JGV Papers in Press. Published February 28, 2014 as doi:10.1099/vir.0.057984-0
1
Journal category: Animal Viruses - Double-strand RNA
2
Genome sequencing identifies genetic and antigenic divergence of porcine
3
picobirnaviruses
4
Krisztián Bányai 1 , Christiaan Potgieter 2,3, Ákos Gellért 4, Balasubramanian Ganesh
5
5
, Maria Tempesta 6, Eleonora Lorusso 6, Canio Buonavoglia 6, Vito Martella 6
6 7
1
8
Academy of Sciences, Budapest, Hungary
9
2
Virology Division, Onderstepoort Veterinary Institute, Onderstepoort, South Africa
10
3
Biochemistry Division, North-West University, Potchefstroom, South Africa
11
4
Agricultural Institute, Centre for Agricultural Research, Hungarian Academy of Sciences,
12
Martonvásár, Hungary
13
5
14
West Bengal, India
15
6
Institute for Veterinary Medical Research, Centre for Agricultural Research, Hungarian
Division of Virology, National Institute of Cholera and Enteric Diseases (NICED), Kolkata,
Dipartimento di Medicina Veterinaria, Università di Bari Aldo Moro, Bari, Italy
16 17 18 19 20
Correspondence to:
Krisztián Bányai, Institute for Veterinary Medical Research, Centre for Agricultural Research, Hungarian Academy of Sciences, Hungária krt 21., H-1143 Budapest, Hungary; E-mail: [email protected] Vito Martella, Dipartimento di Medicina Veterinaria, Università di Bari Aldo Moro, S.p. per Casamassima Km 3, 70010 Valenzano, Bari, Italy; E-mail: [email protected]
21 22 23 24 25 26
Running title: Porcine picobirnavirus genome
27
Word count in summary: 193
28
Word count in text: 2570
29 30
GenBank accession numbers: KF861767-KF861773
1
31
SUMMARY
32
The full-length genome sequence of a porcine picobirnavirus detected in Italy in 2004
33
was determined. The S segment was 1730 nucleotide (nt) in length, coding for a
34
putative RNA-dependent RNA polymerase. Two distinct sub-populations of L
35
segment (LA and LB) were identified in the sample, with the sizes ranging from 2351
36
to 2666 nt. The ORF1 coding for a protein of unknown function, contained repetitions
37
of the ExxRxNxxxE motif in variable number. The capsid protein coding ORF2
38
spanned nt 810-2447 in the LB variants and started at nt 734 in the LA variants.
39
However, a termination codon was present only in one of all LA segment variants.
40
Three dimensional modelling of the porcine PBV capsids suggested structural
41
differences in the protruding domain, tentatively involved as antigens in humoral
42
immune response. Altogether, these findings suggest simultaneous presence of two
43
different picobirnavirus strains sharing the same S segment but displaying genetically
44
diverse L segments. In addition, the sample probably contained a mixture of PBVs
45
with aberrant RNA replication products. Altered structure in the L segments could be
46
tolerated and retained in the presence of functionally integer cognate genes and
47
represent a mechanism of virus diversification.
48
Keywords: Picobirnaviridae, reassortment, pig, zoonosis
2
49
TEXT
50
Picobirnaviruses (PBVs), family Picobirnaviridae, are small (35–41 nm in diameter),
51
non-enveloped viruses displaying a distinctive icosahedral capsid. The PBV genome
52
is bisegmented double-stranded RNA (dsRNA). The larger (L) genome segment is
53
2.2–2.7 kb in length and encodes the capsid protein and a putative protein of
54
unknown function, while the smaller (S) genome segment is 1.2–1.9 kb in length and
55
encodes the viral RNA-dependent RNA polymerase (Delmas, 2011). Nucleotide
56
sequence differences in the S segment permit the classification of PBVs into two
57
major clusters, designated as genogroup I and genogroup II. The intra- and
58
intergenogroup sequence similarities of a short nucleic acid fragment of the S
59
segment used in diagnostic PCR ranges from 49 to 100 % and 28 to 37 %,
60
respectively (Bányai et al., 2003; Rosen et al., 2000).
61
PBVs have been found in the feces and intestinal content and, most recently, in the
62
respiratory tract of vertebrates. While the list of host species with record of PBV
63
infections has increased continuously (Fregolente et al., 2009; Gallimore et al., 1995;
64
Ganesh et al., 2011; Ghosh et al., 2009; Gillman et al., 2013; Green et al., 1999;
65
Haga et al., 1999; Malik et al., 2011; Masachessi et al., 2007, 2012; Pereira et al.,
66
1988; Wang et al., 2007; Woo et al., 2012), an association of PBV infection with
67
enteric or respiratory diseases has not been formally demonstrated. In fact, PBV is
68
often detected together with co-infecting major enteric pathogens or in clinical
69
specimens from inapparently infected hosts. The lack of cell culture or animal model
70
further hinders the recognition of any disease associations. However, studies in
71
immuncompromised hosts have suggested an opportunistic role for PBV in diarrheic
72
infections (Giordano et al., 1998; Grohmann et al., 1993).
73
Transmission of PBV from one host species to another has been suggested in
74
molecular epidemiological and strain characterization studies, based on very close
75
genetic relatedness between human and porcine or human and equine PBV strains
76
(Bányai et al., 2008; Ganesh et al., 2011; Giordano et al., 2011). Given that PBV is
77
frequently found in waste water and surface water, a water-born route of
78
transmission has been also suggested (Ganesh et al., 2011). Indeed, the genetic
79
similarity among heterologous PBV strains has been seen along a short, ~200 bp
3
80
gene fragment of the S segment. Because the PBV genome is segmented and
81
measures ~ 4 kbp, the value of this short fragment to describe molecular
82
epidemiology may be limited. A major obstacle to perform whole genome based
83
studies is the lack of genome sequence data for a reasonably high number of strains.
84
Thus far, the complete genome sequence has been determined only for a human
85
strain, and more recently, for a phocine PBV strain (Wakuda et al., 2005; Woo et al.,
86
2012). In addition, complete or partial L segment sequences have become available
87
for a lapine strain and some human strains (Green et al., 1999; Rosen et al., 2000),
88
while the sequences of a dozen or so S gene fragments longer than 1000 bp has
89
been determined for human, porcine, simian, bovine, and murine PBVs (Ghosh et al.,
90
2009; Phan et al., 2011; Rosen et al., 2000; van Leeuwen et al., 2010; Wang et al.,
91
2012; and unpublished GenBank entries). In this study, sequencing and molecular
92
characterization of the full-length genome of a porcine PBV strain was performed by
93
sequence independent amplification and high throughput sequencing of the genomic
94
dsRNA.
95
During a rotavirus surveillance project in 2004, total RNAs prepared from 37
96
randomly selected fecal samples collected from diarrheic piglets were analyzed by
97
polyacrylamide gel electrophoresis and silver staining to diagnose rotavirus
98
infections. One sample contained bands consistent with picobirnavirus dsRNA (Fig.
99
1a). This porcine PBV strain, designated 221/04-16 was detected from a 45-day-old
100
piglet with diarrhea at a farm in Brescia, Northern Italy, in 2004.
101
Total RNA was extracted from fecal specimen using the commercial TRI-REAGENT–
102
LS (Molecular Research Centre), according to the manufacturer’s protocol. Single
103
stranded RNA (ssRNA) was removed by precipitation with 2 M LiCl (Sigma) at 4 C for
104
16 h followed by centrifugation at 16 000 g for 30 min. dsRNA was purified from the
105
resulting supernatant using a MinElute gel extraction kit (Qiagen). The complete
106
nucleotide sequence was determined for the small (S) and large (L) segments of this
107
strain by sequence-independent amplification and deep sequencing as described
108
previously (Potgieter et al., 2002, 2009; Lambden et al., 1992). Briefly, an anchor
109
primer was ligated to dsRNA and used for reverse transcription and subsequent
110
amplification of virus genome segments. Amplified cDNA products were then
4
111
subjected to sequencing with GS FLX technology at Inqaba biotechTM (Pretoria, South
112
Africa).
113
Lasergene7 software from DNASTAR was used for de novo assembly of the contigs.
114
Files containing the sequence information, quality values and flowgrams (sff files)
115
were loaded into the Seqman 7 programme of the Lasergene software. Contigs
116
resulting from the assembly were checked manually and their consensus sequences
117
were exported as FASTA files. Consensus sequences were aligned to known
118
sequences using MEGALIGN (Lasergene7). Finally, sequences were subjected to
119
BLASTN analysis using the National Center for Biotechnology Information website.
120
The
121
(http://www.ncbi.nlm.nih.gov/gorf/gorf.html). Nucleotide and amino acid sequence
122
alignments were prepared by Multalin (Corpet, 1988) and manually adjusted to
123
codons. Phylogenetic analysis was performed by using MEGA5 (Tamura et al.,
124
2011). Sequences were deposited in GenBank under the following accession
125
numbers: KF861767-KF861773.
126
Small genome segment. The S segment of strain 221/04-16 was 1730 bp in length.
127
In comparison with other full-length PBV S segments, this segment is shorter than
128
those of the human strain Hy005102 (1745 bp), the monkey strain CHN-49/2002
129
(1784 bp), and the bovine strain RUBV-P/IND/20 (1758 bp) and longer than those of
130
the human strains 1-CHN-97 (1696 bp) and GPBV6C1 (1711 bp), and the otarine
131
PBV (1687 bp). The length of the 5’ and 3’ UTRs were 74 and 49 bp, respectively,
132
and were structurally heterogenous when compared with reference strains, except for
133
the common GTAAA and ACTGC pentanucleotides, located at the termini of the RNA
134
segments and conserved among the various PBV strains. The S segment of strain
135
221/04-16 contained a single 1607 nt-long ORF (nt 75 to 1682). Alternative ORFs
136
were also predicted (data not shown). However, the predicted proteins were too short
137
and their presence was not consistently found in the S segments of other strains;
138
therefore these putative ORFs were ignored in this analysis. The protein coded by
139
the S segment ORF was 535 aa long and was predicted to be the viral RNA
140
dependent RNA polymerase (RdRp). It had conservative motifs in aa positions 266-
141
271, 328-333 and 364-366 defining the GDD sequence motif. When comparing the
142
221/04-16 S gene sequence with the Chinese strain SD, a single nucleotide deletion
NCBI’s
ORF
finder
was
used
to
identify
protein
coding
regions
5
143
was observed at position 768 that was compensated at position 848 by a single
144
nucleotide insertion. This insertion caused a codon shift, altering the translation of a
145
28 aa stretch between the two porcine PBVs (Fig. 1c).
146
Phylogenetic analysis was first performed using a ~200 bp fragment of the RdRp
147
used to amplify GGI PBV strains in diagnostic PCR (Rosen et al., 2000). This
148
analysis identified a Chinese porcine PBV strain as presenting the greatest nt
149
sequence identity (95%) with that of the Italian strain. Other relatively closely related
150
sequences (up to 89%) included viral sequences identified from US water samples.
151
The Italian porcine PBV strain shared lower nt similarity to European and American
152
(up to 75%) porcine PBVs and shared comparable similarities with European, Asian,
153
and American human PBVs (up to 88%). Another, ~280 bp long, partially overlapping
154
region of the S segment used to specifically amplify porcine PBVs (Carruyo et al.,
155
2008) was analyzed separately. This analysis demonstrated that the Italian strain
156
shared somewhat greater similarities with Argentinean porcine and human PBV
157
strains (up to 97% nt identity; data not shown). An analysis of the whole ORF of
158
several strains of porcine, human, murine, phocine, and bovine origin identified
159
comparable sequence identity data when these strains were compared along the
160
commonly characterized ~200 bp long fragment. The same observation was made
161
when comparing the Italian strain with two genogroup II strains, 4-GA-91 and
162
GPBV6G2. However, it should be considered that extrapolating nucleotide similarities
163
obtained from a short fragment may not be relevant in each case as exemplified by
164
the codon shift described above. Phylogenetic analysis of an ~1200 bp fragment of
165
selected strains with longer S genome segment sequence confirmed the similarity
166
based prediction of the genetic relatedness between the Italian strain and the
167
Chinese SD strain (Fig. 1b).
168
Large genome segment. Amplification and cloning of the L segment yielded a
169
heterogeneous population of cDNA products, comprising two major types and
170
several minor sequence variants within each type. The major sequence types were
171
designated LA and LB. The LB variants (ie LB-1, LB-3, and LB-4) were 2524 bp in
172
length and had two predicted ORFs at nt 195-794 (ORF1) and 810-2447 (ORF2),
173
encoding proteins of 199 and 545 aa in length, respectively. The LA variants, LA-2,
174
LA-5 and LA-7, respectively, were 2361, 2666 and 2351 bp long. The ORF1 started
6
175
at nt 185 and ended at nt 721, coding for a polypeptide of 175 aa in length. The
176
ORF2 started at nt 734 in all LA sequences. However, while LA-5 was terminated
177
with a stop codon, LA-2 and LA-7 seemed to be pseudogenes without translation
178
termination signals. The protein encoded by the LA-5 sequence was longer (619 aa)
179
than those of other PBV capsid proteins (ie, 590 aa in the lapine PBV strain, 552 aa
180
in the human strain Hy005102 and 575 aa in the otarine strain). The 3’ UTRs ranged
181
from 184 (LA segments) to 194 nt (LB segments). The 5’ UTRs ranged from 76 (LB
182
segments) to 84 nt (LA-5).
183
The ORF1-encoded proteins of the LA and LB clones were extremely diverse, yet
184
retained common genetic signatures. One shared feature was the recently
185
recognized 10-aa long sequence motif (Da Costa et al., 2011), ExxRxNxxxE, where
186
x, can be any amino acid. Four and 6 repeats of this motif was identified in LB clones
187
(e.g. LB-3) and LA clones (e.g. LA-5), respectively. The amino acid distance between
188
two adjacent motifs were variable. When comparing the ORF1s with other reference
189
strains, we found similar diversities in the number and position of the ExxRxNxxxE
190
motif (Fig 1d).
191
Due to the low number of sequence data derived from cognate genes, a meaningful
192
phylogenetic analysis of the capsid protein coding genes could not be performed.
193
However, as the three dimensional structure of a rabbit PBV has been determined by
194
using recombinant capsid protein (Duquerroy et al., 2009), attempts were made to
195
gain insight into the structure of the capsid protein of porcine PBVs by homology
196
modelling. The deduced aa sequences of the LB-3 and LA-7 clones were subjected
197
to protein modeling employing the Prime module of Schrödinger Suite (Schrödinger,
198
LLC, New York, NY, 2013) (www.schrodinger.com) molecular modelling software
199
package using the rabbit picobirnavirus capsid protein X-ray structure (PDB ID code:
200
2VF1) as homo-dimer template. Sequence similarity values of LB-3 and LA-7 with the
201
rabbit capsid protein were 38.3 % and 39.6 %, respectively. The models were refined
202
with the MacroModel energy minimization module of the Schrödinger Suite to
203
eliminate the steric conflicts between the side-chain atoms. The reconstructed
204
porcine PBV virion-like-particles (VLPs) were generated using the Oligomer
205
Generator
206
Molecular graphics and sequence alignment visualization were prepared using VMD
service
of
VIPERdb
(http://viperdb.scripps.edu/oligomer_multi.php).
7
207
version 1.9.1 (Humphrey et al., 1996) and the Multiple Sequence Viewer of the
208
Schrödinger Suite (Schrödinger, LLC, New York, NY, 2013), respectively. Structural
209
differences between the rabbit and the porcine PBV capsid protein (CP) models are
210
visualized in Fig. 2. The homology model for LB-3 proved to be very similar to the
211
rabbit PBV capsid protein (Fig. 2, panels a, c and d). Interestingly, we identified four
212
additional external surface exposed loop regions on the LA-7 pig PBV CP model (Fig.
213
2, panels b, c, d and g). These additional surface loops are located on the most
214
protruded edge of the coat protein, raising the possibility that these peptide insertions
215
may act as potential immune epitopes for B-cell mediated response.
216
The porcine PBV genome described represents the third complete sequence of this
217
highly diverse virus family. Until now only one human and one phocine PBV genome
218
sequence had been determined (Wakuda et al., 2005; Woo et al., 2012). In this study
219
we used the sequence independent amplification of genomic dsRNA and the next
220
generation sequencing strategy. This seemed important to maximize the depth of
221
sequence information and minimize any bias inherent to sequence variation seen in
222
PBV genomes, given that porcine PBVs frequently occur as mixed strain population
223
in the same host and there is evidence indicative of the quasi species nature of PBVs
224
(Bányai et al., 2008).
225
The porcine stool specimen from which strain 221/04-16 was isolated, contained two
226
major PBV subpopulations, each sharing the same RdRp gene and expressing
227
different capsid genes. The conservation of the RdRp gene sequence suggested that
228
one strain may be a reassortant. The likelihood that another RdRp gene was also
229
present but remained hidden in the sample is low, given that sequence independent
230
amplification coupled with deep sequencing were employed which, on one hand
231
shows no bias in recognition of the template genomic dsRNA segment and on the
232
other hand should detect minor sequence variants. This is further illustrated with the
233
truncated L segments of type LA clones. In this study we did not determine whether
234
pseudogenes among the LA sequences were generated as an artifact during RNA
235
processing or the shorter L segments encoding pseudogenes were incorporated in
236
the capsid. However, PBV genes with altered structure (pseudogenes) may be
237
tolerated/maintained if functionally integer cognate genes are retained (Bányai et al.,
8
238
2008) and we were able to determine at least one LA segment variant (LA-5) with a
239
correct ORF architecture in the porcine PBV genome.
240
Sequencing of the whole genome of strain 221/04-16 uncovered another key
241
mechanism likely contributing to the great sequence diversity seen even among PBV
242
strains collected from the same host species (Bányai et al., 2003, 2008; Carruyo et
243
al., 2008; Fregolente et al., 2009; Phan et al., 2011; Rosen et al., 2000; van Leeuwen
244
et al., 2010; Wang et al., 2012). We found further proof that insertions and deletions
245
(in/del) occur in the PBV genome, as evidenced by sequence length heterogeneity
246
seen in the UTRs and the coding regions of both S and L segments. The effect of
247
such mutations on phylogenetic analysis and protein evolution was demonstrated by
248
the observation of a deleterious mutation in the RdRp gene, which was compensated
249
by a single nucleotide insertion several nucleotides downstream (Fig. 1c). This
250
mutation resulted in a highly diverse deduced amino acid sequence in the flanking
251
region. It is possible that, in addition to the accumulation of single point mutations,
252
these compensatory in/del mutations are also common in the PBV genome,
253
generating variants of viral proteins with modified antigenic structure, more efficient
254
replication and/or increased viral fitness (Ganesh et al., 2012).
255
In conclusion, this study provides a more detailed insight into the evolutionary
256
mechanisms of PBVs. How different mechanisms contribute to better adaptation to a
257
homologous host or allow interspecies transmission should be the subject of future
258
studies. Until adequate cell culture and model animals are not established, whole
259
genome sequencing of PBVs from various host species remains the only alternative
260
to investigate this intriguing question.
261
Acknowledgements
262
K. B. was supported by the Momentum program (Hungarian Academy of Sciences).
263
Á. G. was the recipient of a János Bolyai fellowship from the Hungarian Academy of
264
Sciences. The licensing of the Schrödinger Suite software package was funded from
265
the OTKA grant under agreement No.:108793.
266 267
References
9
268
Bányai, K., Jakab, F., Reuter, G., Bene, J., Uj, M., Melegh, B. & Szucs, G. (2003).
269
Sequence
heterogeneity
among
human
picobirnaviruses
270
gastroenteritis outbreak. Arch Virol 148, 2281–2291.
detected
in
a
271
Bányai, K., Martella, V., Bogdán, Á., Forgách, P., Jakab, F., Meleg, E., Bíró, H.,
272
Melegh, B., & Szucs, G. (2008). Genogroup I picobirnaviruses in pigs: evidence
273
for genetic diversity and relatedness to human strains. J Gen Virol 89, 534–539.
274
Carruyo, G. M., Mateu, G., Martínez, L. C., Pujol, F. H., Nates, S. V., Liprandi, F.,
275
& Ludert, J. E. (2008). Molecular characterization of porcine picobirnaviruses and
276
development of a specific reverse transcription-PCR assay. J Clin Microbiol 46,
277
2402–2405.
278 279 280 281
Corpet, F. (1988). Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res 16, 10881–10890. Da Costa, B., Duquerroy, S., Tarus, B., & Delmas, B. (2011). Picobirnaviruses encode a protein with repeats of the ExxRxNxxxE motif. Virus Res 158, 251–256.
282
Delmas, B. (2011). Picobirnaviridae In Virus Taxonomy. Ninth Report of the
283
International Committee on Taxonomy of Viruses, pp. 535-539. Edited by King, A.
284
M. Q., Adams, M. J., Carstens, E. B., Lefkowitz, E. J. Elsevier Inc.
285
Duquerroy, S., Da Costa, B., Henry, C., Vigouroux, A., Libersou, S., Lepault, J.,
286
Navaza, J., Delmas, B., & Rey, F. A. (2009). The picobirnavirus crystal structure
287
provides functional insights into virion assembly and cell entry. EMBO J 28,
288
1655–1665.
289
Fregolente, M. C., de Castro-Dias, E., Martins, S. S., Spilki, F. R., Allegretti, S.
290
M., & Gatti, M. S. (2009). Molecular characterization of picobirnaviruses from new
291
hosts. Virus Res 143, 134–136.
292
Gallimore, C. I., Appleton, H., Lewis, D., Green, J., & Brown, D. W. (1995).
293
Detection and characterisation of bisegmented double-stranded RNA viruses
294
(picobirnaviruses) in human faecal specimens. J Med Virol 45, 135–140.
295
Ganesh, B., Banyai, K., Masachessi, G., Mladenova, Z., Nagashima, S., Ghosh,
296
S., Nataraju, S. M., Pativada, M., Kumar, R., & Kobayashi, N. (2011).
297
Genogroup I picobirnavirus in diarrhoeic foals: can the horse serve as a natural
298
reservoir for human infection? Vet Res 42, 52.
299
Ganesh, B., Bányai, K., Martella, V., Jakab, F., Masachessi, G., & Kobayashi, N.
300
(2012). Picobirnavirus infections: viral persistence and zoonotic potential. Rev
301
Med Virol 22, 245–256. 10
302
Ghosh, S., Kobayashi, N., Nagashima, S., & Naik, T. N. (2009). Molecular
303
characterization of full-length genomic segment 2 of a bovine picobirnavirus
304
(PBV) strain: evidence for high genetic diversity with genogroup I PBVs. J Gen
305
Virol 90, 2519–2524.
306 307
Gillman, L., Sánchez, A. M., & Arbiza, J. (2013). Picobirnavirus in captive animals from Uruguay: identification of new hosts. Intervirology 56, 46-49.
308
Giordano, M. O., Martinez, L. C., Rinaldi, D., Gúinard, S., Naretto, E., Casero, R.,
309
Yacci, M. R., Depetris, A. R., Medeot, S. I., & Nates, S. V. (1998). Detection of
310
picobirnavirus in HIV-infected patients with diarrhea in Argentina. J Acquir
311
Immune Defic Syndr Hum Retrovirol 18, 380–383.
312
Giordano, M. O., Martinez, L. C., Masachessi, G., Barril, P. A., Ferreyra, L. J.,
313
Isa, M. B., Valle, M. C., Massari, P. U., & Nates, S. V. (2011). Evidence of
314
closely related picobirnavirus strains circulating in humans and pigs in Argentina.
315
J Infect 62, 45–51.
316
Green, J., Gallimore, C. I., Clewley, J. P., & Brown, D. W. (1999). Genomic
317
characterisation of the large segment of a rabbit picobirnavirus and comparison
318
with the atypical picobirnavirus of Cryptosporidium parvum. Arch Virol 144, 2457–
319
2465.
320
Grohmann, G. S., Glass, R. I., Pereira, H. G., Monroe, S. S., Hightower, A. W.,
321
Weber, R., & Bryan, R. T. (1993). Enteric viruses and diarrhea in HIV-infected
322
patients. Enteric Opportunistic Infections Working Group. N Engl J Med 329, 14–
323
20.
324
Haga, I. R., Martins, S. S., Hosomi, S. T., Vicentini, F., Tanaka, H., & Gatti, M. S.
325
(1999).
326
(Picobirnavirus) in faeces of giant anteaters (Myrmecophaga tridactyla). Vet J
327
158, 234–236.
328 329
Identification
of
a
bisegmented
double-stranded
RNA
virus
Humphrey, W., Dalke, A., & Schulten, K. (1996). VMD - Visual Molecular Dynamics. J Mol Graphics 14, 33–38.
330
Lambden, P. R., Cooke, S. J., Caul, E. O., & Clarke, I. N. (1992). Cloning of
331
noncultivatable human rotavirus by single primer amplification. J Virol 66, 1817–
332
1822.
333
Malik, Y. S., Chandrashekar, K. M., Sharma, K., Haq, A. A., Vaid, N., Chakravarti,
334
S., Batra, M., Singh, R., & Pandey, A. B. (2011). Picobirnavirus detection in
11
335
bovine and buffalo calves from foothills of Himalaya and Central India. Trop Anim
336
Health Prod 43, 1475–1478.
337
Masachessi, G., Martínez, L. C., Giordano, M. O., Barril, P. A., Isa, B. M.,
338
Ferreyra, L., Villareal, D., Carello, M., Asis, C., & Nates, S. V. (2007).
339
Picobirnavirus (PBV) natural hosts in captivity and virus excretion pattern in
340
infected animals. Arch Virol 152, 989–998.
341
Masachessi, G., Martinez, L. C., Ganesh, B., Giordano, M. O., Barril, P. A., Isa,
342
M. B., Ibars, A., Pavan, J. V., & Nates, S. V. (2012). Establishment and
343
maintenance of persistent infection by picobirnavirus in greater rhea (Rhea
344
Americana). Arch Virol 157, 2075–2082.
345
Pereira, H. G., Flewett, T. H., Candeias, J. A., & Barth, O. M. (1988). A virus with a
346
bisegmented double-stranded RNA genome in rat (Oryzomys nigripes) intestines.
347
J Gen Virol 69, 2749–2754.
348 349
Phan, T. G., Kapusinszky, B., Wang, C., Rose, R.K., Lipton, H.L., & Delwart, E.L. (2011). The fecal viral flora of wild rodents. PLoS Pathog 7, e1002218.
350
Potgieter, A. C., Steele, A. D., & van Dijk, A. A. (2002). Cloning of complete
351
genome sets of six dsRNA viruses using an improved cloning method for large
352
dsRNA genes. J Gen Virol 83, 2215–2223.
353
Potgieter, A. C., Page, N. A., Liebenberg, J., Wright, I. M., Landt, O., & van Dijk,
354
A. A. (2009). Improved strategies for sequence-independent amplification and
355
sequencing of viral double-stranded RNA genomes. J Gen Virol 90, 1423–1432.
356
Rosen, B. I., Fang, Z. Y., Glass, R. I. & Monroe, S. S. (2000). Cloning of human
357
picobirnavirus genomic segments and development of an RT-PCR detection
358
assay. Virology 277, 316–329.
359
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., & Kumar, S. (2011).
360
MEGA5: molecular evolutionary genetics analysis using maximum likelihood,
361
evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28,
362
2731–2739.
363
Van Leeuwen, M., Williams, M. M., Koraka, P., Simon, J. H., Smits, S. L., &
364
Osterhaus, A. D. (2010). Human picobirnaviruses identified by molecular
365
screening of diarrhea samples. J Clin Microbiol 48, 1787–1794.
366
Wakuda, M., Pongsuwanna, Y., & Taniguchi, K. (2005). Complete nucleotide
367
sequences of two RNA segments of human picobirnavirus. J Virol Methods 126,
368
165–169. 12
369
Wang, Y., Tu, X., Humphrey, C., McClure, H., Jiang, X., Qin, C., Glass, R. I., &
370
Jiang, B. (2007). Detection of viral agents in fecal specimens of monkeys with
371
diarrhea. J Med Primatol 36, 101–107.
372
Wang, Y., Banyai, K., Tu, X., & Jiang, B. (2012). Simian genogroup I
373
picobirnaviruses: prevalence, genetic diversity, and zoonotic potential. J Clin
374
Microbiol 50, 2779–2782.
375
Woo, P. C., Lau, S. K., Bai, R., Teng, J. L., Lee, P., Martelli, P., Hui, S. W., &
376
Yuen, K. Y. (2012). Complete genome sequence of a novel picobirnavirus,
377
otarine picobirnavirus, discovered in California sea lions. J Virol 86, 6377–6378.
378 379 380
Figure legends
381 382
Fig. 1. (a) RNA profile of the PBV strain 221/04-16 identified in an Italian pig with
383
diarrhea. The genomic pattern of a Rotavirus A strain electrophoresed in parallel with
384
the PBV strain is also shown. (b) Nucleotide (NT) and amino acid (AA) sequence
385
based phylogeny of the partial S genome segment. The gene fragment used in the
386
phylogeny encodes a ~ 400 aa long polypeptide (position nt 290-1414, reference
387
strain Hy005102). Bootstrap values ≥50 are indicated. GGI and GGII refer to the
388
genogroup assignments of PBV strains. The scale bar (5 nt changes per 100
389
positions) is proportional to the genetic distance. (c) Partial S genome segment of
390
two porcine PBV strains, the Italian 221/04-16 and the Chinese SD. The alignment
391
shows the relative position of the single nucleotide insertion and deletion resulting in
392
remarkable amino acid change within the affected region. Numbers above the
393
alignment refer to the nucleotide positions of 221/04-16. (d) Number and position of
394
the sequence motif ExxRxNxxxE in the ORF1-encoded protein (L segment) of
395
various PBV strains (x can be any amino acid; Hu, human, La, lapine, Ot, otarine, Po,
396
porcine; Hu-1 is VS10, Hu-2 is Hy005102, Po-1 is LA-5, Po-2 is LB-3). Numbers on
397
the right indicate the length of the protein encoded by ORF1, ranging from 106
398
(lapine PBV) to 224 (Hy005102) aa in length. The protein sequences are not aligned.
399 400
Fig. 2. Structure comparison of PBV capsid (ORF2, large segment) proteins and
401
virions. Structure based alignment of the newly described LB-3 and LA-7 pig
402
picobirnavirus mature capsid protein amino acid sequences (a) and (b). The 13
403
background of the sequence alignments reflects the homology levels of the two
404
related PBV capsid protein sequences: identical amino acids are red, similar amino
405
acids are light orange while different amino acids are light pink. The four additional
406
external surface exposed loop regions on LA-7 PBV capsid protein are indicated
407
consequently on panel (b, c, d and g): loop region from aa 262 to 268 is orange, loop
408
region form aa 405 to 411 is green, loop region from aa 454 to 466 is red, while loop
409
region from aa 558 to 563 is purple. The superimposed LB-3 (pink) and LA-7 (cyan)
410
porcine PBV capsid protein and the template rabbit (grey) PBV capsid protein dimers
411
are in cartoon representation (c) and (d). The extra four outer loops on LA-7 PBV
412
VLP are illustrated in licorice representation. Molecular surface representation of the
413
rabbit (e), and the porcine LB-3 (f) and LA-7 PBV (g) virions. The molecular surface
414
is colored by radial extension of the amino acids from the virion centre. Dark blue
415
represents the most protruding capsid protein parts.
416
14
Fig. 1. (b)
(a)
768 |
(c)
848 |
221/04-16 SD
221/04-16 SD
(d)
Fig. 2.
1
Journal category: Animal Viruses - Double-strand RNA
2
Genome sequencing identifies genetic and antigenic divergence of porcine
3
picobirnaviruses
4
Krisztián Bányai 1 , Christiaan Potgieter 2,3, Ákos Gellért 4, Balasubramanian Ganesh
5
5
, Maria Tempesta 6, Eleonora Lorusso 6, Canio Buonavoglia 6, Vito Martella 6
6 7
1
8
Academy of Sciences, Budapest, Hungary
9
2
Virology Division, Onderstepoort Veterinary Institute, Onderstepoort, South Africa
10
3
Biochemistry Division, North-West University, Potchefstroom, South Africa
11
4
Agricultural Institute, Centre for Agricultural Research, Hungarian Academy of Sciences,
12
Martonvásár, Hungary
13
5
14
West Bengal, India
15
6
Institute for Veterinary Medical Research, Centre for Agricultural Research, Hungarian
Division of Virology, National Institute of Cholera and Enteric Diseases (NICED), Kolkata,
Dipartimento di Medicina Veterinaria, Università di Bari Aldo Moro, Bari, Italy
16 17 18 19 20
Correspondence to:
Krisztián Bányai, Institute for Veterinary Medical Research, Centre for Agricultural Research, Hungarian Academy of Sciences, Hungária krt 21., H-1143 Budapest, Hungary; E-mail: [email protected] Vito Martella, Dipartimento di Medicina Veterinaria, Università di Bari Aldo Moro, S.p. per Casamassima Km 3, 70010 Valenzano, Bari, Italy; E-mail: [email protected]
21 22 23 24 25 26
Running title: Porcine picobirnavirus genome
27
Word count in summary: 193
28
Word count in text: 2570
29 30
GenBank accession numbers: KF861767-KF861773
1
31
SUMMARY
32
The full-length genome sequence of a porcine picobirnavirus detected in Italy in 2004
33
was determined. The S segment was 1730 nucleotide (nt) in length, coding for a
34
putative RNA-dependent RNA polymerase. Two distinct sub-populations of L
35
segment (LA and LB) were identified in the sample, with the sizes ranging from 2351
36
to 2666 nt. The ORF1 coding for a protein of unknown function, contained repetitions
37
of the ExxRxNxxxE motif in variable number. The capsid protein coding ORF2
38
spanned nt 810-2447 in the LB variants and started at nt 734 in the LA variants.
39
However, a termination codon was present only in one of all LA segment variants.
40
Three dimensional modelling of the porcine PBV capsids suggested structural
41
differences in the protruding domain, tentatively involved as antigens in humoral
42
immune response. Altogether, these findings suggest simultaneous presence of two
43
different picobirnavirus strains sharing the same S segment but displaying genetically
44
diverse L segments. In addition, the sample probably contained a mixture of PBVs
45
with aberrant RNA replication products. Altered structure in the L segments could be
46
tolerated and retained in the presence of functionally integer cognate genes and
47
represent a mechanism of virus diversification.
48
Keywords: Picobirnaviridae, reassortment, pig, zoonosis
2
49
TEXT
50
Picobirnaviruses (PBVs), family Picobirnaviridae, are small (35–41 nm in diameter),
51
non-enveloped viruses displaying a distinctive icosahedral capsid. The PBV genome
52
is bisegmented double-stranded RNA (dsRNA). The larger (L) genome segment is
53
2.2–2.7 kb in length and encodes the capsid protein and a putative protein of
54
unknown function, while the smaller (S) genome segment is 1.2–1.9 kb in length and
55
encodes the viral RNA-dependent RNA polymerase (Delmas, 2011). Nucleotide
56
sequence differences in the S segment permit the classification of PBVs into two
57
major clusters, designated as genogroup I and genogroup II. The intra- and
58
intergenogroup sequence similarities of a short nucleic acid fragment of the S
59
segment used in diagnostic PCR ranges from 49 to 100 % and 28 to 37 %,
60
respectively (Bányai et al., 2003; Rosen et al., 2000).
61
PBVs have been found in the feces and intestinal content and, most recently, in the
62
respiratory tract of vertebrates. While the list of host species with record of PBV
63
infections has increased continuously (Fregolente et al., 2009; Gallimore et al., 1995;
64
Ganesh et al., 2011; Ghosh et al., 2009; Gillman et al., 2013; Green et al., 1999;
65
Haga et al., 1999; Malik et al., 2011; Masachessi et al., 2007, 2012; Pereira et al.,
66
1988; Wang et al., 2007; Woo et al., 2012), an association of PBV infection with
67
enteric or respiratory diseases has not been formally demonstrated. In fact, PBV is
68
often detected together with co-infecting major enteric pathogens or in clinical
69
specimens from inapparently infected hosts. The lack of cell culture or animal model
70
further hinders the recognition of any disease associations. However, studies in
71
immuncompromised hosts have suggested an opportunistic role for PBV in diarrheic
72
infections (Giordano et al., 1998; Grohmann et al., 1993).
73
Transmission of PBV from one host species to another has been suggested in
74
molecular epidemiological and strain characterization studies, based on very close
75
genetic relatedness between human and porcine or human and equine PBV strains
76
(Bányai et al., 2008; Ganesh et al., 2011; Giordano et al., 2011). Given that PBV is
77
frequently found in waste water and surface water, a water-born route of
78
transmission has been also suggested (Ganesh et al., 2011). Indeed, the genetic
79
similarity among heterologous PBV strains has been seen along a short, ~200 bp
3
80
gene fragment of the S segment. Because the PBV genome is segmented and
81
measures ~ 4 kbp, the value of this short fragment to describe molecular
82
epidemiology may be limited. A major obstacle to perform whole genome based
83
studies is the lack of genome sequence data for a reasonably high number of strains.
84
Thus far, the complete genome sequence has been determined only for a human
85
strain, and more recently, for a phocine PBV strain (Wakuda et al., 2005; Woo et al.,
86
2012). In addition, complete or partial L segment sequences have become available
87
for a lapine strain and some human strains (Green et al., 1999; Rosen et al., 2000),
88
while the sequences of a dozen or so S gene fragments longer than 1000 bp has
89
been determined for human, porcine, simian, bovine, and murine PBVs (Ghosh et al.,
90
2009; Phan et al., 2011; Rosen et al., 2000; van Leeuwen et al., 2010; Wang et al.,
91
2012; and unpublished GenBank entries). In this study, sequencing and molecular
92
characterization of the full-length genome of a porcine PBV strain was performed by
93
sequence independent amplification and high throughput sequencing of the genomic
94
dsRNA.
95
During a rotavirus surveillance project in 2004, total RNAs prepared from 37
96
randomly selected fecal samples collected from diarrheic piglets were analyzed by
97
polyacrylamide gel electrophoresis and silver staining to diagnose rotavirus
98
infections. One sample contained bands consistent with picobirnavirus dsRNA (Fig.
99
1a). This porcine PBV strain, designated 221/04-16 was detected from a 45-day-old
100
piglet with diarrhea at a farm in Brescia, Northern Italy, in 2004.
101
Total RNA was extracted from fecal specimen using the commercial TRI-REAGENT–
102
LS (Molecular Research Centre), according to the manufacturer’s protocol. Single
103
stranded RNA (ssRNA) was removed by precipitation with 2 M LiCl (Sigma) at 4 C for
104
16 h followed by centrifugation at 16 000 g for 30 min. dsRNA was purified from the
105
resulting supernatant using a MinElute gel extraction kit (Qiagen). The complete
106
nucleotide sequence was determined for the small (S) and large (L) segments of this
107
strain by sequence-independent amplification and deep sequencing as described
108
previously (Potgieter et al., 2002, 2009; Lambden et al., 1992). Briefly, an anchor
109
primer was ligated to dsRNA and used for reverse transcription and subsequent
110
amplification of virus genome segments. Amplified cDNA products were then
4
111
subjected to sequencing with GS FLX technology at Inqaba biotechTM (Pretoria, South
112
Africa).
113
Lasergene7 software from DNASTAR was used for de novo assembly of the contigs.
114
Files containing the sequence information, quality values and flowgrams (sff files)
115
were loaded into the Seqman 7 programme of the Lasergene software. Contigs
116
resulting from the assembly were checked manually and their consensus sequences
117
were exported as FASTA files. Consensus sequences were aligned to known
118
sequences using MEGALIGN (Lasergene7). Finally, sequences were subjected to
119
BLASTN analysis using the National Center for Biotechnology Information website.
120
The
121
(http://www.ncbi.nlm.nih.gov/gorf/gorf.html). Nucleotide and amino acid sequence
122
alignments were prepared by Multalin (Corpet, 1988) and manually adjusted to
123
codons. Phylogenetic analysis was performed by using MEGA5 (Tamura et al.,
124
2011). Sequences were deposited in GenBank under the following accession
125
numbers: KF861767-KF861773.
126
Small genome segment. The S segment of strain 221/04-16 was 1730 bp in length.
127
In comparison with other full-length PBV S segments, this segment is shorter than
128
those of the human strain Hy005102 (1745 bp), the monkey strain CHN-49/2002
129
(1784 bp), and the bovine strain RUBV-P/IND/20 (1758 bp) and longer than those of
130
the human strains 1-CHN-97 (1696 bp) and GPBV6C1 (1711 bp), and the otarine
131
PBV (1687 bp). The length of the 5’ and 3’ UTRs were 74 and 49 bp, respectively,
132
and were structurally heterogenous when compared with reference strains, except for
133
the common GTAAA and ACTGC pentanucleotides, located at the termini of the RNA
134
segments and conserved among the various PBV strains. The S segment of strain
135
221/04-16 contained a single 1607 nt-long ORF (nt 75 to 1682). Alternative ORFs
136
were also predicted (data not shown). However, the predicted proteins were too short
137
and their presence was not consistently found in the S segments of other strains;
138
therefore these putative ORFs were ignored in this analysis. The protein coded by
139
the S segment ORF was 535 aa long and was predicted to be the viral RNA
140
dependent RNA polymerase (RdRp). It had conservative motifs in aa positions 266-
141
271, 328-333 and 364-366 defining the GDD sequence motif. When comparing the
142
221/04-16 S gene sequence with the Chinese strain SD, a single nucleotide deletion
NCBI’s
ORF
finder
was
used
to
identify
protein
coding
regions
5
143
was observed at position 768 that was compensated at position 848 by a single
144
nucleotide insertion. This insertion caused a codon shift, altering the translation of a
145
28 aa stretch between the two porcine PBVs (Fig. 1c).
146
Phylogenetic analysis was first performed using a ~200 bp fragment of the RdRp
147
used to amplify GGI PBV strains in diagnostic PCR (Rosen et al., 2000). This
148
analysis identified a Chinese porcine PBV strain as presenting the greatest nt
149
sequence identity (95%) with that of the Italian strain. Other relatively closely related
150
sequences (up to 89%) included viral sequences identified from US water samples.
151
The Italian porcine PBV strain shared lower nt similarity to European and American
152
(up to 75%) porcine PBVs and shared comparable similarities with European, Asian,
153
and American human PBVs (up to 88%). Another, ~280 bp long, partially overlapping
154
region of the S segment used to specifically amplify porcine PBVs (Carruyo et al.,
155
2008) was analyzed separately. This analysis demonstrated that the Italian strain
156
shared somewhat greater similarities with Argentinean porcine and human PBV
157
strains (up to 97% nt identity; data not shown). An analysis of the whole ORF of
158
several strains of porcine, human, murine, phocine, and bovine origin identified
159
comparable sequence identity data when these strains were compared along the
160
commonly characterized ~200 bp long fragment. The same observation was made
161
when comparing the Italian strain with two genogroup II strains, 4-GA-91 and
162
GPBV6G2. However, it should be considered that extrapolating nucleotide similarities
163
obtained from a short fragment may not be relevant in each case as exemplified by
164
the codon shift described above. Phylogenetic analysis of an ~1200 bp fragment of
165
selected strains with longer S genome segment sequence confirmed the similarity
166
based prediction of the genetic relatedness between the Italian strain and the
167
Chinese SD strain (Fig. 1b).
168
Large genome segment. Amplification and cloning of the L segment yielded a
169
heterogeneous population of cDNA products, comprising two major types and
170
several minor sequence variants within each type. The major sequence types were
171
designated LA and LB. The LB variants (ie LB-1, LB-3, and LB-4) were 2524 bp in
172
length and had two predicted ORFs at nt 195-794 (ORF1) and 810-2447 (ORF2),
173
encoding proteins of 199 and 545 aa in length, respectively. The LA variants, LA-2,
174
LA-5 and LA-7, respectively, were 2361, 2666 and 2351 bp long. The ORF1 started
6
175
at nt 185 and ended at nt 721, coding for a polypeptide of 175 aa in length. The
176
ORF2 started at nt 734 in all LA sequences. However, while LA-5 was terminated
177
with a stop codon, LA-2 and LA-7 seemed to be pseudogenes without translation
178
termination signals. The protein encoded by the LA-5 sequence was longer (619 aa)
179
than those of other PBV capsid proteins (ie, 590 aa in the lapine PBV strain, 552 aa
180
in the human strain Hy005102 and 575 aa in the otarine strain). The 3’ UTRs ranged
181
from 184 (LA segments) to 194 nt (LB segments). The 5’ UTRs ranged from 76 (LB
182
segments) to 84 nt (LA-5).
183
The ORF1-encoded proteins of the LA and LB clones were extremely diverse, yet
184
retained common genetic signatures. One shared feature was the recently
185
recognized 10-aa long sequence motif (Da Costa et al., 2011), ExxRxNxxxE, where
186
x, can be any amino acid. Four and 6 repeats of this motif was identified in LB clones
187
(e.g. LB-3) and LA clones (e.g. LA-5), respectively. The amino acid distance between
188
two adjacent motifs were variable. When comparing the ORF1s with other reference
189
strains, we found similar diversities in the number and position of the ExxRxNxxxE
190
motif (Fig 1d).
191
Due to the low number of sequence data derived from cognate genes, a meaningful
192
phylogenetic analysis of the capsid protein coding genes could not be performed.
193
However, as the three dimensional structure of a rabbit PBV has been determined by
194
using recombinant capsid protein (Duquerroy et al., 2009), attempts were made to
195
gain insight into the structure of the capsid protein of porcine PBVs by homology
196
modelling. The deduced aa sequences of the LB-3 and LA-7 clones were subjected
197
to protein modeling employing the Prime module of Schrödinger Suite (Schrödinger,
198
LLC, New York, NY, 2013) (www.schrodinger.com) molecular modelling software
199
package using the rabbit picobirnavirus capsid protein X-ray structure (PDB ID code:
200
2VF1) as homo-dimer template. Sequence similarity values of LB-3 and LA-7 with the
201
rabbit capsid protein were 38.3 % and 39.6 %, respectively. The models were refined
202
with the MacroModel energy minimization module of the Schrödinger Suite to
203
eliminate the steric conflicts between the side-chain atoms. The reconstructed
204
porcine PBV virion-like-particles (VLPs) were generated using the Oligomer
205
Generator
206
Molecular graphics and sequence alignment visualization were prepared using VMD
service
of
VIPERdb
(http://viperdb.scripps.edu/oligomer_multi.php).
7
207
version 1.9.1 (Humphrey et al., 1996) and the Multiple Sequence Viewer of the
208
Schrödinger Suite (Schrödinger, LLC, New York, NY, 2013), respectively. Structural
209
differences between the rabbit and the porcine PBV capsid protein (CP) models are
210
visualized in Fig. 2. The homology model for LB-3 proved to be very similar to the
211
rabbit PBV capsid protein (Fig. 2, panels a, c and d). Interestingly, we identified four
212
additional external surface exposed loop regions on the LA-7 pig PBV CP model (Fig.
213
2, panels b, c, d and g). These additional surface loops are located on the most
214
protruded edge of the coat protein, raising the possibility that these peptide insertions
215
may act as potential immune epitopes for B-cell mediated response.
216
The porcine PBV genome described represents the third complete sequence of this
217
highly diverse virus family. Until now only one human and one phocine PBV genome
218
sequence had been determined (Wakuda et al., 2005; Woo et al., 2012). In this study
219
we used the sequence independent amplification of genomic dsRNA and the next
220
generation sequencing strategy. This seemed important to maximize the depth of
221
sequence information and minimize any bias inherent to sequence variation seen in
222
PBV genomes, given that porcine PBVs frequently occur as mixed strain population
223
in the same host and there is evidence indicative of the quasi species nature of PBVs
224
(Bányai et al., 2008).
225
The porcine stool specimen from which strain 221/04-16 was isolated, contained two
226
major PBV subpopulations, each sharing the same RdRp gene and expressing
227
different capsid genes. The conservation of the RdRp gene sequence suggested that
228
one strain may be a reassortant. The likelihood that another RdRp gene was also
229
present but remained hidden in the sample is low, given that sequence independent
230
amplification coupled with deep sequencing were employed which, on one hand
231
shows no bias in recognition of the template genomic dsRNA segment and on the
232
other hand should detect minor sequence variants. This is further illustrated with the
233
truncated L segments of type LA clones. In this study we did not determine whether
234
pseudogenes among the LA sequences were generated as an artifact during RNA
235
processing or the shorter L segments encoding pseudogenes were incorporated in
236
the capsid. However, PBV genes with altered structure (pseudogenes) may be
237
tolerated/maintained if functionally integer cognate genes are retained (Bányai et al.,
8
238
2008) and we were able to determine at least one LA segment variant (LA-5) with a
239
correct ORF architecture in the porcine PBV genome.
240
Sequencing of the whole genome of strain 221/04-16 uncovered another key
241
mechanism likely contributing to the great sequence diversity seen even among PBV
242
strains collected from the same host species (Bányai et al., 2003, 2008; Carruyo et
243
al., 2008; Fregolente et al., 2009; Phan et al., 2011; Rosen et al., 2000; van Leeuwen
244
et al., 2010; Wang et al., 2012). We found further proof that insertions and deletions
245
(in/del) occur in the PBV genome, as evidenced by sequence length heterogeneity
246
seen in the UTRs and the coding regions of both S and L segments. The effect of
247
such mutations on phylogenetic analysis and protein evolution was demonstrated by
248
the observation of a deleterious mutation in the RdRp gene, which was compensated
249
by a single nucleotide insertion several nucleotides downstream (Fig. 1c). This
250
mutation resulted in a highly diverse deduced amino acid sequence in the flanking
251
region. It is possible that, in addition to the accumulation of single point mutations,
252
these compensatory in/del mutations are also common in the PBV genome,
253
generating variants of viral proteins with modified antigenic structure, more efficient
254
replication and/or increased viral fitness (Ganesh et al., 2012).
255
In conclusion, this study provides a more detailed insight into the evolutionary
256
mechanisms of PBVs. How different mechanisms contribute to better adaptation to a
257
homologous host or allow interspecies transmission should be the subject of future
258
studies. Until adequate cell culture and model animals are not established, whole
259
genome sequencing of PBVs from various host species remains the only alternative
260
to investigate this intriguing question.
261
Acknowledgements
262
K. B. was supported by the Momentum program (Hungarian Academy of Sciences).
263
Á. G. was the recipient of a János Bolyai fellowship from the Hungarian Academy of
264
Sciences. The licensing of the Schrödinger Suite software package was funded from
265
the OTKA grant under agreement No.:108793.
266 267
References
9
268
Bányai, K., Jakab, F., Reuter, G., Bene, J., Uj, M., Melegh, B. & Szucs, G. (2003).
269
Sequence
heterogeneity
among
human
picobirnaviruses
270
gastroenteritis outbreak. Arch Virol 148, 2281–2291.
detected
in
a
271
Bányai, K., Martella, V., Bogdán, Á., Forgách, P., Jakab, F., Meleg, E., Bíró, H.,
272
Melegh, B., & Szucs, G. (2008). Genogroup I picobirnaviruses in pigs: evidence
273
for genetic diversity and relatedness to human strains. J Gen Virol 89, 534–539.
274
Carruyo, G. M., Mateu, G., Martínez, L. C., Pujol, F. H., Nates, S. V., Liprandi, F.,
275
& Ludert, J. E. (2008). Molecular characterization of porcine picobirnaviruses and
276
development of a specific reverse transcription-PCR assay. J Clin Microbiol 46,
277
2402–2405.
278 279 280 281
Corpet, F. (1988). Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res 16, 10881–10890. Da Costa, B., Duquerroy, S., Tarus, B., & Delmas, B. (2011). Picobirnaviruses encode a protein with repeats of the ExxRxNxxxE motif. Virus Res 158, 251–256.
282
Delmas, B. (2011). Picobirnaviridae In Virus Taxonomy. Ninth Report of the
283
International Committee on Taxonomy of Viruses, pp. 535-539. Edited by King, A.
284
M. Q., Adams, M. J., Carstens, E. B., Lefkowitz, E. J. Elsevier Inc.
285
Duquerroy, S., Da Costa, B., Henry, C., Vigouroux, A., Libersou, S., Lepault, J.,
286
Navaza, J., Delmas, B., & Rey, F. A. (2009). The picobirnavirus crystal structure
287
provides functional insights into virion assembly and cell entry. EMBO J 28,
288
1655–1665.
289
Fregolente, M. C., de Castro-Dias, E., Martins, S. S., Spilki, F. R., Allegretti, S.
290
M., & Gatti, M. S. (2009). Molecular characterization of picobirnaviruses from new
291
hosts. Virus Res 143, 134–136.
292
Gallimore, C. I., Appleton, H., Lewis, D., Green, J., & Brown, D. W. (1995).
293
Detection and characterisation of bisegmented double-stranded RNA viruses
294
(picobirnaviruses) in human faecal specimens. J Med Virol 45, 135–140.
295
Ganesh, B., Banyai, K., Masachessi, G., Mladenova, Z., Nagashima, S., Ghosh,
296
S., Nataraju, S. M., Pativada, M., Kumar, R., & Kobayashi, N. (2011).
297
Genogroup I picobirnavirus in diarrhoeic foals: can the horse serve as a natural
298
reservoir for human infection? Vet Res 42, 52.
299
Ganesh, B., Bányai, K., Martella, V., Jakab, F., Masachessi, G., & Kobayashi, N.
300
(2012). Picobirnavirus infections: viral persistence and zoonotic potential. Rev
301
Med Virol 22, 245–256. 10
302
Ghosh, S., Kobayashi, N., Nagashima, S., & Naik, T. N. (2009). Molecular
303
characterization of full-length genomic segment 2 of a bovine picobirnavirus
304
(PBV) strain: evidence for high genetic diversity with genogroup I PBVs. J Gen
305
Virol 90, 2519–2524.
306 307
Gillman, L., Sánchez, A. M., & Arbiza, J. (2013). Picobirnavirus in captive animals from Uruguay: identification of new hosts. Intervirology 56, 46-49.
308
Giordano, M. O., Martinez, L. C., Rinaldi, D., Gúinard, S., Naretto, E., Casero, R.,
309
Yacci, M. R., Depetris, A. R., Medeot, S. I., & Nates, S. V. (1998). Detection of
310
picobirnavirus in HIV-infected patients with diarrhea in Argentina. J Acquir
311
Immune Defic Syndr Hum Retrovirol 18, 380–383.
312
Giordano, M. O., Martinez, L. C., Masachessi, G., Barril, P. A., Ferreyra, L. J.,
313
Isa, M. B., Valle, M. C., Massari, P. U., & Nates, S. V. (2011). Evidence of
314
closely related picobirnavirus strains circulating in humans and pigs in Argentina.
315
J Infect 62, 45–51.
316
Green, J., Gallimore, C. I., Clewley, J. P., & Brown, D. W. (1999). Genomic
317
characterisation of the large segment of a rabbit picobirnavirus and comparison
318
with the atypical picobirnavirus of Cryptosporidium parvum. Arch Virol 144, 2457–
319
2465.
320
Grohmann, G. S., Glass, R. I., Pereira, H. G., Monroe, S. S., Hightower, A. W.,
321
Weber, R., & Bryan, R. T. (1993). Enteric viruses and diarrhea in HIV-infected
322
patients. Enteric Opportunistic Infections Working Group. N Engl J Med 329, 14–
323
20.
324
Haga, I. R., Martins, S. S., Hosomi, S. T., Vicentini, F., Tanaka, H., & Gatti, M. S.
325
(1999).
326
(Picobirnavirus) in faeces of giant anteaters (Myrmecophaga tridactyla). Vet J
327
158, 234–236.
328 329
Identification
of
a
bisegmented
double-stranded
RNA
virus
Humphrey, W., Dalke, A., & Schulten, K. (1996). VMD - Visual Molecular Dynamics. J Mol Graphics 14, 33–38.
330
Lambden, P. R., Cooke, S. J., Caul, E. O., & Clarke, I. N. (1992). Cloning of
331
noncultivatable human rotavirus by single primer amplification. J Virol 66, 1817–
332
1822.
333
Malik, Y. S., Chandrashekar, K. M., Sharma, K., Haq, A. A., Vaid, N., Chakravarti,
334
S., Batra, M., Singh, R., & Pandey, A. B. (2011). Picobirnavirus detection in
11
335
bovine and buffalo calves from foothills of Himalaya and Central India. Trop Anim
336
Health Prod 43, 1475–1478.
337
Masachessi, G., Martínez, L. C., Giordano, M. O., Barril, P. A., Isa, B. M.,
338
Ferreyra, L., Villareal, D., Carello, M., Asis, C., & Nates, S. V. (2007).
339
Picobirnavirus (PBV) natural hosts in captivity and virus excretion pattern in
340
infected animals. Arch Virol 152, 989–998.
341
Masachessi, G., Martinez, L. C., Ganesh, B., Giordano, M. O., Barril, P. A., Isa,
342
M. B., Ibars, A., Pavan, J. V., & Nates, S. V. (2012). Establishment and
343
maintenance of persistent infection by picobirnavirus in greater rhea (Rhea
344
Americana). Arch Virol 157, 2075–2082.
345
Pereira, H. G., Flewett, T. H., Candeias, J. A., & Barth, O. M. (1988). A virus with a
346
bisegmented double-stranded RNA genome in rat (Oryzomys nigripes) intestines.
347
J Gen Virol 69, 2749–2754.
348 349
Phan, T. G., Kapusinszky, B., Wang, C., Rose, R.K., Lipton, H.L., & Delwart, E.L. (2011). The fecal viral flora of wild rodents. PLoS Pathog 7, e1002218.
350
Potgieter, A. C., Steele, A. D., & van Dijk, A. A. (2002). Cloning of complete
351
genome sets of six dsRNA viruses using an improved cloning method for large
352
dsRNA genes. J Gen Virol 83, 2215–2223.
353
Potgieter, A. C., Page, N. A., Liebenberg, J., Wright, I. M., Landt, O., & van Dijk,
354
A. A. (2009). Improved strategies for sequence-independent amplification and
355
sequencing of viral double-stranded RNA genomes. J Gen Virol 90, 1423–1432.
356
Rosen, B. I., Fang, Z. Y., Glass, R. I. & Monroe, S. S. (2000). Cloning of human
357
picobirnavirus genomic segments and development of an RT-PCR detection
358
assay. Virology 277, 316–329.
359
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., & Kumar, S. (2011).
360
MEGA5: molecular evolutionary genetics analysis using maximum likelihood,
361
evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28,
362
2731–2739.
363
Van Leeuwen, M., Williams, M. M., Koraka, P., Simon, J. H., Smits, S. L., &
364
Osterhaus, A. D. (2010). Human picobirnaviruses identified by molecular
365
screening of diarrhea samples. J Clin Microbiol 48, 1787–1794.
366
Wakuda, M., Pongsuwanna, Y., & Taniguchi, K. (2005). Complete nucleotide
367
sequences of two RNA segments of human picobirnavirus. J Virol Methods 126,
368
165–169. 12
369
Wang, Y., Tu, X., Humphrey, C., McClure, H., Jiang, X., Qin, C., Glass, R. I., &
370
Jiang, B. (2007). Detection of viral agents in fecal specimens of monkeys with
371
diarrhea. J Med Primatol 36, 101–107.
372
Wang, Y., Banyai, K., Tu, X., & Jiang, B. (2012). Simian genogroup I
373
picobirnaviruses: prevalence, genetic diversity, and zoonotic potential. J Clin
374
Microbiol 50, 2779–2782.
375
Woo, P. C., Lau, S. K., Bai, R., Teng, J. L., Lee, P., Martelli, P., Hui, S. W., &
376
Yuen, K. Y. (2012). Complete genome sequence of a novel picobirnavirus,
377
otarine picobirnavirus, discovered in California sea lions. J Virol 86, 6377–6378.
378 379 380
Figure legends
381 382
Fig. 1. (a) RNA profile of the PBV strain 221/04-16 identified in an Italian pig with
383
diarrhea. The genomic pattern of a Rotavirus A strain electrophoresed in parallel with
384
the PBV strain is also shown. (b) Nucleotide (NT) and amino acid (AA) sequence
385
based phylogeny of the partial S genome segment. The gene fragment used in the
386
phylogeny encodes a ~ 400 aa long polypeptide (position nt 290-1414, reference
387
strain Hy005102). Bootstrap values ≥50 are indicated. GGI and GGII refer to the
388
genogroup assignments of PBV strains. The scale bar (5 nt changes per 100
389
positions) is proportional to the genetic distance. (c) Partial S genome segment of
390
two porcine PBV strains, the Italian 221/04-16 and the Chinese SD. The alignment
391
shows the relative position of the single nucleotide insertion and deletion resulting in
392
remarkable amino acid change within the affected region. Numbers above the
393
alignment refer to the nucleotide positions of 221/04-16. (d) Number and position of
394
the sequence motif ExxRxNxxxE in the ORF1-encoded protein (L segment) of
395
various PBV strains (x can be any amino acid; Hu, human, La, lapine, Ot, otarine, Po,
396
porcine; Hu-1 is VS10, Hu-2 is Hy005102, Po-1 is LA-5, Po-2 is LB-3). Numbers on
397
the right indicate the length of the protein encoded by ORF1, ranging from 106
398
(lapine PBV) to 224 (Hy005102) aa in length. The protein sequences are not aligned.
399 400
Fig. 2. Structure comparison of PBV capsid (ORF2, large segment) proteins and
401
virions. Structure based alignment of the newly described LB-3 and LA-7 pig
402
picobirnavirus mature capsid protein amino acid sequences (a) and (b). The 13
403
background of the sequence alignments reflects the homology levels of the two
404
related PBV capsid protein sequences: identical amino acids are red, similar amino
405
acids are light orange while different amino acids are light pink. The four additional
406
external surface exposed loop regions on LA-7 PBV capsid protein are indicated
407
consequently on panel (b, c, d and g): loop region from aa 262 to 268 is orange, loop
408
region form aa 405 to 411 is green, loop region from aa 454 to 466 is red, while loop
409
region from aa 558 to 563 is purple. The superimposed LB-3 (pink) and LA-7 (cyan)
410
porcine PBV capsid protein and the template rabbit (grey) PBV capsid protein dimers
411
are in cartoon representation (c) and (d). The extra four outer loops on LA-7 PBV
412
VLP are illustrated in licorice representation. Molecular surface representation of the
413
rabbit (e), and the porcine LB-3 (f) and LA-7 PBV (g) virions. The molecular surface
414
is colored by radial extension of the amino acids from the virion centre. Dark blue
415
represents the most protruding capsid protein parts.
416
14
Fig. 1. (b)
(a)
768 |
(c)
848 |
221/04-16 SD
221/04-16 SD
(d)
Fig. 2.
