genetic characterization of sabin types 1 and 3 poliovaccine virus following serial passage in the...

Biologicals (1992) 20, 15-26

Genetic Characterization of Sabin Types 1 and 3 Poliovaccine Virus Following Serial Passage in

the Human Intestinal Tract

G. Contreras,* K. Dimock,l: J. Furesz,* C. Gardell,* D. Hazlett,* K. Karpinski,t G. McCorkle,* and L. Wu*§

*Bureau of Biologics and tBureau of Drugs Research, Drugs Directorate, Health Protection Branch, Health and Welfare, Canada ~.Department of Microbiology, University of Ottawa

§Department of Microbiology and Immunology, The Medical College of Pennsylvania, 3300 Henry Ave., Philadelphia, PA, 19129

Abstract. Poliovirus isolates types 1 and 3 were obtained from five and seven successive passages respectively, in infants who had been fed monovalent OPV in two separate clinical trials conducted in 1960. The purpose of these trials was to answer the question how much the vaccine virus would revert to its original neurovirulent phenotype following multiplication in the intestinal tract. Human passages were performed either by contact exposure or by feeding the excreted virus while the infants were maintained in isolation. Several virus isolates were obtained at each passage level. Infants participating in both studies showed no symptoms of disease. Antigenic studies (McBride, van Wezel) and protein analysis (PAGE) of the isolates, reported earlier from this laboratory, had shown that the isolates remained vaccine-like, although isolates from the later passages revealed some differences. Monkey neurovirulence test results showed that for both types 1 and 3 viruses the loss of attenuation of the vaccine strain upon passage was gradual, although the loss was faster for type 3. Examination of the oligonucleotide maps demonstrated that the oligonucleotide configuration of the isolates remained the same as for the vaccine strain but there was an increase of individual spot differences with increasing passage. The nucleotide sequence analysis of selected regions of the virus genomes revealed that there was no change from a G to A in nucleotide 480 of type 1 isolates; however, nucleotide 476 changed from a U to an A in type 1 passages 3, 4 and 5. Conversely, for type 3 the change of nucleotide 472 from a U to a C changed at the early first passage (4 clays following administration of ©PV), and remained a C in the six following passages; type 3 nucleotide 2034 did not change in the first passage from a U to a C, but it became a C in all further passages tested. The nucleotide changes mentioned for both virus types remained stable in successive passages. However, there was another nucleotide change for type 3 from a U to a C at pposition 1973 only for passages 5 and 6 which reverted to a U for passages 7L and 7LL. Study of selected human passage virus strains could further contribute to the identification of the critical nudleotides that are responsible for the attenuation of these two polio types of vaccine viruses.

Introduction

Epidemic and endemic poliomyelitis has been elimi- nated in many countries conducting active immunization programs, a clear demonstration of the effectiveness of poliomyelitis vaccines. Today, the population of these countries is well over 1 billion

Corresponding author: J. Furesz, Bureau of Biologics, Virus Building, Tunney's Pasture, Ottawa, Ontario, Canada K1A OL2.

and the vaccine most used has been Sabin's live, a t tenuated oral poliovirus vaccine (OPV). When the Sabin vaccine strains were initially being tested in humans during the late 1950s, the important question arose; how much would the virus revert to its original neurovirulence following multiplication in the human intestinal tract? Clinical trials involving human passage of the vaccine virus were designed in an at tempt to answer this question. Two such trials were conducted in Canada in 1960, one for type 1, in

1045-1056/92/010015 + 12 $03.00/0 © 1992 The International Association of Biological Standardization

16 G. Contreras et aL

which five successive passages were achieved,' and the other for type 3 in which seven successive passages were accomplished. 2 Although several virus isolates were obtained from each subject, in all cases the infections, with either types 1 or 3, were asymptomatic. Antigenic characterization using the McBride kinetic neutralization test, reported in 1966, 3 and studies with monoclonal antibodies, 4 and together with protein analyses, 5 showed tha t the isolates remained vaccine-like, although isolates of the later passages revealed minor antigenic differences. With the recently acquired knowledge of the complete nucleotide sequences of the genomes of several poliovirus strains, investigators have correlated single nucleotide changes with either at tenuation or virulence of the virus as measured in monkeys, s-9 The human passage virus isolates are very suitable to further investigate these correlations.

In this paper, we present monkey neurovirulence data on the human passage isolates types 1 and 3, along with a few wild poliovirus strains; we also examine the differences in their oligonucleotide maps and the nucleotide sequences of selected regions of their genomes.

Mater ia ls and methods

Virus strains Virus strains from human passages 3, 4 and 5

were available from the clinical trial of Sabin type 1 OPV conducted in Montreal. ' Since no first human passage virus was available to us from this study, we tested another first passage virus, strain VST2, obtained in 1962 from a healthy vaccine recipient 15 days following immunization. Several type 3 isolates from each of the seven human passages were available from a clinical trial of Sabin type 3 OPV conducted in Quebec City. 2 Briefly, the 1960 trials were performed as follows: healthy unimmunized infants aged 8-12 months were kept in an isolated area before and after immunization with a monotypic OPV; the first infant was fed the vaccine virus, either type 1 or type 3, and then stool samples were collected for several weeks; the second passage was carried out either by 'contact' exposure of the second infant to the first one or by direct feeding of the second infant with the virus excreted in the stool of the first infant; stool samples were collected from the second subject also for several weeks. Successive passages were performed similarly from infant to infant either by 'contact' exposure or direct feeding of stool virus, five and seven times for types 1 and 3 respectively. Isolation of infants at each passage

level was strictly observed. The total cumulative number of days in the intestinal tract for each isolate tested are indicated in Tables 2 and 4 for types 1 and 3 respectively. Infants participating in both studies showed no symptoms of disease. Other poliovirus strains that did not originate from the clinical studies are listed in Table 1. Isolates were received frozen and subcultured once or twice in Hep-2 cells in our laboratories. Viruses were t i trated in a plaque assay using Cercopithecus monkey kidney cells and titers were expressed in plaque forming units (pfu) ml. 3

Monkey neurovirulence test (MNT). The test was performed following World Health Organization (WHO) requirements. '° The evaluation of the test is based on monkeys (Macaca fascicularis) that show histopathology typical of poliomyelitis neuronal damage. Briefly, the severity of neuronal damages found in the histological sections of the lumbar cord, cervical cord and brain stem of each monkey was scored as already described, u The specific poliovirus lesions were often quite different on the left and right sides of the section, so more accurate readings were obtained by scoring them by hemisections. Individual monkey lesion scores (LS) were calcu- lated as follows:

[sum of lumbar score] [sum of cervical score] L S = +

No. of hemisections No. of hemisections

[sum of brain score] + + 3

No. of hemisections

The LS for all positive monkeys were used to calcu- late a mean lesion score (MLS) for the virus being tested. Positive monkeys were those showing at least one specific neuronal lesion.

The WHO procedure for evaluating a vaccine lot requires a minimum of 12 and 20 monkeys for types 1 and 3, respectively. However, we had observed that the number of test monkeys for less at tenuated viruses could be reduced without jeopardizing the validity of the assay. H Consequently, an average of nine and 13 monkeys were used for each virus strain of types 1 and 3, respectively.

Oligonucleotide mapping The method of Pederson and Haseltine 12 was used

with slight modification. Calf intestinal phosphatase t reatment was omitted, since it was unnecessary, given tha t the 5' end of the genome is covalently bound to the vpg protein. Poliovirus RNA for

Characterization of poliovaccine 17

Table 1. Reference sabin and 'wild' poliovirus strains studied

Name/number Isolation data of strain Date place origin

Type 1 Sabin strain 1956 U.S.A.

NA4 1977 U.S.A. VST2 1962 Canada

'Wild' strains Mahoney 1941 U.S.A.

148 1962 Canada 1606 1982 India

855 1988 Canada

Type 3 Sabin 3

NCI

1956 U.S.A.

1979 U.S.A.

'Wild' strains Leon 1937 U.S.A. Sauket t 1950 U.S.A.

367 1960 Canada 1034 1961 Canada 1061 1984 Finland

LS-c, 2ab/KP2 vaccine strain derived from Mahoney Reference vaccine strain Asymptomatic recipient (15 days following OPV)

Healthy children Nonparalytic poliomyelitis Vellore, strain provided by Dr J. John Paralytic poliomyelitis, in contact with visitors from India

Leon 12atb/KP3 vaccine strain derived from Leon Reference vaccine strain

Fatal bulbospinal poliomyelitis Paralytic poliomyelitis Non-paralytic poliomyelitis Healthy child Paralytic poliomyelitis

oligonucleotide mapping was extracted from purified virus as described for sequencing. Genomic RNA was digested with RNase T1, labelled using polynu- cleotide kinase and 7~2P-ATP and subjected to 2-D electrophoresis as described by Lee and Fowlks '3 and Yoneyama et al. TM

Reverse transcriptase sequence analysis

CsCl-purified poliovirus was diluted in 10 mM NaPO4, pH 7.0 and pelleted at 124 000 g for 6 h in the Beckman SW41 rotor at 4°C. Virus pellets were resuspended in 300 pl of 100 mM NaC1, 10 mM EDTA, 50 mM NaAc, pH 5.1 containing 0.5% SDS and 300 pg/ml-' proteinase K and incubated for 30 min at 37°C. Samples were then extracted twice with phenol/CHC1Jisoamyl alcohol (24/1). Poliovirus genomic RNA was precipitated at -20°C following the addition of 2-5 volumes of ethanol. RNA was pelleted, washed with ethanol, dried, resuspended at a concentration of 1 mg/ml-t and frozen in aliquots of 2.5 pg at -80°C. Sequence analysis of poliovirus RNA was carried out as described by Geliebter 15 with minor modifications.

The following oligonucleotides (negative sense) were used to prime sequencing reactions:

(a) 5'-GTAGTCGGTTCCGCCAC-3 (polio 1 residues 544-528) and polio 3 residues 547-531);

(b) 5'-ACATGACGTTCACTGCG-3' (polio 1 residues 3488-3472);

(c) 5'-AAGAGGTCTCTATTC CACAT-3' (polio 1 residues 3504-3485);

(d) 5'-TAGTTCAGTACTTCCCC-3' (polio 3 residues 2079-2063).

Polymerase chain reaction (PCR) ampfification, cloning and sequence analysis

Sequence analysis following PCR amplification and cloning was used to confirm nucleotide assign- ments that were ambiguous by direct sequence analysis of virus RNA. Poliovirus cDNA synthesis was carried out as described by Geliebter, '5 in the absence of dideoxynucleotide triphosphates, using the negative sense oligonucleotides as primers and varying amounts of poliovirus RNA as the template. The final reaction volume was 25 pl. Aliquots of the


cDNA synthesis reaction mixtures were diluted at least four-fold and regions of the poliovirus genome were amplified using the polymerase chain reaction. 16

The primers used for positive strand DNA synthesis were:

(a) 5'-GCGTTGCGCTCAGCACT-3' (polio 1 residues 278-294; polio 3 residues 280-296);

(b) 5'-ACACATCAGAGTCTGGT-3' (polio 1 residues 3271-3288);

(c) 5'-GATAGACACCATGATTCC-3' (polio 3 residues 1906-1923).

Samples were deproteined by phenol/CHC13 extraction and DNA was precipitated with ethanol. Following repair with the Klenow fragment of DNA polymerase I, ~7 the DNA was ligated into the SmaI site of plasmid pGEM7Zf+ and used to transform E. coli DH5a. Dideoxynucleotide sequence analysis of recombinant plasmids was carried out using Sequenase (U.S. Biochemicals).

Resul ts

Type 1 strains

Neurovirulence. Virus strains obtained from passages 1, 3, 4 and 5 had been multiplying in the

intestinal tract of infants for a total of 15, 21, 28 and 35 days (Table 2). The mean lesion score (MLS) of passage (1.76) was significantly different (P = 0.05) from the type 1 reference vaccine (1.14). A larger increase in MLSs relative to the reference was noted for passages 3, 4 and 5 (P < 0.001). However, all four human passage isolates showed MLSs lower than those of the two 'wild' poliovirus strains (Mahoney and strain 855) tested concurrently. It should be noted that the standard deviations are smaller for the less a t tenuated viruses. As shown in Table 2, the virus dose for the 'wild' strains was only 1/1000 and 1/100, respectively, of the dose of the passage isolates. This clearly indicated that while the passage isolates had lost a significant degree of attenuation, reversion to full neurovirulence did not take place.

Oligonucleotide maps

The oligonucleotide maps for the type 1 human passage isolates (Fig. 1) showed the same spot distribution pattern as the Sabin vaccine strain. There were eight spot differences between the maps of the vaccine and the Mahoney virus strain from which Sabin derived his vaccine; however, the general pattern of the maps was the same for both strains. Although the spot pattern is the same, differences in individual spots in the maps of the passage isolates

Table 2. Monkey neurovirulence test results on poliovirus type 1 human passage isolates and selected 'wild' strains

Virus strains

Cumulative no. of Neurovirulence test* days in human No. of Lesion Scores

gut monkeys mean SDt

Human passage No. 1 (Strain VST2) 15 10 1.76 0.60 No. 3 21 9 2.31 0.57 No. 4 28 9 2.27 0-24 No. 5 35 9 2.72 0.23

Reference vaccine NA4 Total: 26 tests NAt Two concurrent tests NA

266 1.28 0-54 28 1.14 0-51

'Wild' strains Mahoney NA 7 3.21 0-21 855 (Canada, 1988) NA 10 3.21 0.13

* Monkeys received 5.6 +- 0.2 log~) pfu for each virus strain except for Mahoney (2.6 log.)) and 855 (3-6 loglo).

t SD = Standard Deviation. $ NA = not applicable.


. • 4 f i r ' . . ~ r Q , ~ e o . . £ , - ,

" s ° . ~ " ~ ' : ~ o " "

: ' ~ - - , Qo ~e

Mahoney Sabin 1 Passage 1

e ,D q lLm/•

Passage 3 Passage 5 Strain 148

Figure l. Oligonucleotide maps for poliovirus type 1 strains. Spot differences between Sabin vaccine and the other strains are shown with three symbols: an arrow for a new spot; a circle for a missing spot; a fractured oval and an arrow for migration of a spot.

are present; there is one spot difference for passage 1, two for passage 3 and eight for passage 5. 'Wild' strain 148 (non-paralytic case, Canada 1962) showed a completely different map (Fig. 1). The observations made from the maps are summarized in Table 3.

Nucleot ide s e q u e n c e s

The nucleotide sequences for type 1 human passage isolates and other selected strains were determined for bases 407-524, inclusive, of the 5'-non- coding region (Fig. 2). The numbering system of

Nomoto et al. Is has been followed. The VST2 strain (Table 1) used as passage 1 is identical to Sabin 1 except it has two mutations at positions 441 (A to U) and at position 482 (C to U). The change from A in Mahoney to a G in Sabin 1 strain in position 480 is credited as an indicator of attenuation. 8 This change is shown in the sequences for this region: 480 is a G in Sabin 1 and remains a G in the four human passage isolates. Base 480 is an A in the Mahoney strain, as well as in the three 'wild' strains (148, Canada 1962; 1606, India 1982; 855, Canada 1988, probably originated in India). However, there is one

Tab le 3. Comparison of oligonucleotide maps of type 1 poliovirus strains using Sabin 1 as baseline

Virus strains Pa t te rn

Oligonucleotide maps

Spot differences

Additions Deletion Migrations Cumulative

Human passage No. 1 No. 3 1 Same No. 5

Mahoney Same 148 (Canada 1962) Different

0 1 0 1 1 1 0 2 7 1 0 8 3 5 0 8

NA

NA = Not applicable.


STRAIN NUCLEOTIDE S~U~CE

410 420 430 440 450 460

SABIN i AACAAGGUGUGAAGAGCCUAUUGAGCUACAUAAGAAUCCUCCGGCCCCUGAAUGCGGCUA

NA4 ............................................................

HUMAN PASSAGE:

i .................................. U .........................

3 ............................................................

WILD:

bIANONEY: ............................................................

148 ............................. C-G---U ........................

855 ............................. C ..... G ........................

1606 ............................. C ..... G ........................

470 480 490 500 510 520

SABIN i AUCCC]L%CCUCGGGGCAC~UGGUCACAAACCAGUGAUUGGCCUGUCGUAACGCGCAAG

NA4 ..........................................................

HUMAN PASSAGE:

1 ............... U ..........................................

3 ......... A ................................................

......... A ................................................

......... A ................................................

WILD:

M~d~ONEY: ............. A ............................................

148 ......... A---A ..... C---UG ........ CA-C-A-U ............. U---

855 .... U .... A---A ....... CCUU--GU .... A-GG-A--U ................

1606 ......... '---'---'---'''--''--- A-GG-A--U ................

Figure 2. Nucleotide sequences of the 5' non-coding region for poliovirus type 1 strains.

change in nucleotide 476, from a U in passage 1, also in Sabin 1 (and in Mahoney) to an A in human passage isolates 3, 4 and 5, all three of which also have higher lesion scores than passage 1. This nucleotide change is also present in the three 'wild' strains; it could, perhaps, play a role in the loss of at tenuat ion detected in the last three passage isolates. The three 'wild' strains show other changes at bases 488-491 and 500-508 and in other scattered areas of this region of the genome. Figure 3 shows the region of the genome from bases 3354-3471 which includes sequences coding for part of the VPI capsid and par t of protein 2A. The human passage isolates did not show changes from Sabin 1 nucleotides, which, in

STRAIN NUCLEOTIDE SEOU~CE

3360 3370 3380 3390 3400 3410 SABIN 1 CACCCCUCUCCACCAAGGAUCUGACCACAUAUGGAUUCGGACACCAAAACAAAGCGGUGUA

NA4 .............................................................

HUMAN PASSAGE:

4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

WILD:

MAHONEY: .............................................................

855 ................ A--C .............. G-A---G ..... G ........ A .....

1606 ............. U-'A--C .............. G-AU--G ..... G ........ A .....

3420 3430 3440 3450 3460 3470

SABIN I CACUGCAGGUUACA~UUUGCAACUACCAUUUGGCCACUCAGGAAGAUUUGCAAAA

NA4 .........................................................

HUMAN PASSAGE:

i ......................................................... ] .........................................................

W I L D :

MAHONEY: .........................................................

855 ---A ..... C ..... G ........ U--U--CC ...... Cn ...... C ..........

1606 ......... n---n .......... U ..... CC---n--C .......... C-U .....

Figure 3. Nucleotide sequences of the VP1/2A coding region for poliovirus type 1 strains; n indicates base assign- ment unclear.

this region of 120 bases, were identical with the Mahoney virus. The three 'wild' strains showed several base changes compared to Sabin. It should be noted that some of the changes were shared by strain 855 (Toronto isolate) with strain 1606 (India isolate).

Type 3 strains

Neurovirulence. The type 3 vaccine strain was seri- ally, passaged seven times in infants (Table 4). Viruses were isolated ei ther within 10 days ('early' strains) or 3-4 weeks ('late' strains) after transmis- sion of the virus to the infant. One virus from passage 7 isolated at 109 days, once the infant had left the isolation area, has been called '7 late-late' (7LL). The isolates tested for neurovirulence were from passage levels 1, 2, 3, 5 and 7 (Table 5). Two isolates, one 'early' and one 'late', were tested for passages 1 and 2 and one 'late' isolate was tested for passages 3, 5 and 7. One 'late-late' isolate was tested for passage 7 (7LL). The cumulative number of days in the intestinal t ract for each isolate tested is also indicated in Table 5. An increased mean lesion score (MLS), 2.22, was obtained after only 4 days in the human gut passage 1E. Fur the r increases (P < 0.001) relative to isolate 1E were obtained after day 18 of passage 1 and for isolates 2E and 2L. The MLSs remained at a level of approximately 2.8 for isolates 2E to 7L. A fur ther increase in the MLS was detected in passage 7LL which had been 109 days in the intestinal tract. The MLS of isolate 7LL suggested that it had regained a degree of virulence similar to tha t of 'wild' strains. This was confirmed when isolate 7LL was inoculated with a smaller virus dose, 103'6 pfu, identical to test 'wild' strains. The MLS remained high (2.72), similar to those found with

Tab le 4. Human passage strains type 3

Cumulative no. of days Human passage no. in human gut

1 Early* 4 1 Late 18 2 Early 14 2 Late 28 3 Late 42 4 Late 91 5 Late 71 6 Late 92 7 Late 104 7 Late-Late 109

* 'Early' and 'late' strains as defined in the text.


Table 5. Monkey neurovirulence test results on poliovirus type 3 human passage isolates and selected 'wild' strains

Cumulative no. of Neurovirulence test

days in human Virus No. of Lesion Scores gut dose* monkeys mean SDt

Human passage No. 1E 4 No. 1L 18 No. 2E 14 No. 2L 28 No. 3L 42 No. 5L 71 No. 7L 104 No. 7LL 109 No. 7LL 109

Reference vaccine Total: 22 tests NA$ Four concurrent tests NA

'Wild' strains 1034 (Canada 1961) NA 1061 (Finland 1984) NA Saukett (U.S.A. 1950) NA Leon (U.S.A. 1937) NA

105"6

103,6

11 2.22 O.36 11 2.84 0.28 12 2.70 0.27 12 2.79 0.22 14 2.73 0.24 14 2-99 0.22 15 2.57 0.27 14 3.18 0-16 14 2.72 0.07

105̀ 6 273 1"41 0"75 105.6 55 1.21 0"67

103"6 14 3-22 0.11 14 2.60 0.19 14 2.73 0.22 14 2.27 0.62

* Virus dose in pfu ± 0.2 log. t Standard Deviation. $ Not applicable.

Saukett and 1061 viruses (Table 5). It should be noted that the Sabin 3 vaccine strain was derived from the Leon virus isolated from a boy who had died of bulbospinal poliomyelitis; however, the latter showed the lowest MLS among the four 'wild' strains tested (Table 5). Strain 1034, isolated from a heal thy boy, has the highest MLS and may be the most viru- lent, of the four strains in monkeys.

Oligonucleot ide m a p s

The oligonucleotide maps for the type 3 human passage isolates (Figs 4(a) and 4(b) showed a similar spot distribution pattern as the vaccine strain, and the 'wild' Leon strain from which the vaccine originated. The map of the early first passage isolate (4 days) is identical with the vaccine strain. The map of the late first passage isolate (18 days) showed three spot differences from Sabin 3: two additions and one deletion. A similar observation was reported by Minor et al.;'9 there was no change from the Sabin 3 oligonucleotide map of viruses isolated during the first few days post-feeding trivalent OPV, but there

were changes in maps of later isolates from an infant vaccine recipient. '9 Human passages 3 and 5 showed three and four spot differences respectively, when compared to the vaccine map (Table 6). However, 'late-late' passage 7 (109 days) showed nine differences when compared to Sabin 3. Two 'wild' strains were included for comparison: Leon, which has the same distribution pattern as the vaccine strain, with no spot differences, and strain 1034 (isolated from an asymptomatic boy in Canada in 1961), with an entirely different spot distribution pattern.

Nucleotide sequences

Nucleotide sequences for human passage isolates and other type 3 strains were determined for bases 410-527, inclusive, of the 5'-non-coding region (Fig. 5). The numbering system of the nucleotides was according to Stanway et al. 2° The sequence results showed that base 472 has changed from a U in the vaccine virus to a C after only 4 days of the first passage in the human gut; this change was also present in all other human passage isolates and in the four

22 G. Cont re ras et aL

(a)

- "

. - " . ~ - - P ' ~ O . . . : . , . ,--: 4 ' . " "

. . ~ . . - ~ ' ~ - ~"

Leon " Sabin 3

d e

Early Passage I Late Passage t

. - ~ , f ~ , - . . : / . .f . - -

. , ~ l ~ , ~ ' " : '~ " : ' = " " " 5 - " , , I I * L ' . - - . " ~ - -----~. - . -

~ Sabin 3 " Late Passage 3

,t abe:F, . : - y l~=d". - - -

D e o ~ / b o ° " ~ . °

Late Passage 5 Late-Late Passage 7 Strain 1034

Figures 4(a), (b). Oligonucleotides for poliovirus type 3 strains. Spot differences to Sabin vaccine are shown by the same symbols used for Fig. 1.

'wild' strains tested. The change from a U in Sabin 3 to a C is credited as an indicator of increased neurovirulence. 6 In the 'wild' strains (Saukett, 1034,1061 and 367), there are base changes in positions close to the two areas mentioned above as well as in bases 488-498. Figure 6 summarizes the results on the region of the genome from bases 1945 to 2062 of the 5' end. This comprises part of the

region coding for the VP3 capsid protein and which has also been shown to be associated with monkey neurovirulence for type 3 viruses, v There, are addi- tional base changes, compared to the vaccine strain, in the human passage viruses; a change from U at position 2034 in the Sabin strain to cytosine has occurred in all human passage isolates except passage 1E; position 2034 was also a U in the Leon


Table 6. Comparison of oligonucleotide maps of type 3 poliovirus strains using sabin 3 as baseline

Oligonucleotide maps Spot differences

Virus strains Pattern Additions Deletions Migrations Cumulative

Human passage No. 1E ] 0 0 0 0

No. 1L t 2 1 0 3 No. 3L Same 1 1 1 3 No. 5L 3 0 1 4 No. 7LL 6 2 1 9

Leon Same 0 0 0 0 1034 (Canada 1961) Different NA*

* Not applicable.

STRAIN NUCLEOTIDE SEOtr~4CE

410 420 430 440 450 460

SABIN 3 AACAGGGUGUGAAGAGCCUAUUGAGCUACAUGAGAGUCCUCCGGCCCCUGAAUGCGGCUA

NCI ............................................................

HUMAN PASSAGE:*

IE IL

2E

2L

3L

4L

5L

6L

7L

7LL

WILD:

LEON

SAUKE'FF

1061

1 0 3 4

3 6 7

470 480 490 500 510 520

SABIN ) AUUCUAACCAUGGAGCAGGCAGCUGCAACCCAGCAGCCAGCCUGUCGUAACGCGCAAG

NCI ...........................................................

HUI~ PASSAGE:

IE IL

2E 2L

)L

4L

St

6L

7L

7LL

WILD:

LEON

Saukect 1061 1034

367

.... A .......................................................

.... A ............... C .......................................

.... A ......................... AA ............................

.... A ......................... AA ............................

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

--C-C .... 0C ......... OAUCA---G .... UGAUUGA .......... U---U ....

--CUD ..... C ........ U''UCA---A .... UGAUUU ....................

--C .............. A-UGA ...... A ........ U---U ........ U ........

--C ................ UGA ...... A ............ U ........ U ........

Figure 5. Nucleotide sequences of the 5'-non-coding region" for poliovirus type 3 strains; * = 'early' and 'late' strains, as defined in the text.

strain but a C in the Saukett strain. This change from U in Sabin 3 to a C is credited as an indicator of increased neurovirulence 7 and causes a Phenylalanine codon in VP3 to be changed to a Serine codon. It is noteworthy that a change in position 2035 from a U in the Sabin strain to a C only occurred in human passages 5L, 6L, 7L and 7LL; but it does not alter the amino acid sequence of VP3 beyond the change resulting from the mutation at 2034. There is another mutation from a U to a C in position, 1973 in human passage isolates 5L and 6L, the mutation at 1973 reverted to a U for passages 7L

SABIN

NCI

HUMAN

IE

IL

2E

2L

3L

4L

5L

6L

7L

7LL

WILD:

LEON

SAUKETT

1061

NUCLEOTI DE S EQtr~CE

1950 1960 1970 1980 1990 2000

3 GAGAAACACAAUGGACAUGU ACAGAGU UACUCUGAGCGACAGUGCCGAU C U AUCGCAACCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PASSAGE : " .............................................................

.............................................. A ..............

.............................................................

.............................................................

.............................................................

.............................................................

............................ C ................................

............................ C ................................

.............................................................

.............................................................

.............................................................

...... U ................. G--C .................................

- - -G - -U - -C ................. G-GGU .......... C - - -A-C- -GAGCG -U - - -

SABIN

NCI HUMAN

IE

IL

2E

2L

3L

4L

5L

6L

7L

7LL

WILD:

LEON

SAUKETT

1 0 6 1

2010 2020 2030 2040 2050 2060

3 A U U U U G U G C U U G U C A C U A U C C C C A G C A U U U G A U C C G C G C U U G U C A C A C A C C A U G C U U . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PASSAGE: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . CC . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . CC . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . CC . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . CC . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - -C . . . . . . . . . . G . . . . . U - -G . . . . C . . . . . . . . . . . . . . G . . . . . . . . . . . . .

- -C- -A- --C .... C--G--A- -G- - -GCA- -C- -U ..... A ........ U ..... A

Figure 6. Nucleotide sequences of the VP3 coding region for poliovirus type 3 strains; * = 'early' and 'late' strains, as defined in the text.

and 7LL. There were also differences in several positions between the Sabin strain and the 'wild' strains, Saukett and 1061. Several of the nucleotide changes in strain 1061, including those at positions 2033, 2034 and 2035 alter coding, whereas in the Saukett strain only the mutation at 2034 alters coding.

D i s c u s s i o n

Neurovirulence is a complex property of polioviruses that may depend on various characteristics of the genotype of the virus and on many different condi- tions in the host, either permissive or restrictive of


virus multiplication and spread. Previous antigenic studies and protein analyses of the human passage isolates had shown stability in their properties upon serial passage and that most of the original genetic structure of the isolates was well conserved2 -5 These observations were supported by the excellent worldwide performance of the Sabin oral poliovaccine that was found to be one of the safest vaccines used against infectious diseases. However, with the use of nucleotide sequence analysis of poliovirus strains in the last 5 years, evidence has been ac- cumulated that pointed to regions in the virus genome, 2' or even to single nucleotides, tha t clearly seem to play a role in determining virulence or attenuation of polioviruses. 6-9

Investigators conducting the 1960 trials in Canada ',2 performed limited monkey neurovirulence tests on some of the human passage strains. They reported that with increasing passage there was some loss of attenuation for the monkey. The monkey neurovirulence test results presented in this paper showed that, for both types 1 and 3, the loss of at tenuation of the vaccine strain upon passage in the gut was gradual, although the loss was faster for type 3. The increase of lesion scores for type 3 following multiplication of 4 days in the intestinal tract was higher than that observed for type 1 isolated after 15 days in the gut. It is noteworthy that the neurovirulence of type 1 human passage strains did not reach the "degree of that observed in the 'wild' strains, whereas the human passage 7 virus strains of type 3 was found to be as neurovirulent as the 'wild' strains. This faster change for type 3 was also observed when 30 infants were fed trivalent OPV for the first time at 3 months of age, and the viruses excreted for 28 days were analysed. 22 The apparent stability of type 1 vaccine strain during first human passage correlates well with the lack of change in nucleotide 480 in the 5'- non-coding region which remained a G, as in the Sabin 1 strain; in contrast, base 472 of type 3 first passage at 4 days in the human gut changed from a U to a C. It is of interest that, for type 1, base 480 remained a G not only for passage 1 but also for passages 3, 4 and 5, possibly confirming its role in main- taining attenuation as postulated by Kawamura et

a l . 8 and reported by Dunn et a l . 22 However, the change for type 1 in position 476 from a U to an A in passages 3, 4 and 5, but not in passage 1, could be correlated with the increased monkey neurovirulence observed for passages 3, 4 and 5 compared to passage 1. Position 476 is also an A in the three 'wild' strains tested.

A similar picture emerges from the single nucleotide changes observed for type 3 passage isolates. Besides the mutation of nucleotide 472, 6 addi- tional point mutations observed in isolates tha t have been multiplying for more days in the gut could play a role in explaining their higher lesion scores. The change from a U to a C in position 2034 reported by Westrop et a l . 7 was detected in passage 1 'late' at 18 days in the intestinal tract and in all subsequent passage isolates; however, base 2034 remained a U in passage 1 'early' at 4 days in the gut.

It is noteworthy that the nucleotide changes expe- rienced by types 1 and 3 following human passage were stable, both in the 5'-non-coding region for types 1 and 3 (U to an A in position 476 and a U to a C in position 472 for types 1 and 3 respectively) and for the coding region of type 3 (U to a C in position 2034). There was an unstable change for type 3 in position 1973 which changed from a U to a C for passages 5 and 6, but then reverted to a U for passages 7L and 7LL. The stability of a nucleotide change probably means tha t it generates some advantage for the multiplication of the mutated virus in the human gut. If this would be the case it is difficult to understand what pressures generate the reversion.

Research with models for the secondary structure of poliovirus RNA suggest that the 5'-non-coding region is important for the virus neurovirulence. 23 Other regions of the poliovirus genome that were not examined in this study might also contribute to the changes to virulence upon human passage; a cur- rent project dealing with the sequence in other regions of selected strains will a t tempt to answer this question.

For both poliovirus vaccine types there was a good correlation between an increase in the number of base substitutions on passage with an increase in the number of spot differences in the oligonucleotide maps, but conserving the general distribution pattern of the spots. Other investigators had also observed that despite changes in single spots, the general distribution pattp.rn in the type 3 oligonucleotide maps was maintained in several viruses isolated over a period of 52 days from an infant vaccine recipient. '9 Conversely, when poliovirus strains other than the vaccine viruses were analysed in this laboratory by both procedures, it became clear tha t virus strains showing several differences in their nucleotide sequences yielded oligonucleotide maps with entirely different spot patterns. Changes in the oligonucleotide maps occurred more rapidly for type 1 (seven spot differences in 35 days) than for type 3


(nine spot differences in 109 days) upon human passage. These rates of change are greater than those found in 'wild' strains isolated from one epidemic. 24 The significance of these findings is difficult to inter- pret at this time.

The diversity of the poliovirus genome has been suspected for many years, based on different biological properties of various strains. The diversity was first observed as intratypic difference in their antigenic structure, 3'4 then by the analysis of viral pro- teins, 5,24 by oligonucleotide mapping, 5.24 and recently confirmed by sequence analysis of the virus RNA.lS These newly developed assays were successfully applied to characterize virus strains isolated from paralytic patients for determining their 'vaccine' or 'wild' origin and to monitor the spread of viruses responsible for poliomyelitis epidemics across the world.24. 25

It is evident from these data tha t there are several poliovirus genotypes compatible with neurovirulence as shown by the different sequences of the human passage and other types 1 and 3 viruses which were assessed for monkey neurovirulence. Similar information is being investigated by other laboratories using mutants ofpoliovirus which differ from the temperature-sensi t ive parental Sabin 1 strain by only a few nucleotide changes. 9 However, in addition to the virus characteristics one has to look at the response of the human host, which depends on many variables: age, gender, genetic predisposition, humoral and cellular immune status etc. The com- bination of virus and host factors makes it difficult to predict the outcome of poliovirus infections; good example are the Mahoney (type 1) and 1034 (type 3) strains which, although highly neurovirulent in monkeys, were isolated from heal thy children. A recent observation on the response of inbred mouse strains to type 2 polioviruses showed that the susceptibility for developing paralysis after intracere- bral inoculation with a t tenuated strain W-2 varied in different mouse strains; there was no correlation of the different responses with the mouse's haplo- type. 26

The emerging picture of the study of the human passage viruses is simultaneously one of change, but also stability. Moreover, different virus genotypes may be compatible with the same expression in the human host. However, complete sequencing of the genome of selected human passage virus strains already in progress, could fur ther identify the critical nucleotides tha t are responsible for the attenuation of polioviruses.

Acknowledgements We are grateful to Dr Olen Kew for providing

some of the poliovirus strains and valuable information; also to Dr Earl Brown for sharing his experience on nucleotide mapping and sequencing of RNA viruses. We acknowledge the excellent technical assistance of Silvie Rivard, Frances Kane and Gaston Labelle.

References

1. Pavilanis V, Lussier G, Foley AR, Charbonneau JH, Di Franco E, Dubreuil R. A study of Sabin type 1 oral vaccine virus during five human passages. Rev Can Biol 1964; 23: 265-276.

2. MacLeod DRE. The neurovirulence of the type 3 Sabin vaccine strain after passage in infants. Can J Public Health 1966; 57: 37.

3. Furesz J, Armstrong RE, Moreau P, Yarosh W, Nagler FP. Antigenic studies on Sabin types 1 & 3 poliovaccine virus during 1 to 7 passages in the human intestinal tract. Am J Epidemiol 1966; 83: 501-508.

4. Wachmann B, Henning D, Armstrong R, Boucher DW, Contreras G, Furesz J. Use of monoclonal antibodies prepared with Sabin vaccine viruses for the characterization of poliovirus strains isolated in Canada. Dev Biol Stand 1984; 57: 141-149.

5. Contreras G, Furesz J, Wu L-X Wachmann B McCorkle G. Genetic stability of Sabin types 1 and 3 poliovaccine virus following serial passages in the human intestinal tract. Paper presented at the 5th International Conference on Comparative and Applied Virology, Lake Louise, Alberta, Canada, May 1986.

6. Evans DMA, Dunn G, Minor PD, Schild GC, Cann AJ, Stanway G, Almond JW, Currey K, Maizel Jr JV. Increased neurovirulence associated with a single nucleotide change in a non-coding region of the Sabin type 3 poliovaccine genome. Nature 1985; 314: 548-550.

7. Westrop GD, Wareham KA, Evans DMA, Dunn G, Minor PD, Magrath DI, Taffs F, Marsden S, Skinner MA, Schild GC, Almond JW. Genetic basis of attenuation of the Sabin type 3 oral poliovirus vaccine. J Virol 1989; 63: 1338-1344.

8. Kawamura N, Kohara M, Abe S, Komatsu T, Tago K, Arita M, Nomoto A. Determinants in the 5' noncoding region of poliovirus Sabin 1 RNA that influence the attenuation phenotype. J Virol 1989; 63:1302-1309.

9. Christodoulou C, Colbere-Garapin F, Macadam A, Taffs LS, Mardsen S, Minor PD, Horaud F. Mapping of mutations associated with neurovirulence in monkeys infected with Sabin 1 poliovirus revertants selected at high temperature. J Virol 1990; 64: 4922--4929.

10. WHO requirements for poliomyelitis vaccine (oral). WHO Expert Committee, on Biological Stand- ardization. WHO Tech Rep Ser 1983; 687:130-134 and 158-162.

11. Contreras G, Furesz J, Karpinski K, Grinwich K, Gardell C. Experience in Canada with the new revised

26 G. Contreras et al.

monkey neurovirulence test for oral poliovirus vaccine. J Biol Stand 1988; 16: 195-205.

12. Pederson FS, Haseltine WA. A micromethod for detailed characterization of highmolecular weight RNA. Methods Enzymol, 1980; 65: 680-687.

13. Lee YF, Fowlks ER. Rapid in vitro labelling procedures for two-dimensional gel fingerprinting. Anal Biochem 1982; 119: 224-235.

14. Yoneyama T, Hagiwara A, Hara M, Shimojo H. Alteration in oligonucleotide fingerprint patterns of the viral genome in poliovirus type 2 isolated from paralytic patients. Infect Immun 1982; 37: 46-53.

15. Geliebter J. Dideoxynucleotide Sequencing of RNA and uncloned cDNA. Focus 1987; 9: 5-8.

16. Kawasaki ES. Amplification of RNA In: PCR Protocols: a Guide to Methods and Applications. New York: Academic Press 1990; 21-27.

17. Maniatis T, Fritsch EF, Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor. New York: Cold Spring Harbor Laboratory 1982; p. 394.

18. Nomoto A, Omata T, Toyoda H, Kuge S, Horie H, Kataoka Y, Genba Y, Nakano Y, Imura N. Complete nucleotide sequence of the attenuated poliovirus Sabin 1 strain genome. Proc Natl Acad Sci USA 1982; 79: 5793-5797.

19. Minor PD, John A, Ferguson M, Icenogle JP. Antigenic and Molecular Evolution of the Vaccine Strain of Type 3 Poliovirus during the Period of Excretion by a Primary Vaccinee. J Gen Virol 1986; 67: 693-706.

20. Stanway G, Cann AJ, Hauptmann R, Hughes P, Clarke LD, Mountford RC, Minor PD, Schild GC Almond JW. The nucleotide sequence of poliovirus

type 3 Leon 12 alb: comparison with poliovirus type 1. Nucleic Acids Res 1983; 11: 5629-5643.

21. Omata T, Kohara M, Kuge S, Konatsu T, Abe S, Semler BL, Kameda A, Itoh H, Arita M, Wimmer E, Nomoto A. Genetic Analysis of the Attenuation Phenotype of Poliovirus Type 1. J Virol 1986; 56: 348-358.

22. Dunn G, Begg NT, Cammack N, Minor PD. Virus excretion and mutation by infants following primary vaccination with oral poliovaccine from two sources. J Med Virol 1990; 32: 92-95.

23. Skinner MA, Racaniello VR, Dunn G, Cooper J, Minor PD, Almond JW. New Model for the secondary structure of the 5' non-coding RNA of poliovirus is supported by biochemical and genetic data that also shows that RNA secondary structure is important in neurovirulence. J Mol Biol 1989, 207: 379-392.

24. Nottay BK, Kew OM, Hatch MH, Heyward JT, Obijeski JF. Molecular variation of type 1 vaccine- related and wild polioviruses during replication in humans. Virology 1981; 108: 405-423.

25. Rico-Hesse R, Pallansch MA, Nottay BK, Kew OM. Geographic distribution of wild poliovirus type genotypes. Virology 1987; 160: 311-322.

26. Jubelt B, Ropka SL, Goldfard S, Waltenbauch C, Oates RP. Susceptibility and resistance to poliovirus- induced paralysis of inbred mouse strains. Virology 1991, 65: 1035-1040.

Received for publication 2 August 1991; accepted 15 October 1991.

genetic characterization of sabin types 1 and 3 poliovaccine virus following serial passage in the...

Documents