret.

With the first three kits (Streptex, Slidex, and Strepslide) E.cecorum and

E. pseudoavium were negative for the group D antigen, which would be expected because these two species of enterococci are group D-negative (2,

6). Birch et al. (1) stated that half of the £. faecalis isolates tested possessed both the group D and group G antigen. In addition to 7 E. faecalis isolates (2 from water, 2 from pork, and 3 from food), 2 isolates of E. malodoratus (from water) were consistently group D- and G-positive. The other instances of group G positives probably were anomalous results. The Streptex kit

(Wellcome) was 98.5% accurate for the identification of group D-positive enterococci and fecal streptococci. All procedures were straightfonmrard, easy to perform, and required a minimum of time and attention. The Bacto Strep

Grouping kit (Difco) acid extraction procedure, on the other hand, was more time consuming because of the number of extractants used, and the kit was inaccurate for identifying group D isolates. Both the Strepslide (NCS) and

Slidex (Vitek) kits were acceptable in their abilities to identify group D enterococci and fecal streptococci from clinical and environmental sources. 147

ACKNOWLEDGMENT

Supported by the U.S. Department of Agriculture grant to the Food Safety Consortium. Journal Paper No. J-15097 of the Iowa Agriculture and

Home Economics Experiment Station, Ames; project 2991. 148

LITERATURE CITED

1. Birch, B. R., G. L. Keaney, and L. A. Ganguii. 1984. Antibiotic susceptibility and biochemical properties of Streptococcus faecalis strains

reacting with both D and G antisera. J. Clin. Pathol. 37:1289-1292.

2. Collins, M. D., R. R. Facldam, J. A. E. Farrow, and R.

Williamson. 1989. Enterococcus raffinosus sp. nov., Enterococcus

solitarius sp. nov. and Enterococcus pseudoavium sp, nov. FEMS Microbiol. Lett. 57:283-288.

3. Facklam, R. R., and J. A. Washington III. 1991. Streptococcus and related catalase-negative gram-positive cocci, pp. 238-257. In A. Balows,

W. J. Hausler, K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (eds.),

Manual of Clinical Microbiology. 5th ed. American Society for

Microbiology, Washington, D. C.

4. Knudtson, L. M. 1992. Classification of enterococci and their roles in

spoilage of pork products and as sanitary indicators in pork processing.

Ph.D. thesis. Iowa State University, Ames.

5. Lancefield, R. 0. 1933. A serological differentiation of human and other

groups of hemolytic streptococci. J. Exp. Med. 57:571-595. 149

6. Williams, A. M., J. A. E. Farrow, and M. D. Collins. 1989. Reverse transcriptase sequencing of 16S ribosomai RNA from Streptococcus cecorum. Lett. Appl. IVIicrobiol. 8:185-189. 150

GENERAL SUMMARY AND DISCUSSION

Over the past six years, a revised classification of the streptococci and enterococci has emerged that is based primarily on molecular techniques such as 16S rRNA sequencing and DNA-DNA hybridization studies. As a result of these studies, the genus Streptococcus was divided into three genera: Enterococcus, Lactococcus, and Streptococcus (132). The genus

Enterococcus now contains 19 species, including some species transferred from the genus Streptococcus, new species that previously belonged to other species, and new species not previously classified.

The first part of this dissertation involved devising classification schemes to differentiate the species of the genus Enterococcus and

Streptococcus bovis and Streptococcus equinus. An identification system using conventional tube tests exclusively was introduced in 1989 by Facklam and Collins (53). However, conventional tube tests are cumbersome, costly and labor intensive. Furthermore, our studies showed that their test results could not be accurately reproduced. A comparison of my tube test results with the published results of Facklam and Collins (53) revealed 17 discrepancies among 87 instances where comparisons could be made. Other researchers (17) also have had difficulties reproducing Facklam and Collins' results. When tube test results were compared with results obtained with the

API or MicroScan panels, 28 and 24 discrepancies, respectively, were observed. These results indicate that Facklam and Collins' classification, based on tube tests, should not be used to identify isolates tested with these panels. There were also discrepancies (12) between the API Rapid Strep 151

panel and the MicroScan Pos ID panel. Some of these were possibly a result of differences in methodologies between the two panels, but others could not be explained.

A classification scheme was devised to identify the thirteen species of

EnterococcusXhaX belonged to the genus at the beginning of this project, as well as S. bovis and S. equinus. Flow charts were fashioned of key tests that could be used to differentiate these species. Three supplemental tests were required to differentiate all of the species: pigmentation, motility and sucrose fermentation. The enterococci are important as potential indicators of fecal pollution

(7,108) because they are readily isolated in large numbers from human and animal feces. The possibility of using enterococci as indicators of fecal pollution by using differences in species distribution in different hosts as a means to pinpoint the source of pollution has been suggested (7, 87,124,

126). Although none of the enterococci can be considered absolutely host specific, some species do show a degree of host specificity (70). In swine,

Enterococcus faecium is the predominant organism of the intestinal tract, whereas Enterococcus faecalis predominates in the intestinal tract of humans

(7, 87,124,126). With this in mind, the second part of my dissertation focuses on using my newly devised classification scheme to identify enterococcal isolates from hog carcasses during different stages in the slaughter process.

Fresh, expired and spoiled processed pork sausage also were examined.

Enterococci were enumerated and isolates were identified from these samples to ascertain the possible fecal pollution present at different stages of pork processing, and to pinpoint the origin(s) of that pollution. Results from 152

this study showed that enterococcal counts at different stages of the pork process varied significantly at different plants. The polishing and back scraping machines, as well as the scalding tank, contributed to higher enterococcal counts in two plants. Counts from all three plants were lower after the final rinse. In one plant counts increased significantly after a 24-hour chill, whereas counts in the other two plants remained low. Since enterococci are thermoduric organisms (64), even slight increases in the cooler temperatures could have contributed to this large increase in enterococcal counts. Since significant contamination occurred at this stage, it is a critical control point for this plant. There are differences in many aspects of the slaughter process at the three plants, which explains why HACCP plans must be devised for each individual plant (146). The most interesting finding of this study is the overall predominance of Enterococcus faecalis in all three plants and at each processing stage as well. Because enterococci are associated with the intestinal tracts of pigs, enterococci isolated from a hog carcass in a slaughtering plant presumably would arise from fecal contamination of the carcasses by hog fecal matter; therefore, £ faecium should predominate. The predominance of E. faecalis indicates that hogs slaughtered in these three plants during the 4-month sampling period had an intestinal flora in which E faecalis predominated, or that the enterococcal contamination arose from other sources. Once introduced into a slaughtering plant, enterococci can become established, and the subsequent contamination of a food product, or carcass, does not necessarily indicate fecal pollution. If this is the case, the E faecalis might have been introduced by air contamination, human contamination (E. faecalis predominates in the human intestine), or improper 153

cleaning of surfaces or equipment. Further studies are being conducted by other investigators to pinpoint possible sources of this contamination.

Both fresh and expired pork sausage yielded low enterococcal counts that were predominantly E. faecalis. When expired pork samples were examined, high numbers of lactobacilli grew on the enterococcal media resulting in falsely high counts and masking enterococcal colonies. Spoiled pork sausage, on the other hand, yielded an almost pure culture of E. faecium.

It is possible that the contamination of hog carcasses with fecal matter occurs infrequently: when it does occur, contamination of a single carcass may be sufficient to instigate spoilage of sausage or hams made from the carcass.

Many media have been devised to isolate and enumerate enterococci.

KF streptococcal (KF) agar (86) contains, among other ingredients, sodium azide, bromcresol purple, and 2, 3, 5-triphenyltetrazolium chloride (TTC). The sodium azide selects for enterococci (72), and the reduction of tetrazolium

(TTC) differentiates E. faecalis, which strongly reduces tetrazolium, from other enterococci which reduce tetrazolium weakly if at all. Fluorescent gentamicin- thallous-carbonate (fGTC) agar contains gentamicin and thallous acetate (45) as the selective agents, and NaHCOa, Tween 80 and KH2PO4 to stimulate the growth of group D streptococci (90). The incorporation of amylose azure and 4-methylumbelliferyl-a-D-galactoside allowed differentiation of fecal streptococci into three categories (97). These two media were compared for their abilities to enumerate enterococci and fecal streptococci in pork, beef, and poultry products. In every instance, enterococcus counts made with fGTC agar were significantly higher than counts on KF agar. Testing of known species of enterococci on each medium revealed that all tested species of 154

enterococci and fecal streptococci grew on fGTC agar, whereas three species

(E. cecorum, S. bovis, and S. equinus) failed to grow on KF agar. Since these species were not isolated from the meat samples with any frequency, their inability to grow on KF agar could not account for the differences in counts obtained by using the two agars. Identities of the enterococci isolated from the pork, beef, and poultry samples, revealed that only Enterococcus durans and

Enterococcus casseliflavus were isolated on fGTC agar but not KF agar. The percentages of other enterococci isolated were similar between the two media. One explanation for the difference in counts is that a larger proportion of injured enterococci grew on fGTC agar than on KF agar. FGTC agar is as good as, if not better than, KF agar to enumerate enterococci and fecal streptococci in foods. Some care must be exercised, however, when using fGTC agar. When samples heavily contaminated with lactobacilli were plated on fGTC agar, enterococcal colonies could not be discerned among a heavy background of Lactobacillus colonies.

Recent attention has focused on enterococci because of their remarkable and increasing resistance to antimicrobial agents (114). This resistance allows them to survive in environments in which antimicrobial agents are heavily used. Antimicrobials are given to food animals, including swine, to improve growth rate and feed conversion. It has been suggested that antimicrobial drugs to which food animals are exposed provide selective pressure that leads to the appearance of antimicrobial resistant strains in food animals (23). Enterococci and fecal streptococci collected from pork, water, and clinical infections were tested for resistance to 27 antibiotics. Enterococci isolated from pork samples showed less resistance than enterococci isolated 155

from water or clinical material. If food animals harbor resistant bacteria as suggested, levels of resistance from pork isolates would be expected to be at least as high as those from other sources. When antimicrobial resistance patterns of different species of enterococci were compared, some trends were seen. Almost all species of enterococci showed some degree of resistance to the aminoglycosides, cephalosporins, the semisynthetic penicillins, and clindamycin. Since all enterococci carry intrinsic resistance to these classes of antimicrobials, these results were expected. Enterococcus faecium appears to carry the highest amount of resistance to the p-lactam antimicrobials, especially ampicillin, whereas E. faecalis strains appeared to be more often resistant to the synergistic affects of the antibiotic combinations tested.

The majority of streptococci, enterococci, and lactococci possess group specific antigens which are usually carbohydrate structural components of the cell wall. Group D antigens, however, are glycerol teichoic acids that are buried deep in the cell wall. Because of the intracellular nature of these teichoic acids, identification of group D isolates is difficult. Lancefield discovered that these antigens could be extracted in soluble form and identified by precipitin reactions with appropriate antisera. The precipitin reaction has largely been replaced by methods in which latex beads are coated with antisera to specific group antigens. These latex particles agglutinate strongly in the presence of homologous antigen, and are easy to visualize. In the final paper of this dissertation, four streptococcal grouping kits (Streptex, Slidex, Strepslide, and Bacto) were compared for their abilities to accurately type the group D enterococci. Isolates from culture collections. 156

water, clinical infections, pork, and other foods were tested. The Streptex

(Wellcome) kit results yielded the fewest false-negative results for group D antigen and no false-positive results were observed. When the Slidex (Vitek) and Strepslide (NCS) kits were used, some false-positive results were observed. Weak agglutination reactions with antisera specific for other groups also were observed with the latter two kits. The Bacto strep grouping kit

(Difco) employed a different procedure for identifying groups B and D which resulted in clumps and agglutination in every group B and D test, regardless of the group of the isolate. I recommend that the Difco kit not be used to type presumptive group D isolates. 157

LITERATURE CITED

1. Amstein, C. F., and P. A. Hartman. 1973. Differentiation of some enterococci by gas chromatograpfiy. J. Bacteriol. 113:38-41.

2. Anderson, M. E., H. E. Huff, H. D. Naumann, R. T. Marshall, J. Damaré, R. Johnston, and M. Pratt. 1987. A research note: Evaluation of swab and tissue excision methods for recovering microorganisms from washed and sanitized beef carcasses. J. Food Prot. 50:741-743.

3. Andrewes, F. W., and T. J. Horder. 1906. A study of the streptococci pathogenic for man. Lancet 2:708-713.

4. Barnes, E. M. 1956. IVIethods for the isolation of faecal streptococci (Lancefield group D) from bacon factories. J. Appl. Bacteriol. 19:193-203.

5. Barnes, E. M. 1956. Tetrazolium reduction as a means of differentiating S. faecalisirom S. faecium. J. Gen. Microbiol. 14:57-68.

6. Barnes, E. M. 1964. Distribution and properties of serological types of Streptococcus faecium, Streptococcus durans and related strains. J. Appl. Bacteriol. 27:461-470.

7. Barnes, E. M., and M. Ingram. 1955. The identity and origin of faecal streptococci in canned hams. Ann. Inst. Pasteur, Lille 7:115-119.

8. Barnes, E. M., G. C. Mead, C. S. Impey, and B. W. Adams. 1978. The effect of dietary bacitracin on the incidence of Streptococcus faecalis subspecies liquifacierjs and related streptococci in the intestines of young chicks. Br. Poult. Sci. 19:713-723.

9. Bartley, C. H., and L. W. Slanetz. 1960. Types and sanitary significance of fecal streptococci isolated from feces, sewage and water. Am. J. Public Health 50:1545-1552.

10. Beighton, D., J. M. Hardie, and R. A. Whiley. 1991. A scheme for the identification of the viridans streptococci. J. Med. Microbiol. 35:367- 372.

11. Bentley, R. W., J. A. Leigh, and M. D. Collins. 1991. Intrageneric structure of Streptococcus based on comparative analysis of small- subunit rRNA sequences. Int. J. Syst. Bacteriol. 41:487-494. 158

12. Besnier, J. M., C. Leport, A. Buré, and J. L. Vildé. 1990. Vancomycin-aminoglycoside combinations in therapy of entdocarditis caused by Enterococcus species and Stretococcus bovis. Eur. J. Clin. Microbiol. Infect. Dis. 9:130-133.

13. Betzl, D., W. Ludwig, and K. H. Schleifer. 1990. Identification of lactococci and enterococci by colony hybridization with 23S rRNA- targeted oligonucleotide probes. Appl. Environ, Microbiol. 56:2927-2929.

14. Black, E. M., and F. V. Buskirk. 1973. Gentamicin as a selective agent for the isolation of p haemolytic streptococci. J. Clin. Pathol. 26:154-156.

15. Bridge, P. D., and P. H. A. Sneath. 1982. Streptococcus gallinarum sp. nov. and Streptococcus oralis sp. nov. Int. J. Syst. Bacteriol. 32:410- 415.

16. Bridge, P. D., and P. H. A. Sneatli. 1983. Numerical taxonomy of Streptococcus. J. Gen. Microbiol. 129:565-597.

17. Bryce, E. A., S. J. V. Zemcov, and A. M. Clarke. 1991. Species identification and antibiotic resistance patterns of the enterococci. Eur. J. Clin. Microbiol. Infect. Dis. 10:745-747.

18. Campbell, J. J. R., and I. 0. Gunsalus. 1944. Citric acid fermentation by streptococci and lactobacilli. J. Bacteriol. 48:71-76.

19. Carey, W. E., G. M. Dack, and E. Myers. 1931. An institutional outbreak of food poisoning possibly due to a streptococcus. Proc. Soc. Exptl. Biol. Med. 29:214-215.

20. Chenoweth, C., and D. Schaberg. 1990. The epidemiology of enterococci. Eur. J. Clin. Microbiol. Infect. Dis. 9:80-89.

21. Chirurgi, V. A., S. E. Oster, A. A. Goldberg, M. J. Zervos, and R. E. McCabe. 1991. Ampicillin-resistant Enterococcus raffinosus 'm an acute-care hospital: case-control study and antimicrobial susceptibilities. J. Clin. Microbiol. 29:2663-2665.

22. Clewell, D. B. 1990. Movable genetic elements and antibiotic resistance in enterococci. Eur. J. Clin. Microbiol, Infect. Dis. 9:90-102.

23. Cohen, Wl. L., and R. V. Tauxe. 1986. Drug-resistant Salmonella in the United States: An epidemiological perspective. Science 234:964- 969. 159

24. Collins, M. D., R. R. Facklam, J. A. E. Farrow, and R. Williamson. 1989. Enterococcus raffinosus sp. nov., Enterococcus solitarius sp. nov. and Enterococcus pseudoavium sp. nov. FEMS Microbiol. Lett. 57:283-288.

25. Collins, M. D., J. A. E. Farrow, and D. Jones. 1986. Enterococcus mundtii sp. nov. Int. J. Syst. Bacteriol. 36:8-12.

26. Collins, M. D., D. Jones, J. A. E. Farrow, R. Kilpper-Bâiz, and K. H. Schleifer. 1984. Enterococcus avium nom. rev., comb, nov.; £ casseliflavus nom. rev., comb, nov.; E. durans nom. rev., comb, nov.; E. gallinarum comb, nov.; and E. malodoratus sp. nov. Int. J. Syst. Bacteriol. 34:220-223.

27. Collins, M. D., U. M. Rodrigues, N. E. Pigott, and R. R. Facklam. 1991. Enterococcus dispar sp. nov, a new Enterococcus species from human sources. Lett. Appl. Microbiol. 12:95-98.

28. Collins, M. D., A. M. Williams, and S. Wallbanks. 1990. The phylogeny of Aerococcus an6 Pediococcus as determined by 16S rRNA sequence analysis: description of Tetragenococcus gen. nov. FEMS Microbiol. Lett. 70:255-262.

29. Colman, G., and L. C. Ball. 1984. Identification of streptococci in a medical laboratory. J, Appl. Bacteriol. 57:1-14.

30. Cooper, K. E., and F. M. Ramadan. 1955. Studies in the differentiation between human and animal pollution by means of faecal streptococci. J. Gen. Microbiol. 12:180-190.

31. Daly, J. A., N. L. Clifton, K. C. Seskin, and W. M. Gooch III. 1991. Use of rapid, nonradioactive DNA probes in culture confirmation tests to detect Streptococcus agalactiae, Haemophilus influenzae, and Enterococcus spp. from pediatric patients with significant infections. J. Clin. Microbiol. 29:80-82.

32. Daly, J. A., and M. L. Rufener. 1987. Evaluation of the Strep-A-Chek technique for presumptive identification of group A p-hemolytic streptococci and group D enterococci. Diagn. Microbiol. Infect. Dis. 7:215- 218.

33. Dealler, S. F., L. Campbell, K. G. Kerr, J. McGoldrick, K. A. Flannigan, and P. M. Hawkey. 1989. Reliable five-minute test strip method for identification of Streptococcus pyogenes. Eur. J. Clin. Microbiol. Infect. Dis. 8:308-310. 160

34. Deibel, R. H. 1964. The group D streptococci. Bacterid. Rev. 28:330- 366.

35. Deibel, R. H., D. E. Lake, and J. C. F. Niven. 1963. Physiology of the enterococci as related to their taxonomy. J. Bacteriol. 86:1275-1282.

36. Deibel, R. H., and H. W. Seeley. 1974. Streptococcaceae, pp. 490- 509. In R. E. Buchanan , and N. E. Gibbons (eds.), Bergey's manual of determinative bacteriology. 8th ed. Williams and Wilkins Co., Baltimore,

37. Devriese, L. A., K. Ceyssens, and F. Haesebrouck. 1991. Characteristics of Enterococcus cecorum strains from the intestines of different animal species. Lett. Appl. Microbiol. 12:137-139.

38. Devriese, L. A., K. Ceyssens, J. Hommez, R. Kilpper-Bâiz, and K. H. Schleifer. 1991. Characteristics of different Streptococcus suis ecovars and description of a simplified identification method. Vet. Microbiol. 26:141-150.

39. Devriese, L. A., K. Ceyssens, U. M. Rodrigues, and M. D. Collins. 1990. Enterococcus columbae, a species from pigeon intestines. FEMS Microbiol. Lett. 71:247-251.

40. Devriese, L. A., G. N. Dutta, J. A. E. Farrow, A. van de Kerckhove, and B. A. Phillips. 1983. Streptococcus cecorum, a new species from chickens. Int. J. Syst. Bacteriol. 33:772-776.

41. Devriese, L. A., J. Hommez, R. Wijfels, and F. Haesebrouck. 1991. Composition of the enterococcal and streptococcal intestinal flora of poultry. J. Appl. Bacteriol. 71:46-50.

42. Devriese, L. A., L. Laurier, P. D. Herdt, and F. Haesebrouck. 1992. Enterococcal and streptococcal species isolated from faeces of calves, young cattle and dairy cows. J. Appl. Bacteriol. 72:29-31.

43. Devriese, L. A., A. van de Kerckhove, R. Kilpper-Bâiz, and K. H. Schleifer. 1987. Characterization and identification of Enterococcus species isolated from the intestines of animals. Int. J. Syst. Bacteriol. 37:257-259.

44. Dickson, J. S. 1990. Comparison of homogenization by blending or stomaching on the recovery of Listeria monocytogenes from beef tissues. J. Food Soi. 55:655-657. 161

45. Donnelly, L. S., and P. A. Hartman. 1978. Gentamicin-based medium for the isolation of group D streptococci and application of the medium to water analysis. Appl. Environ. Microbiol. 35:576-581.

46. Dykes, G. A., T. E. Cloete, and A. von Holy. 1991. Quantitation of microbial populations associated with the manufacture of vacuum- packaged, smoked Vienna sausages. Int. J. Food Microbiol. 13:239-248.

47. Ellopoulos, G. M., and C. T. Eliopoulos. 1990. Therapy of enterococcal infections. Eur. J. Clin. Microbiol. Infect. Dis. 9:118-126.

48. Etheridge, M. E., R. H. Yolken, and S. L. Vonderfecht. 1988. Enterococcus hirae implicated as a cause of diarrhea in suckling rats, J. Clin. Microbiol. 26:1741-1744.

49. Facklam, R. R. 1972. Recognition of group D streptococcal species of human origin by biochemical and physiological tests. J. Clin. Microbiol. 23:1131-1139.

50. Facklam, R. R. 1973. Comparison of several laboratory media for presumptive identification of enterococci and group D streptococci. Appl. Microbiol. 26:138-145.

51. Facklam, R. R., G. S. Bosley, D. Rhoden, A. R. Franklin, N. Weaver, and R. Schulman. 1985. Comparative evaluation of the API 20S and AutoMicrobic Gram-Positive Identification systems for non-Beta- hemolytic streptococci and aerococci. J. Clin. Microbiol. 21:535-541.

52. Facklam, R. R., and R. B. Carey. 1985. Streptococci and aerococci, pp. 154-175. In E. H. Lennette, A. Balows, W. J. Hausler Jr., and H. J. Shadomy (eds.). Manual of clinical microbiology. 4th ed. American Society for Microbiology, Washington, D. C.

53. Facklam, R. R., and M. D. Collins. 1989. Identification of enterococcus species isolated from human infections by a conventional test scheme. J. Clin. Microbiol. 27:731-734.

54. Facklam, R. R., J. F. Padula, L. G. Thacker, E. 0. Wortham, and B. J. Sconyers. 1974. Presumptive identification of group A, B, and D streptococci. Appl. Microbiol. 27:107-113.

55. Facklam, R. R., D. L. Rhoden, and P. B. Smith. 1984. Evaluation of the Rapid Strep system for the identification of clinical isolates of Streptococcus species. J. Clin. Microbiol. 20:894-898. 162

56. Facklam, R. R., L. G. Thacker, B. Fox, and L. Eriquez. 1982. Presumptive identification of streptococci with a new test system. J. Clin. Microbiol. 15:987-990.

57. Facklam, R. R., and J. A. Washington ill. 1991. Streptococcus and related catalase-negative gram-positive cocci, pp. 238-257. In A. Balows, W. J. Hausler, K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (eds.), Manual of Clinical Microbiology. 5th ed. American Society for Microbiology, Washington, D. C.

58. Farrow, J. A. E., and M. D. Collins. 1985. Enterococcus hirae, a new species that includes amino acid assay strain NCDO 1258 and strains causing growrth depression in young chickens. Int. J. Syst. Bacteriol. 35:73-75.

59. Farrow, J. A. E., D. Jones, B. A. Phillips, and M. D. Collins. 1983. Taxonomic studies on some group D streptococci. J. Gen. Microbiol. 129:1423-1432.

60. Fliss, I., R. E. Simard, and A. Ettriki. 1991. A research note: comparison of three sampling techniques for microbiological analysis of meat surfaces. J. Food Sci. 56:249-250, 252.

61. Fontana, R., P. Canepari, IW. M. Leo, and G. Satta. 1990. Mechanisms of resistance of enterococci to beta-lactam antibiotics. Eur. J. Clin. Microbiol. Infect. Dis. 9:103-105.

62. Garg, S. K., and B. K. Mital. 1991. Enterococci in milk and milk products. Crit. Rev. Microbiol. 18:15-45.

63. Garvie, E. I. 1976. Hybridization between the deoxyribonucleic acids of some strains of heterofermentative lactic acid bacteria. Int. J. Syst. Bacteriol. 26:116-122.

64. Gordon, C. L. A. 1991, Thermal susceptibility of Streptococcus faecium strains isolated from frankfurters. Can. J. Microbiol. 37:609-612.

65. Gray, J. W., D. Stewart, and S. J. Pedler. 1991. Species identification and antibiotic susceptibility testing of enterococci isolated from hospitalized patients. Antimicrob. Agents Chemother. 35:1943-1945. 163

66. Grayson, M. L., G. M. Eliopoulos, C. B. Wennersten, K. L. Rouff, K. Klimm, F. L. Sapico, A. S. Bayer, and R. C. Moellering Jr. 1991. Comparison of Enterococcus raffinosus wWh Enterococcus avium on the basis of penicillin susceptibility, penicillin- binding protein analysis, and high-level aminoglycoside resistance. Antimicrob. Agents Chemother. 35:1408-1412.

67. Gross, K. C., M. P. Houghton, and L. B. Senterfit. 1975. Presumptive speciation of Streptococcus bovis and other group D streptococci from human sources by using arginine and pyruvate tests. J. Clin. Microbiol. 1:54-60.

68. Hammonds, S. J., T. J. Easterbrook, and M. B. Mclllmurry. 1991. Differentiation of enterococci from other group D streptococci by means of a specific monoclonal antibody. FEMS Microbiol. Lett. 82:91-94.

69. Hannay, C. L. 1950. The serological identity of a yellow-pigmented Streptococcus. J. Gen. Microbiol. 4:294-297.

70. Hartman, P. A., R. H. Deibel, and L. M. Sieverding. 1992. Enterococci, pp. 523-531. In C. Vanderzant, and D. F. Splittstoesser (eds.), Compendium of methods for the microbiological examination of foods. 3rd ed. American Public Health Association, Washington, D. C.

71. Hartman, P. A., G. W. Reinbold, and D. S. Saraswat. 1965. Indicator organisms-a review. II. The role of enterococci in food poisoning. J, Milk Food Technol. 28:344-350.

72. Hartman, P. A., G. W. Reinbold, and D. S. Saraswat. 1966. Media and methods for isolation and enumeration of the enterococci. Adv. Appl. Microbiol. 8:253-289.

73. Hechelmann, H., and R. Kasprowiak. 1991. Microbiological criteria for stable products. Fleischwirtschaft 71:1303-1308.

74. Hegazi, F. 2. 1991. Factors influencing the synthesis of an extracellular Proteinase by Enterococcus faecalis subsp. liquifaciens. Nahrung 35:841-848.

75. Herman, D. J., and D. N. Gerding. 1991. Screening and treatment of infections caused by resistant enterococci. Antimicrob. Agents Chemother. 35:215-219.

76. Hitchcock, C. H. 1924. Precipitation and compliment fixation reactions with residue antigens in the non-hemolytic streptococcus group. J. Exp. Med. 40:575-581. 164

77. Hoffman, S. A., and R. C. Moellering Jr. 1987. The Enterococcus: "putting the bug in our ears". Ann. Intern. Med. 106:757-761.

78. Jacob, A. E., G. J. Douglas, and S. J. Hobbs. 1975. Self- transferable plasmids determining the haemolysin and bacteriocin of Streptococcus faecalis war. zymogenes. J. Bacteriol. 121:863-872.

79. Jayarao, B. M., S. P. Oliver, K. R. Matthews, and S. H. King. 1991. Comparative evaluation of Vitek Gram-Positive Identification system and API Rapid Strep system for identification of Streptococcus species of bovine origin. Vet. Microbiol. 26:301-308.

80. Jones, D. 1978. Composition and differentiation of the genus Streptococcus, pp. 1-49. In F. A. Skinner, and L. B. Quesnel (eds.), Streptococci. Academic Press, Inc., London.

81. Jones, D., M. J. Sackin, and P. H. A. Sneath. 1972. A numerical taxonomic study of streptococci of serological group D. J. Gen. Microbiol. 72:439-450.

82. Kalina, A. P. 1970. The taxonomy and nomenclature of enterococci. Int. J. Syst. Bacteriol. 20:185-189.

83. Kandler, O., K. H. Schleifer, and R. Dandl. 1968. Differentiation of Streptococcus faecalis Andrewes and Horder and Streptococcus faecium Orla-Jensen based on the amino acid composition of their murein. J. Bacteriol. 96:1935-1939.

84. Kaufhold, A., and P. Ferrieri. 1991. Isolation of Enterococcus mundtii from normally sterile body sites in two patients. J. Clin. Microbiol. 29:1075-1077.

85. Kaufhold, A., R. Lutticken, and U. Schwien. 1989. Few-minutes tests for the identification of group A streptococci and enterococci with chromogenic substrates. Zentralbl. Bakteriol. 272:191-195.

86. Kenner, B. A., H. F. Clark, and P. W. Kabler. 1961. Fecal streptococci. I. Cultivation and enumeration of streptococci in surface waters. Appl. Microbiol. 9:15-20.

87. Kjellander, J. 1960. Enteric streptococci as indicators of fecal contamination of water. Acta Pathol. Microbiol. Scand., Suppl. 136, 48:9- 124. 165

88. Krumwiede, M. D., Jr., and J. S. Pratt. 1914. Observations on the growth of bacteria on media containing various aniiin dyes. J. Exp. Med. 19:20-27.

89. Kusuda, R., K. Kawai, F. Salati, C. R. Banner, and J. L. Fryer. 1991. Enterococcus seriolicida sp. nov., a fish pathogen. Int. J. Syst. Bacteriol. 41:406-409.

90. Lachica, R. V. F., and P. A. Hartman. 1968. Two improved media for isolating and enumerating enterococci in certain frozen foods. J. Appl. Bacteriol. 31:151-156.

91. Lambert, A. D., J. P. Smith, and K. L. Dodds. 1991. Shelf life extension and microbiological safety of fresh meat-a review. Food Microbiol. 8:267-297.

92. Lancefietd, R. C. 1933. A serological differentiation of human and other groups of hemolytic streptococci. J. Exp. Med. 57:571-595.

93. Langston, C. W., J. Guttierez, and C. Bouma. 1960. Motile enterococci {Streptococcus faecium var. mobilis N) isolated from grass silage. J. Bacteriol. 80:714-718.

94. Lasta, J., and R. Fonrouge. 1988. Significance of samples taken for bacterial counts from reduced areas of bovine carcasses. J. Food Prot. 51:214-217.

95. Lewis, C. M., and lUI. J. Zervos. 1990. Clinical manifestations of enterococcal infection. Eur. J. Clin. Microbiol. Infect. Dis. 9:111-117.

96. Liilard, H. S. 1988. Comparison of sampling methods and implications for bacterial decontamination of poultry carcasses by rinsing. J. Food Prot. 51:405-408.

97. Littei, K. J., and P. A. Hartman. 1983. Fluorogenic selective and differential medium for isolation of fecal streptococci. Appl. Environ. Microbiol. 45:622-627.

98. Louie, M., A. E. Simor, S. Szeto, IVI. Patei, B. Kreiswirtii, and D. E. Low. 1992. Susceptibility testing of clinical isolates of Enterococcus faecium and Enterococcus faecalis. J. Clin. Microbiol. 30:41-45. 166

99. Ludwig, W., E. Seewaldt, R. Kilpper-Bâiz, K. H. Schleifer, L. Magrum, C. R. Woese, G. E. Fox, and E. Stackebrandt. 1985. The phylogenetic position of Streptococcus and Enterococcus. J. Gen. Microbiol. 131:543-551.

100. Maki, D. G., and W. A. Agger. 1988. Enterococcal bacteremia: clinical features, the risk of endocarditis, and management. Medicine (Baltimore) 67:248-269.

101. Mallmann, W. L. 1940. A new yardstick for measuring sewage pollution. Sewage Works J. 12:875-878.

102. Martinez-Murcia, A. J., and M. D. Collins. 1991. Enterococcus sulfureus, a new yellow-pigmented Enterococcus species. FEMS Microbiol. Lett. 80:69-74.

103. McKenzie, D. A. 1941. The use of thallium acetate glucose broth in the diagnosis of streptococcal mastitis. Vet. Rec. 53:473-480.

104. Moellering, R. C., Jr. 1990. The enterococci: an enigma and a continuing theraputic challenge. Eur. J. Clin. Microbiol. Infect. Dis. 9:73- 74.

105. Moellering, R. C., Jr., O. M. Korzeniowski, M. A. Sande, and C. B. Wennersten. 1979. Species-specific resistance to antimicrobial synergism in Streptococcus faecium and Streptococcus faecalis. J. Infect. Dis. 140:203-208.

106. Mogollon, J. D., C. Pijoan, M. P. Murtaugh, E. L. Kaplan, J. E. Collins, and P. P. Cleary. 1990. Characterization of prototype and clinically defined strains of Streptococcus suis by genomic fingerprinting. J. Clin. Microbiol. 28:2462-2466.

107. Morgan, I. R., F. Krautil, and J. A. Craven. 1985. A comparison of swab and maceration methods for bacterial sampling of pig carcasses. J. Hyg., Camb. 95:383-390.

108. Moussa, R. S. 1965. Species differentiation of faecal and nonfaecal enterococci. J. Appl. Bacteriol. 28:466-472.

109. Mundt, J. 0.1963, Occurrence of enterococci in animals in a wild environment. J. Appl. Microbiol. 11:136-139.

110. Mundt, J. O. 1982. The ecology of the streptococci. Microb, Ecol. 8:355-369. 167

111. Mundt, J. O. 1986. Enterococci and lactic acid streptococci, pp. 1063- 1066. In P. H. A. Sneath, N. S. Mair, M. E. Sharpe, and J. G. Holt (eds.), Sergey's manual of systematic bacteriology. Williams and Wilkins Co., Baltimore.

112. Mundt, J. O., and W. F. Graham. 1968. Streptococcus faecium var. casseliflavus nov. var. J. Bacteriol. 95:2005-2009,

113. Mundt, J. O., A. H. Johnson, and R. Khatchikian. 1958. Incidence and nature of enterococci on plant materials. Food Res. 23:186-193.

114. Murray, B. E. 1990. The life and times of the Enterococcus. Clin. Microbiol. Rev. 3:46-65.

115. Murray, B. E., K. V. Singh, J. D. Heath, B. R. Sharma, and G. M. Weinstock. 1990. Comparison of genomic DMAs of different enterococcal isolates using restriction endonucleases with infrequent recognition sites. J. Clin. Microbiol. 28:2059-2063.

116. Niven, C. F., Jr. 1955. Significance of streptococci in canned hams. Ann. Inst. Pasteur, Lille 7:120-123.

117. Nortje, G. L., E. Swanepol, R. T. Naude, W. H. Holzapfei, and P. L. Steyn. 1982. Evaluation of three carcass surface microbial sampling techniques. J. Food Prot. 45:1016-1017.

118. Nowlan, S. S., and R. H. Deibel. 1967. Group Q streptococci. I. Ecology, serology, physiology, and relationship to established enterococci. J. Bacteriol. 94:291-296.

119. Oliver, D. R., B. L. Brown, and D. B. Clewell. 1977. Characterization of plasmids determining haemolysin and bacteriocin production in Streptococcus faecalis 5952. J. Bacteriol. 130:948-950.

120. Pierson, M. D., and J. D. A. Corlett (ed.). 1992. HACCP. Principles and applications. Van Nostrand, Reinhold, N. Y.

121. Pompai, R., F. Beriutti, M. C. Thailer, A. ingianni, G. Cortis, and B. Dainelii. 1992. Enterococcus flavescens sp. nov., a new species of enterococci of clinical origin. Int. J. Syst. Bacteriol. 42:365- 369.

122. Pompel, R., G. Lampis, F. Beriutti, and M. C. Thaiier. 1991. Characterization of yellow-pigmented enterococci from severe human infections. J. Clin. Microbiol. 29:2884-2886. 168

123. Pompei, R., M. C. Thaller, F. Pittaluga, O. Flore, and G. Satta. 1992. Analysis of bacteriolytic activity patterns, a novel approach to the taxonomy of enterococci. Int. J. Syst. Bacterid. 42:37-43.

124. Fourcher, A. M., L. A. Devriese, J. F. Hernandez, and J. M. Delattre. 1991. Enumeration by a miniaturized method of Escherichia coli, Streptococcus bovis and enterococci as indicators of the origin of faecal pollution of waters. J. Appl. Bacterid. 70:525-530.

125. Raibaud, P., M. Caulet, J. V. Galpin, and G. Mocquot. 1961. Studies on the bacterial flora of the alimentary tract of pigs. II. Streptococci: selective enumeration and differentiation of the dominant group. J. Appl. Bacterid. 24:285-306.

126. Ramadan, F. M., and M. S. Sablr. 1963. Differentiation studies of fecal streptococci from farm animals. Can. J. Microbiol. 9:443-450.

127. Rappaport, T., K. P. Sawyer, and I. Nachamkin. 1988. Evaluation of several commercial biochemical and immunologic methods for rapid identification of gram-positive cocci directly from blood cultures. J. Clin. Microbiol. 26:1335-1338.

128. Rodrigues, U., and M. D. Collins. 1990. Phylogenetic analysis of Streptococcus saccharolyticus based on 16S rRNA sequencing. FEMS Microbiol. Lett. 71:231-234.

129. Rouff, K. L 1990, Recent taxonomic changes in the genus Enterococcus. Eur. J. Clin. Microbiol. Infect. Dis. 9:75-79.

130. Ruoff, K. L, L. D. L. Maza, M. J. Murtagh, J. D. Spargo, and M. J. Ferraro. 1990. Species identities of enterococci isolated from clinical specimens. J. Clin. Microbiol. 28:435-437.

131. Schaberg, D. R., and M. J. Zervos. 1986. Intergeneric and interspecies gene exchange in gram-positive cocci. Antimicrob. Agents Chemother. 30:817-822.

132. Schlelfer, K. H., and R. Klipper-Bâiz. 1984. Transfer of Streptococcus faecalis and Streptococcus faecium to the genus Enterococcus nom. rev. as Enterococcus faecalis comb. nov. and Enterococcus faecium comb. nov. Int. J. Syst. Bacterid. 34:31-34.

133. Schlelfer, K. H., and R. KlIpper-Bâlz. 1987. Molecular and chemotaxonomic approaches to the classification of streptococci, enterococci and lactococci: a review. Syst. Appl. Microbiol. 10:1-19. 169

134. Schmidhuber, S., W. Ludwig, and K. H. Schleifer. 1988. Construction of a DNA probe for the specific identification of Streptococcus oralis. J. Clin. IVIicrobiol. 26:1042-1044.

135. Schutz, F. 1991. IVIonitoring the contamination rate of pigs on the killing line. Fleischwirtschaft 71:57-60.

136. Sharpe, M. E., and B. G. Fewins. 1960. Serological typing of strains of Streptococcus faecium and unclassified group D streptococci isolated from canned hams and pig intestines. J. Gen. Microbiol. 23:621-630.

137. Shattock, P. M. F. 1955. The identification and classification of Steptococcus faecalis and some associated streptococci. Ann. Inst. Pasteur, Lille 7:95-100.

138. Sherman, J. M. 1937. The streptococci. Bacterid. Rev. 1:3-97.

139. Sherman, J. M., and H. U. Wing. 1937. Streptococcus durans n. sp. J. Dairy Sci. 20:165-167.

140. Shiaes, D. M., and B. Binczewski. 1990. Enterococcal resistance to vancomycin and related glycopeptide antibiotics. Eur. J. Clin. Microbiol. Infect. Dis. 9:106-110.

141. Slanetz, L. W., and C. H. Bartley. 1957. Numbers of enterococci in water, sewage and feces determined by the membrane filter technique with an improved medium. J. Bacteriol. 74:591 -595.

142. Slifkin, Ul., and G. M. Gil. 1983. Rapid biochemical tests for the identification of groups A, 8, C, F, and G streptococci from throat cultures. J. Clin. Microbiol. 18:29-32.

143. Snijers, J. M. A., M. H. W. Janssen, G. E. Gerats, and G. P. Cortiaensen. 1984. A comparative study of sampling techniques for monitoring carcass contamination. Int. J. Food Microbiol. 1:229-236.

144. Taylor, R. F., M. Ikawa, and W. Chesbro. 1971. Carotenoids in yellow-pigmented enterococci. J. Bacteriol. 105:676-678.

145. Thian, T. S., and P. A. Hartman. 1981. Gentamicin-thallous- carbonate medium for isolation of fecal streptococci from foods. Appl. Environ. Microbiol. 41:724-728. 170

146. Tompkin, R. B. 1990. The use of HACCP in the production of meat and poultry products, J. Food Prot. 53:795-803.

147. Tritz, D. M., P. C. Iwen, and G. L. Woods. 1990. Evaluation of MicroScan for identification of Enterococcus species. J. Clin. Microbiol. 28:1477-1478.

148. Vaughan, D. H., W. S. Riggsby, and J. O. Mundt. 1979. Deoxyribonucleic acid relatedness of strains of yellow-pigmented group D streptococci. Int. J. Syst. Bacteriol. 29:204-212.

149. Visconti, A., C. Sielewicz, and L. Visconti. 1989. Rapid and simplified methods for the identification of gram-positive bacteria: a brief review, pp. 800-808. Ink. Balows, R. C. Tilton, and A. Turano (eds.), Rapid methods and automation in microbiology and immunology. Brixia Academic Press, Brescia, Italy.

150. Welch, D. F. 1991. Applications of cellular fatty acid analysis. Clin. Microbiol. Rev. 4:422-438.

151. Whittenbury, R. 1965. The differentiation of Streptococcus faecalis and S. faecium. J. Gen. Microbiol. 38:279-287.

152. Williams, A. M., J. A. E. Farrow, and M. D. Collins. 1989. Reverse transcriptase sequencing of 16S ribosomal RNA from Streptococcus cecorum. Lett. Appl. Microbiol. 8:185-189.

153. Williams, A. M., U. M. Rodrigues, and M. D. Collins. 1991. Intergeneric relationships of enterococci as determined by reverse transcriptase sequencing of small-subunit rRNA. Res. Microbiol. 142:67-74.

154. Woese, C. R. 1987, Bacterial evolution. Microbiol. Rev. 51:221-271. 171

AKNOWLEDGMENTS

First, I must thank my major professor, Paul A. Hartman. He allowed me to find my own way, while never failing to support me. I would also like to thank my committee members, Eisa Murano, Bonita Glatz, Lynn Knipe, and Peter Pattee, for all their help and support. I thank Chris Guyers, our lab technician, for getting up at 5 AM to swab hog carcasses so I didn't have to do it.

Special thanks to my friends, both grad students and others, who stuck with me even when I disappeared from the social circle for long periods of time (like at prelims and thesis writing time): and to my family, who sounded excited when I talked about my latest experiment, even though they had no idea what I was talking about!

To my husband Kevin, because no matter what happens during my professional career; I received something even more valuable than my Ph.D. during my graduate school career, my wedding ring. 172

APPENDIX A: MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM PORK CARCASSES SAMPLED IN THREE PLANTS 173

TABLE Al : Pork carcass mean counts (of duplicate plates) from plant A RUN& LOIN STAGE MEDIUM (LEAN) HAM SPECIES IDENTIFIED RUN1 STAGE 1 fGTC 3.9X10^ 1.4X10 E. solitarius KF <50 <50 No enterococci

STAGE 2 fGTC 7.5X10^ 1.4X10 No enterococci KF <50 <50 No enterococci

STAGE 3 fGTC 3.0X10^ 1.4X10 E faecalis KF 7.9X10'' 4.6X10 E. faecalis

STAGE 4 fGTC 1.3X10^ E. faecalis, E. malodoratus KF 1.5X10^ E. faecalis

RUN 2 STAGE 1 fGTC 3.1 X10^ 7.3X10 E. faecalis KF 9.0X10^ 9.0X10 E. faecalis

STAGE 2 fGTC <50 <50 No enterococci KF <50 <50 No enterococci

STAGE 3 fGTC 6.6X10^ 2.3X10 Unidentified enterococci KF 1.4X10^ 3.4X10 E. faecalis

STAGE 4 fGTC 3.0X10^ E. faecium, durans, pseudoavium KF 1.9X10^ E. faecalis, E. pseudoavium

RUN 3 STAGE 1 fGTC 2.7X10^ 1.5X10 E. faecalis KF 1.8X10^ 7.5X10 E. faecalis

STAGE 2 fGTC <50 6.3X10 E. faecalis KF <50 6.3X10 E. faecalis

STAGE 3 fGTC <50 1.0X10 E. faecalis KF <50 4.2X10 E. faecalis

STAGE 4 fGTC 1.1 XIO^ E. faecalis, faecium, malodoratus KF 6.2X10^ E. faecalis 174

TABLE A2: Pork carcass mean counts (of duplicate plates) from plant B RUN& LOIN STAGE MEDIUM (LEAN) HAM SPECIES IDENTIFIED RUN1 STAGE 1 fGTC 7.9X10^ <50 No enterococci KF <50 <50 No enterococci

STAGE 2 fGTC <50 1.3X10^ No enterococci KF <50 <50 No enterococci O C 5 - U O X T M C O STAGE 3 fGTC 5.5X10^ M E faecalis, E. faecium O C C X KF <50 E. faecalis

STAGE 4 fGTC 2.5X10^ E. faecalis, faecium, casseliflavus KF 1.7X10'^ E. faecalis, E.pseudoavium '

RUN 2 - O X T

STAGE 1 fGTC O O <50 No enterococci X C C KF <50 No enterococci

STAGE 2 fGTC 1.7X10^ 1.4X10^ E. faecalis, E. faecium KF 7.9X10^ 7.9X10^ E. faecalis, E. faecium

STAGE 3 fGTC 7.9 Xio] 6.3 Xio] E. faecalis, E. faecium KF 7.9X10^ 6.3X10^ E. faecalis

STAGE 4 fGTC 9.2 Xio] E. faecalis, E. malodoratus KF 6.3X10^ E. faecalis — 1

RUNS O O o C C X — T O STAGE 1 fGTC 1.1 xicf O E. faecalis o X KF <50 C C No enterococci

STAGE 2 fGTC <50 6.3 xio] No enterococci KF <50 6.3X10'' No enterococci

STAGE 3 fGTC 1.4X10^ <50 E. faecalis, E. faecium KF <50 <50 No enterococci

STAGE 4 fGTC 7.1 Xio] E. faecalis KF 5.6X10^ 175

Table A3: Pork carcass mean counts (of duplicate plates) from plant C RUN & LOIN STAGE MEDUIM (LEAN) HAM SPECIES IDENTIFIED RUN 1 STAGE 1 fGTC 2.7X103 6.8X10 3 No enterococci KF 3.6X1O2 8.7X10 2 No enterococci M O O C — O C C X 7 M

fGTC O O STAGE 2 C No enterococci — 1.6X1O2 O C C X KF T 6.3X10 2 No enterococci

STAGE 3 fGTC <50 2.4X10 2 No enterococci KF 6.3X102 1.2X10 2 E. faecalis

STAGE 4 fGTC 2.6X102 E. faecalis KF 7JX10l E. faecalis

RUN 2 STAGE 1 fGTC 1.3X104 1.1 xio4 E. faecalis KF 1.2X10^ 6.0X10 2 E. faecalis

STAGE 2 fGTC 2.9X102 7.2X10 2 No enterococci KF 7.9X10"' 1.1 X102 No enterococci

STAGE 3 fGTC <50 5.1 X102 E. faecalis KF <50 2.7X10 2 E. faecalis

STAGE 4 fGTC 5.5X102 E. faecium, faecalis, solitàrius KF 4.2X102 E. faecalis

RUNS STAGE 1 fGTC 9.3X102 1.2X10 3 E. faecalis KF 1.0X102 <50 E. faecalis

STAGE 2 fGTC 8.4X102 1.8X1O2 No enterococci KF 1.0X102 <50 No enterococci

STAGE 3 fGTC <50 <50 E. faecalis KF <50 <50 E. faecium

STAGE 4 fGTC 1.1 X102 E. faecalis KF 6.7X101 E. pseudoavium 176

APPENDIX B: MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM TWO PORK PROCESSING PLANTS 177

TABLE B1: Mean counts (of duplicate plates) from pork processing plant A

KF fGTC TOTAL STREPTOCOCCAL ENTEROCOCCAL SPECIES SAMPLE COUNT COUNT COUNT IDENTIFIED

FRESH PORK ND <50 2.0X10'=^ 100% E. faecalis SAUSAGE 7.5X10; <50 1.0X10^ No enterococci 1.0X10® <50 1.0X10^ No enterococci 1.2X10® <50 1.0X10^ No enterococci

SPOILED PORK 2.5X10® 1.7X10® 2.4X10® 100% E. faecium SAUSAGE 2.5 XIO® 2.2X10® 2.0X10® 100% E. faecium 1.2X10® 3.3.x 10® 1.1X10® 67% E. faedum 33% E. durans 4.0X10^ 1.4X10® 2.0X10^ 67% E.faecium 33% E. faecalis 2.2X10® <50 3.6X10® No enterococci 7.4X10^ <50 6.1 X10^ No enterococci 2.8X10; 3.1 X104 4.3X10® 100% E. faecium 3.0x10" I.OXIO'^ 6.4X10® 50% E. faecium 50% £. facealis 3.2X10^ 4.5X10® 8.8X10® 80% E faecium 20% E, faecalis 2.8X10^ 1.7X10® 1.4X10® 83% E. faecium 17% E. faecalis 3.9X10® 3.5 XIO 5 2.0X10® 100% E faecium 2.0X10' <50 9.8X10® 50% E faecium 50% E faecalis 178

TABLE B2: Mean counts (of duplicate plates) from processing plant B

KF fGTC TOTAL STREPTOCOCCAL ENTEROCOCCAL SPECIES SAMPLE COUNT COUNT COUNT IDENTIFIED

FRESH PORK 1.5 X 3.5 xio; 2.0 X 10 q 100% E faecalis SAUSAGE 2.0 X 2.0 X 10' 2.0X10 6W0 E. faecalis - 33% E hirae 1.1 X 2.5 xio; 1.2X10p m%E. faecalis 1.2X 2.5X10' 9.0 X 10 4 100% E faecalis 5.9 X <50 4.7 X 10 . No enterococcl 6.7 X <50 3.6 X 10 . No enterococcl 9.5 X 8.5X10' 6.0 X 10 50% E faecalis 50% E malodoratus 5.0 X 4.4X10' 1.7X10 100% E faecalis 5 FRESH PORK 4.6 X 2.1 X 10) 5.8 X 10 c No enterococcl SAUS. (LINK) 6.7 X 3.8X10' 6.5 X 10 100% E malodoratus p EXPIRED PORK 1.1 X 7.0x10; 1.0 X 10 g No enterococcl SAUSAGE 1.6 X 9.0X10; 1.0 X 10 p No enterococcl 2.1 X 1.0X10' 1.2X10 g m%E. faecalis 2.7 X <50 , 1.1 X10 g No enterococcl 2.7 X 1.0X10' 1.7 XIO g m%E. faecalis 1.1 X <50 1.1 XIO g No enterococcl 2.2 X <50 . 1.5X 10 y No enterococcl 1.1 X 8.1 X 10: 7.5 X 10 y No enterococcl 1.0X 5.2X10^ 8.8 X 10 y No enterococcl 9.8 X 3.9X10: 6.2 X 10 y No enterococcl 1.0X 3.0X10; 5.5 X 10 y No enterococcl 1.2X 5.7X10' 7.4 X 10 g No enterococcl 1.5X <50 6.8 X 10 y No enterococcl 3.0 X <50 9.6 X 10 g No enterococcl 3.3 X Sa <50 2.1X10 No enterococcl Q EXPIRED PORK 4.6 X 1.0X10 1.4 X 10 g No enterococcl SAUS. (LINK) 1.5X S: 3.0X10^ 4.2X10 50% E. faecalis 50% E pseudoavium

PORK TRIM 1.6 X 0® 4.2X1of 1.4X10^ 100% E faecalis (42%) 4.8 X 0^ 4.8 XIO* 1.4X10 No enterococcl

PORK TRIM 1.1 X 07 5.3X10^ 2.6X10® 100% E faecalis (72%) 6.5 X 0^ 1.6X10° 1.9X10 33% E faecalis 33% E faecium 33% E malodoratus 179

APPENDIX C: COUNTS AND ENTEROCOCCAL IDENTIFICATIONS FROM RETAIL SAMPLES 180

Table Cl : Mean counts (of duplicate plates) from retail samples Total fGTC KF Retail Sample Count Count Count Species ID Pork Hormel chopped ham ND 1.5X10 5 <50 S. bovis Fresh pork sausage ND 8.9X10^ 9.6X10^ S. salivarius Hillshire farms fresh bratworst ND 7.0X10% 2.0X10^ No enterococci Mild Italian pork sausage ND 1.4X104 4.0X10^ E faecium Boneless pork loin ND 6.0X104 3.1X10^ E. faecalis Iowa pork chop ND 1.0X102 <50 No enterococci Fareway bratworst ND 3.5X10^ 5.6X10^ No enterococci JD hot pork sausage ND 2.0X10% <50 E faecalis Purnell's country pork sausage ND 2.0X10% i.oxio; No enterococci Farmland pork sausage ND 1.7X10 3 4.0X10% No enterococci JD sage pork sausage 7.5X10^ 1.0X102 <50 No enterococci JD regular pork sausage 1.0X10^ 1.0X10% <50 No enterococci Oldham's whole hog pork sausage 5.6X10^ 5.6X10% <50 E faecalis JD turkey and pork sausage 1.2X10^ 1.0X10% <50 No enterococci

Beef Signature ground beef (93% lean) 1.4X10^ 1.4X10^ 3.1X10^ E.durans/S. bovis Ground Beef (90% lean) 7.7X10; 3.1X10] 2.9X10 3 E.durans/malocloratus Ground Beef (85% lean) 2.0X10^ 9.3xio; 2.3X10^ E malodoratus Ground Beef (70% lean) 3.9X10j 4.3X10^ 2.3X10^ E.durans/faecalis Banquet Beef pot pie i.oxio; 2.5X10% <50 No enterococci Swanson Beef pot pie 3.5X10^ 1.0X10% 1.0X10^ E faecalis Armour beef & bacon pattie <50 <50 <50 No enterococci Poultry Chicken breast tenders 2.1X10^ <50 <50 No enterococci Whole frier (chicken) 2.0X10^ 3.0X10^ <50 No enterococci Country pride ground chicken s.zxio'^ 5.9X10^ 1.0X10^ E faecalis Banquet chicken pot pie 1.4X10^ <50 <50 No enterococci Swanson chicken pot pie 7.5X10^ <50 <50 No enterococci Banquet turkey pot pie 6.0X10^ 1.0X10% <50 No enterococci Swanson turkey pot pie 1.0X10^ <50 <50 No enterococci Longacre ground turkey 1.9X10"^ 8.6X10^ 6.0X10^ E.faecalls/faecium Armour ground turkey 1.4X104 1.3X10^ <50 No enterococci Louis Rich ground turkey 2.7X104 7.9X10^ <50 E faecalis

Cheese Hy-Vee mild Cheddar cheese 7.4X10'^ 2.3X10"^ <50 No enterococci Kraft Havarti cheese 3.0X10^ 4.4X10^ 1.4X10^ No enterococci Kraft Swiss cheese 2.2X10® 3.4X101 2.8X10 J E faecium Kraft natural gouda cheese ND 3.0X10^ 2.5X10^ E durans 181

APPENDIX D: MEAN COUNTS AND ENTEROCOCCAL IDENTIFICATIONS ' ' FROM WATER SAMPLES 182

Table D1: Mean counts from water samples mENTERO mENDO enterococci SOURCE coliforms/100ml /100ml SPECIES IDENTIFIED LAKE SAMPLES Lake Laverne 56 34 E. faecium JACUZZI SAMPLES ATFC 0 0 Gateway 0 0 Starllte 0 0 RIVER SAMPLES College Creek-1 TNTC 184 E. faecium E. faecalis College Creek-2 4520 234 E. faecium DM River -1 200 820 E. faecium E. faecalis DM Rlver-5 10 20 E. faecium E. faecalis E. casseliflavus DM River-6 1320 520 E. faecium E. faecalis K Creek 720 240 E. durans E. hirae Racoon River-10 470 720 E. faecium E. faecalis Skunk River -1 700 740 E. faecium E. mundtii E. casseliflavus Skunk River-2 2380 1200 E. gallinarum Squaw Creek-1 237 1000 E. faecium E. faecalis E. casseliflavus Squaw Creek-2 2150 1270 E. malodoratus Sugar Creek 366 700 E. faedum E mundtii E. durans Worle Creek-1 810 780 E faecalis E. hirae Worle Creek-2 1370 1870 E. hirae E. gallinarum WELL SAMPLES Mills Co.-Residence 0 0 Tama Co.-Residence 0 0 Red Rock R-87-4 0 0 Red Rock 5-RB 0 0 Red Rock 23-R 0 0 Red Rock 5-RA 0 0 Red Rock 30-0 0 0 Red Rock 29-0 0 0 Corn Field Wells -6B 6ft. 645 3 E. faecalis -6B 8ft. 0 2 E. faecalis -68 10ft. 20 5 E. faedum E. malodoratus -6B12ft. 0 11 Staphylococcus -8B 6ft. 0 38 Staphylococcus -8B 8ft. 0 10 -88 10ft. 0 1 -88 12ft. 0 1 E. casseliflavus 183

APPENDIX E; SOURCE AND IDENTIFICATIONS OF CLINICAL ISOLATES FROM MERCY HOSPITAL MEDICAL CENTER DES MOINES, IOWA 184

TABLE E1: Source and identification of clinical isolates

Isolate Species Isolate Species Source # Source Identified # Identified

1 Urine E. faecalis 26 Genital E. faecalis 2 Genital E. faecalis 27 Genital E. faecalis 3 Urine E. faecalis 28 Urine E. faecalis 4 Urine E. faecalis 29 Urine E. faecalis 5 Urine E. faecium 30 Urine E. faecalis 6 Urine E. faecalis 31 Urine E. faecalis 7 Urine E. faecalis 32 Unknown E. faecalis 8 Urine E. faecalis 33 Urine E. faecalis 9 Urine E. faecalis 34 Unknown E. malodoratus 10 Urine E. faecalis 35 Urine E. faecalis 11 Urine E. faecium 36 Vaginal E. faecalis 12 Urine E. faecalis 37 Urine E. faecalis 13 Urine E. faecium 38 Urine E. faecalis 14 Bile E. faecium 39 Urine E. faecalis 15 Urine E. faecalis 40 Urine E. faecalis 16 Urine E. faecalis 41 Urine E. faecalis 17 Genital E. faecalis 42 Unknown E. faecalis 18 Urine E. faecalis 43 Unknown E. faecium 19 Urine E. faecalis 44 Blood E. faecalis 20 Urine E. faecalis 45 Urine E. faecalis 21 Urine E. faecalis 46 Unknown E. faecalis 22 Urine E. faecalis 47 Urine E. faecalis 23 Urine E. faecalis 48 Unknown E. faecalis 24 Urine E. faecalis 49 Urine E. faecalis 25 Urine E. faecalis 50 Urine E. faecalis