Enumeration and comparative distribution of coliphages in foods

MISSING IMAGE

Material Information

Title:
Enumeration and comparative distribution of coliphages in foods
Uncontrolled:
Coliphages in foods
Physical Description:
ix, 178 leaves : ; 28 cm.
Language:
English
Creator:
Kennedy, James E., 1947-
Publication Date:

Subjects

Subjects / Keywords:
Bacteriophages   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1985.
Bibliography:
Includes bibliographical references (leaves 161-177).
Statement of Responsibility:
by James E. Kennedy, Jr.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000555372
notis - ACY0262
oclc - 13568895
System ID:
AA00003402:00001

Full Text











ENUMERATION AND COMPARATIVE DISTRIBUTION
OF COLIPHAGES IN FOODS
















BY

JAMES E. KENNEDY, JR.


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1985


































TO LAURA AND DYLAN
WITH LOVE AND GRATITUDE














ACKNOWLEDGMENTS


I wish to express my sincere appreciation to Drs. James L.

Oblinger and Cheng-i Wei for their guidance, assistance, encouragement

and friendship as my chairperson and co-chairperson, respectively,

during this research. To Drs. J. L. Oblinger and C. I. Wei I express

my special thanks and respect for providing leadership and support

under unique and often difficult circumstances during this project.

I also extend my gratitude to the other members of my committee,

Drs. J. A. Koburger, G. Bitton, S. R. Farrah and L. B. Bailey, for

their cooperation, advice and support during the course of my graduate

program.

Appreciation is also expressed to Dr. R. C. Littell for his advice

concerning the statistical analyses. Special thanks are also extended

to Mrs. Janet Eldred for her excellent work in typing this dissertation

and to Mr. Walter Jones for his assistance with the graphic components

of this dissertation. I also want to extend my sincere gratitude to

the Food Science and Human Nutrition Department for providing me the

opportunity to conduct this research and for the academic and

professional experience which I have received during my career as both

a graduate student and staff member in the department.

Finally, my sincere gratitude is extended to my wife, Laura, and

my son, Dylan, for their love, patience and inspiration, and, with whom

the difficult times were overcome; and to my parents, Mr. and Mrs.


iii








James E. Kennedy, Sr., for their continuing faith and love throughout

all of my endeavors.














TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS ................................................. iii

ABSTRACT ....................................................... viii

INTRODUCTION ............................ ....................... 1

LITERATURE REVIEW ............................................... 7

Biological and Ecological Characteristics of
Coliphages ............................................... 7

Biological Properties and Classification .............. 7
Occurence and Distribution in the Environment ......... 11
Host Range and Specificity ............................ 16
Effects of Temperatures on Coliphage Ecology .......... 23

Coliphages as Indicators of Fecal Contamination ............ 25

Overview of Classic Bacterial Indicators .............. 25
Coliphages as Indicators of Enteric Bacteria in
Water and Wastewater ................................ 26
Coliphages as Indicators of Enteric Viruses in
Water and Wastewater ................................ 29
Coliphages as Indicator Organisms in Foods ............ 33

Comparative Physicochemical Stability of Coliphages ........ 34

General Overview ...................................... 34
Effects of High Temperature ........................... 35
Effects of Low Temperatures and Other
Physicochemical Factors ............................. 41
Resistance to Disinfection ............................ 45

Basic Methodology for Coliphage Recovery from Foods ........ 46

General Overview ...................................... 46
Techniques and Principles of Virus Adsorption-
Elution ............................................. 47
Physical Methods of Suspension, Clarification
and Concentration ................................... 52
Assay of Coliphages ................................... 54








Page

STUDY 1
METHODOLOGY FOR ENUMERATION OF COLIPHAGES IN FOODS .............. 55

Introduction ............................................... 55
Materials and Methods ...................................... 57

Preparation of T2 and MS2 Coliphage Stocks ............ 57
Samples and Sample Preparation ....................... 58

General .......................................... 58
Preparation of inoculated samples ................ 59

Preparation of Sample Suspension ...................... 61
Clarification of Sample Suspension .................... 62
Suspending Media and/or Eluents ....................... 64
Coliphage Assays ..................................... 67

Host bacteria cultures ........................... 67
Coliphage assay medium ........................... 68
Coliphage assay procedures ....................... 69

Statistical Analyses .................................. 71

Results .................................................... 71

Sample Clarification Techniques ....................... 71
Sample Suspension Techniques .......................... 74
Eluent Composition and pH ............................. 81

Basal eluents .................................... 81
Chaotropic agents ............................... 83
Sample suspension pH ............................. 87
Recovery of T2 and MS2 from various foods ........ 88

Discussion ................................................. 92
Summary and Conclusions ................................... 100

STUDY 2
CHARACTERIZATION AND COMPARATIVE DISTRIBUTION OF
COLIPHAGES IN VARIOUS FOODS ..................................... 103

Introduction .............................................. 103
Materials and Methods ...................................... 106

Samples ............................................... 106
Bacteriological Analyses ............................. 107
Coliphage Analyses .................................... 108

Sample preparation ............................... 108
Host bacteria .................................... 108









Page

Coliphage assay .................................. 108
Physiological characterization of coliphages ..... 109

Correlation Determinations ............................ 110

Results .................................................... 111

Recovery of Coliphages and Bacteria from Foods ........ 111
Comparison of Host Strains ............................ 118
Relationships between Coliphage and Bacterial
Indicators ......................................... 120
Temperature of Infectivity ............................ 131

Discussion ................................................. 134
Summary and Conclusions .................................... 143

SUMMARY AND CONCLUSIONS ......................................... 148

APPENDIX ........................................................ 155

LITERATURE CITED ................................................ 161

BIOGRAPHICAL SKETCH ............................................ 178


vii















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


ENUMERATION AND COMPARATIVE DISTRIBUTION
OF COLIPHAGES IN FOODS

By

James E. Kennedy, Jr.

December, 1985

Chairperson: James L. Oblinger
Co-Chairperson: Cheng-i Wei
Major Department: Food Science and Human Nutrition

Studies were conducted to develop expedient, rapid and efficient

methods for enumeration of coliphages in foods and to investigate the

comparative distribution and ecological specificity of coliphages in

various foods.

In the first study, the effects of eluent composition, pH and

chaotropic agents on the recovery of T2, MS2 and indigenous coliphages

from various foods were investigated. Methods of sample liquifaction

and sample-suspension clarification were evaluated for recovery of

coliphages and applicability to various foods. Centrifugation and

polypropylene mesh filtration were more effective than glass wool

filtration with regard to speed, simplicity and coliphage recovery.

Blending, stomaching and shaking methods were comparable for

liquifaction of samples and subsequent release of coliphages from

various foods. Complex basal eluents, i.e., EC medium or 1% casein,

were generally more effective than less complex eluents, i.e.,


viii








phosphate buffer, for recovery of coliphages from foods. Chaotropic

agents, i.e., sodium trichloroacetate, urea, Tween 80, Triton X-100,

sodium nitrate and sodium chloride, generally did not enhance recovery

of coliphages. The recovery of coliphages was generally not affected

by the pH of sample-suspensions over a range of pH 6 to 9. Using EC

medium eluent and blending, recovery rates ranged from 48 to 81% for T2

and from 58 to 100% for MS2, depending upon the food type.

In the second study, a total of 120 food samples (12 products)

were analyzed for coliphages using methods developed in the first

study. E. coli, fecal coliforms, coliforms, aerobic plate counts and

Salmonella determinations were also conducted. In addition, coliphages

isolated from various foods were characterized with regard to tempera-

ture of infectivity. Coliphages were detected (1 10 PFU/100g) in 56%

of samples and 11 of 12 products. At a level of at least 30 organisms

per 100g, coliphages, E. coi, fecal coliforms and coliforms were

recovered in 43, 43, 68 and 81% of samples with overall mean recoveries

of 13, 19, 93 and 4300 organisms per lOOg, respectively. Highly

significant correlations (p < 0.00001) between coliphages, E. coli,

fecal coliforms and coliforms were found overall. The large percentage

of high temperature type coliphages found in most products suggested a

fecal and/or animal origin for many coliphages recovered from foods.














INTRODUCTION


The economic, legal and public health significance of micro-

biological quality assurance with regard to monitoring the sanitary

quality of foods at various stages of procurement and processing is

well documented (Hauschild and Bryan, 1980; Levy and McIntire, 1974;

U.S.D.A., 1985). Although a number of procedures have been developed

and become well established for assessing the microbial quality and

safety of foods and water, more expedient and reliable methods continue

to be sought. The use of generally non-pathogenic bacteria of fecal

origin, i.e., Escherischia coli and/or coliform bacteria, were

introduced as indicators of enteric pathogens in water in the late

nineteenth century (Schardinger, 1892; Smith, 1895). On the basis of

their successful and established use for water quality testing,

coliform bacteria and other indicator organisms such as fecal

streptococci have been widely used in monitoring the sanitary quality

and/or processing adequacy of foods (Buttiaux and Mossel, 1961; Jay,

1978; Mossel, 1982; Thatcher and Clark, 1968). Although analyses for

indicator organisms may offer considerable advantages with regard to

time, simplicity and cost over those for specific pathogens in food and

water on a routine basis, they may not preclude testing for pathogens

in many situations. Additionally, the standard method (U.S. Food and

Drug Administration, 1978) for determination of specific fecal

indicators such as E. coli requires up to 10 days for completion.





2


The need for development of more inexpensive and rapid procedures for

enumeration of specific fecal indicators, e.g., E. coli or fecal

coliforms, in foods has recently been addressed by various workers

(Feng and Hartman, 1982; Holbrook et al., 1980; Petzel and Hartman,

1985; Powers and Latt, 1979; Silliker, 1982).

Another group of organisms which have received increased attention

in recent years as rapid and simple indicators of fecal pollution and

enteric pathogens in water and wastewater are coliphages, i.e.,

bacteriophages of E. coli (Buras and Kott, 1966; Fannin et al., 1977;

Hilton and Stotzky, 1973; Kenard and Valentine, 1974; Kott et al.,

1974; Scarpino, 1975, 1978; Simkova and Cervenka, 1981; Stetler, 1984;

Vaughn and Metcalf, 1975; Wentsel et al., 1982; Zaiss, 1981).

Coliphage analyses can be completed in 6 to 24 hours and offer a rapid

and economical alternative to conventional analyses of water or waste-

water for E. coli or enteric pathogens. Additionally, several studies

suggest that coliphages resemble enteric viruses with regard to

environmental persistence and are generally superior to coliform

bacteria as indicators of viral inactivation or removal following water

and wastewater treatment (Burns and Sproul, 1967; Chaudhuri and

Englebrecht, 1970; Durham and Wolf, 1973; Glass and O'Brien, 1980;

Grabow et al., 1978; Kott et al., 1974; Lothrop and Sproul,

1969; Stetler, 1984). Coliphage models have also been used to evaluate

thermal destruction of viruses in processed shellfish (DiGirolamo

et al., 1972) and in composting systems (Burge et al., 1981).

Coliphages generally occur in environments with which their host

bacteria, E. coli, are commonly associated. As with E. coli,

coliphages can be recovered in relatively high concentrations from the








feces of numerous warm blooded animals including man (Dhillon et al.,

1976; Osawa et al., 1981; Smith and Crabb, 1961). Coliphages also can

be found in raw sewage, sewage effluents, natural waters, aquatic sedi-

ments and drinking water in association with fecal pollution, e.g.,

coliform bacteria or enteric pathogens (Buras and Kott, 1966; Giraldi

and Donati, 1976; Kenard and Valentine, 1974; Kott et al., 1974;

Scarpino, 1978; Simkova and Cervenka, 1981; Stetler, 1984; Wentsel

et al., 1982; Zaiss, 1981).

Highly significant correlations (p < 0.001) reported between

coliphage and fecal coliforms (Kenard and Valentine, 1974; Wentsel

et al., 1982) as well as between coliphages and enteric viruses

(Stetler, 1984) in various water systems suggest the validity of

coliphages as indicators of fecal pollution and/or enteric viruses in

these waters. However, the acceptability of coliphage indicators has

been questioned by some investigators with regard to host specificity

(Hilton and Stotzky, 1973; Seeley and Primrose, 1982; Vaughn and

Metcalf, 1975) and the possible non-fecal origin of some coliphages or

their alternate host bacteria (Cliver and Salo, 1978; Parry et al.,

1981; Seeley and Primrose, 1980, 1982; Vaughn and Metcalf, 1975).

However, many of these problems can be overcome by use of appropriate

host bacteria (Havelaar and Hogeboom, 1983, 1984; Ignazzitto et al.,

1980; Primrose et al., 1982; Seeley and Primrose, 1982) and recent

studies suggest that the relative incidence of coliphages in water and

wastewater having alternate bacterial hosts of non-fecal origin is low

(Dhillon and Dhillon, 1972; Havelaar and Hogeboom, 1983; Primrose

et al., 1982; Zaiss, 1981). Although male or F plasmid-specific

coliphages can infect E. coli as well as other Enterobacteriaceae








carrying an F plasmid, the F-pili required for adsorption of these

coliphages is not produced by bacteria below 300C (Birge, 1981;

Primrose et al., 1982; Seeley and Primrose, 1982). Thus, replication

of these coliphages is generally limited to the gut of warm blooded

animals.

Whereas coliphage distribution and their use as indicator

organisms in wastewater and water have been extensively investigated,

very little information is available concerning the distribution of

coliphages or their relationship to sanitary quality in various foods.

The recovery of bacteriophages of E. coli, Salmonella typhi and

S. paratyphi from raw milk was reported as early as 1937 (Lipska,

1937). More recently, coliphages have been recovered from fresh

oysters (Kott and Gloyna, 1965; Vaughn and Metcalf, 1975), fresh meats

and cooked luncheon meats (Kennedy et al., 1984). Kott and Gloyna

(1965) reported a high correlation between levels of coliphage and

coliforms in oysters obtained from water stations at various distances

from a sewage outfall. Vaughn and Metcalf (1975) found that coliphages

were widely distributed in oysters and/or shellfish-growing waters but

observed no relationship between the occurrence of coliphages and

enteric viruses in oysters, estuarine water or sediment samples.

Coliphages were recovered from 100% of fresh chicken and pork sausage

samples tested as well as 33% of delicatessen meats by Kennedy et al.

(1984) with some positive correlation observed between levels of

coliphage and fecal coliforms.

Methodology for recovery of coliphages and viruses from the

environment has been extensively developed and evaluated in recent

years (Seeley and Primrose, 1982). Likewise, numerous procedures for








detection of viruses in various foods have been developed (Cliver

et al., 1983a, b). However, methodology for enumeration of coliphages

and other bacteriophages in foods has not been investigated for optimal

recovery efficiency or ease of application to various food types

(Kennedy et al., 1984; Kott and Gloyna, 1965; Whitman and Marshall,

1971a). Development of methods for enumeration of coliphages in foods

which are expedient, efficient and broadly applicable to various foods

is needed to study the distribution and ecology of coliphages in foods

as well as to provide rapid and simple procedures suitable for routine

use. The demonstrated value of coliphages as indicators of fecal

contamination in water and wastewater (Kenard and Valentine, 1974;

Stetler, 1984; Wentsel et al., 1982) as well as their frequent recovery

and association with coliform bacteria in a limited number of foods

examined (Kennedy et al., 1984; Kott and Gloyna, 1965) warrants further

investigation of coliphage occurrence and ecology in foods. Therefore,

the objectives of these investigations were as follows:

1. To develop and evaluate methodology for expedient and

optimal recovery of coliphages in various types of

foods. The effects of elution media composition and pH

of elution media as well as method of sample liquifaction

and sample-suspension clarification were investigated for

recovery of T2, MS2 and indigenous coliphages from

representative foods.

2. To determine the comparative distribution of coliphages

and fecal indicator bacteria as well as total bacteria

and Salmonella in a wide variety of foods using optimal

methodology developed. Comparative recovery of coliphages




6


using three different E. coli host bacteria was deter-

mined and representative coliphages were isolated and

physiologically characterized according to temperature of

infectivity.














LITERATURE REVIEW


The following discussion will review research concerning the

relationship between coliphages and fecal contamination, e.g., enteric

indicator bacteria and pathogens, in natural waters, wastewater and

water or wastewater treatment as well as in foods. In order to place

coliphages in perspective with regard to their sanitary significance in

water and foods, it is also necessary to address their biological

nature, their ecological characteristics and their comparative

physicochemical stability in relation to other microorganisms.

Additionally, the basic methodology and underlying principles for

recovery of bacteriophages and viruses from the environment and from

foods will be discussed in reference to development of procedures for

rapid, expedient and efficient recovery of coliphages from foods.


Biological and Ecological
Characteristics of Coliphages


Biological Properties and Classification


Coliphages, i.e., bacteriophages infecting Escherichia coli,

comprise a morphologically and genetically diverse group of viruses.

Bacteriophages are most commonly classified according to their

morphology or structure and the nature of their genome (Ackermann,

1978a, b, c; Ackermann et al., 1978; Bradley, 1967). Their genome

or nucleic acid can be either deoxyribonucleic acid (DNA) or








ribonucleic acid (RNA) and may be either single stranded or double

stranded, linear or circular and consist of one or several pieces.

The nucleic acid of bacteriophages is centrally located and is

surrounded by a protein coat or capsid which is comprised of protein

subunits or capsomeres. The capsids of coliphages can be classified

morphologically as cubic, filamentous or tailed. Cubic phages have

isometric icosohedral capsids and resemble spheres in general appear-

ance, whereas filamentous phages have helical arrays of capsomeres and

resemble filaments. Tailed phages have an elongated or isometric,

icosohedral head and a filamentous tail structure. Tailed phages and

some cubic phages adsorb to somatic or cell wall components of the

bacterial host whereas filamentous and other cubic phages adsorb

specifically to F-pili of bacteria carrying the F plasmid and are

collectively referred to as male-specific coliphages. A comparative

summary of morphological and nucleic acid properties of basic groups of

coliphages and representative enteric viruses is presented in Table 1.

The physical similarities between the cubic RNA coliphages and many

enteroviruses have been noted by various workers with regard to the

potential of these coliphages as indicators of enterovirus inactivation

and distribution in wastewater and water (Kott et al., 1974; Metcalf,

1978; Seeley and Primrose, 1982). The classification, basic proper-

ties, genetics and biology of bacteriophages have been extensively

reviewed elsewhere (Ackermann, 1978a, b, c; Ackermann et al., 1978;

Bradley, 1967; Dulbecco and Ginsberg, 1980).











































0
U


Z -o
z



















00
0
-4
0


0
2:


I


x

--4 )
-4-4







I !
I I





co -r1
CO *H
o../c./


CO .-1

OH p
U r-l
.0 CO

:3 -W
U



0U
00
'0 4H









0000 -










-. H-


II C
I 0


I-4
I I








00 1
I I

I- ,-


,0

1O 1-4
c= 4,-4


0
00 C
c cc
o0 0
*r--I 0

*.r ?0
CO '-4 0
0 H 0





a co
U




S0 0
4-' 0,C


I I
I I


01n
I
I I
I c


O
r- I
I I
I
,0



(. CO
Wc *-


0
C
.C 0



0)0
4-. .



'.-4 U)
Sr-I


*H -


CO r
0)


04-
I CO
0 =
c )


4-4 r-4
0 1



,--4
S P.



0 I.
41





0)
-M
CO




0 j
4j
E n

co 0
-4 -M
cc


U
.H
4--
CO
0
rC


0)
0)


0

CO *H
0.
U

00 0
ca -0
r-4 CO


u




4-4
*C




0 *
0 *l
C I





0) 0
e m
0

0 0

CO
U m
-3


EO
0)C



















N
*H





X
0
(-4


*rlir-1i
U

,- *I
U U





















0
I-l
0
.0
0
$-4





0





0.



0
>0


.4- 4

U 0
(U

C:6 $
S.I


g- r -- 0
























0
U

-4

0
*0

U U u
.0 .0 .0
3 3 3
u u u






0 0 0
-I t- -


-o
0
0)
C
c*-







-4
II





.-o






*a






$4
0




(1



0 ,-
o- -
v0













0O ON



c 0







0 0
II
>-o





















6 0 04
ca ao











0 0 04
0 4 *


II 0 0




^3
-4 00J *r


0 c o .o c
** a) *Hr 4-4

r 30 00*H < 0
w 2 -H .: > 00
r-q U () 1 .o (o
>- 0 > 0 *H *H
.*i (00aC -Y 41 -4 w-4
> C *H -4 *--I
1- CO 4- Ca >
U 0 :3 U (o 3 ca
*rH 00 X 0. w- 4 -
i. *H O 0) 0 0
a) 1O>U 3D >
0-







Occurrence and Distribution in the Environment


Since bacteriophages of a given bacterial species can generally be

isolated from the same environment in which their host bacteria are

prevalent, coliphages have been recovered from environments in which

E. coli resides or is present as a contaminant. Both coliphages

(Dhillon et al., 1976; Osawa et al., 1981) and E. coli (Geldreich

et al., 1962; Mushin and Ashburner, 1964; Smith and Crabb, 1961) occur

in varying densities in the intestinal contents or feces of many warm

blooded animals. In addition, coliphages and E. coli or other enteric

bacteria are generally present wherever fecal pollution occurs, e.g.,

sewage, wastewater, natural waters and aquatic sediments, and are

capable of persistence and some degree of proliferation in these

extraenteral environments depending upon suitable physical and chemical

factors (Anderson, 1957; Buttiaux and Mossel, 1961; Laliberte and

Grimes, 1982; Parry et al., 1981; Scarpino, 1978; Seeley and

Primrose, 1980, 1982; Vaughn and Metcalf, 1975).

Few studies have addressed the distribution of coliphages in the

feces of warm blooded animals in quantitative terms with no concomitant

enumeration of coliphages and E. coli in feces being reported (Dhillon

et al., 1976; Osawa et al., 1981). The reported distribution of

E. coli and coliphages in the intestinal contents of warm blooded

animals is summarized in Table 2. Based upon a relatively limited

number of studies, the incidence and density of coliphages in the feces

of most warm blooded animals is generally less than that of E. coli

(see Table 2). It should be noted that the number of coliphages

enumerated in a given sample is dependent upon the strain of E. coli



















4-0 0O O 0
r-1- '- -. i






WW1t1 I c 4W1, I Cl- 1 w


0 0 0
S-l Co Co -i Ca Ca Ca :- a
-H 00 CO M H o o Ca r-H Ca
=00=000=0







C -4O-q 0 0 0 c, rt
C 0r-l cONr-
1-4


r-L rlf r r~-: 'to v
I I I I I\ I I I
Or-r-- or--r->i-rcor-=


00 U
CO C
C =
U
0





00

3cn
CO

4-1 Go
0
0
OC )



















U
00 0
Sr-















-H
w 0



















00
EO










1-
0


0-- O -

C'N 0\ W --4 CM -4 0M -4






0a0 -o H Ca0 04 cc 0H














O0 )00) 00 0
C C U- -I ,-0 0 a-\ -4


CN r
I
I I I
I
No
OM


CM4

U)
a) -S W

--4-4 U U U U

a) U U W" -- )
- ( ^ 0 0 0



Q)U 4- 44 0 -
a4-4 4-0 >

U) a)( Co
00 > 0
H 0 0
0L u a


(U
Q) 4,-)



r- 0
0 u
c c

U
r-1
14 Ca '
CO C
*HM a'-
HC %'4-H U) W
,'- .,-( U)-- 4U) U) a) a) a
^-S4-1 w CO w wU w 0
U) 0 0 C0 0 W 4-I U U
a) a) a) U U 4l-' a) a)
u u41 w = 4-4 4-
) ) = 44.4 4-4 -H >i-/'-
4A4 4-4 -H ^'-.--' 5.
) 1-4 ca
0 0 0 3
uH 00
Q0 0 0


00
*H


Q)


u
-o




1-
c
*-H


0



-H







host used in the assay and, therefore, recoveries of coliphages

reported in different studies are often difficult to compare (Havelaar

and Hogeboom, 1983; Ignazzitto et al., 1980). The relative distribu-

tion of specific coliphages in the feces or intestinal contents of

animals was also investigated by Dhillon et al. (1976) and Osawa

et al. (1981). Osawa et al. (1981) recovered RNA coliphages from all

types of warm blooded animals with the highest incidence found in pigs

and cows whereas Dhillon et al. (1976) found no male-specific and/or

RNA coliphages in the feces of various warm blooded animals. Various

RNA coliphages displayed a habitat preference in that certain sero-

logical groups were prevalent in animal feces and others in human feces

(Osawa et al., 1981). Likewise, pX174-type and S13-type coliphages

were found only in pigs and cows, respectively, whereas Tl-type

coliphages were isolated in both cows and pigs (Dhillon et al., 1976).

Temperate coliphages have also been isolated from the feces of warm

blooded animals (Dhillon et al., 1976).

Coliphages of numerous E. coli host strains have been recovered

ubiquitously from raw sewage in numbers ranging from approximately 103

to 108 phages per 100 ml (Bell, 1976; Buras and Kott, 1966; Dhillon and

Dhillon, 1974; Dhillon et al., 1970; Havelaar and Hogeboom, 1983;

Ignazzitto et al., 1980; Kenard and Valentine, 1974; Kott et al., 1974;

Ware and Mellon, 1956). Coliphages have also been routinely recovered

from sewage after various degrees of treatment, e.g., activated sludge,

trickling filters, oxidation ponds, aerosols near sewage treatment

facilities and chlorinated final effluent (Bitton et al., 1981;

Dias and Bhat, 1965; Durham and Wolf, 1973; Fannin et al., 1977;

Glass and O'Brien, 1980; Ignazzitto et al., 1980; Kennedy et al., 1985;







Kott et al., 1974). Mean coliphage to coliform ratios of 1:358

(Ware and Mellon, 1956) and 1:100 (Buras and Kott, 1966) as well as

coliphage to fecal coliform ratios of 1:52 (Pretorius, 1962) and 1:82

(Bell, 1976) have been reported for raw sewage. However, there was

generally considerable variation in these ratios among individual

samples in most studies and E. coli hosts differed between studies.

Levels of coliphages and coliforms generally decrease during successive

stages of sewage treatment but ratios of coliphages to coliforms or

fecal coliforms increase during treatment due to the greater resistance

of coliphages to environmental stress and chlorination (Bell, 1976;

Burns and Sproul, 1967; Fannin et al., 1977; Ignazzitto et al., 1980).

No apparent net multiplication of coliphages nor reduction of E. coli

or coliform populations by coliphage has been observed during sewage

treatment despite the ability of coliphage to replicate and inhibit

growth of specific bacterial hosts under laboratory conditions (Dias

and Bhat, 1965; Pretorius, 1962; Ware and Mellon, 1956).

Coliphages have been isolated from numerous aquatic environments

receiving various degrees of fecal pollution. They have been recovered

from river water (Bell, 1976; Hilton and Stotzky, 1973; Kenard and

Valentine, 1974; Parry et al., 1981; Primrose and Day, 1977; Simkova

and Cervenka, 1981; Zaiss, 1981), river sediments (Zaiss, 1981),

streams (Kott et al., 1974; Primrose et al., 1982), estuarine water

(Kott, 1966; Vaughn and Metcalf, 1975) and estuarine sediments (Vaughn

and Metcalf, 1975). Coliphages were also isolated in treated, potable

water by Kott et al. (1974) and Wentsel et al. (1982) but not by

Stetler (1984). Kott et al. (1974) recovered 2 to 8 coliphages and

2 to 350 coliforms per 100 ml of tap water from a community during an







enteric disease outbreak in Israel. Coliphages were not found in tap

water samples free of coliforms. Wentsel et al. (1982) was also able

to detect coliphages in potable water samples at a water treatment

plant using a non-quantitative, enrichment technique.

Relatively little information is available concerning the compara-

tive distribution of coliphages and indicator organisms or enteric

pathogens in various foods despite the interest in their indicator

status in water and wastewater. As early as 1937, bacteriophages of

the "coli-typhi-paratyphi group," i.e., E. coli, Enterobacter

aerogenes, Salmonella typhi and S. paratyphi, were found in raw milk

along with E. coli and other coliform bacteria (Lipska, 1937). No

quantitative determinations of coliphages or bacteria in raw milk were

reported. More recently, coliphages of E. coli B were recovered from

oysters obtained from water stations at various distances from a sewage

outfall and from market oysters in numbers ranging from 20 to 2,000 and

from < 2 to 17 per lOOg, respectively (Kott and Gloyna, 1965). Vaughn

and Metcalf (1975) found coliphages to be widely distributed in shell-

fish growing waters as well as in oysters and sediments from those

waters although quantitative assessments of coliphages were not

reported. Coliphages of E. coli C have recently been found in 100% of

fresh chicken and pork samples and 33% of luncheon meat samples in

numbers up to 25,000, 3,500 and 540 coliphage per lOOg in chicken, pork

and cooked luncheon meats, respectively (Kennedy et al., 1984).

Reported relationships between coliphages and indicator bacteria or

enteric pathogens in foods and water will be addressed in a following

section.








Host Range and Specificity


The acceptability of coliphages as indicators of fecal pollution

or enteric pathogens has been questioned with regard to host

specificity and/or lack of appropriate E. coli host strains (Hilton

and Stotzky, 1973; Scarpino, 1978; Seeley and Primrose, 1982; Vaughn

and Metcalf, 1975). A given type of bacteriophage generally has a

relatively limited range of susceptible bacteria; i.e., most phages are

genus-specific, with some phages being species specific, or strain-

specific within a species (Adams, 1959; Ackermann et al., 1978).

However, some bacteriophages have been isolated which infect bacteria

across genus boundaries in some families, e.g., Enterobacteriaceae and

Pseudomonaceae (Adams, 1959; Ackermann et al., 1978; Dhillon and

Dhillon, 1972; Hertman, 1964; Lazarus and Gunnison, 1947; Smith and

Burrows, 1961), while at least one phage has been isolated which

apparently may attack bacteria of different families (Goze and

Lapchine, 1971). When considering the host range of bacteriophages, it

is essential to distinguish between (1) "somatic" phages, i.e., those

which adsorb to cell wall components of the host other than pili,

(2) male-specific phages, i.e., those which adsorb to F-pili of

Enterobacteriaceae carrying an F plasmid, and (3) various plasmid

specific phages, i.e., those specific for bacteria carrying certain

drug resistant or colicinogenic plasmids (Ackermann et al., 1978;

Primrose et al., 1982; Seeley and Primrose, 1982).

Since the F plasmid or sex factor F can be transferred to many

but not all strains of E. coli as well as to other members of

Enterobacteriaceae, it is not surprising that adsorption and








replication of male-specific coliphages on strains of Salmonella

(Havelaar and Hogeboom, 1984; Primrose et al., 1982), Shigella

(Kitano, 1966) and Proteus (Horiuchi and Adelburg, 1965) carrying an

F plasmid derived from E. coli has been demonstrated. However,

neither Shigella nor Proteus carrying an F plasmid will permit

formation of plaques by male-specific coliphage due to low efficiency

of adsorption. It should also be noted that the F-pili required for

adsorption of male-specific coliphages are only synthesized by actively

growing bacteria at temperatures above 300C (Birge, 1981; Primrose

et al., 1982; Seeley and Primrose, 1982). Thus, it has been suggested

that the replication and/or origin of these phages is primarily limited

to the gut of warm blooded animals (Primrose et al., 1982). There is

also some evidence for adsorption and multiplication of MS2, a male-

specific coliphage, on strains of E. coli carrying certain conjugative

plasmids other than the F plasmid, e.g., resistance plasmids (R

factors) and colicinogens (Meynell and Datta, 1966a, b). Thus, the

conjugative pili and/or phage receptors determined by some resistance

plasmids and colicinogens appear the same as that determined by the

F plasmid. However, unlike the F plasmid, the transfer operon for

expression of conjugative pili by resistance plasmids or colicinogens

is normally repressed by fertility inhibition genes (Birge, 1981).

Additionally, certain resistance plasmids inhibit F-pili synthesis

where the F plasmid is in the same cell, and thereby protect the

bacteria from infection by male-specific phages. Other plasmids, i.e.,

colicin I plasmids, code for I-pili which are distinct from F-pili and

confer sensitivity to a different set of phages than do F-pili (Birge,

1981). The incidence of E. coli strains in feces, wastewater or the








environment which are susceptible to infection by male-specific phages

is not clear. Meynell and Datta (1966b) found that six of 26 E. coli

strains freshly isolated from clinical sources were able to support

replication but not plaque formation of a male-specific coliphage, MS2,

and that all susceptible E. coli strains carried colicinogens.

Dhillon et al. (1970) reported that none of 700 freshly isolated

E. coli strains from sewage were able to support plaque formation by

MS2 although none were tested for replication of MS2 in broth culture.

Despite these observations, male-specific coliphages are widespread in

wastewater and natural waters, and often comprise a majority of the

total coliphages recovered (Dhillon et al., 1970; Dhillon and Dhillon,

1974; Havelaar and Hogeboom, 1984; Primrose et al., 1982; Seeley and

Primrose, 1979).

Other plasmid specific bacteriophages have been isolated which are

specific for bacteria carrying certain drug resistance plasmids of

Pseudomonas spp., i.e., P-1, N and W type plasmids (Bradley, 1974;

Bradley and Rutherford, 1975; Olsen et al., 1974; Stanisich, 1974; Wong

and Bryan, 1978). These plasmids are transmissible to various other

gram negative bacteria including Enterobacteriaceae and Acinetobacter

spp. Groups P, N and W plasmid-specific phages often have lipid con-

taining capsids and include cubic DNA phages with double capsids, cubic

RNA phages, tailed DNA phages and filamentous DNA phages. Because of

the apparent promiscuity of these plasmids in some environments, the

associated phages have no relationship to fecal pollution (Seeley and

Primrose, 1982) and have been recovered in high titers from river water

using Salmonella and Pseudomonas mixed hosts carrying the appropriate

plasmid (Primrose et al., 1982). It should be noted that these








plasmid-specific phages can only be propagated and/or detected on host

bacteria which carry the appropriate plasmid and not on common E. coli

hosts.

Somatic bacteriophages infective for more than one genera of

Enterobacteriaceae (Dhillon and Dhillon, 1972; Hertman, 1964; Lazarus

and Gunnison, 1947; Smith and Burrows, 1961) as well as for strains of

more than one family (Goze and Lapchine, 1971) have been isolated.

Some phages of Pasteurella (Yersinia) pestis have been reported

infective for certain strains of E. coli, Salmonella and Shigella but

not for strains of Klebsiella, Proteus, Serratia, Enterobacter,

Vibrio, Pseudomonas, Alcaligenes, Chromobacterium or various gram

positive bacteria (Hertman, 1964; Lazarus and Gunnison, 1947; Smith and

Burrows, 1961). Dhillon and Dhillon (1972) isolated phages which were

polyvalent for various strains of E. coli, Shigella sonnei and

Salmonella typhimurium, but found that none of these polyvalent phages

were active on strains of Klebsiella or Serratia. A tailed bacterio-

phage of Pseudomonas maltophilia isolated from sewage was reported to

have a remarkable host range which included strains of other species of

Pseudomonas, as well as strains of Moraxella, Serratia marcescens,

Enterobacter aerogenes and E. coli. Lysis of test bacteria by

bacteriocins or other lysins derived from the original host bacteria

during phage propagation can contribute to the observed polyvalence of

some phages and should be excluded by obtaining formation of individual

plaques (Ackermann et al., 1978). This possibility cannot be com-

pletely ruled out in some of the aforementioned investigations based

upon methods reported.








The true incidence of coliphages having hosts other than E. coli

in various environments is not clear, but recent studies suggest that

their incidence in water and wastewater is relatively low (Havelaar and

Hogeboom, 1984; Primrose et al., 1982; Zaiss, 1981) and that coliphages

having alternate hosts other than other pathogenic enteric bacteria,

e.g., Shigella and Salmonella, in wastewater are rare (Dhillon and

Dhillon, 1972). Results of Havelaar and Hogeboom (1984) and Primrose

et al. (1982) indicate that there is little overlap in the host range

of somatic E. coli and Salmonella phages and that somatic phages

polyvalent for E. coli and Salmonella comprise less than 1% of total

coliphage recovered from natural water and wastewater. In addition,

Primrose et al. (1982) isolated no phages polyvalent for E. coli and

Pseudomonas other than P-group plasmid specific phages from river

water. Zaiss (1981) tested 20 coliphages recovered from river water

and sediments for their infectivity on genera other than E. coli and

found that they were all restricted to strains of E. coli; none of the

coliphages were infective for strains of Enterobacter aerogenes,

Klebsiella pneumoniae, Serratia marcescens, Salmonella typhimurium,

Proteus vulgaris, Pseudomonas aeruginosa, Alcaligenes eutrophus or

various gram positive bacteria. Delisle and Levin (1969) reported that

the host range of 138 Pseudomonas bacteriophages isolated from sea

water, sewage and spoiled fish was restricted to strains of Pseudomonas

and Alcaligenes and did not include strains of E. coli, Enterobacter

aerogenes, Serratia marcescens or Salmonella typhimurium. Likewise,

Whitman and Marshall (1971a) found that phages of Pseudomonas or

Enterobacteriaceae isolated from refrigerated foods were not cross

reactive for bacterial isolates across family boundaries.







As previously mentioned, the numbers of coliphage recovered in a

given sample are a function of the E. coli host used as well as various

other factors, e.g., the age of the host bacteria, the temperature of

incubation, and composition of the assay medium (Havelaar and Hogeboom,

1983; Primrose et al., 1982). The choice of an appropriate host may

depend upon the objectives of a study, i.e., the recovery of all

possible coliphages or that of a particular group of coliphages (Seeley

and Primrose, 1982). With regard to detecting maximal numbers of

coliphages, wild-type or freshly isolated E. coli strains have been

reported inferior to laboratory strains for enumerating indigenous

coliphages in wastewater (Dhillon and Dhillon, 1974; Bell, 1976), river

water (Hilton and Stotzky, 1973) and estuarine water (Vaughn and

Metcalf, 1975). This observation has been attributed to the usual

presence of a complete 0-antigen complex in wild-type E. coli which

can mask many phage receptors located deeper within the core of the

lipopolysaccharide (Havelaar and Hogeboom, 1983). Thus, rough or semi-

rough laboratory strains of E. coli lacking a complete 0-antigen

complex are generally more productive hosts. Laboratory strains of

E. coli commonly used for coliphage assays in water and wastewater

include female strains of E. coli B (Bell, 1976; Bitton et al., 1981;

Buras and Kott, 1966; Dhillon et al., 1970; Havelaar and Hogeboom,

1983; Ignazzitto et al., 1980; Kennedy et al., 1985; Ware and Mellon,

1956), E. coli K-12 (Dhillon et al., 1970; Primrose et al., 1982;

Simkova and Cervenka, 1981; Zaiss, 1981) and E. coli C (Dhillon et al.,

1976; Havelaar and Hogeboom, 1983; Ignazzitto et al., 1980; Kenard and

Valentine, 1974; Kennedy et al., 1985; Stetler, 1984; Wentsel et al.,

1982). Male strains of E. coli having additional receptors, i.e.,








F-pili, such as E. coli A 19 (Glass and O'Brien, 1980; Stetler, 1984),

E. coli C-3000 (Fannin et al., 1977; Gerba et al., 1978) and male

derivatives of E. coli K-12 (Dhillon et al., 1970; Durham and Wolf,

1973; Fannin et al., 1977; Havelaar and Hogeboom, 1983; Ignazzitto

et al., 1980; Kott et al., 1974; Parry et al., 1981; Primrose

et al., 1982; Seeley and Primrose, 1979) have also been used for assay

of total coliphages in water and wastewater.

Few comprehensive studies have compared different host strains of

E. coli for enumeration of coliphages in water or wastewater.

E. coli C has been found to recover more coliphages from sewage and

water than many other host strains tested including E. coli B, E. coli

K-12, and even male derivatives of E. coli K-12 (Havelaar and Hogeboom,

1983; Ignazzitto et al., 1980; Kennedy et al., 1985; Stetler, 1984).

Gerba et al. (1978) reported higher recoveries of coliphages from

wastewater with E. coli C-3000 than with E. coli B while slightly

higher recoveries were reported for E. coli C-3000 than for E. coli

K-12 HfrD (male strain) by Fannin et al. (1977). The higher plaque

counts obtained with E. coli C host have been attributed to the lack

of DNA restriction and modification systems, which can prevent penetra-

tion of phage DNA into the cell, as well as the presence of more

receptors for different types of coliphages than other E. coli strains

(Havelaar and Hogeboom, 1983; Seeley and Primrose, 1982). Although no

universal plating host for all coliphages has been isolated, the

enumeration of coliphages can be optimized with regard to maximal

production of plaques using broad-range hosts such as E. coli C or a

male strain of E. coli C (Havelaar and Hogenboom, 1983; Seeley and

Primrose, 1982). E. coli C is also the designated host for a standard








test method for determination of coliphages in water (ASTM, 1983).

Host systems have also been developed for selectively enumerating male-

specific coliphages (Havelaar and Hogeboom, 1984; Primrose et al.,

1982) as well as group P plasmid-specific bacteriophages (Primrose

et al., 1982).


Effects of Temperatures on Coliphage Ecology


The age, origin and degree of fecal contamination in water based

upon the presence of E. coli or their bacteriophages may be difficult

to establish in environments capable of supporting their concomitant

proliferation (Parry et al., 1981; Seeley and Primrose, 1980; Primrose

et al., 1982; Vaughn and Metcalf, 1975). Distinct, physiological types

of coliphages have been recognized with regard to the effect of

temperature on their ability to multiply or produce plaques on E. coli

hosts (Parry et al., 1981; Seeley and Primrose, 1980). Parry et al.

(1981) categorized coliphages in river water as low temperature (LT) or

high temperature (HT) types depending upon their ability to replicate

exclusively below or above incubation temperatures of 28-300C. These

workers suggested that the ratio of coliphage counts on plates

incubated at 370C to those on plates incubated at 220C might be useful

to indicate the age of fecal pollution in water, and that only HT

coliphages be used as viral pollution indicators. In a more comprehen-

sive investigation, Seeley and Primrose (1980) observed three physio-

logical types of coliphages based upon the effect of temperature on

their infectivity as follows: (1) HT phages, plating in the range of

25-450C, (2) LT phages, plating in the range of 15-300C, and (3) mid-

temperature (MT) phages, plating in the range of 15-420C. The relative








distribution of these types of coliphages closely reflected the

temperature and/or degree of fecal contamination of the environments

from which they were isolated. For example, the feces of warm-blooded

animals contained only HT and MT phages with an average percentage

ratio (HT:MT) of 63:37 whereas the percentage ratio of HT:MT:LT phages

in raw sewage was 9:84:7, reflecting the ability of ambient sewage

temperature to favor the growth of MT and LT phages. Furthermore, the

proportion of LT phages was highest in unpolluted river water near the

river's source, decreased in river water near sewage outfalls and

increased once again as the river self-purified downstream. The

HT:MT:LT percentage ratio was 5:42:53 at the source of the river and

8:78:16 immediately downstream from two sewage outfalls.

The results of Seeley and Primrose (1980) suggested that the

temperature of natural water would favor replication of LT and MT

phages with the percentage of HT phages being correspondingly lower and

that the ratio of MT to LT phages would increase with influence of

domestic or agricultural wastes. These observations were supported by

Primrose et al. (1982) who reported that only HT coliphages were

recovered from lake water during months when temperatures were 100C or

less whereas LT coliphages were recovered in higher numbers than HT

phages during warmer months, i.e., water temperatures greater than

15C, when replication of LT phages was possible. Male-specific

coliphages as a group have been suggested as more direct indicators of

fecal pollution than total coliphages in natural water since their

infectivity is restricted to temperatures greater than 300C and,

generally the gut of warm blooded animals (Primrose et al., 1982;

Seeley and Primrose, 1982).







Coliphages as Indicators of Fecal Contamination


Overview of Classic Bacterial Indicators


The examination of potable water for generally non-pathogenic

indicator organisms of fecal origin, i.e., E. coli and/or related

lactose fermenting, coliform bacteria which would presumably indicate

the presence of fecal contamination was introduced in the late

nineteenth century (Schardinger, 1892; Smith, 1895). Coliform bacteria

have long since been established as indicators of fecal pollution in

water and have also been widely used in assessing the sanitary quality

and processing adequacy of various foods (Buttiaux and Mossel, 1961;

Jay, 1978; Mossel, 1982; Thatcher and Clark, 1968). Since many members

of the coliform group are indigenous to the extraenteral environment,

fecal coliforms and E. coli have been adapted as more direct indicators

of fecal contamination or enteric organisms in water and foods.

Various other bacteria or groups of bacteria, e.g., fecal streptococci,

the family Enterobacteriaceae, Clostridium perfringens and Bacteriodes

spp., have also received attention as fecal indicators in water and

food, depending upon the system being examined. The sanitary signifi-

cance of particular indicator bacteria in food products is generally

unique to the type of product with regard to its origin, the degree of

processing or handling to which it is exposed, its packaging and its

storage conditions (Buttiaux and Mossel, 1961; Jay, 1978; Mossel, 1982;

Thatcher and Clark, 1968). For example, the presence of coliforms and

E. coli in many raw commodities is often unavoidable and represents an

acceptable component of the initial microflora whereas their presence

in heat processed or pasteurized foods would not be acceptable. Due to








the uneven distribution of enteric pathogens in foods, quantitative

relationships between bacterial indicators and enteric pathogens in

foods or food safety have not been demonstrated in foods as they have

in some water systems (Geldreich and Bordner, 1971; Miskimin et al.,

1976; Solberg et al., 1977). However, it has been suggested that the

presence of inordinately high levels of indicator bacteria in a given

food product may indicate grossly contaminated raw materials as well as

a lack of sanitation in handling, distribution and/or storage (Buttiaux

and Mossel, 1961; Mossel, 1982). The presence of coliform indicator

bacteria in processed foods indicates grossly contaminated raw

materials, inadequate processing or post-processing contamination as a

result of insanitary packaging or handling conditions.


Coliphages as Indicators of Enteric Bacteria in Water and Wastewater


In a recent review of bacteriophages as potential indicators of

enteric pathogens in water and wastewater, Scarpino (1978, p. 218)

states that

correlations appear to exist in fresh and marine waters
between fecal bacterial pathogens such as Salmonella and
Shigella species and fecal indicator bacteria such as
E. coli and their bacteriophages.

Associations between fecal pollution levels, i.e., coliform levels, and

the presence of bacteriophage infecting the coli-dysentery-typhoid-

paratyphoid group of bacteria in natural water was first noted over 50

years ago (Gildemeister and Watanabe, 1931; Schlossmann, 1932). The

public health significance of bacteriophages of E. coli and of enteric

pathogens in water was also demonstrated by subsequent studies

(Dienert, 1934, 1944, 1945; Etrillard and Lambert, 1936; Guelin,








1948a, b, 1950). Many of these early researchers concluded that

bacteriophages, i.e., coliphages, were more reliable indicators of

sewage pollution than E. coli since they could often be recovered in

the presence of enteric pathogens when E. coli could not.

More recent studies have supported early observations and

quantitatively examined the relationship between levels of coliphage

and fecal pollution, e.g., coliforms or fecal coliforms, in wastewater

and water (Kenard and Valentine, 1974; Wentsel et al., 1982). Kenard

and Valentine (1974) reported a correlation coefficient of 0.95

(p < 0.001) between coliphages (E. coli C host) and fecal coliforms

recovered from 35 samples of wastewater and surface water. Linear

regression analyses indicated that the ratios of coliphages to fecal

coliforms and coliphages to coliforms were 0.7:1 and 1:25, respec-

tively, in five different sources of surface water and one source of

sewage effluent, across a broad range of contamination levels. Based

upon these results, Kenard and Valentine (1974) suggested that predic-

tions of fecal coliform levels in water could be made by enumeration of

coliphages with results available within 8 hours. In a more comprehen-

sive examination of natural waters across the U.S., highly significant

correlation coefficients (p < 0.001) were found between coliphages

(E. coli C host) and fecal or total coliforms recovered from 600 water

samples (Wentsel et al., 1982). As few as six coliphages per 100 ml

of water were enumerated within 6 hours. Slopes of regression lines

between coliphages and fecal coliforms in the studies of Kenard and

Valentine (1974) and Wentsel et al. (1982) were noted by the latter

authors who suggested that the correlation between coliphage and fecal

coliform levels in natural waters was constant over time, geographical








location and type of water. The relatively constant ratios between

coliphages and coliforms or fecal coliforms in numerous sources of

water reported by Kenard and Valentine (1974) and Wentsel et al. (1982)

are somewhat surprising in light of various studies which indicate the

greater resistance of coliphages and other viruses to inactivation in

water and wastewater than coliform bacteria (Bell, 1976; Durham and

Wolf, 1973; Fannin et al., 1977; Hejkal, 1985; Ignazzitto et al., 1980;

Kapuscinski and Mitchell, 1982). Thus, ratios of coliphages to

coliform bacteria have generally been found to increase during sewage

treatment and/or over time in isolated wastewater and water systems.

Some workers have also questioned the reliability of coliphage or

bacteriophage indicators in natural waters with regard to apparent

fluctuations in host specificity of indigenous coliphages (Hilton and

Stotzky, 1973; Vaughn and Metcalf, 1975). Inconsistent correlations

between bacteriophages of wild-type E. coli or Enterobacter aerogenes

hosts and coliform levels in river water were observed over a 4 month

period by Hilton and Stotzky (1973). Vaughn and Metcalf (1975) also

noted fluctuations in host specificity between three E. coli hosts

including a wild-type host for coliphages recovered from estuarine

water over a period of 3 years. It should be noted that broad-range

E. coli hosts (Havelaar and Hogeboom, 1983; Ignazzitto et al., 1980)

were not used in the studies of Hilton and Stotzky (1973) and Vaughn

and Metcalf (1975), thereby limiting the types of coliphages which

could be detected.

Bacteriophages of enteric bacteria other than E. coli have

also been considered as indicators of fecal contamination or

presence of specific pathogenic bacteria in environmental, food and








clinical samples. Bacteriophages of Bacteriodes spp., a predominant

intestinal genus of many warm blooded animals including man, have

recently been recovered from water, wastewater and aquatic sediments as

well as the feces of humans and various animals (Tartera and Jofre,

1985). Levels of Bacteriodes bacteriophages were found to correlate

with the degree of fecal pollution, e.g., coliform and fecal coliform

levels, in surface waters, wastewater and aquatic sediments. Giraldi

and Donati (1976) reported that the detection of bacteriophages of

salmonellae, shigellae, vibrios and E. coli in polluted waters closely

correlated with the detection of the respective bacteria using conven-

tional methods. A similar approach has been applied to diagnosis of

El Tor Vibrio carriers by detection of bacteriophages specific for

El Tor Vibrio in the stools of patients (Takeya et al., 1965). The

phage method was more sensitive than direct testing of stools for El Tor

Vibrio. Another bacteriophage technique for indirect detection of

Vibrio cholera in wastewater, vegetables and stools is based upon

multiplication of two types of bacteriophage specific for most strains

of Vibrio cholera in sample enrichment cultures (Sechter et al., 1975).

A similar technique for indirect detection of Salmonella in milk

detection of plaques on lawns of bacteria derived from selective

enrichment cultures by a highly virulent salmonellae-specific bacterio-

phage has also been described (Hirsh and Martin, 1984).


Coliphages as Indicators of Enteric Viruses in Water and Wastewater


Coliphages have received considerable attention as expedient

indicators of enteric viruses in water and wastewater (Fannin et al.,

1977; Glass and O'Brien, 1980; Grabow et al., 1978; Kott et al., 1974;








Scarpino, 1975, 1978; Simkova and Cervenka, 1981; Stetler, 1984; Vaughn

and Metcalf, 1975). Relationships between levels of coliforms or

coliphages and enteric viruses in wastewater and water are difficult to

assess due to the extreme fluctuations in the incidence and levels of

viruses entering wastewater as compared to more ubiquitous intestinal

microorganisms (Scarpino, 1978). However, correlations between occur-

rence or levels of coliphages and enteric viruses in wastewater and

natural waters have been reported (Kott et al., 1974; Simkova and

Cervenka, 1981; Stetler, 1984). Kott et al. (1974) reported the

simultaneous recovery of enteroviruses and coliphages (E. coli B host)

in various wastewater samples as well as samples of river, stream and

drinking water; human enteric viruses were absent when coliphages were

absent. Ratios of coliphages to enteric viruses varied over time with

ratios ranging from 1:1 to 103:1 in natural water and 10 :1 to 105:1 in

wastewater; coliphages levels were relatively constant as compared to

levels of enteric viruses (Kott et al., 1974). In addition, coliphages

and enteroviruses were simultaneously detected in tap water samples in

a ratio of 10:1, respectively, during periods of enteric disease out-

breaks in a small community. Parallel seasonal fluctuation patterns

and persistence of coliphages and enteric viruses in river water

receiving treated sewage effluent (Stetler, 1984) and in various

surface waters receiving fecal and chemical pollution (Simkova and

Cervenka, 1981) have also been reported. Stetler (1984) found that

enterovirus and coliphage counts obtained with two broad-range E. coli

hosts were significantly correlated (p < 0.01) in river water providing

a source for a water treatment plant as well as in water samples taken

at various stages of water treatment. Enterovirus counts were more








strongly correlated with coliphage counts than with coliforms, fecal

coliforms, fecal streptococci or standard plate counts for all types of

water (Stetler, 1984). On the other hand, a very inconsistent rela-

tionship between the incidence of coliphages and enteric viruses in

shellfish and corresponding estuarine water and sediments was reported

by Vaughn and Metcalf (1975). However, it is difficult to compare the

coliphage recoveries of Vaughn and Metcalf (1975) to those of Kott

et al. (1974), Simkova and Cervenka (1981) or Stetler (1984) because

the former workers were using less sensitive coliphage recovery tech-

niques than the latter workers. Vaughn and Metcalf (1975) also

demonstrated the ability of indigenous coliphages to multiply in the

presence of E. coli B in estuarine water. These observations and the

ability of some coliphages, e.g., LT or MT types (Parry et al., 1981;

Seeley and Primrose, 1980) to multiply in suitable E. coli hosts at

ambient water temperatures, must be considered in regard to use of

coliphages as viral indicators in some situations.

Both coliphages and enteric viruses are generally more resistant

to inactivation by chlorination (Burns and Sproul, 1967; Clark and

Kabler, 1954; Durham and Wolf, 1973; Fannin et al., 1977; Kott et al.,

1974; Liu et al., 1971; Scarpino, 1975) and various environmental

stresses (Bell, 1976; Fannin et al., 1977; Hejkal, 1985; Ignazzitto

et al., 1980; Kapuscinski and Mitchell, 1982) than are coliform and

other enteric bacteria. For example, coliphages have been recovered

consistently in the absence of coliforms in chlorinated sewage effluent

while Hepititis A virus as well as other enteric viruses have been

recovered from drinking water which contained acceptable chlorine

residual and was free of coliform bacteria (Durham and Wolf, 1973;







Hejkal et al., 1982; Marzouk et al., 1980; Payment, 1981; Peterson

et al., 1983). Various studies suggest that coliphages as a group are

superior to coliform bacteria as indicators of viral inactivation

during water (Stetler, 1984) or wastewater treatment (Burns and Sproul,

1967; Durham and Wolf, 1973; Fannin et al., 1977; Kott et al., 1974;

Lothrop and Sproul, 1969). Stetler (1984) reported significant corre-

lations between enterovirus and coliphage counts but not between

enterovirus and bacterial indicator counts at various stages of water

treatment, i.e., source, post-sedimentation, post-sand filtration and

finished, at a municipal water treatment plant. Glass and O'Brien

(1980) found that inactivation rates and kinetics of enteroviruses and

coliphages were not statistically different (p > 0.01) during activated

sludge treatment and also suggested that coliphages were acceptable

indicators of viral inactivation in sewage treatment. Similar survival

patterns of coliphages and enteroviruses in estuarine waters and

oysters have also been reported (Vaughn and Metcalf, 1975). Noting the

similarities between male-specific, RNA coliphages and enteric viruses

with regard to size, structure and stability, various investigators

have suggested their use as viral indicators in shellfish and shellfish

waters (Metcalf, 1978) as well as models of viral inactivation in

natural waters (Primrose et al., 1982; Seeley and Primrose, 1982),

sewage treatment (Kott et al., 1974), water treatment (Cramer et al.,

1976) and composting (Burge et al., 1981). The relative physico-

chemical stability of bacteriophages, viruses and bacteria will be more

broadly addressed in a subsequent section.








Coliphages as Indicator Organisms in Foods


Relatively little information is available concerning the occur-

rence of coliphages or the relationship between their presence and that

of enteric bacteria or viruses in various types of foods. As early as

1937, bacteriophages of E. coli, Enterobacter aerogenes, Salmonella

typhi and S. paratyphi were detected in raw milk along with various

coliform bacteria although the correlation or relationship between

phages and bacteria was not reported (Lipska, 1937). Kott and Gloyna

(1965) reported a strong correlation between coliphages (E. coli B

host) and coliform MPN counts, over a range of 10-103 phages and 10 -

10 coliforms per lOOg of oysters placed in estuarine water stations at

various distances from a sewage outfall. A relatively consistent ratio

of 10 coliforms to one coliphage was found in oysters from three sampl-

ing stations. Coliphages were also recovered in market oysters, but in

lower levels ranging from less than two to 17 per 100 g of oyster meat

(Kott and Gloyna, 1965). Coliphages of three E. coli hosts including

E. coi B and a wild-type E. coli were widely recovered in oysters and

shellfish growing waters (Vaughn and Metcalf, 1975). Vaughn and

Metcalf (1975) did not report levels of phage or enteroviruses

recovered from oysters and did not test samples for bacterial

indicators.

The recovery of coliphages (E. coli C host) from 100% of fresh

chicken and pork sausage samples as well as from 33% of cooked

delicatessen meats has recently been reported (Kennedy et al., 1984).

A positive correlation was observed between coliphage and fecal

coliform levels for 18 fresh meat samples over a range of approximately








0.1 to 102 coliphages and less than 10 to 103 fecal coliforms per gram.

However, ratios of coliphage to fecal coliforms varied considerably for

a given food type and mean ratios for each food type differed by one to

two orders of magnitude. The host specificity of coliphages and their

comparative distribution in relation to levels or occurrences of

E. coli and other bacterial indicators as well as the types of

coliphages occurring in various foods has not been reported.


Comparative Physicochemical Stability of Coliphages


General Overview


Bacteriophages and other microorganisms in foods are subject to

inactivation by a number of physiochemical factors, e.g., temperature

extremes, pH, osmotic pressure, water activity and various inorganic

and organic compounds added to, or naturally comprising, foods (Jay,

1978). In addition, the rates of inactivation are influenced by

interactions between these factors and other food components such as

fats, sugars and salts. E. coli and coliphages are also capable of

proliferation in the extraenteral environment and possibly in various

foods under suitable physicochemical, ecological and nutritional

conditions (Anderson, 1957; Buttiaux and Mossel, 1961; Laliberte and

Grimes, 1982; Seeley and Primrose, 1980; Vaughn and Metcalf, 1975;

Primrose et al., 1982). The relative stability of coliphages in

various food systems is important when considering their use as

indicator organisms for enteric viruses or bacteria in foods. Very

little information is available concerning the stability of coliphages

or other bacteriophages in food environments as a function of








processing, food composition or storage conditions (DiGirolamo et al.,

1972; DiGirolamo and Daley, 1973). Studies of viral stability in some

food systems as well as basic studies on the relative effects of

various physicochemical factors on bacteriophages and viruses can pro-

vide some perspective. However, it is often difficult to predict the

stability of an organism in foods based upon studies of related groups

or species because of the variation in stability reported with even

closely related organisms and the complexity of the food environment

(Cliver and Salo, 1978; Jay, 1978). Because of the paucity of informa-

tion concerning the physicochemical stability of coliphages in food

systems, relevant studies of human viruses in this regard will also be

discussed.


Effects of High Temperature


The thermal inactivation of viruses, bacteriophages and bacteria

generally follows first order kinetics although biphasic inactivation

plots are often observed with increasing time (Hiatt, 1964; Stumbo,

1973). The rate of inactivation of microorganisms is influenced by a

number of factors including the nature and number of organisms and

various characteristics of the suspending medium such as water activity

and pH as well as the presence of solutes, fats and proteins (Jay,

1978; Cliver and Salo, 1978). The thermostability of viruses and

bacteriophages generally increases with increasing salt concentrations

and/or decreasing water activity but is salt specific (Adams, 1949;

Burnet and McKie, 1930; Dimmock, 1967; Lark and Adams, 1953; Wallis and

Melnick, 1962a; Wallis et al., 1965). Proteinaceous and/or more

complex media such as nutrient broth enhance the thermal stability of








coliphages as compared to water or simple buffers (Adams, 1949; Burnet

and McKie, 1930; Lark and Adams, 1953). Bovine serum albumin, nutrient

broth, skim milk and cystine have a similar protective effect for

polioviruses exposed to heat (Wallis and Melnick, 1961). Reducing

substances such as cystine, cysteine, dithiothreitol and glutathione

have also been reported to increase the thermostability of entero-

viruses (Halsted et al., 1970; Pohjanpelto, 1958). Various components

of food also have been shown to contribute to increased heat resistance

of viruses and phage. A bacteriophage of Streptococcus cremoris was

more heat resistant in milk than culture broth at 650C (Koka and

Mikolajcik, 1967). The heat resistance of polioviruses was reported to

be enhanced by increasing levels of fat in ground beef (Filppi and

Banwart, 1974) and by collagen in phosphate buffered saline (Milo,

1971). Enteroviruses (Kaplan and Melnick, 1952, 1954), adenoviruses

and reoviruses (Sullivan et al., 1971c) were less heat resistant in

milk than in ice cream mix which contains higher levels of fat, sugar

and milk solids than milk. The heat resistance of phages and viruses

is also influenced by the pH of the suspending medium with the effect

of pH varying according to the type or strain of virus (Daoust et al.,

1965; Friedman and Cowles, 1953; Janda and Vonka, 1963; Rafajko and

Young, 1964; Wallis and Melnick, 1962b; Wilkowske et al., 1954).

Extreme variation in relative heat sensitivity has been

demonstrated in viruses, particularly bacteriophages, and considerable

diversity has been noted among strains of the same group or serological

type of virus. A comparative summary of thermal resistance reported

for various bacteriophages, animal viruses and representative bacteria

or bacterial spores, presented in Table 3, illustrates this point.

















-4 -
V) Lr1



r01 ON0
mM n

4-4 o
ca co
wcc



00O'0> I
a -,t co4-1I
ca c (ON (L)W
Ca 0a ) m a)>

0ri r -H
r-O r-- W r-1 ,


5-4mm
0 0 "a0 co







C S
G- *H -








00





*.0 M.a.0
-11 CO
*CO OO C (1




r"1 r-1 %O 0 -








'0 0 .o CO
C *rl 3

* On 11.i


0U
u


Q)
SI









>

0)







(U








E-







.H
540
0 3






C
(H

























-W
c-




00


E-l E-l E-l 4 ,-l



*-r4
1
0 0
) U
00 *-H
s &
0 i-
54 -)
C) U
U M
COu
W 0
a u
co


/-S



0'0 .0
-41-1



*L0 CUTO







00 0000 I c


to M ca,.H
00000:3


c E
-*H C-

00
00
II I

^o


c C
*M -H
EE



vIv
I -4
C 0 I


000000
,,o r- ,o,. oo '.o


o 0c. 0



HH.
c/) En u
&-4 E- &-


00
.-4 ,.


*
4-> 4-14->^-^ 4- 4.)
En2 (1 (A2 -4 (12 (a
-4 -04 C -1 CO -




U)
CO0


icrI
CO 0 m U
0 0 -
6 oc 0 0
0 :3 IV 3C 4-)
3) (U4-4 4-14 Q)
a) CO a
EM *o o -

(< 12 |
CIHIIC/


- -4




i-i r-4 Ca


0 0





0 0 o
000












S I









' 0 3


JS
0
'-4
*.0
22 2 H-i
- -l -1 .


*
1-1
44-4
U -4/- (0-
U 5-- C-
%-,^ I-, -


0 4)3 4-
E o r-4 co
C) Co r-4 00
w- w-4'Ha)
o u E
C/2IO
C/31M
enen


.c .ca

00 4-
O E-

* >-














-4-.

ON u U
-4 -4-
r-- r- c -






ON 0aa%
.-~4 '-l %- .1


-)d W~ s.4 $.4 ca cc1


u () c) ) -H > >
0 00O0 Ml .q H
0 C 00 -i r-
E 3 3 3 r-I r-1 -1
0 0 0 *H w








C *r a -H
*H*2rC5s-4
c rice i



CN C-4
c@ **** e*

I o I T 0
C) 0 I% CM *

%0 00 -4 c0 -t, C- C


U
C
1)
r-4





90)
'-4
a)









40)

0>
uI =





0)1
U)



0)
C4









Of
C> C,..

I-l0
0) 5-1
0. (
C13 u
(1) %1.
SQ)






E-





r-i



0)














i-l
CO


















ca
o0
0) _


000



Lr in ro Lro


XX
40 4-4e
0) ON ON ON 'a)
4-4 ON as ON 0) e E
4-4 CO CO
:3 rL -H

H 0000 00)
5S4 .) 5. U U






(D (0
IIIowH


*r4 *H- rI H 3
> > > > w
0 0 00 "
*H *i-4 *-l r. >
H 0
0 0 0o0
0.4 (14 A4 --t


cO
r-4



.0 (m C m 0 c 0
) c- c- .-i -4 rca c rc -








w1.-== =.= o'-= ,!
C 0 l



0) )




3 0D cU CO cO cO -C ( l .C
Oi 00;QQ4cQOC/


CCCCcccCe
-H rq *r- *g *r *Hg *H H C

Nc00a-Tcn mco mt oN
OOOOOM NrCnr-n








VVn000i ii00m00


*- *H -H -H


0u 01 0) 0
0000
0000
444
3>> S3


rM
00



CO
ca r-I



.C

U
,-H
Un


U) U) 1 -S


0 ,0 U


cU 4-C Q4- n
0) 0)M 3 -
U U
U) ca ca u
U r-4 1-4 0U
S 0 ) 0) 0 U
000 V


0 r-4 0 0 0< ca
u 4 o E S Wu 4-4
0, CO CO 4-
M fL I v-- 5-)
-z cac C|


0
'0

in) or



























0)
u

















E-I
()






Co





Co. c

















a)



r-














00
a





















Co


0 0
0 0
4-4 4-4


C Co

ca
O Co


> 0



--H
Co



L l o


oH 1 0
Co 0 3H





0.0 U 4.)
U- Co


II
CO CO
H 04




0
a c
4-o

o 4-4




0
Uo u






0II





w 0
o C


Q. 0)

II 4-4
>-1


2 (















o
.c u






Co





0
Sa
4 C
rU 0)






0
1-4
o 0

CU

-H 0
a 00
4-j *
0





o0 CO
U Q




( 4-1
0-


CO

0
P4 a
H0 U



*r4 0
0M
E--


I


S0


0




u
U




o
Q) *-4


41
l-i 0



3 U




C OP





O C






W r
0 C
44 J


0 E
0 -










.o
o
CO C
E )


Ca



0cE
-o



0



C

e Co

o a
0'
a







(0


*C'
u >
a )

0 r 0

a )


Co 0


> 0


u


0
>








.H
U


C
0


4-4




Co
*.-4











t-
CC
4-4




CO

*H
4-4

w.
0


CO










0
C4J
r-
0











ca
I-l
C
0)





0-
in
cu
>-
CO

en








The thermoduric nature of various phages has been demonstrated in broth

media and suggests that some bacteriophages, including coliphages, are

capable of surviving various heating and/or pasteurization processes,

depending upon initial numbers of phage in the food and the type of

food (Table 3). Some enteroviruses (DiGirolamo et al., 1970; Filppi

and Banwart, 1974; Kaplan and Melnick, 1952, 1954; Sullivan et al.,

1975) and thermoduric bacteria, e.g., Streptococcus faecalis,

S. faecium, Micrococcus sp., Microbacterium sp. and Lactobacillus sp.

(Dabbah et al., 1971b; Ott et al., 1961; Shannon et al., 1970;

Thomas et al., 1967) have been reported to survive various thermal

processes in foods but may not be as heat resistant as certain phages,

based upon comparative studies (Table 3).

Very little data are available concerning the heat stability of

coliphages or other bacteriophages in food systems. Using coliphage T4

as a model for viral inactivation, DiGirolamo et al. (1972) reported

high uptake of T4 by crabs in water and subsequent survival rates of 21

and 2.5% in crabs after 5 and 20 minutes of boiling, respectively. The

corresponding internal temperatures of crabs after 5 and 20 minutes of

boiling were 70 and 84C, respectively. Coliphages have been recovered

from cooked meat products although the concomitant recovery of fecal

coliform bacteria indicated that post-processing contamination did

occur (Kennedy et al., 1984). Thus, coliphages recovered from cooked

meats could reflect their thermostability and/or post-processing

contamination (Kennedy et al., 1984).

Thermostability studies of enteroviruses in various foods may be

of some interest in relation to analogous stability of some coliphages

since enteroviruses and cubic RNA phages are similar in size, structure







and morphology. Survival rates of poliovirus in oysters ranged from

approximately 7 to 13% after stewing, steaming, frying or baking for

8, 30, 8 and 20 minutes, respectively. The internal temperatures of

oysters following these processes were 75, 94, 100 and 900C,

respectively (DiGirolamo et al., 1970). The D-values at 700C of

poliovirus in ground beef subjected to thermal processing were 2.7 and

4.4 minutes in the presence of 3 and 27% fat, respectively (Filppi and

Banwart, 1974). Sullivan et al. (1975) reported survival rates 1.4

and 1.7% for poliovirus 1 and coxsackie virus B-2 in rare, broiled

hamburgers heated to an internal temperature of 60C over 7.9 minutes

and quickly cooled. However, no virus was detected if the hamburgers

were allowed to maintain 60C for 3 minutes after cooking or if they

were broiled to 71 or 76.70C, corresponding to medium and well done,

respectively.


Effects of Low Temperatures and Other Physicochemical Factors


The inactivation of microorganisms including viruses and

bacteriophages as a result of freezing, storage at subfreezing

temperatures and thawing is influenced by physicochemical factors

similar to those affecting heat stability. Substances such as egg

white, sucrose, corn syrup, glycerol and meat extracts protected

various food-borne bacteria from freeze inactivation whereas lowering

pH tended to increase inactivation (Georgala and Hurst, 1963).

Likewise, Tl and T7 coliphages were not freeze-inactivated when

suspended in nutrient broth, casein, peptone or beef extract but were

significantly inactivated in Tris buffer (Schiffenbauer and Calderon,

1985). However, the presence of 0.1% egg albumin protected Tl but not








T7 from freeze inactivation. Differences in freeze sensitivity between

strains of phages and suspending media were also noted for P. fragi

phages where phage "psl" was 70% inactivated by freezing and storage

at -200C for 24 hours in skim milk with phage "wy" undergoing no

inactivation (Whitman and Marshall, 1971b). On the other hand, neither

P. fragi phage was affected by freezing and storage at -20C for 24

hours in trypticase soy broth or ground beef extract.

Since bacteriophages, animal viruses and bacteria as well as

related strains within these groups may vary greatly in their resis-

tance to freeze inaction, it is difficult to generalize about the

comparative freeze resistance of these groups. Among bacteria, the

cocci are generally more freeze and low temperature resistant than gram

negative rods, with endospores being virtually unaffected by freezing

(Jay, 1978). For example, survival of fecal streptococci and coliform

organisms in chicken gravy at -210C for 91 days was approximately 100

and 2%, respectively (Kereluk and Gunderson, 1959). Inactivation rates

of enterococci after 14 or 15 days of storage at 1C were approximately

0 and 0 to 25% in trypticase soy broth, TSB (Patterson and Jackson,

1979), and ground beef (Goepfert and Kim, 1975), respectively, whereas

inactivation rates of E. coli under similar conditions were approxi-

mately 15 and 43 to 69% in TSB and ground beef, respectively.

Inactivation rates of salmonellae at 1 to 20C over 14 days of storage

were 26 to 38% and 78 to 95% on beef plate surfaces (Kennedy et al.,

1980) and ground beef (Goepfert and Kim, 1975), respectively.

Several studies suggest that many viruses and bacteriophages can

survive frozen and refrigerated storage in foods for extended periods.

The inactivation rate of coliphage T4 in cooked and uncooked crabs was








65 and 23%, respectively, after 30 days of storage at -200C whereas

the inactivation rate after 5 days at 80C was 60 and 71% for cooked

and uncooked crabs, respectively (DiGirolamo and Daley, 1973).

S. aureus phage 80 in oyster plasma held at 50C for 5 days was not

detectably inactivated (Feng, 1966). The survival of poliovirus in

frozen oysters held at -17.50C was 91, 40 and 10% after 14, 42, and 84

days, respectively, with 40% of poliovirus surviving storage at 5C

for 15 days (DiGirolamo et al., 1970). In another study, no reduction

of poliovirus viability occurred in six of seven oyster lots after 28

days of storage at 50C (Tierney et al., 1982). Titers of coxsackie-

virus A9 in ground beef were reduced by only 16% during 8 days of

storage at 40C (Herrmann and Cliver, 1973). Poliovirus 1 and

coxsackieviruses B1 and B6, added to various commercially frozen or

convenience foods, maintained 5 to 100% of viability after 4 to 5

months of storage at -200C or 2 weeks at 100C ranging from approxi-

mately 27 to 37% and 0.5 to 2% for non-acid and acidic foods,

respectively (Heidelbaugh and Giron, 1969). Cliver et al. (1970) also

noted that acidity significantly enhanced the inactivation of entero-

viruses in low moisture foods held at 5C or 22 to 24C depending upon

the protein and salt content of the food. It should be noted that

coliphages are generally less resistant to extremes in acidity or

alkalinity than are enteroviruses (Sabatino and Maier, 1980).

The stability of enteroviruses in various foods held at room

temperature has also been investigated and generally found to be less

than the corresponding stability at refrigeration or subfreezing

temperatures (Cliver et al., 1970; Heidelbaugh and Giron, 1969;

Herrmann and Cliver, 1973; Lynt, 1966). The influence of spoilage








bacteria or their metabolites on the inactivation of enteroviruses in

foods has also been investigated (Cliver and Herrmann, 1972; Cliver

et al., 1970; Heidelbaugh and Giron, 1969; Herrmann and Cliver, 1973).

Although protoeolytic enzymes derived from bacteria in growing cultures

have been reported to inactivate enteroviruses (Cliver and Herrmann,

1972), the presence of high levels of spoilage bacteria does not seem

to significantly affect enteroviruses in foods undergoing spoilage at

refrigeration or room temperatures (Cliver et al., 1970; Heidelbaugh

and Giron, 1969; Herrmann and Cliver, 1973). The inactivation of

phages or viruses by various food additives has not been extensively

investigated, although poliovirus was apparently inactivated by sodium

bisulfite added to commercially prepared cole slaw (Lynt, 1966).

Little information is available concerning the stability of

viruses and phages with regard to other food processes such as

fermentation, freeze-drying or irradiation. Approximately 10% of

L. plantarum phages (Nes and Sorheim, 1984) and 14% of coxsackie-

viruses (Herrmann and Cliver, 1973) survived 12 and 1 days, respec-

tively, in fermenting meat saugage products. Freeze drying or freeze

dehydration of various foods reduced the titer of poliovirus by three

to four log cycles in one study (Heidelbaugh and Giron, 1969) and two

log cycles in another (Cliver et al., 1970). However, it was noted

that polioviruses were quite stable in freeze dried foods for 15 weeks

at 50C following an initial reduction of 2 log cycles during the freeze

drying process (Cliver et al., 1970). With regard to preservation of

foods by irradiation, viruses are generally much more resistant to

gamma radiation than bacteria or endospores with the possible exception

of Micrococcus radiodurans (Rowley et al., 1978). It has been








suggested that the use of radappertization, a radiation process

designed to inactivate spores of C. botulinum by 12 log cycles, could

result in survival of viruses and resistant bacteria in meat products

(Rowley et al., 1978; Sullivan et al., 1971a, b). However, the

inactivation of these organisms is thought to be accomplished by a

pre-irradiation heat process used to inactivate autolytic enzymes

(Rowley et al., 1978).


Resistance to Disinfection


Viruses and coliphages are generally more resistant to inactiva-

tion by various environmental stresses (Bell, 1976; Fannin et al.,

1977; Hejkal, 1985; Ignazzitto et al., 1980; Kapuscinski and Mitchell,

1982) and disinfection (Burns and Sproul, 1967; Clark and Kabler, 1954;

Fannin et al., 1977; Kott et al., 1974; Liu et al., 1971; Scarpino,

1975) than are vegetative bacteria. A wide range of sensitivity to

chlorine action has been reported among different types of coliphage

(Kott et al., 1974; Scarpino, 1975; Shah and McCamish, 1972) and

viruses (Englebrecht et al., 1980; Liu et al., 1971). Some studies

indicate that coliphages f2 and MS2 are more resistant to chlorine

inactivation than polioviruses and other types of coliphage (Kott

et al., 1974; Shah and McCamish, 1972). Coxsackievirus A2 was found

to be 7 to 46 times more resistant to chlorine inactivation than

E. coli (Clark and Kabler, 1954) and most of the 20 strains of entero-

viruses tested by Liu et al. (1971) were approximately 10 times more

resistant to chlorine inactivation than enteric bacteria. On the

other hand, adenovirus type 3 has been found equivalent to E. coli

in chlorine susceptibility (Clark et al., 1956; Liu et al., 1971).








Many bacteriophages and viruses are also more resistant than enteric

bacteria to inactivation by various environmental factors. Coliphages

survive better than coliform bacteria and/or E. coli in aerosols

emitted by wastewater treatment facilities (Fannin et al., 1977), in

raw sewage and in primary and activated-sludge effluents (Bell, 1976;

Ignazzitto et al., 1980), in river water (Hejkal, 1985) and in sea

water (Kapuscinski and Mitchell, 1982). It is probable that coliphages

and viral contaminants could be quite persistent in food processing

environments subjected to inadequate disinfection and/or cleaning.

Additionally, conditions suitable for proliferation of bacterial

contaminants, e.g., E. coli, in food processing or handling areas and

equipment could result in replication of contaminating coliphages if

susceptible E. coli hosts are among those contaminants.


Basic Methodology for Coliphage
Recovery from Foods


General Overview


Few attempts have been made to enumerate indigenous coliphages or

other bacteriophages in foods and the procedures reported have not been

evaluated or refined with regard to recovery efficiency and applica-

bility for various food types (Kennedy et al., 1985; Kott and Gloyna,

1965; Whitman and Marshall, 1971a). Development of coliphage enumera-

tion methods which are simple, economical and efficient as well as

broadly applicable to various foods is necessary in order to investi-

gate the distribution of coliphages and their possible role as

indicator organisms in foods. Numerous procedures have been developed

and evaluated for quantitative recovery of animal viruses and








coliphages from the environment (Bitton et al., 1979, 1982; Seeley and

Primrose, 1982) as well as for the recovery of animal viruses from

foods (Cliver et al., 1983a, b). Many of these techniques and related

physicochemical principles provide a basis for development of methods

for the efficient recovery of coliphages from foods.

Recovery of coliphages or viruses from soils, aquatic sediments

and food materials can be divided into four general steps as follows:

(1) suspension and liquifaction of sample with elution or desorption

of viruses from the test material, (2) clarification of the resulting

sample suspension to obtain a sample preparation free of particulate

material, (3) concentration of clarified sample extract if virus levels

are too low for direct assay, and (4) testing or assay of sample

extract or concentration for viruses (Bitton et al., 1979, 1982;

Cliver et al., 1983a, b; Seeley and Primrose, 1982; Wait and Sobsey,

1983). General techniques which are used for each phase of virus

recovery from environmental or food materials and which are applicable

to expendient recovery of coliphages from foods will be discussed.


Techniques and Principles of Virus Adsorption-Elution


Viruses including bacteriophages are biocolloid particles which

are negatively charged at pH values near or above neutrality and have

amphoteric electrical properties, e.g., carboxyl- and amino-groups,

governing their sorptive behavior toward biological or non-biological

surfaces (Bitton, 1975, 1980). The initial adsorption of a virus to a

surface is a function of Brownian motion, electrostatic forces,

electrical double layer phenomena and hydrophobic forces (Bitton, 1975,

1980; Farrah et al., 1981b; Valentine and Allison, 1959). As a








result, their adsorption to or desorption from surfaces is influenced

by pH, the presence of organic compounds, the concentration and type of

salt and the presence of chaotropic agents (Bitton, 1975, 1980; Farrah

et al., 1981a, b; Mix, 1974; Valentine and Allison, 1959; Wallis

et al., 1979).

If the pH of the suspending medium is increased above the iso-

electric point of a given virus (pH 6 to 8), viruses and most solids,

e.g., soils and membrane filters, have a net negative charge and will

repel each other, resulting in enhanced desorption of viruses (Bitton,

1975, 1980; Kessick and Wagner, 1978; Mandel, 1971; Seeley and

Primrose, 1982; Sobsey and Jones, 1979). On the other hand, at lower

pH values near or below the isoelectric point of a given virus,

adsorption of viruses to a surface may be enhanced due to net opposite

charges on the viruses and solid surfaces. The presence of certain

salts, e.g., di- and trivalent cations, permits adsorption of viruses

to solids at relatively high pH values that would generally preclude

adsorption or minimize attractive forces (Bitton, 1975, 1980; Mix,

1974; Valentine and Allison, 1959; Wallis et al., 1979). This effect

has been attributed to the promotion of electrostatic interactions

between viruses and surfaces by (1) the cations acting as salt bridges

between the negative charges of the virus and solid surface (Kessick

and Wagner, 1978; Mix, 1974), (2) the cations reducing the size of the

double layer of charged ions that surround charged particles in solu-

tion (Valentine and Allison, 1959), or (3) the cations increasing the

positive charge on the filter (Kessick and Wagner, 1978).

Hydrophobic interactions between viruses and solids have recently

been shown to be at least as important as electrostatic interactions in







the sorptive association of viruses and coliphages with solids (Farrah,

1982; Farrah et al., 1981b; Shields and Farrah, 1983; Wait and Sobsey,

1983). Both hydrophobic and electrostatic interactions affect the

adsorption of viruses to solid surfaces, e.g., membrane filters, with

hydrophobic interactions dominating at higher pH values and electro-

static interactions dominating at lower pH values. Chaotropic salts,

e.g., sodium trichloroacetate, sodium thiocyanate, guanidium hydro-

chloride and sodium nitrate, as well as unionized compounds, e.g.,

Tween 80, urea and ethanol, disrupt hydrophobic interactions and

thereby promote elution of viruses from solids whereas antichaotropic

salts, e.g., magnesium sulfate, magnesium chloride and sodium sulfate,

promote hydrophobic interactions and/or adsorption of viruses to

solids.

Soluble organic compounds such as proteins generally reduce the

extent of viral adsorption to surfaces and thereby promote desorption

or elution of viruses from surfaces although these interactions may

vary with the type of organic material and the surface material

(Bitton, 1975, 1980). This effect has been hypothesized to be a result

of competition between viruses and organic molecules for specific

adsorption sites on surfaces (Bitton, 1980; Lipson and Stotzky, 1984).

In addition, Wait and Sobsey (1983) have suggested that smaller

proteins or organic molecules may intercalate between larger viruses

and adsorptive sites on surfaces and reduce the extent of attractive

forces between the virus and surface.

Based upon these observations, desorption or elution of viruses

from solid materials, e.g., membrane filters, sludge, soils and aquatic

sediments, has been effected by lowering the ionic strength of the








suspending media, addition of soluble proteins to the media, raising

the pH of the medium, addition of chaotropic compounds to the medium or

by a combination of these techniques (Bitton, 1980; Bitton et al.,

1979, 1982; Farrah et al., 1981a, b; Seeley and Primrose, 1982; Wait

and Sobsey, 1983; Zaiss, 1981). Proteinaceous materials used for

elution of viruses from solid surfaces include beef extract, casein,

non-fat dry milk, fetal calf serum, lysine, glycine and tryptose

phosphate broth. Bitton et al. (1979) found that solutions of 0.5%

isoelectric casein or 0.5% non-fat dry milk (pH 9) were more efficient

than 3% beef extract (pH 9), tryptose phosphate broth (pH 7), 10% fetal

calf serum (pH 9), muck solution (pH 9.0) or 0.25 M glycine + 0.05 M

EDTA (pH 11.0) for elution of poliovirus from soil. Eluents containing

chaotropic agents, e.g., 4 M urea + 0.05 M lysine (pH 9), 0.6 M

trichloroacetic acid + 0.02 M glycine (pH 9.0) and 1% casein + 0.1%

Tween 80 (pH 9), provided effective elution of polioviruses from fresh

water and marine sediments (Bitton et al., 1982). Eluents such as 3%

beef extract (pH 9.0), 0.5% casein (pH 9.0) and 4.0 M urea + 1.0%

lysine were effective in eluting poliovirus from sewage sludge (Farrah

et al., 1981b). Another chatropic agent of lower potency, i.e., sodium

nitrate, incorporated into 3% beef extract at pH 5.5 to 7.5 was

reported to be effective for elution of enteroviruses from highly

organic estuarine sediments (Wait and Sobsey, 1983).

A similar approach has been adopted for elution of animal viruses

from foods with suspending media consisting of buffer components,

salts, acids or bases, proteins and various combinations of these

components (Cliver et al., 1983a, b). For example, 0.08 or 0.09 M

glycine (pH 8.8) solutions have been used as a suspending and/or








eluting medium for detection of viruses in various foods, e.g.,

oysters, clams, ground beef, frankfurters, chicken salad, lettuce and

cake (Kostenbader and Cliver, 1973; Landry et al., 1980). Modified

tissue culture medium (Eagle's MEM, pH 8.5) has been used for eluting

viruses from ground beef (Sullivan et al., 1970; Tierney et al., 1973)

while a number of suspending media, e.g., 1% dry milk + 10% NaCi

(pH 9), glycine NaCI (pH 7.5) and 0.1% NaCl (pH 9.0), have been used

to elute viruses from shellfish (Metcalf et al., 1980; Sobsey et al.,

1978; Sullivan and Peeler, 1982). Neither comprehensive comparisons of

suspending media for their effectiveness in eluting viruses from foods

nor the application of many eluents, e.g., chaotropic compounds, used

in environmental techniques to virus elution from foods has been

reported.

The influence of various factors, e.g., ionic composition, pH and

chaotropic agents, on the adsorption or desorption of viruses from

various food materials has not been extensively investigated. Finance

et al. (1981) found that maximal poliovirus adsorption to food

particles generally occurred at the natural pH of the food and varied

according to the type of food and/or the food surface. For example,

between pH 4.0 and 7.0, the optimal and minimal adsorption pH was 4.0

and 6.0, respectively, for fish whereas the optimal and minimal

adsorption pH was 6.0 and 4.0, respectively, for oysters. On the other

hand, the adsorption of viruses to ground beef or mussels in suspension

was not influenced by pH between pH 4.0 and 7.0. Likewise, elution of

viruses from ground beef and oysters was similar at pH 7.5 or pH 9.0

whereas elution of viruses from fish or mussels was higher at pH 9.0

than at pH 7.5. The concentration of salt did not influence the








elution of viruses from oysters, ground beef or mussels between 0.1 and

1.6% NaCi whereas the elution of viruses from fish was directly

proportional to the concentration of NaC1.


Physical Methods of Suspension, Clarification and Concentration


Food samples must be physically suspended and/or liquified in a

suspending medium in order to achieve desorption or extraction of

viruses from food particles. This has generally been accomplished by

stirring, shaking or homogenation, e.g., blending, procedures for

varying lengths of time with care being taken to avoid heating of the

suspension (Cliver et al., 1983a, b). The appropriate liquifaction

technique may be based upon the type of food to be tested. For

example, a rinse and/or shaking technique may be applicable to fruits,

nuts or vegetables but not to animal tissues, shellfish or prepared

foods. Metcalf et al. (1980) reported that high speed homogenation was

necessary to disaggregate shellfish tissue masses for elution of animal

viruses. Acoustic energy, i.e., sonication, has been reported to

improve the efficiency of virus elution from solids in wastewater and

sludge (Hejkal et al., 1981; Wellings et al., 1976) and from tissue in

shellfish (Metcalf et al., 1980) but not from magnetite particles in

lakewater and wastewater concentrates (Bitton et al., 1981). Neither

types nor times of physical liquification techniques have been compre-

hensively evaluated or compared for optimal elution of viruses from

various foods.

Clarification of liquified food suspensions to obtain a

pariculate-free sample for virus assay has been achieved by

centrifugation, filtration, chemical methods of phase separation







and adsorption-elution (Cliver et al., 1983a, b). Centrifugation is

the most commonly used technique for clarification of crude food

suspensions and is generally expedient and economical depending upon

the type of food being tested. However, glass wool and woven fiber-

glass filtration techniques were found to be more efficient with regard

to subsequent recovery of polioviruses than low speed centrifugation

for clarification of ground beef sample-suspensions (Tierney et al.,

1973). The glass wool filtration technique developed by Larkin et al.

(1975) has been used for clarification of food sample-suspensions prior

to assay for coliphages, although the technique proved tedious and

required modification for certain types of foods (Kennedy et al.,

1984).

Concentration procedures, i.e., reduction of liquid volume from

the sample preparation without loss of viruses, are often necessary to

detect low numbers of viruses in a large volume of sample. A number of

concentration techniques, e.g., adsorption-elution from filters or

flocculants, ultracentrifugation, ultrafiltration rind two-phase polymer

separation, have been applied for environmental materials and foods

(Bitton, 1980; Cliver et al., 1983a, b; Seeley and Primrose, 1982).

Cliver et al. (1983a, p. 252) states that concentration of food sample

extracts for virus recovery is ". justified only if the cost of

concentrating is less than that of testing the entire extract volume

in cell cultures." The relatively high numbers of coliphages reported

in foods (Kennedy et al., 1984; Kott and Gloyna, 1965) as well as the

development of simple and direct coliphage assay techniques for large

volumes of sample (Kenard and Valentine, 1974; Kott and Gloyna, 1965;








Wentsel et al., 1982) seem to obviate the need for a laborious and

costly concentration step for enumeration of coliphages in foods.


Assay of Coliphages


The agar layer method (Adams, 1959) is the most widely used basic

technique for enumeration of bacteriophages in liquid samples. An MPN

technique for enumeration of coliphages in sewage, water and shellfish

also has been reported (Kott, 1966; Kott and Gloyna, 1965). It is

generally necessary to eliminate the indigenous bacterial flora in

environmental or food samples which may interfere with host bacterial

growth or with development of visible plaques on host bacterial lawns.

Pretreatment of samples with chloroform or membrane filtration as well

as incorporation of antibiotics or selective agents in the phage assay

media have been used to eliminate bacterial interference (Adams, 1959;

Gerba et al., 1978; Glass and O'Brien, 1980; Havelaar and Hogeboom,

1983; Kennedy et al., 1984, 1985; Vaughn and Metcalf, 1975). A selec-

tive assay medium, EC medium, was effective in eliminating bacterial

interference and in assaying of coliphages in food sample preparations

containing relatively high levels of indigenous bacteria (Kennedy

et al., 1984). As previously discussed, the enumeration of coliphages

in water and wastewater can be optimized with regard to maximal plaque

counts by using appropriate broad-range hosts (Havelaar and Hogeboom,

1983; Ignazzitto et al., 1980; Seeley and Primrose, 1982). Different

E. coli hosts have not been compared for recovery of coliphages from

foods.














STUDY 1
METHODOLOGY FOR ENUMERATION OF
COLIPHAGES IN FOODS


Introduction


Coliphages have received increased attention and support in recent

years as expedient and inexpensive indicators of fecal pollution and/or

enteric pathogens in water and wastewater (Kenard and Valentine, 1974;

Kott et al., 1974; Scarpino, 1978; Simkova and Cervenka, 1981; Stetler,

1984; Wentsel et al., 1982; Zaiss, 1981). Likewise, methodology for

quantitative recovery of coliphages and other viruses from the environ-

ment has been extensively investigated over the past several years

(Bitton et al., 1979, 1982; Seeley and Primrose, 1982; Wait and Sobsey,

1983). Although numerous methods for recovery of animal viruses in

various foods have been developed (Cliver et al., 1983a, b), few

attempts have been made to enumerate indigenous bacteriophages in foods

and the procedures reported were not evaluated or refined in relation

to recovery efficiency or applicability to various types of foods

(Kennedy et al., 1984; Kott and Gloyna, 1965; Whitman and Marshall,

1971a). Development of coliphage enumeration methods which are rapid,

simple, economical and efficient as well as broadly applicable to

different foods is necessary to investigate the distribution of

coliphages and their possible role as indicator organisms in foods.

The physicochemical principles related to the sorptive behavior of

viruses as well as the methodology applied to the recovery of viruses








from environmental materials (Bitton et al., 1979, 1982; Farrah et al.,

1981a, b; Wait and Sobsey, 1983) and foods (Cliver et al., 1983a, b;

Finance et al., 1981) provide a basis for development of efficient

techniques for recovery of coliphages from foods. Techniques used for

elution or desorption of viruses and phages from solid materials, e.g.,

membrane filters, soil, sludge or aquatic sediments, generally involve

lowering the ionic strength of the eluent or sample suspension, addi-

tion of soluble proteins or other organic compounds to the suspension,

raising the pH of the suspension, adding chaotropic agents to the

suspension or various combinations of these techniques (Bitton, 1980;

Bitton et al., 1979, 1982; Farrah, 1982; Farrah and Bitton, 1978, 1979;

Farrah et al., 1978, 1981a, b; Seeley and Primrose, 1982; Shields and

Farrah, 1983; Wait and Sobsey, 1983; Zaiss, 1981). Similar approaches

have been used for desorption and/or recovery of animal viruses from

foods with eluents and/or suspending media consisting of various con-

centrations of salts, acids or bases, buffers, proteins and various

combinations of these components (Cliver et al., 1983a, b). Little

information is available concerning the comparative effects of eluent

composition and pH of suspending medium on the elution of viruses from

foods (Finance et al., 1981) with no information available concerning

the effects of chaotropic agents on the recovery of viruses from foods.

A number of methods, e.g., shaking, stirring and blending or

homogenization, have been used for physical suspension or liquifaction

of food samples in order to release animal viruses from foods (Cliver

et al., 1983a, b; Metcalf et al., 1980) although those methods have

not been standardized or extensively evaluated for broad application

to various foods. Several techniques for clarification of food








sample-suspensions prior to virological assay, e.g., centrifugation,

filtration, chemical phase-separation and adsorption-elution, have been

reported (Cliver et al., 1983a, b). Tierney et al. (1973) found

the glass wool or woven fiber glass filtration technique to be more

efficient than other techniques, including centrifugation, for clari-

fication of ground beef suspensions and subsequent recovery of

polioviruses. A modification of the glass wool filtration technique

(Larkin et al., 1975; Tierney et al., 1973) was used to clarify various

food suspensions for subsequent assay of coliphages, although the

technique was time consuming for many samples (Kennedy et al., 1984).

The objectives of this study were to develop and evaluate pro-

cedures for practical and efficient recovery of coliphages from various

types of foods. The effects of eluent composition, eluent or sample-

suspension pH and incorporation of chaotropic agents into eluents as

well as methods of sample suspension and clarification on the recovery

of T2, MS2 and indigenous coliphages from various foods was

investigated.


Materials and Methods


Preparation of T2 and MS2 Coliphage Stocks


All bacteriological media and reagents were obtained from Difco

Laboratories (Detroit, MI) and Fisher Scientific Co. (Springfield, NJ),

respectively, unless otherwise indicated. Stock cultures of T2 and MS2

were supplied by S. R. Farrah (Microbiology and Cell Science Depart-

ment, University of Florida). A fresh stock culture of each phage was

prepared for use in all subsequent studies. The original stock culture








suspensions, in Tryptic Soy Broth (TSB), were serially diluted in

sterile TSB and plated out with appropriate E. coli hosts, i.e.,

E. coli B (ATCC 11303) for T2 and E. coli C-3000 (ATCC 15597) for MS2,

using the agar layer method (Adams, 1959) with TSB as a basal medium.

Plates were incubated at 370C for 18 h and the soft agar-layer of

plates having approximately 99% confluent lysis was scraped off and

mixed with 1 ml of chloroform. The resultant mixture was mixed with

10 ml of sterile TSB. The phage agar-layer chloroform TSB mixture was

centrifuged (Sorvall centrifuge, DuPont, Co., Wilmington, DL) for

5 min in a sterile glass centrifuge tube (50 ml) at a relatively low

speed (2000-3000 X g) using an SS-34 rotor and the chloroform-free

supernatant was recovered and centrifuged at a higher speed (12000 X g)

for 10 min to separate any remaining bacterial and agar debris from the

phage-TSB suspension. The phage-TSB suspensions were transferred to

sterile 18 x 150 mm screw cap tubes and stored at 1-20C. The titer of

these stock suspensions of T2 and MS2 were approximately 10 plaque

forming units (PFU) per ml.


Samples and Sample Preparation


General


Samples, i.e., fresh chicken breasts, fresh ground beef, fresh

pork sausage, canned corned beef and frozen mixed vegetables, were

obtained at retail markets in the Gainesville, FL, area. Samples were

transported to the laboratory in an insulated container containing ice

and were held at 1-20C in the laboratory until time of analysis (18-36

h). For a given experimental trial, a homogenous sample lot (1-2 kg)








of a particular food was prepared by mixing and/or finely chopping with

aseptic technique. A sample lot was divided into an appropriate number

(16-30) of subsamples (50-100g) of uniform composition for triplicate

treatment comparisons. Frozen mixed vegetables were removed form

cartons, spread on sterile aluminum foil sheets and thawed at 50C for

4-8 h prior to mixing and dividing into subsamples. Subsamples were

held in sterile Whirlpak bags (Fisher Scientific) at 1-20C until time

of analysis. An outline of the general methods protocol is presented

in Figure 1.


Preparation of inoculated samples


For a given experiment and/or sample lot, inocula of T2 and MS2

phages were prepared by taking a 0.1 ml aliquot of the stock phage

suspension (ca. 1011 PFU/ml) and serially diluting the suspension in

sterile TSB broth. MS2 and T2 suspensions were diluted by a factor of

10- and 10-6, respectively. The titers of the resulting inocula were

approximately 107 and 105 PFU/ml for MS2 and T2, respectively. A given

subsample (50g), prepared as previously described, was inoculated drop-

wise with 0.5 ml of the final inoculum-suspension of each phage. Thus,

the rate of inoculation was 1.0 ml phage suspension per lOOg of sample

(ca. 107 MS2 and 105 T2 per lOOg). After inoculation, each subsample

was aseptically mixed by hand and held in sterile Whirlpak bags at

1-20C for 4-6 h prior to analysis. For each experiment, phage inocula

and inoculated samples were assayed as described in another section.











SAMPLE



preparation of sample lots and subsamples


inoculation with T2 and MS2 coliphages as indicated



SAMPLE SUSPENSION



50g subsample + lOOg eluent (various compositions, pH values)


blending, shaking or stomaching



SAMPLE CLARIFICATION



--glass wool filtration, polypropylene mesh filtration or

centrifugation



COLIPHAGE ASSAY



agar layer method with appropriate E. coli host

plaque enumeration


Figure 1. General protocol for enumeration of coliphages in foods.








Preparation of Sample Suspensions


Subsamples were diluted 1:2 (wt:wt) in a given eluent or suspend-

ing medium. Thus, 50g of sample would be diluted with lOOg of eluent

or lOOg of sample with 200g of eluent. This dilution factor allowed

minimal dilution of indigenous coliphages for maximal detection

sensitivity in subsequent assays and provided a suspension from which

an adequate volume and degree of clarification could be obtained with

various food types. The 1:1 (wt:vol) dilution recommended in a

virological procedure for ground beef (Larkin et al., 1975) was not

suitable with regard to the aforementioned considerations.

After subsamples were diluted with a given eluent, they were

suspended or liquified by one of three of the following general

techniques: (1) blending at 2000 or 4,500 RPM in sterile glass

containers (500 ml or 1200 ml capacity as deemed appropriate) using

a model 5010 laring blender (Waring Products Division, New Hartford,

CT) attached to a type 3PN116B Powerstat Variable Autotransformer

(Superior Electric Co., Bristol, CT) for controlling blending speed,

(2) stomaching in sterile plastic bags using a Coleworth Stomacher

400 (Dynatech Laboratories, Inc., Alexandria, VA), or (3) shaking

in sterile plastic bags on a Model G2 rotary Shaker (New Brunswick

Scientific Co., Inc., New Brunswick, NJ) at 300 RPM. Samples suspended

by shaking generally required some kneading to break up clumps of

sample prior to shaking; all tissue samples were finely chopped as

previous mentioned.

For each suspending technique, various times of sample processing

or suspending were compared for recovery of indigenous coliphages from








chicken as well as for elution of T2 and MS2 from chicken, ground beef,

corned beef and mixed vegetables with blending and shaking techniques.

Additionally, relatively slow (ca. 2000-2500 RPM) and fast (ca. 4000-

4500 RPM) blending speeds were compared for their effect on release

or elution of T2, MS2 and indigenous coliphages from chicken as well

as elution of T2 and MS2 from ground beef, corned beef and mixed

vegetables. The lower blending speed resulted in a sample suspension

which was compatible with the rapid, polypropylene-mesh clarification

technique (described in another section) which afforded rapid compari-

sons of large numbers of subsamples in a single work day. The higher

blending speed resulted in a greater degree of sample fragmentation

and/or homogenization in the sample-suspensions; adequate volumes of

clarified suspension for many samples could be obtained only by a

centrifugation technique (described in another section). The effect

cf sonication for 2 min at 125 watts (Model B-220 Ultrasonic Cleaner,

Branson CLeaning Equipment Co., Shelton, CT) on the release of

indigenous coliphages from suspensions of fresh chicken prepared by

blending (2000 RPM, 10 min), shaking (300 RPM, 30 min) and stomaching

(10 min) was also examined. For studies comparing different methods

and/or times of sample processing, EC medium was used as an eluent.


Clarification of Sample Suspensions


The following general techniques were evaluated for clarification

of sample-suspensions in relation to speed, application and subsequent

recovery of coliphages from the suspensions: (1) filtration of sample

suspensions through 5 g sterile Pyrex glass wool (Fisher Scientific

Co.), which was fitted into a large polypropylene funnel (Top








diameter = 150 mm, stem length = 30 mm, stem O.D. = 28 mm) and pre-

treated with 30 ml of sterile EC medium to minimize possible adsorption

of coliphages to the filter itself; methods were similar to those

previously described (Kennedy et al., 1984; Larkin et al., 1975),

(2) rapid filtration of sample suspension by drawing portions of the

suspension through a wide mouth 10 or 11 ml pipet fitted with a

sterile, "disposable filter tip" consisting of a 5 cm piece of latex

tubing (5/16 in. x 3/32 in., Fisher Scientific Co.) capped with

500 pm polypropylene mesh (Spectramesh, Spectrum Medical Industries,

Los Angeles, CA); construction of the "Filter tips" was similar to that

described by Peterkin and Sharpe (1981) except the mesh was affixed to

the tubing with nickel-chromium wire (B&S gauge 24; Fisher Scientific,

Co.) rather than epoxy, and (3) centrifugation (5 min) of sample-

suspensions in 250 ml capacity sterile, polycarbonate centrifuge

bottles (Nalge/Syborn Co., Rochester, NY) at approximately 3000 X g in

a Sorvall centrifuge (DuPont Co., Wilmington, DL) using a GSA rotor.

Funnels fitted with glass wool filters were individually placed

into sterile 250 ml graduated cylinders and the suspensions allowed to

drain. After 20 min, the food slurry and/or suspension left on the

filter were compressed using sterile tongue depressors as described by

Larkin et al. (1975) to express remaining liquid from the filter and

slurry. With most blended sample-suspensions, e.g., fresh chicken,

ground beef or pork sausage, the glass wool filters became rapidly

blocked and expression of remaining liquid became difficult. This

difficulty was also observed in a previous study (Kennedy et al.,

1984). The clarified suspension was also diluted by the EC medium (30

ml) used to treat the glass wool filter. Other than for comparative







studies with the other clarification techniques, glass wool filtra-

tion was not used in any other comparisons due to its inherent

impracticality.

Both 250 and 500 Um polypropylene mesh sizes were evaluated for

ease of application with various sample-suspensions and for recovery

of coliphages (data not presented). Coliphage recoveries were not

affected by mesh size. The 500 pm mesh allowed more rapid filtration

of samples suspensions and became blocked less easily than the 250 Pm

mesh while still providing adequate clarification of suspension for the

coliphage assay. Centrifugation at approximately 3000 X g (5 min) was

required for adequate sedimentation of sample debris and/or clarifica-

tion of various sample-suspensions. The relative effects of higher

centrifuge speeds (up to 12,000 X g) and longer times of centrifuging

(10 min) on the recovery of coliphages was also examined. No differ-

ences in coliphage recovery were noted between suspensions centri-

fuged at 3000 X g or 12,000 X g for 5 or 10 min (data not presented).

Either the polypropylene mesh (500 pm) or the centrifugation (3000 X g,

5 min) techniques were used in various comparisons as considered

applicable and expedient.


Suspending Media and/or Eluents


For studies evaluating basal eluents, eluent composition and

eluent pH on the recovery of coliphages from foods, sample-suspensions

were prepared by blending (2000 or 4500 RPM, 5 or 10 min) and clarified

by polypropylene mesh (500 4m) or centrifugation (3000 X g, 5 min) as

indicated.








Several basal eluents were compared for desorption and/or

recovery of indigenous coliphages from fresh chicken. These included

Butterfield's phosphate buffer (APHA, 1984) at pH 7.5, 0.1% peptone at

pH 7.0, 3% beef extract at pH 7.3, 1% casein (Sigma Chemical Co.,

St. Louis, MO), at pH 7.2, 0.1 M lysine (Sigma Chemical Co.) at pH 7.2,

Tryptose Phosphate (TP) broth at pH 7.2, EC medium at pH 6.9 and

MacConkey broth at pH 7.3. EC medium, 1% casein, TP broth, 0.1 M

lysine, and Butterfield's phosphate buffer also were used in various

other comparisons. Eluents were sterilized by autoclaving at 1210C for

15 min. The pH of various eluents was adjusted in some studies to

achieve a particular pH value in the resulting sample-suspension. The

pH of eluents and/or sample suspensions was adjusted with 1.0 N NaOH or

1.0 N HC1 as indicated. A flat surface combination electrode (Fisher

Scientific Co.) and a Model 130 Corning pH Meter (Corning Glass Works,

Medfield, MA) were used for pH measurements of eluents as well as agar

media surfaces.

Two means of adjusting the pH of sample-suspensions were initially

evaluated for use in subsequent experiments. The pH could be adjusted

directly by addition of 1 N NaOH to the sample-suspension as the

suspension was being blended or stirred. This technique was tedious

and time consuming in that equilibration of suspension took up to 5 min

and the electrode required extensive cleaning and/or sanitizing after

each suspension adjustment to prevent cross contamination. Another

technique involved preadjusting the pH of the eluent to obtain a

specific pH level in the resulting sample-suspension (ranging from pH

6.0 to 10.0). The pre-set pH values for eluents was determined for a

given food type by preparing a minimal number of sample suspensions







with eluent at different pH values and plotting the subsequent pH

values of sample-suspensions (after blending) against the pH value of

the eluent. Thus, a standard curve for a given product was used to

estimate the pH required in an eluent required to obtain a desired

pH in the resulting sample-suspension. Representative values are

presented in Table A-I (Appendix). These relationships were checked

for accuracy for every sample lot and/or comparative study and

adjustments made if necessary. Adjustment of sample suspension pH

by the latter technique resulted in consistently higher recoveries

of indigenous coliphages than the direct adjustment technique in

preliminary investigations (data not presented). The latter technique

was, therefore, used for studies concerning the effects of sample-

suspension pH on coliphage elution from foods. The pH of phage assay

plates prepared from sample-suspensions of various pH levels was also

determined and found to vary by less than 0.2 pH units due to pH of

eluent.

Various concentrations of chaotropic salts of strong, e.g., sodium

trichloroacetate (NaTCA), or weak potency, e.g., sodium chloride (NaC1)

and sodium nitrate (NaNO3), as well as other chaotropic agents such as

urea or detergents, e.g., Tween 80 and Triton X-100 (Eastman Kodak Co.,

Rochester, NY) were incorporated into various eluents to investigate

their effect on the desorption and/or recovery of coliphages from

foods. Bile salts number 3 (Difco Laboratories) also were incorporated

into a basal eluent at a concentration of 0.15% (equivalent to that in

EC medium) for some comparative studies. These compounds generally were

added to the eluents after sterilization in order to avoid possible

degradation of the compounds or interactions with other components.








Background contamination with regard to coliphages or bacteria as a

result of preparing the eluent formulations was monitored and found to

be absent, i.e., below detection levels. Control phage-assay plates

containing corresponding concentrations of these compounds were pre-

pared with sample suspensions of basal eluents in order to determine

the effect of these compounds on the growth of the host bacteria and

the ability of the coliphages to produce plaques on resulting host

lawns.


Coliphage Assays


Host bacteria cultures


E. coli B (ATCC 11303), E. coli C-3000 (ATCC 15597) and E. coli C

(ATCC 13706) were used as host bacteria for the assay of T2, MS2 and

indigenous coliphages, respectively. Host cultures were maintained in

TSB containing 10% (v/v) glycerin at -20C following the general pro-

cedures described by ASTM (1983) and Havelaar and Hogeboom (1983).

However, for a given set of analyses, the thawed host cultures (TSB,

10% glycerin) were not used directly for preparation of coliphage assay

plates but were used to inoculate appropriate volumes of sterile EC

medium (1 ml of TSB-glycerin culture per 10 ml of EC medium) for

production of host inocula. The EC medium cultures were incubated at

35C for approximately 6 h for use in coliphage assays. The host

cultures were found to grow more rapidly in the EC based assay medium

if grown in this manner than by direct inoculation with the TSB-

glycerin cultures. Recoveries of coliphages by both methods of host

inoculation were similar (data not presented). Additionally, the large








volumes of host culture used for these studies would have required an

impractical amount of freezer space and glassware using direct inocula-

tion with thawed cultures.


Coliphage assay medium


EC medium was used for preparation of all coliphage assay media

basically as described in previous studies (Kennedy et al., 1984,

1985). EC medium with 1.5% agar was used for preparation of bottom-

layer plates (approximately 10 ml per plate) in 100 x 15 mm plastic

petri dishes (Fisher Scientific Co.). These plates were dried for 24-

48 h at 250C before use. For sample aliquots of 1 or 2 ml, overlay

medium was prepared with EC medium adjusted to 0.75% agar; this medium

was boiled and dispensed (3 ml/tube) into 18 x 125 mm screw cap tubes

and autoclaved at 1210C for 10 min. For sample aliquots of 4 ml,

overlay medium was adjusted to 1.5% agar, dispensed (1.5 ml/tube), and

prepared as indicated. Thus, the agar content of final overlays,

including sample aliquots, overlay medium and host cultures (0.3 ml)

was 0.4 to 0.5% agar; total volumes of overlay ranged from 4.3 ml (1 ml

sample aliquot) to 5.8 ml (4 ml sample aliquot). Prior to analyses,

the overlay medium was heated to liquification by either steaming (ca.

105C, 5 min) or placing tubes in a boiling water bath (10-15 min).

The overlay medium was then maintained at 450C 0.20C in a Blue M

Magniwhirl water bath (Blue M Electric Co., Blue Island, IL).

Where sample-suspensions were prepared with eluents other than EC

medium, the composition of the overlay medium was adjusted with bile

salts (bile salts No. 3, Difco Laboratories), NaCl and Tryptose so that








the composition and selective activity of the final overlay mixtures

were similar regardless of the eluent used in a given study.


Coliphage assay procedures


Clarified sample-suspensions were assayed for coliphages using the

basic agar layer method of Adams (1959) with modifications as described

previously (Kennedy et al., 1984). For indigenous coliphage assays

clarified sample-suspensions were assayed without dilution whereas

sample-suspension were serially diluted in sterile EC medium by

transferring 1.0 ml aliquots into 9.0 ml sterile EC medium blanks for

T2 and MS2 assays. For recovery of indigenous coliphages from chicken,

2 ml of sample-suspension were mixed by hand for ca. 10 sec with 0.3 ml

of the E. coli C host culture in the tube containing 3 ml (0.75% agar)

overlay medium (450C) and overlaid on the basal agar plate; five plates

were prepared for each subsample and/or treatment replicate. The pro-

cedure was identical for recovery of indigenous coliphages from other

foods except that 4.0 ml of suspension were prepared with 1.5 ml (1.5%

agar) overlay medium in order to obtain increased sensitivity. The

sensitivity obtained with 5 x 4.0 ml and 5 x 2.0 ml sample aliquots was

10 and 20 coliphages per lOOg of food, respectively. For determina-

tions of T2 and MS2 in inoculated samples, 1 ml aliquots of the 10l

and 10-3 dilutions for T2 and MS2, respectively, were prepared and

plated in triplicate as indicated. Since MS2 does not produce plaques

on E. coli B used for T2, and, T2 was sufficiently diluted to eliminate

interference with MS2 on E. coli C-3000 at the higher dilution, both of

these coliphages could be inoculated and recovered from the same sample

using this design. For each experiment, the T2 and MS2 inocula used








for preparing the subsamples were held at 1-20C and assayed at the time

of sample analysis by serially diluting in EC medium and plating as

indicated. In addition, respective uninoculated samples were assayed

with E. coli B and C-3000 to monitor indigenous coliphages. All assay

plates were allowed to solidify (ca. 30 min, 250C), inverted and incu-

bated at 350C for 16-18 h. Plaque forming units (PFU) were counted

using a Quebec Colony Counter Model 3325 (American Optical Co.,

Buffalo, NY) or using oblique lighting.

Since all subsamples were diluted 1:2 (wt:wt) in eluents, e.g.,

50g sample + 100g eluent, counts of indigenous coliphage (PFU) per lO0g

food sample were calculated by multiplying the average PFU/ml of

sample-suspension by 200. The levels of T2 or MS2 could be calculated

in this fashion although it was more convenient to calculate their

percentage recovery for various comparisons by dividing the average

PFU/ml of sample-suspension by the calculated average PFU/ml of T2 or

MS2 in the inoculum and corrected for dilution by the eluent used to

prepare the sample-suspension. For example, if 0.5 ml of a 105 PFU/ml

inocula of T2 was used to incoulate 50g of sample which was mixed with

lOOg (or ml) of eluent, the maximal titer of T2 in the eluent would be

(105/mi) x (0.5 ml) (100 ml) = 5 x 102/ml at 100% recovery. Thus,

coliphage counts of samples were calculated upon the basis of eluent

volume rather than the total weight of the sample-suspension as is done

with most bacteriological analyses of foods (APHA, 1984). The accuracy

of these calculations was confirmed empirically (data not presented).

For example, calculations of MS2 percentage recovery based upon total

weight of suspensions (chicken, ground beef) ranged from 135 to 152%

whereas those based upon eluent volume ranged from 90 to 101%.








Likewise, these calculations were confirmed for determining the level

of indigenous coliphages in chicken by comparing the calculated PFU/

total-eluent-volume as previously described to the sum of coliphages

assayed in the supernatants of repeatedly resuspended and centrifuged

sample pellets (data not presented).


Statistical Analyses


Indigenous coliphage counts were converted to PFU per 100g of

sample and T2 and MS2 counts were converted to percentage recoveries

for subsequent statistical comparisons. Data were analyzed using

several procedures including the paired t-test, a one-way analysis of

variance or a factorial design, as deemed appropriate for a given

experiment or comparison (Montgomery, 1984). Means of coliphage counts

were compared using Duncan's multiple range test when indicated.


Results


Sample Clarification Techniques


Previous observations in this laboratory (Kennedy et al., 1984)

indicated that the glass wool filtration technique for clarification of

blended sample-suspensions was too tedious for routine analysis of

large numbers of samples and was not effective for rapid clarification

of the relatively large volumes of sample-suspension required for many

comparisons in the present studies (see Materials and Methods).

Therefore, more rapid and practical procedures for clarification of

various sample-suspensions were investigated.








The polypropylene mesh filtration technique, previously

described for clarification of relatively small volumes of food

sample-suspensions (Kennedy et al., 1985; Peterkin and Sharpe, 1981),

was evaluated for clarification of 10 to 20 ml volumes of sample-

suspensions for subsequent coliphage assays (see Materials and

Methods). Rapid clarification of 20 to 25 ml of many sample-

suspensions could be achieved using the 500 pm polypropylene-mesh,

although sample-suspensions prepared by blending speeds in excess of

approximately 4000 to 5000 RPM could not be clarified by this technique

due to the viscosity of the suspension. In addition, clarification of

volumes greater than 20-25 ml of sample-suspension prepared by any

method required two or more sterile mesh "filter-tips" and, thus, was

impractical and time consuming for some types of comparisons and/or

samples. On the other hand, larger volumes of clarified sample-

suspension could readily be obtained by centrifugation at relatively

low speeds (3000 X g) for 5 min (see Materials and Methods). Excluding

material preparation, total time required for clarification of a

chicken or pork sausage sample-suspension was approximately 2, 10 and

30 min for the polypropylene mesh, centrifugation and glass wool

clarification methods, respectively.

The polypropylene mesh, centrifugation and glass wool clarifica-

tion techniques were compared for recovery of indigenous coliphages

from fresh chicken and pork sausage (Table 4). The highest and lowest

numbers of coliphages per ml of clarified sample suspension were found

with centrifugation and glass wool clarification, respectively.

The mean recoveries of coliphage from suspensions of chicken were

significantly higher (p < 0.05) when clarified by mesh filtration or








Table 4. Comparison of methods for clarification of blended sample
suspensions with regard to recovery of indigenous coliphages
using EC medium as an eluent.1

2
Coliphage Recovery


Method Sample Suspensions Food Samples4


Chicken Pork Chicken Pork


(PFU/ml) (103 PFU/100 g)


Glass Wool 14.0b 10.0a 3.56b 2.71a
a a a a
Polypropylene Mesh 18.5 12.8 3.96 2.41
a a a a
Centrifugation 20.1 15.4 4.03 3.08



1Sample suspensions prepared by blending (4500 RPM, 10 min).

Values represent means of three subsamples (lOOg each). Within each
column, means having the same superscripts were not significantly
different (p > 0.05).
3
Glass wool = filtration through 5g glass wool; polypropylene mesh =
filtration through 500 pm polypropylene mesh; centrifugation =
3000 X g, 5 min.

For the glass wool method, the volume of EC eluent used to treat the
filter was included as part of the total volume for calculation of
PFU/1OOg of food.








centrifugation than by glass wool filtration; differences were not

significant (p > 0.05) with pork sausage suspensions. Although sample-

suspensions clarified by the glass wool technique were diluted to some

extent by the EC medium used for pretreatment, these data indicated

that the glass wool technique results in less sensitivity in terms of

coliphage detection than centrifugation or mesh filtration. When the

values for PFU per ml of sample-suspension were converted to PFU per

lOOg of sample, similar results were obtained for chicken and pork

samples. For the glass wool technique, the dilution factor (1.15:1)

associated with EC medium used for filter pretreatment was incorporated

into the calculations for coliphage per lOOg of sample.

Since no significant differences (p > 0.05) were noted between the

mesh filtration and centrifugation techniques in terms of coliphage

recovery, either of these techniques were used for clarification of

sample-suspensions in other comparisons as considered rapid and applic-

able for a given food sample and/or experiment.


Sample Suspension Techniques


Based upon methods reported for sample liquifaction or suspension

for recovery of coliphages (Kennedy et al., 1984) or animal viruses

(Cliver et al., 1983a, b) from foods and the types of equipment

generally available for microbiological analyses of foods (APHA, 1984),

procedures involving blending, shaking and stomaching were evaluated

for recovery of coliphages from foods. The effect of time of sample-

suspension was determined for blending (ca. 2000 RPM), stomaching and

shaking (ca. 300 RPM) for the recovery of indigenous coliphages from

fresh chicken (Table 5). For samples suspended by blending (study 1),








Table 5.


Effect of sample processing time on recovery of indigenous
coliphages from fresh chicken using EC medium as an
eluent. 1,2


Coliphage Recovery

Processing
Time Blending (2000 RPM)
Stomaching Shaking
(300 RPM)
Study 1 Study 2


min ------------------- 10 3 PFU/100g -------------------


1 1.37b --- 3.12 --a

2 1.43 b -- 3.14

5 1.31b 9.95a 3.46a 1.99c

10 2.66 a 11.03a 3.86 a 2.58bc

15 --- 10.01a ---

20 1.48 b 11.49 a 3.12ab

a
30 --- 3.35


Values
having
0.05).


represent means of three subsamples. Within each column, means
the same superscript(s) were not significantly different (p >


Sample suspensions in blending time comparisons were clarified by
polypropylene mesh filtration whereas those in stomaching and shaking
comparisons were clarified by centrifugation (3000 X g, 5 min).








blending for 10 min resulted in a significantly higher recovery (p <

0.05) of coliphages than blending for 1, 2, 5 or 20 min. In another

study (study 2), however, there were no significant differences (p >

0.05) noted as a result of blending times of 5, 10, 15 or 20 min (Table

5). Recovery of indigenous coliphages from chicken was generally

greater with increasing time of stomaching or shaking. There were no

significant differences (p > 0.05) in coliphage recovery between 1, 2,

5 or 10 min of stomaching, although the mean recovery was highest with

10 min stomaching. For subsamples suspended by shaking (300 RPM),

significantly higher (p < 0.05) recovery of coliphages were observed

after 30 min than after 5 or 10 min; there was no significant differ-

ence (p > 0.05) in recovery between samples shaken for 20 min or 30 min

or between those shaken for 20 min or 10 min.

Based upon the preceding studies (Table 5), variations of the

three methods, i.e., blending, stomaching and shaking, of sample-

suspensions were compared for the recovery of indigenous coliphages

from fresh chicken as well as the recovery of T2 and MS2 coliphages

from chicken, ground beef, corned beef and mixed vegetables (Tables 6

and 7). Since faster blending speeds may be indicated for disaggrega-

tion of some food samples, i.e., meat tissues or prepared foods, a

blending speed of approximately 4500 RPM was also included in these

comparisons.

For recovery of indigenous coliphages from chicken, blending at

4500 RPM (10 or 20 min) was generally superior to blending at 2000 RPM

(10 or 20 min); this difference was significant (p < 0.05) in study 1

but not in study 2 (Table 6). Highest recoveries of indigenous

coliphages were noted for sample-suspensions prepared by shaking








Table 6. Comparison of sample processing methods for recovery of T2,
MS2 and indigenous coliphages from fresh chicken using EC
medium as an eluent.1,2


Indigenous
Time Coliphages
Method3 of T2 MS2
Processing
Study 1 Study 2


--- min -- -- 103 PFU/1OOg -- % recovery -


Blending 10 3.24c 4.52a 79.8a 70.7a
(slow) 20 2.92 5.30 74.2 73.9


Blending 10 4.52b 5.41 a 62.6a 78.1
(fast) 20 --- 5.64 --- ---

a a a
Shaking 20 --- 5.33 83.3a 74.2a
(300 RPM) 30 5.30 5.73 74.4 69.8


Stomaching 10 4.12b 5.20a 70.2a 70.2a



IValues represent means of three subsamples. Within each column,
means having the same superscript(s) were not significantly different
(p > 0.05).

2Sample suspensions clarified by centrifugation (3000 X g, 5 min).

3Blending (slow) = 2000 RPM; Blending (fast) = 4500 RPM.








(30 min) with significantly higher (p < 0.05) recoveries observed in

study 1. There were no significant differences (p > 0.05) between time

of processing for any suspension technique for recovery of indigenous

coliphages and no significant differences (p > 0.05) among all suspen-

sion techniques in study 2. It is important to qualify these results

with regard to the shaking procedure as applied in these studies, in

that the chicken sample lots were finely chopped before analysis of

subsamples. This was a tedious but necessary treatment required to

obtain relatively homogenous sample lots for the experimental compari-

sons but may have negated some inherent advantages associated with

blending or stomaching with regard to disaggregation of the sample

tissues. For routine analysis of various foods, the samples would

probably not be subjected to this degree of chopping, thereby limiting

the efficacy of a shaking technique.

There were few consistent or significant differences (p < 0.05)

between various sample suspension techniques for percentage recovery of

T2 or MS2 coliphages from incoculated chicken, ground beef, corned beef

or mixed vegetables (Tables 6 and 7). Mean recoveries of T2 were

generally higher with shaking or slow blending techniques than with

fast blending or stomaching techniques. However, significant differ-

ences (p < 0.05) were noted only for ground beef with shaking (20 or 30

min) and slow blending (10 min) resulting in significantly higher

(p < 0.05) recoveries of T2 than fast blending (10 min) or stomaching

(10 min) (Table 7). On the other hand, higher recoveries of MS2 from

chicken and ground beef were observed with fast blending (10 min)

than shaking or slow blending methods although differences were

not significant (p > 0.05). Moreover, there were no significant
















0





0
*-,
4-4
U)
(1)
CO
















co
r-4
0
u
+14
C/)



cu


























0
0..
o






0













0 o
S)
0
'M
CO)

0



ca





u
0U)






Ul






-H 1

0







4-4
-o


UN


c






CO














0
4-'















0 C0
0






(U
CO



V
0
a)











o N
C-E







N

) )











aU














a4









-


00 00 *H~~
*H *H -" C
*t3 0 4-j r-I Co
S0 C -d 0 E
Q) CU Co O 0 0
r-4 In 4-4 = n 41
cc %- '1= CU) '-, M


I
I0 CO CO









S CO1\00 -I
I 00 00






I oo


>1 '. cli rN
>

u



0clc co
c 00 '0

I I1- c






I .
I 00 0

I 00 0











I







I 0
C1O 0
1I- 00 00



oo *




I O
C W 0
*H 0 0




1 O


c c0 ca
E) .-1 00
*
-NO 0


rca M- co




c40 CO C
H0" N




'O 0













'00 M
*










Co co








NO -




mm .
c1 co ca
*















N-4 N
a'0 N
CT- rn co


m C1 n













00 c0














c0 C0 C
"O 0 -4
M *


U(







CO










0
U


r)




4-4
Q 4-4

a)



-4

E
c

DCO
EU

u




*-H
0
O

r- c
U"0
CU

a()







U) E
a) $4
a o
-4-


r-4-

C



Ex
0 *
0
0 0
CO 0
N -

o 11
v m



4-1
CO 4-II
















co
0 CU











3 0
0 -

-4 C
c o


DO


40






0 a








differences (p > 0.05) between all sample-suspension techniques for

recovery of MS2 from any food tested. Trends for recoveries of MS2

from chicken and ground beef were similar to those of indigenous

coliphages from chicken with regard to method of sample processing.

Overall recovery rates for MS2 and T2 were similar for most foods

although different trends in recovery as related to suspension

techniques were observed as previously discussed. Recoveries of T2 or

MS2 were generally higher from inoculated chicken, ground beef and

mixed vegetables than from corned beef. Optimal recoveries for T2

ranged from 70% (shaking, 30 min) for corned beef to 93% (shaking, 20

min) for ground beef whereas those for MS2 ranged from 65% (shaking,

30 min) for corned beef to 89% (slow blending, 10 min) for mixed

vegetables.

Since sonication has been found effective for enhancing the

release or elution of enteroviruses from suspensions of shellfish

(Metcalf et al., 1980) and sewage-sludge (Wellings et al., 1976), the

effect of sonication on the elution of indigenous coliphages from

chicken-suspensions prepared by slow blending (10 min), shaking (30

min) or stomaching (10 min) was investigated (Table A-2, Appendix).

Sonication of sample-suspensions for 2 min at 125 Watts had no

significant effect (p > 0.05) on the recovery of indigenous coliphages

from chicken-suspensions. Longer sonication times were not evaluated

due to excessive heating of sample-suspensions and increased overall

time of analysis.








Eluent Composition and pH


Basal eluents


Several solutions of varying complexity with regard to content of

protein or protein-type compounds, ionic strength, buffering capacity

and selective detergents, e.g., bile salts, were compared for elution

and/or recovery of T2, MS2 and indigenous coliphages from fresh chicken

(Table 8). In study 1, the pH of resulting sample-suspensions varied

from approximately 6.0 (0.1 M lysine, Butterfield's phosphate buffer,

and 0.1% peptone buffer) to 6.5 (TP broth, EC medium and MacConkey

broth). For study 2 and inoculated-sample studies, sample-suspensions

were adjusted to a common pH, i.e., 6.3 0.1, in order to minimize any

possible pH effects.

The recovery of indigenous coliphages generally coincided with

increasing protein content of the eluent (Table 8). For example, EC

medium, TP broth, MacConkey broth and 1% casein eluents resulted in

significantly higher (p < 0.05) recoveries of indigenous coliphages

from chicken than Butterfield's phosphate buffer or 0.1% peptone

buffer in study 1. Butterfield's buffer and 0.1% peptone buffer were

evaluated because of their common usage in bacteriological analyses

of foods (APHA, 1984). Likewise, significantly higher (p < 0.05)

recoveries of indigenous coliphages were obtained with EC medium, 1%

casein or 0.1 M lysine eluents than with Butterfield's buffer (study

2). The highest mean recovery of T2 from fresh chicken was obtained

with EC medium eluent; recoveries were significantly higher (p < 0.05)

with EC medium than with Butterfield's buffer or 0.1 M lysine. On the

other hand, there were no significant differences (p > 0.05) between








Table 8. Comparison of various eluents for the recovery of T2, MS2 and
indigenous coliphages from fresh chicken.1,2,3,4


Indigenous
Coliphages
Eluents T2 MS2

Study 1 Study 2


-- 103 PFU/100g -- -- % recovery --


Butterfield's buffer 0.69d 3.35b 31.8b 90.5a

0.1M Lysine 1.28abc 8.03 a 40.7 b 88.5 a

1% Casein 1.52 a 10.13 a 53.1 ab 85.0 a

EC medium 1.73a 9.75a 78.0a 88.5a

0.1% Peptone 0.93cd

3% Beef extract 1.04bcd

Tryptose phosphate broth 1.59a --

MacConkey broth 1.42ab


Values represent means of three subsamples. Within
having the same superscripts were not significantly
0.05).


each column, means
different (p >


Subsamples for study 1 and study 2 were derived from different sample
lots; subsamples for T2 and MS2 comparisons were derived from the same
sample lot.

In study 1, sample suspensions were prepared by blending (2000 RPM, 5
min) and clarified by polypropylene mesh filtration; for other
comparisons, sample suspensions were prepared by blending (4500 RPM,
10 min) and clarified by centrifugation (3000 X g, 5 min).

In study 1, pH of eluents was that of the media formulation, 7.2 for
0.1% peptone, 0.1M Lysine, 1% casein and 7.5 for Butterfield's
phosphate buffer; in study 2 as well as T2 and MS2 comparisons, pH of
eluents was preadjusted to obtain pH values in suspensions equal to
that with EC medium, i.e., 6.3 0.2.








EC medium, 1% casein, 0.1% lysine and Butterfield's buffer for elution

of MS2 from chicken. The results observed for MS2 in this experiment

were not typical of those noted in subsequent experiments comparing EC

medium and 0.1 M lysine.


Chaotropic agents


The effects of various concentrations of chaotropic agents and/or

detergents incorporated into basal eluents of differing complexity on

the recovery of indigenous coliphages from fresh chicken were examined

(Table 9). The pH of TP broth and 0.1 M lysine eluents was 7.2 with

the pH of resulting sample suspensions being ca. 6.0 and 6.5 for 0.1 M

lysine and TP broth eluents, respectively. The possible interactive

effect of chaotropic agents and pH on coliphage elution from chicken

was examined in a subsequent experiment. All concentrations of

choatropic agents used in these comparisons allowed growth of E. coli

host in the overlay medium comparable to a basal medium control.

The recovery of indigenous coliphages from chicken was not

significantly (p < 0.05) affected by sodium trichloroacetate (NaTCA)

concentrations of 0.02 to 0.20 M in TP broth or 0.1 M lysine; the

recovery of coliphages with 0.4 M NaTCA in either eluent was signifi-

cantly less (p < 0.05) than at any other concentration or with the

basal eluent (Table 9). The decreased recovery of coliphages with

0.4 M NaTCA corresponded to decreased plaque production of coliphages

from a basal eluent sample-suspension which was assayed with overlay

medium adjusted to a NaTCA concentration corresponding to 0.4 M NaTCA

in the eluent-suspensions (data not presented). Thus, the plaque

production by the indigenous coliphages in chicken was inhibited by the








Table 9.


Effect of various chaotropic agents incorporated into
eluents on the recovery of indigenous coliphages from fresh
chicken. 1,2,3,4


Coliphage Recovery Coliphage Recovery


Incorporated Incorporated
Compound 0.1M Lys TP broth Compound 0.1M Lys TP broth


--- 103 PFU/100g -- --- 103 PFU/100g ---

NaTCA Urea

None 4.06a 3.37 a None 2.85a 72.2a
0.02M 3.67a 3.44a 0.1M 3.47a 60.6ab
0.08M 3.63a 3.13a 0.4M 3.07a 67.7ab
0.20M 3.38a 3.10a 1.0M 3.11a 57.1b
0.40M 1.83 1.17 2.0M 3.12 48.9

NaNO NaCI
b a a a
None 3.49 4.27a None 1.90a 2.02a
0.1M 4.23a 4.39b 0.1M 2.23a 2. 10b
0.5M 1.13c 1.21b 0.5M 1.75a 1.33b
I.OM 0.97c 0.95 I.OM 1.63 1.45

Tween 80 Triton X-100

None 1.34a 1.54a None 26.93a 25.60
0.1% 1.61a 1.39a 0.1% 27.34a 26.01a
0.5% 1.69a 1.64a 0.5% 24.07a 22.47a
1.0% 1.37 a 1.42


Values represent means of three subsamples. Within each column, for
a given compound, means having the same superscript(s) were not
significantly different (p > 0.05).
2
0.1M Lys = 0.1M lysine buffer, pH 7.2; TP = tryptose phosphate broth,
pH 7.2; NaTCA = sodium trichloroacetate.
3
In NaTCA and urea comparisons, subsamples for 0.1M Lys and TP broth
eluents were not derived from the same sample lot whereas subsamples
for 0.1M lys and TP broth eluents were derived from the same sample
lot in each of the other comparisons.

Sample suspensions prepared by blending (2000 RPM, 5 min) and
clarified by polypropylene mesh filtration.








concentration of NaTCA in the overlay medium at an eluent concentration

of 0.4 M NaTCA using these dilution and assay procedures. Likewise,

the recovery of indigenous coliphages was not significantly increased

(p > 0.05) by the addition of various concentrations of urea (0.1 to

2.0 M) to the basal eluents. In fact, the recovery of coliphages with

TP broth + 2.0 M urea was significantly less (p < 0.05) than that with

TP broth alone. The decrease in coliphage recovery associated with TP

broth + 2.0 M urea did not reflect a similar reduction in plaque pro-

duction with overlay medium containing a corresponding concentration

of urea (data not presented). The observed results, i.e., decreased

recovery of coliphages, can not be readily explained on the basis of

these data.

The addition of sodium nitrate (NaNO3) or sodium chloride (NaCI)

to basal eluents at concentrations of 0.1, 0.5 and 1.0 M generally did

not result in improved recovery of indigenous coliphages from chicken

over that of the basal eluent (Table 9). Although the recovery of

coliphages with 0.1 M lysine + 0.1 M NaNO3 was significantly higher

(p < 0.05) than with 0.1 M lysine alone, this difference was not

considered to have a practical significance since this effect was not

seen with TP broth or in subsequent studies (Table 10). Additionally,

recoveries with 0.5 or 1.0 M NaNO3 in either basal eluent were

significantly less (p < 0.05) than those with basal eluent alone.

Concentrations of NaNO3 in overlay medium corresponding to 0.5 or 1.0 M

NaNO3 in the eluents also inhibited plaque production by indigenous

coliphages from basal eluent suspensions (data not presented). The

recovery of coliphages generally decreased with increasing concentra-

tions of NaCl in basal eluents -ilthough no significant differences








Table 10. Effect of pH and eluent composition on the recovery of
indigenous coliphages from fresh chicken. 1,2,3,4


Coliphage Recovery


Eluents pH of Sample Suspension


6.0 6.5 7.0 8.0 9.0 10.0


--------------- 103 PFU/1OOg ------------------


Study 1


0.1M Lys

TP broth


1.59a


1.71a 1.55a 1.51a 1.53a


3.11 a


2.59a 2.55a 0.73 b


Study 2


0.1M Lys


0.1M Lys
(0.1M NaN03)

0.1M Lys
(0.4M urea)

0.1M Lys
(0.5% Tween 80)


a
1.21

1.45a


1.21a


1.40

1.23a


1.46a


1.40a


1Values represent means of three subsamples.


For study 1, within each


row, means having the same superscripts were not significantly
different (p > 0.05). For study 2, all means were not significantly
different (p > 0.05).

20.1M Lys = 0.1M lysine buffer; TP broth = tryptose phosphate broth.

3In study 1, subsamples for 0.1M Lys and TP broth eluents were not
derived from the same sample lot whereas all subsamples in study 2
were derived from the same sample lot.

4Sample suspensions prepared by blending (2000 RPM, 5 min) and
clarified by polypropylene mesh filtration.








(p > 0.05) were noted with 0.1 M lysine eluent. The recovery of

coliphages with TP broth + 0.5 or 1.0 M NaCI was significantly less

(p < 0.05) than with TP broth alone or TP broth + 0.1 M NaCl. The

addition of NaCl to overlay medium corresponding to 0.5 or 1.0 M in

eluents did not effect plaque production by indigenous coliphages from

basal eluent suspensions (data not presented).

The addition of detergents such as Tween 80 or Triton X-100 to

the basal eluents in concentrations of 0.1 to 1.0% had no significant

effect (p > 0.05) on the recovery of indigenous coliphages from fresh
chicken (Table 9). The effects of Tween 80 on recovery of coliphages

from various samples was also examined in subsequent studies.


Sample suspension pH


Using 0.1 M lysine as an eluent, the recovery of indigenous

coliphages from fresh chicken was not significantly (p > 0.05) affected

by the pH of sample-suspension over a range of approximately 6.0 to

10.0 (Table 10, study 1). Using TP broth as an eluent, there were no

significant differences (p > 0.05) between sample-suspension pH values

of 6.5, 8.0 and 9.0 but significantly less (p < 0.05) coliphages were

recovered at a sample-suspension pH of 10.0 than at other pH values.

The observed decrease in recovery of coliphages in sample-supsensions

at pH 10 with TP broth but not with 0.1 M lysine could be attributed to

the higher pH value of the TP broth (pH 12.4) required to attain a

sample-suspension with pH 10 than that of the 0.1 M lysine (pH 11.4)

(Table A-1, Appendix). Although coliphages are inactivated by pH

values in excess of pH 11.5 in less than 1 min (Sabatino and Maier,

1980), the relative inactivation of indigenous coliphages by exposure








to pH values of eluent and/or sample-suspensions in these experiments

can not be ascertained on the basis of these data. As previously

mentioned (see Materials and Methods), direct adjustment of sample-

suspension pH values with 1.0 N sodium hydroxide resulted in

correspondingly lower recoveries of coliphages than preadjustment of

eluent pH values to obtain desired pH values in sample-suspensions

(data not presented).

Since eluents incorporating chaotropic agents and/or detergents

have been successfully applied to elution of enteroviruses from marine

or aquatic sediments (Bitton et al., 1982) and sewage-sludge (Farrah

et al., 1981b) at a pH value of 9.0, the effects of representative

chaotropic agents in 0.1 M lysine eluent on the recovery of indigenous

coliphages from chicken-suspension at pH 6.0 and 8.0 were studied

(Table 10, study 2). The recovery of coliphages from sample-

suspensions at pH 6.0 and 8.0 were not significantly different (p >

0.05) for 0.1 M lysine, 0.1 M lysine + 0.1 M NaN03, 0.1 M lysine +

0.4 M urea or 0.1 M lysine + 0.5% Tween 80. Likewise, the recovery of

coliphages using any 0.1 M lysine formulation was not significantly

higher (p > 0.05) than 0.1 M lysine alone, regardless of the sample-

suspension pH. These results generally supported findings in previous

experiments with regard to chaotropic agents (Table 9). The effects of

Tween 80 on the recovery of MS2 and T2 coliphages from various foods

was also investigated in subsequent experiments.


Recovery of T2 and MS2 from various foods


Since EC medium eluent generally resulted in higher recoveries of

T2 and indigenous coliphages from fresh chicken than other eluents








tested, various formulation and pH values of EC medium and a less

complex basal eluent, i.e., 0.1 M lysine, were compared for recovery of

T2 and MS2 from chicken (Table 11) as well as fresh ground beef, canned

corn beef and thawed mixed vegetables (Table 12). EC medium (pH 7.0)

resulted in significantly higher (p < 0.05) or equivalent recoveries of

indigenous coliphages from chicken as well as T2 or MS2 from all foods

tested than any formulation of 0.1 M lysine. Significantly higher (p <

0.05) recoveries of T2 and MS2 from chicken and of T2 from ground beef

were obtained with EC eluent (pH 7.0) than with 0.1 M lysine eluent (pH

7.2). Few significant differences (p < 0.05) were noted in coliphage

recovery in relation to eluent pH. Recoveries of T2 from fresh chicken

and ground beef were significantly higher (p < 0.05) using 0.1 M lysine

at pH 8.5 than 0.1 M lysine at pH 7.2; recoveries of T2 from these

foods with 0.1 M lysine (pH 8.5) were significantly less (p < 0.05)

than with EC eluent (pH 7.0). No significant differences (p > 0.05)

were observed for recovery MS2 or indigenous coliphages as a result of

eluent pH and recoveries of T2 with EC eluent were not significantly

affected (p > 0.05) by eluent pH (7.0 or 8.5). Corresponding pH values

of sample-suspensions varied with the eluent and food type in these

experiments (Table A-I, Appendix).

The effect of 0.15% bile salts incorporated into 0.1 M lysine

eluent on the recovery of coliphages from various foods was also

examined (Tables 11, 12). Bile salts (0.15%) contained in EC medium

as a selective agent for E. coli and/or gram negative bacteria was

suspected to have a potential chaotropic and/or eluting effect in food

sample-suspensions based upon previous experiments (Table 8). The

addition of 0.15% bile salts to 0.1 M lysine eluent had no significant







Table 11. Comparison of eluent composition and pH on recovery of T2,
MS2 and indigenous coliphages from fresh chicken.1,2


3 Indigenous
Eluent Coliphages T2 MS2


-- 103 PFU/100g -- -- % recovery --

a c b
0.1M Lysine 16.3 38.1 81.7
(pH 7.2)

0.1M Lysine 17.9a 34.6c 82.3 b
(0.5% Tween 80, pH 7.2)

0.1M Lysine 20.0a 34.4c 80.8 b
(0.15% bile salts, pH 7.2)

0.1M Lysine 16.3a 52.1b 84.7 b
(pH 8.5)

EC medium 22.7a 65.6a 100.6a
(pH 7.0)

EC medium 18.7a 54.9b 89.9ab
(0.5% Tween 80, pH 7.0)

EC medium 20.8a 69.8 a 91.9 ab
(pH 8.5)


IValues represent means of three subsamples. Within each column, means
having the same superscript(s) were not significantly different (p >
0.05).
2All sample suspensions were prepared by blending (4500 RPM, 10 min)
and clarified by centrifugation (3000 X g, 5 min).

3Bile salts = "bile salts mixture number 3" (Difco Laboratories,
Detroit, MI).














0



0
(-4
CO







0











E-
142





00
Ca




















0.
I-I






















0
U

















4-o

C.
H



























*0















CO
41
0













r-i
















004
U


ca
0

0.
-o
c




4-4
























- 41



i-o

H-


(12
ca
-o








(4-4
00










0)










0
(U
x









a)







0



CQ

k-o



0
0







T3








0

0
U


N








N









CN
cl









E-










CN
N











CN
E-


ca c0
I -




I *4 -



I N









ca ca
I





C14 L)I
a)

0
QU












CO
oo r
*

I 0 CO






I U
C0 -1

I NN








CN


I N 0
IC

00
*W


00

3 N00
*H HI


*o
X 0

O C;


CN

0



(n InV
0 0 In
S 0

r-i C

-.) 00 .
a.r-4 a)4 E- a) .
C .: 3
CH *Ho 2
02- 02 0
>.I >- *H *HM- -H
>1 'Orn -P46, -4
) a) a)
X *X E E E0 E
*C 3 3
O 0 E 0


al ca cm c
*
I In C: r,
%0 -' '0 CO
00 r oo %0




ca Ca cm m c
\1I 0 00 CY\









C1 0 In) t)
en *a o M

\0o '.0 in N.


m^ co co
na CO
Vn --T


U
0


001 O 00
O Ca Ca
<0> -st In
r- 0 0






C.)
C4 CO \
N \D -T
r~- m' m'
3-~ .0 '-~


S-N




4

(J



0r


00





C
C
0
CO



CO


r-





Q)





0
O

















E
Mo


0
C





COt


0=






-U a

4-H










il 00
o -H


0 >
M





3
4-1

0 C






(U




r --4
-) 0


0


c4

0
'n


0














"-
0

















t3
O
4-1


















Ca



0
-o






























0
0










































CO
C-4
0..

0I-
0



01
-4-






























-H co
0.
a)




















*l-l
en


0
,--
C
ca
Q)