Journal of Virological Methods, 9 (1984) 87-98

Elsevier

JVM 00326

87

SILVER STAINING OF DNA RESTRICTION FRAGMENTS FOR THE RAPID

IDENTIFICATION OF ADENOVIRUS ISOLATES: APPLICATION DURING

NOSOCOMIAL OUTBREAKS

MARTHA BROWN, MARTIN PETRIC and PETER J. MIDDLETON

Department of Virology, The Hospitalfor Sick Children, 555 University Avenue, Toronto, Ontario M5G 1x8.

Canada

(Accepted 13 April 1984)

The ultr.asensitive photochemical silver stain for nucleic acids, described by Beidler et al. (1982). has been

applied to the detection of adenovirus restriction fragments as a relatively rapid technique for the

identification of virus isolates. In this study, restriction enzyme cleavage analysis was used to characterize

adenovirus isolates from what appeared to be two nosocomial outbreaks. The first outbreak was thus

shown to include two clusters of patients, and involved two serotypes Ad7c and Ad40. The second outbreak

was unrelated and involved Ad35. Although restriction analysis does not replace serum neutralization as a

routine method for typing adenoviruses, it is a much more rapid means of discriminating between different

patient isolates, providing a current rather than retrospective analysis of a nosocomial outbreak. During the

first outbreak, restriction analysis identified two distinct adenovirus serotypes from one patient - Ad7c

from a nasopharyngeal aspirate and Ad41 from a stool specimen. Restriction analysis is also valuable for

the sub-typing of virus isolates. In this study, the Ad40 and Ad41 isolates were shown to be variants of the

respective prototype strains.

adenovirur nosocomial infection silver stain restriction patterns identification

INTRODUCTION

The value of restriction enzymes in the epidemiological analysis of virus infections

is well recognized (Summers, 1980; Wade11 et al., 1980). Restriction enzymes have

been used largely for the identification of different virus strains among members of the

herpesvirus group, including herpes simplex types 1 and 2 (Buchman et al., 1978;

Linneman et al., 1978; Roizman and Togman, 1983), cytomegalovirus (Spector, 1983)

and varicella-zoster (Straus et al., 1983). Restriction enzymes have also been useful in

the study of adenoviruses. Mulder et al. (1974) initially demonstrated that four

different adenovirus serotypes each had a distinct EcoRI cleavage pattern. Wade11 et

al. (1980) subsequently found that not only did each serotype have a characteristic

cleavage pattern with respect to a given restriction enzyme but that different isolates of

the same serotype could have related but distinct restriction patterns. Thus, character-

Olh6-0934/84/$03.00 Q 1984 Elsevier Science Publishers B.V

88

ization of adenovirus isolates by restriction analysis is the method of choice for

establishing the non-identity of different virus isolates.

DNA restriction patterns are usually detected by staining with ethidium bromide.

Recently, Beidler et al. (1982) described a simple photochemical silver stain for the

ultrasensitive detection of DNA in polyacrylamide gels. In this study, the silver stain

was coupled with restriction analysis of viral DNA from clinical isolates to follow the

transmission of adenovirus within our hospital. Because the silver stain is highly

sensitive (100 X more sensitive than ethidium bromide), very little DNA is required for

this analysis. Sufficient DNA can be obtained from high salt extracts of cells infected

directly with the patient specimen, thus eliminating the need for serial passage or for

purification of the virus. Results are available within 2 days from the time the cells are

harvested thus facilitating the current rather than retrospective analysis of an out-

break.

MATERIALS AND METHODS

Ceils and virus The cells used in this study were Graham’s 293 cells, a continuous line of human

embryonic kidney cells transformed with sheared adenovirus type 5 DNA (Graham et

al., 1977). The 293 cells were obtained from Dr. F. Graham at passage level 43 and

were used in this study between passage levels 50 and 80.

Prototype strains of adenovirus types 7,7a, 40 (Dugan) and 41 (Tak) were purchas-

ed from the American Type Culture Collection and were propagated in 293 cells.

Clinical specimens were suspended in culture medium (Eagle’s minimal essential

medium supplemented with 0.8 mM arginine and 10% fetal calf serum) for inoculation

of cell monolayers.

Preparation of viral DNA Virus preparations (clinical specimens and prototype strains) were inoculated onto

l-day-old subconfluent monolayers of 293 cells in 60 mm Petri dishes. When the

cytopathic effect was well-developed, cells were scraped into the culture fluid using a

rubber policeman, poured into a centrifuge tube and pelleted at 75 X g for 5 min. The

cell pellet was resuspended in 0.3 ml 20 mM Tris, pH 7.5, 10 mM EDTA and poured

into a siliconized microcentrifuge tube (1.5 ml capacity). Viral DNA was isolated by a

modification of the Hirt procedure (1967) as follows: sodium dodecyl sulphate (SDS)

was added to a concentration of 0.6% along with Proteinase K (Boehringer Mann-

heim) (final concentration 250 &I). Following incubation for 1 h at 37°C 5 M NaCl

was added to a final concentration of 1 M and the tube was held at 4°C overnight. The

microcentrifuge tube was placed in an International ultracentrifuge SB283 bucket and

spun at 17,000 X g for 30 min to pellet the high molecular weight cellular DNA. The

supernatant was poured into another siliconized microcentrifuge tube, extracted once

with phenol saturated with Tris-HC1 pH 8.1 and once with chloroform: isoamyl

89

alcohol (24 : 1). Following precipitation with ethanol for 1 h at -70°C, the DNA was

collected by centrifugation at 12,000 X g in an Eppendorf microcentrifuge. The pellet

was suspended in 10 mM Tris-HCl, pH 7.4, 10 mM MgCl,, digested with ribonuclease

(25 ug/ml) (Sigma, heated at 90°C for 10 min to destroy deoxyribonuclease activity)

for 1 h at 37°C and the DNA re-precipitated with ethanol. After drying, the DNA

pellet was suspended in 25 pl 10 mM Tris-HCl, pH 7.4 and stored at -20°C. Aliquots

of 1 or 2 ul were used for restriction analysis. The use of endonuclease-free Proteinase

K in this procedure is critical since contaminating endonucleases in other protease

preparations cause breakdown of the high molecular weight DNA which results in a

background smear on the gel after staining, thus obscuring the restriction fragments.

Restriction enzyme digestion Restriction enzymes Hind III, Sma I, Bgl II and Barn HI were purchased from

Bethesda Laboratories (BRL) and digestion conditions were those described by the

manufacturer. A Hind III digest of h DNA was also purchased from BRL.

Gel electrophoresis Samples were electrophoresed through 5% polyacrylamide gels (polyacrylamide :

bisacrylamide 30 : 0.8) in 0.5 X Tris/borate/EDTA, pH 8.3 (1X TBE: 89 mM Tris, 90

mM boric acid, 2.5 mM disodium EDTA). Electrophoresis was at constant voltage for

approximately 1000 V-h. In one experiment, samples were electrophoresed through a

horizontal 1.2% agarose gel in 0.5 X TBE for 400 V-h.

Staining of gels Polyacrylamide gels were stained with silver using the photochemical method

described by Beidler et al. (1982). The viral DNA obtained from 1 X 60 mm Petridishis

sufficient for lo-20 tracks on a gel. An aliquot of a Hind III digest of h DNA (0.01 ug)

was included in each gel as a control for the intensity of staining.

Neutralization tests The adenovirus isolates were serotyped in micro-neutralization tests using 293 cells.

Specific antisera against adenovirus types l-7 were purchased from Microbiological

Associates, antisera against types 8-31 were obtained from National Institutes of

Health, Bethesda, MD, and antisera against types 32-39 were provided by The Center

for Disease Control, Atlanta, Georgia. Antiserum to adenovirus 41 was generously

provided by Dr. Goran Wadell, Umea, Sweden.

RESULTS

Description of outbreaks Two adenovirus outbreaks (I and II) recently occurred in our hospital. They were

unrelated in time and location and the index case was identified in each outbreak. An

90

outline of the patients involved in the two outbreaks, along with their symptoms and

specimens collected, is presented in Table 1.

The index case in the first outbreak was an 8-mth-old male presenting with a

pneumonia not responding to antibiotics. Two siblings were in the convalescent phase

of a respiratory infection and the mother had bilaterial conjunctivitis. Adenovirus was

isolated in 293 cells from the nasopharyngeal secretions of patient 1, stool specimens

from patient 1 and both siblings, and a conjunctival swab from the mother.

Patient 2, an 8-mth-old male who had been in room contact with the index case, was

discharged, and readmitted after 5 days with a pneumonia which proved fatal 9 days

after readmission. Adenovirus was isolated from a nasopharyngeal aspirate, a stool

specimen and lung biopsy.

Patient 3, a 7-wk-old male, who was in room contact with patient 2 in the intensive

care unit developed gastroenteritis followed by a respiratory infection. Adenovirus

was detected in the stool by direct negative contrast electron microscopy and was

isolated in cell culture. Adenovirus was also isolated from a nasopharyngeal aspirate.

A second cluster of cases involving two I-yr-old females (patients 4 and 5) and a

3-mth-old male (patient 6) occurred on a cardiac ward. All three patients presented

with gastroenteritis. Adenovirus was detected in stools of these patients by electron

microscopy and by isolation in cell culture. Two of these patients had been in the

TABLE 1

Adenovirus isolates from two nosocomial outbreaks

Patient no. Symptoms Specimen Restrictiot?

pattern

Serotype

Outbreak I

(patients

1-6)

Outbreak II

(patients

7-8)

I pneumonia NPS A 1

sibling a RI stool A 7

sibling b RI stool A 7

Mother of I conjunctivitis cs A I

2 pneumonia NPS A I

lung biopsy A 7

stool A 7

3 pneumonia NPS A 7

stool B 41

4 enteritis stool c 40

5 enteritis stool C 40

6 enteritis stool C 40

7 RI NPS D 35

8 RI stool D 35

a Isolates with identical restriction patterns are listed with the same letter (A, B, C, D).

NPS = Nasopharyngeal secretions: RI = respiratory infection; CS = conjunctival swab.

91

intensive care unit but not in room contact with patients 2 and 3. A chronogram in Fig.

1 summarizes the hospital stay of these six patients.

The second unrelated episode of adenovirus pneumonia (outbreak II, Table 1)

occurred 2 mth later in a perinatal ward. Following the death of a 2-wk-old girl

(patient ‘7) with adenovirus pneumonia, one of the contacts (patient 8, a 7-wk-old boy)

also developed pneumonia. Adenovirus was isolated in culture from a nasopharyn-

geal aspirate from patient 7 and from a stool specimen from patient 8.

Since patient 2 had been in room contact with patient 1, it was possible that patient

2’s infection had been acquired nosocomially. In an effort to rule this out, restriction

patterns of the two virus isolates were compared. Similarly, the isolates from patient 3

were compared with those of patients 1 and 2. It was also possible that patient 4

acquired the infection while in the intensive care unit and carried the virus to the

cardiac ward, where patients 5 and 6 were infected. The isolates from patients 4,5 and

6 were therefore compared to those from patients 1, 2 and 3.

Similarly, patient 8 appeared to have acquired the infection from patient 7 and thus

it was important to compare these isolates as well.

Identjfication of virus isolates

All the isolates from the index patient (patient 1) and his family contacts had

identical DNA restriction patterns when cleaved with Hind III (Fig. 2A, tracks b-f)

aqd with Sma I (Fig. 2B, tracks b-f). Moreover, the isolates from patient 2 (Fig. 2A

and B, tracks g, h) also had DNA restriction patterns identical to those of the isolates

from patient 1.

No. of days 1 10 20 30 40 50 60

I I I I I I

Patient 1 ,___..___mm_- J ,

1 2

b__ _ _ _ _/ f----h

JJ. 3

4

Fig. I. Chronogram of adenovirus outbreak showing patient contact. Time spent in isolation ward I---).

intensive care unit _, cardiac ward) ----I. Arrow (1) denotes day on which adenovirus positive specimen

was collected. Patient 3 was admitted to hospital on day of birth, 1 mth prior to the beginning of the

outbreak. In the case of patient 3, the first arrow represents collection of the stool specimen, the second

arrow represents collection of the nasopharyngeal aspirate.

92

Hind III Sma I

A abcdefghiikl

B abcefahii kl

Fig. 2. Hind III and Sma I digests of adenovirus DNA preparations from patients 1, 2 and 3. (A) Hind.111

digests (a) 1 DNA, 0.01 pg. (b) stool, sibling a of patient 1, (c)stool, sibling b of patient 1, (d) eye swab,

mother of patient I, (e) NPS, patient 1, (I) stool, patient 1, (g) NPS, patient 2, (h) stool, patient 2, (i) NPS,

patient 3, (j) stool, patient 3, (k) adenovirus type 41. prototype strain Tak. (I) adenovirus type 40, prototype

strain Dugan. (B) Sma I digests (a) Hind III digest of h DNA, 0.01 ug (b-l) same as in (A). NPS,

nasopharyngeal secretions.

Adenovirus was isolated from a nasopharyngeal aspirate from patient 3 after two

passages in 293 cells. The DNA restriction pattern (Fig. 2A and B, track i) was

identical to that of the isolates from patient 2. Patient 3 subsequently experienced

respiratory difficulties. A chest X-ray 12 days following initial exposure to patient 2

and 10 days after collection of the nasopharyngeal specimen showed increased right

lung density consistent with an adenovirus infection. Adenovirus isolated from the

stool of patient 3 had a DNA restriction pattern distinct from that of the nasopharyn-

geal specimen (Fig. 2A and B, tracks i, j). This is consistent with the hypothesis that

patient 3 was infected with two distinct adenoviruses.

All adenovirus isolates in the second cluster of cases (patients 4, 5 and 6) had

identical DNA restriction patterns (Fig. 3, tracks c-e). These patterns differed from

those of isolates from patients 1 and 2 and from that of the stool isolate of patient 3.

Similarly, isolates from patients 7 and 8 had identical restriction patterns when

cleaved with four different enzymes: Hind III, Sma I, Bgl II and Sst I (Fig. 4). These

isolates were serotyped by conventional microneutralization tests, using 293 cells, as

adenovirus type 35.

93

Hind III_

obcdef

Smrr t

hiiklm

Fi_e. 3. Hind III and Sma I digests ofadenovirus DNA preparations from patients 3,4, 5 and 6. All isolates

were from stool specimens. (a) Hind III digest of 2. DNA, 0.0 1 gig, (b.h) adenovirus type 40, prototype strain

Dugan, (c,i) patient 6, (dj) patient 5, (e,k) patient 4, (f,I) patient 3, (g,m) adenovirus type 41. prototype

strain Tak.

Isolates from patients 1 and 2 and the respiratory isolates from patient 3 were typed

by microneutralization tests as adenovirus 7. Barn HI digestion was used to identify the

isolates as prototype 7,7a, 7b or 7c, as described by Wade11 et al. (1981). The Barn HI

cleavage pattern of the type 7 isolates was consistent with the published Barn HI

pattern for type 7c and distinct from that of types 7, 7a and 7b (Fig. 5).

94

X Hinditl Sma I 891 II Sst I X

Fig. 4. Restriction endonuclease digests of adenovirus DNA isolated from patients 7 and 8. (a) Hind III

digest of )i DNA, 0.01 pg. (b,c) Hind 111, (d,e) Sma I, (f,g) Bgl II, (h,i) Sst I.

The stool isolate of patient 3 was neutralized by antiserum to adenovirus 41, a

fastidious enteric adenovirus. The isolates from patients 4, 5 and 6 were identified as

adenovirus 40 (also a fastidious enteric adenovirus) by their restriction patterns which

were characteristic of the prototype strain (Dugan) (Fig. 3, tracks b and h).

DISCUSSION

Restriction analysis demonstrated that the virus infecting patients 1, 2 and 3 (i.e.

Ad7c) was different than that affecting patients 4, 5 and 6 (i.e. Ad40). Thus, what

95

abcdef

Fig. 5. Sub-typing of Ad 7 isolates from patients 1,2 and 3. DNA was cleaved with Barn HI and fragments

were separated on a 1.2% agarose gel. Bands were visualized by staining with ethidium bromide. (a$) Hind

III digest of 1 DNA (0.25 pg), (b) Ad 7, (c) Ad 7a, (d) NPS, patient I, (e) NPS, patient 2. NPS =

Nasopharynpeal secretions.

appeared to be a single outbreak was in fact two separate outbreaks (Ia and Ib).

The identical restriction patterns of type 7c isolates from patients 1,2 and 3 (Fig. 2A

and B) and the contacts between these patients (Fig. 1) strongly suggest the nosoco-

mial transmission of the virus from patient 1 to 2 and from patient 2 to 3. Two other

nosocomial outbreaks of adenovirus type 7 (one of them type 7b, the other not

subtyped) have recently been reported (Fee et al., 1983; Straube et al., 1983).

The virus involved in outbreak Ib (patients 4, 5 and 6) was Ad40, one of the

fastidious enteric adenoviruses. The other fastidious enteric adenovirus (Ad41) was

isolated from the stool of patient 3 durin g an episode of gastroenteritis.

96

Restriction analysis using four different enzymes (Hind III, Sma I, Bgl II and Sst I)

(Fig. 4) failed to show any difference between the Ad35 isolates from patients 7 and 8,

implying nosocomial transmission of the virus from patient 7 to 8.

Thus, the value of restriction analysis for the characterization of adenovirus isolates

is illustrated in this study. Coupled with the highly sensitive silver stain of Beidler et al.

(1982), results are available within two days from the time the infected cells are

harvested. The sensitivity of the silver stain is illustrated in Figs. 2-5 by the tracks

containing a known amount (0.01 ug) of h DNA digested with Hind III. Because only

a small amount of viral DNA is required, it can usually be obtained from the cells

inoculated with the original specimen without serial passage of the virus and without

the need to purify the virus. This is a significant advantage of the technique since

amplification of the virus by serial passage can lead to misdiagnosis of the serotype

associated with the infection especially when stool specimens are involved (Brown et

al., submitted for publ.). In another study, four stool specimens were shown to contain

one adenovirus serotype in low proportion relative to another yet the minor serotype

was selectively amplified by serial passage in cell culture so that it became the only

species detectable (Brown et al., submitted for publ.).

A silver stain technique (Sammons et al., 1981) has also been used for the detection

of rotavirus RNA segments (Follett and Desselberger, 1983). More recently, Whitton

et al. (1983) described the general applicability of the silver stain (Sammons et al.,

1981) for detection of single-stranded and double-stranded RNA and DNA in polya-

crylamide gels. The silver stain technique described by Beidler et al. (1982) and used in

the study reported here, requires less time (2 h as opposed to 5 h) and fewer reagents

than that described by Sammons et al. (1981) yet is highly sensitive.

Three significant findings arose from this study in which restriction analysis was

used to characterize adenovirus isolates from clinical specimens.

(1) Two distinct viruses were isolated from patient 3 - Ad7c from a nasopharyngeal

aspirate and Ad41 from a stool specimen. This is relevant in terms of the suggested

association of the fastidious enteric adenoviruses, types 40 and 41, with respiratory

symptoms (Yolken et al., 1982; Uhnoo et al., 1983). Earlier data had demonstrated

that in contrast to adenovirus types l-39, types 40 and 41 appeared to be associated

solely with enteric infections and not with respiratory infections (Petric et al., 1982).

The results reported here indicate that caution must be exercised in associating

respiratory symptoms with Ad41 enteric infections and that before Ad40 and Ad41

can be implicated in respiratory disease, they will have to be identified specifically in

the respiratory tract. The fact that the cells inoculated with the nasopharyngeal

specimen from patient 3 were passaged twice before cytopathic effect became evident,

reflects the very low concentration of virus in the specimen. Since the specimen was

taken only 2 days after the patient’s initial exposure to patient 2 and prior to

symptoms of respiratory disease, it is su,, Doested that the virus was recovered from the

nasopharynx during the incubation period.

(2) The Ad7 isolates from outbreak Ia (patients 1, 2 and 3) were subtyped by

97

restriction analysis as Ad7c. It has been reported that of the type 7 variants (Ad7, 7a,

7b, 7c), type 7b has been responsible for most cases of severe adenovirus type 7 disease

since 1969 whereas type 7c was the circulating strain during the 1960’s (Wade11 et al.,

1981; Straube et al., 1983). Based on the fact that type 7b was also responsible for an

adenovirus epidemic in 1956, it was suggested that types 7b and 7c may alternate over

the decades as the causative agents of severe Ad7 infections (Wade11 et al., 1981). It is

possible, therefore, that the identification of type 7c in this recent outbreak marks

another conversion in the circulating strain of Ad7. It is relevant that an adenovirus

isolate from a patient not involved in this outbreak but with disease symptoms very

similar to those of patient 1, was also identified as type 7c (results not shown),

indicating that Ad7c is currently endemic in the community. In the outbreak described

here, the virus caused pneumonia in two patients, one of whom died. The virus spread

within the family of the index patient, causing respiratory disease in two siblings and

bilaterial conjunctivitis in the mother. The index patient also developed encephalopa-

thy during the course of his disease. Encephalopathy has been documented as an

extrapulmonary manifestation of adenovirus type 7 pneumonia (Huttunen, 1970;

Ladisch et al., 1979; Kim and Gohd, 1983).

(3) The three Ad40 isolates and the single Ad41 isolate were variants of the

respective prototype strains (Fi g. 2j, k; Fig. 3b, c). The fact that the difference in Ad40

was seen only with Hind III digestion (Fi g. 3b, c) and not with Sma I digestion (Fig.

3h, i) emphasizes the importance of using more than one enzyme for comparison of

isolates. The smaller restriction fragments are better visualized with the silver stain

than wi1.h ethidium bromide, thus improvin g the comparison of enzyme digests.

Because of the large number of adenovirus serotypes (41 at present) (De Jong et al.,

1983) and the variability in restriction patterns within each serotype, restriction

endonuclease fingerprinting is not suitable for the routine typing of adenovirus

isolates. However, because it can detect genotypic differences between isolates of the

same serotype, this is the method of choice for precise typing and for assessing the

identity of virus isolates. Coupled with silver staining of the restriction fragments, this

technique requires very little viral DNA and provides a relatively rapid and precise

means of virus identification.

ACKNOWLEDGEMENTS

This ,work was supported by a grant (MA 6401) from the Medical Research Council

of Canada. The co-operation of Dr. Lee Ford-Jones, Department of Infectious

Diseases, The Hospital for Sick Children, is acknowledged.

REFERELNCES

Beidler, J.L., P.R. Hilliard and R.L. Rill, 1982, Anal. Biochem. 126, 374.

Buchman, T.G., B. Riozman, G. Adams and B.H. Stover, 1978, J. Infect. Dis. 138, 488

98

De Jong, J.C., R. Wigand, A.H. Kidd, G. Wadell, J.G. Kapsenber,, - C.J. Muzerie, A.G. Wermenbol and

R.G. Firtslaff, 1983, J. Med. Virol. 11, 215.

Fee, M.A., E. Charney, S.A. Plotkin, H.M. Friedman, P. Pershing, A. O’Hara and C. Ferrer, 1983, Morbid.

Mortal. Weekly Rep. 32, 258.

Follett, E.A.C. and V. Desselberger. 1983, J. Med. Viral. II, 39.

Graham, F.L., J. Smiley, W.C. Russell and R. Nairn, 1972, J. Gen. Virol. 36, 59.

Hirt, B., 1967, J. Mol. Biol. 26, 365.

Huttunen, L., 1970, &and. J. Infect. Dis. 2, 151.

Kim, K.S. and R.S. Gohd, 1983, Arch. Nemo]. 40, 58.

Ladisch, S., F.H. Lovejoy, J.C. Hierholzer, M.N. Oxman, D. Strieder, G.F. Vawter, N. Finer and M.

Moore, 1979, J. Pediatr. 95, 348.

Linneman, C.C. Jr., L.J. Light, T.G. Buchman, J.L. Ballard and B. Roizman, 1978, Lancet i, 964.

Mulder, C., P.A. Sharp, H. Delius and V. Petersson, 1974, J. Viral. 14, 68.

Petric, M., S. Krajden, N. Dowbnia and P.J. Middleton, 1982, Lancet i, 1074.

Roizman, B. and M. Togman, 1983, Curr. Top. Microbial. Immunol. 104, 273.

Sammons, D.W., L.D. Adams and E.E. Nishizawa, 1981, Electrophoresis 2, 135.

Spector, S.A., 1983, Lancet i, 378.

Straube, R.C., M.A. Thompson, R.B. VanDyke, G. Wadell, J.D. Connor, D. Wingard and S.A. Spector,

1983, J. Infect. Dis. 147, 814.

Straus, S.E.. J. Hay, H. Smith and J. Owens, 1983, J. Gen. Virol. 64. 1031.

Summers, W.C., 1980, Yale J. Biol. Med. 53, 55.

Uhnoo, I., G. Wadell, L. Svensson and M. Johansson, 1983, Dev. Biol. Stand. 53, 311.

Wadell, G., M.L. Hammarskjold, G. Winberg, T.M. Varsanyi and G. Sundell, 1980, Ann. N.Y. Acad. Sci.

16, 354.

Wadell, G., J.C. de Jong and S. Wolontis, 1981, Infect. Immun. 34, 368.

Whitton, J.L., F. Hundley, B. O’Donnell and V. Desselberger, 1983, J. Viral. Methods 7, 185.

Yolken, R.H.. F. Lawrence, F. Leister, H.E. Takiff and S.E. Straus, 1982, J. Pediatr. 101, 21.