_Arehives_

Vfrology

Arch Virol (1991) 120:43-55

© Springer-Verlag1991 Printed in Austria

Genetic diversity and similarity among mammalian rotaviruses in relation to interspecies transmission of rotavirus O. Nakagomi 1 and T. Nakagomi 2

Departments of 1Laboratory Medicine, and 2Microbiology, Akita University School of Medicine, Akita, Japan Accepted January 20, 1991

Summary. To address the question whether there was any molecular evidence for interspecies transmission of rotaviruses from one animal species to another, genetic relationships among human and animal rotaviruses were examined by a series of hybridization experiments in which genomic RNAs from 14 rotavirus strains derived from. seven different host species were hybridized with the [32p]_ labelled transcription probes prepared from 11 strains representing rotaviruses from those seven host species. In general, higher level of homology among most, if not all, of the cognate gene segments that allowed classification into the same genogroup was shared among rotaviruses recovered from the same animal species but this level of homology was not found among rotavirus strains derived from different host species. However, such a high level of homology that was usually found among rotaviruses recovered from the same animal species was detected between feline rotavirus strain Cat97 and canine rotavirus strain K9 as well as between human rotavirus strain AU-1 and feline rotavirus strain FRV-1. The sharing of closely related genetic constellation of most of the 11 gene segments (genogroup) by rotaviruses recovered from different animal species provided molecular evidence that interspecies transmission of rotaviruses occurred in nature at least recently in the evolutionary history.

Introduction Group A rotaviruses, family Reoviridae, genus Rotavirus, have been recovered from the young of many mammalian species including humans and implicated as major causative agents of diarrhea [ 14]. An earlier study by plaque reduction neutralization assays depicted serotypic diversity and similarity of rotaviruses of mammalian and avian origin under a unified scheme [10]. Epitopes resided on the major outer-capsid glycoprotein, VP7, are primarily responsible for the

44

O. Nakagomi and T. Nakagomi

serotype specificity determined by neutralization assays with hyperimmune sera [4]. This serotype has been proposed to term G type (for gtycoprotein) [4, 11]. The number of G types has recently increased to 11 including 10 types from mammalian rotaviruses and one from avian rotaviruses [t, 3, 10, 18, 19, 30, 36, 39]. Nucleic acid hybridization assays under stringent conditions have been used to examine the genetic background of those serotypically diverse rotavirus strains from various mammalian origin. Earlier works done by Schroeder et al. with RNA-cDNA hybridization [33] and by Flores et at. with RNA-RNA hybridization [5, 7] demonstrated that homology is greater among strains derived from the same animal species than among strains derived from different animal species. Furthermore, extensive sequence analysis of the VP7 genes of rotavirus strains from a broad spectrum of host species [31] and sequence analysis on the trypsin-cleavage region of the VP4 gene from various rotavirus strains [13] have lent a strong support to the concept of the restricted host range of rotavirus. In addition, an epidemiologic study in Panama of farm workers with frequent contact with cattle which indicated that rotaviruses were transmitted within family units but not to or from animals [32], providing further support for the existence of the species barrier between human and animal rotaviruses. On the other hand, molecular evidence for interspecies transmission of animal rotaviruses to humans has emerged through the analysis with RNA-RNA hybridization techniques of human rotavirus strains with an unusual combination of serotype (G3), subgroup (I) and RNA pattern (long) [25, 29]. For example, AU228 strain which was isolated from a child with a clinical history of contact with a cat was shown by RNA-RNA hybridization to be more closely related to feline rotaviruses than to any other human rotaviruses [28]. Further molecular evidence is accumulating which suggests that interspecies transmission of rotaviruses from one animal species to another is taking place in nature [9, 27]. Apart from its evolutionary interest, the issue of the zoonotic origin of novel human strains as whole virions or by genetic reassortment has a significant medical importance in the development of rotavirus vaccines by "Jennerian" and "modified Jennerian .... approaches. In this study, we have carried out a series of hybridization experiments in which genomic RNAs from 14 rotavirus strains derived from seven different host species were hybridized with the probes prepared from 11 strains representing rotaviruses from those seven host species.

Materials and methods Viruses

Cell-culture adapted human and animal rotavirus strains used in this study are summarized in Table 1.

45

Genetic homology among mammalian rotaviruses Table 1. Rotavirus strains used in this study Strain

Species

G type

Subgroup

Reference

Wa KUN AU- 1 FRV- 1 Cat97 RS 15 K9 BI HI23 SA 11 RRV OSU NCDV KK3

human human human cat cat dog dog foal foal vervet monkey rhesus monkey pig cattle cattle

1 2 3 3 3 3 3 3 3 3 3 5 6 l0

II I I I I I I nd nd I I I I I

38 15 26 28 2 22 8 12 12 16 35 37 20 23

nd Not determined Preparation of dsRNA Genomic double-stranded RNA (dsRNA) was extracted with phenol-chloroform from the partially purified virions which were prepared from infected MA104 cells by pelleting them at 38,000 rpm for 3 h in a Hitachi RP42 rotor and then by sedimentation through 30% (wt/vol) sucrose at 38,000 rpm for 3 h in a Hitachi RPS40T rotor.

Preparation of ssRNA transcripts Single-stranded RNA (ssRNA) probes were prepared by in vitro transcription of rotavirus genomic RNA in the presence of [32p]GTP as described previously [24].

RNA-RNA hybridization RNA-RNA hybridization in solution was performed as previously described [24]. Briefly, genomic RNA was denatured by 2 rain of incubation at 100 °C, followed by quenching on ice for 2 rain. 32p-labelled probes (20,000 cpm for each denatured dsRNA) were added; and hybridization was allowed to occur at 65°C for 16h in a buffer containing 5raM Tris acetate, 150 mM NaC1, 1 mM EDTA, and 1% sodium dodecyl sulfate (pH 7.5). The resulting hybrids were fractionated on a 10% potyacrylamide gel. The gels were stained with ethidium bromide and then exposed to X-Omat AR films (Eastman Kodak Co., Rochester, N.Y.).

Results Analysis o f the results obtained by R N A - R N A hybridization in solution u n d e r stringent conditions requires the u n d e r s t a n d i n g o f its inherent properties a n d limitation. First, the hybridization c o n d i t i o n used in this study was calculated to allow up to t 8 % nucleotide sequence m i s m a t c h [24]. Secondly, since hybrids were separated on a p o l y a c r y l a m i d e gel after hybridization in solution, hom o l o g o u s b a n d s were identified by bands that c o m i g r a t e d with the c o r r e s p o n d -

46

O. Nakagomi and T. Nakagomi

ing genomic R N A segments, whereas hybrids consisting of a lower degree of homology can be observed as aberrantly migrating bands with lesser intensity. The exact identification of the gene segment involved in a hybrid is thereby very difficult except that there is a unique hybrid and a unique empty space of a homologous hybrid. Nevertheless the number of hybrids (and their relative intensity) formed between genomic R N A from a rotavirus strain and a probe is considered to be indicative of relative and overall gene homology between two strains. For simplicity and comparison, we arbitrarily defined a minimum level of homology between two strains as formation of two hybrid bands or less, a low level of homology as formation of three to four hybrid bands, a medium level of homology as formation of five to seven hybrid bands, and a high level of homology as formation of eight hybrids or more. Denatured genomic R N A s from 14 mammalian rotavirus strains recovered from seven different host species were hybridized to the 3ap-labelled transcription probes which were prepared from 11 rotavirus strains representing rotaviruses of those seven m a m m a l i a n host origin. The rotavirus strains from which probes were made included three h u m a n rotaviruses Wa, K U N , and AU-1 each of which are k n o w n to represent three distinct h u m a n rotavirus genogroups [24], two feline rotaviruses FRV-1 and Cat97, a canine rotavirus K9, an equine rotavirus 32p. p r o b e s

from=

Wa

KUN

humans

AU-1 FRV -1

cats

~0-2

bands

dogs

~3-4

bands

foals

~5-7

bands

Cat97 L. e3

RS15 K9

ztY BI U

H123

SA11

monkeys ~ 8 - 1 1

bands

e=

RRV OSU

pig

NCDV

cattle

KK-3

Fig. 1. Relative homology among rotavirus strains derived from 7 different mammalian species. The relative homology is schematically expressed by the number of hybrid segments formed between genomic RNAs and [32P]-labelled transcription probes. We arbitrarily defined a minimum level of homology between two strains as formation of two hybrid bands or less, a low level of homology as formation of three to four hybrid bands, a medium level of homology as formation of five to seven hybrid bands, and a high level of homology as formation of eight hybrids or more

Genetic homology among mammalian rotaviruses

47

dsRNA

A

1

23 4

78 9

10 11

B

1 23

4

78

9

lO 11

I

Probe

~ I

Wa

IL

KUN

_

_

J

AU'I

Fig. 2. Hybridization patterns obtained between genomic RNAs from Wa, KUN, and AU-1, and the [32p]-labelled transcription probes made from these three strains. A Ethidium bromide stained gels under UV light illumination. B Corresponding autoradiograms. RNA segments are indicated to the left of each panel

B!, two simian rotaviruses SA11 and RRV, a porcine rotavirus OSU and a bovine rotavirus KK-3. Figure 1 shows the summarized result of the crosshybridization with regard to relative homology among rotavirus strains. When genomic RNAs from a panel of 14 rotavirus strains were hybridized to the probes prepared from three hmnan rotaviruses, Wa, KUN, and AU-1 (Figs. 2 and 3a-c), a high degree of homology was found only between the genomic RNA from FRV-1 and the AU-1 probe. The AU-t probe also showed a medium to low level of homology with feline and canine strains. By contrast, the Wa and K U N probes did not exhibit any significant homology with other animal rotavirus strains except a low level of homology between the Wa probe and porcine strain OSU. A corresponding reciprocal hybridization pattern was recognized when the genomic RNA from AU-1 was hybridized with the probe prepared from feline rotavirus FRV-1 (Fig. 3d). In addition, the FRV-1 probe showed a medium level of homology with canine rotavirus RS15, whereas it showed only a low level of homology with feline rotavirus strain Cat97 and canine rotavirus strain K9. Hybridization of genomic RNAs from the 14 rotavirus strains with the probes made from feline rotavirus strain Cat97 and canine rotavirus strain K9 yielded similar results (Fig. 3e and f). Thus, the Cat97 probe formed eight or

48 o. Nakagomi and T. Nakagomi: Genetic homology among mammalian rotaviruses more hybrids with two canine strains RS15 and K9, whereas it showed only a minimum homology with another feline strain FRV-1 (Fig. 3e). Similarly, the K9 probe produced eight or more hybrids with genomic RNAs not only from another canine strain RS15 but also from feline strain Cat97 (Fig. 31). While it showed a low level of homology with human AU-1 and feline FRV-1, the K9 probe displayed no significant homology with other rotaviruses. When the BI probe was hybridized to the genomic RNAs from the panel of rotaviruses, only dsRNA from another equine strain HI23 yielded largest number of hybrids (five) (Fig. 3g). Two monkey rotaviruses SA11 and RRV produced quite different hybridization patterns. While the SA11 probe yielded minimum numbers of hybrids with genomic RNAs from any of the 14 rotavirus strains including RRV (Fig. 3h), the RRV probe showed a low level of homology with a number of rotaviruses from different animal species including AU-1, FRV-1, Cat97, RS 15, K9, BI, HI23, NCDV, and KK3 (Fig. 3i). When hybridization was performed with the probe prepared from porcine strain OSU, it was revealed that the only rotavirus strain which showed a significant level of homology with the OSU probe was human rotavirus Wa (Fig. 3j). Genomic RNAs from other rotavirus strains exhibited no homology with the OSU probe. When the KK-3 probe was hybridized to the genomic RNAs from the panel of rotaviruses, a high degree of homology was found only with the genomic RNA from another bovine strain NCDV (Fig. 3k). A low level of homology was noted between the KK-3 probe and genomic RNA from equine strain BI (Fig. 3k), confirming the reciprocal hybridization result between genomic RNA from KK-3 and the BI probe (Fig. 3g).

Discussion

Understanding the extent of genetic relatedness among human and animal rotaviruses has provided insight into rotavirus evolution. Like other viruses that have segmented RNA genome such as influenza A virus, both antigenic "shift" which is the result of gene reassortment and "drift" which is caused by successive accumulation of mutations are anticipated to occur during the course of viral evolution [5, 6, 33, 34]. Furthermore, the ubiquitous nature of rotavirus suggests that rotavirus strains from one animal species may infect other animal

Fig. 3 a-k. Hybridization patterns obtained between a the Wa probe, b the KUN probe,

c the AU-1 probe, d the FRV-1 probe, e the Cat97 probe, f the K9 probe, g the BI probe, h the SAll probe, i the RRV probe, j the OSU probe, k the KK-3 probe and genomic RNAs from various mammalian rotavirus strains indicated on the top of the lanes. A Ethidium bromide stained gels under UV light illumination. B Corresponding autoradiograms. RNA segments are indicated to the left of each panel

A 1 23 4

789

lO 11

B

B

1

I

23 4

23 4

78

789

9

10 11

10

Wa

probe

KUN

b

probe

A 1

1

23 4

23 4

789 789 10 10 11

B

B

1

1

23 4

23 4

789 789 lO

lo

....

C

AU-1

probe

m

~

.....

d

FRV-1

probe

A 1

1

23 4

23 4

5

5

6

6

789

78

10

10 11

11

B 1 2 3 4

789

10

11

tl

K9

probe

Cat 97 probe

,~

~