PENG Yi ZUO Shu-Qing JIANG Jia-Fu YANG Hong CAO Wu-Chun**

(1. School of Public Health, Central South University, Changsha, Hunan 410078, China; 2. State Key Laboratory of Pathogen and Biosecurity, Beijing Institute of Microbiology and Epidemiology, Beijing 100071, China)

Abstract Hantaviruses are maintained by persistently infected rodents, with incidental infection of humans. Previous work about the long-term infected Vero-E6 cell cultures by Seoul virus (SEOV) showed that deletions at the 3′ termini of virus RNAs accumulated during the infection, which may play an important role for virus persistence infection. However, changes of SEOV RNA rodent hosts and patients naturally infected were not clear. In this study, total RNA was extracted from blood clot of SEOV positive patients and lung tissues of SEOV positive Rattus norvegicus captured in field respectively. Variability of the termini of SEOV S segment was determined by RNA sequences from different clones, and modeled for secondary RNA structures. Termini diversity could be detected for SEOV S sequence from natural infected rodent hosts and patients. There were 5′ end, 3′end and both ends nucleotide sequence deletions observed, especially the typical types of DI RNAs (RNAs with large genome deletions) at both ends could be detected in rodent hosts, which was distinct from what was observed in long-term Vero-E6 cell cultures, in which only short 3′ terminal sequence missing could be detected. While, diversities for clones from patients was lower than that from rodent hosts, except for one clone from a patient had 3′ end nucleotide missing. Secondary RNA structures deduced from clones with different termini sequence missing changed significantly. Variations of terminal sequences from rodent hosts appeared to be distinctive with those from patients, which might attribute to the influences by host factors and it might be important to viral persistence in rodent hosts.

Key words Seoul virus; Terminal sequence; Variants; Quasispecies; Genome deletions

Hantaviruses(HV) comprise a genus in the familyBunyaviridaeand possess a negative-sense RNA genome that consists of three segments. The three segments are designated as the large (L), medium (M) and small (S) segments, encoding RNA dependent RNA polymerase, two glycoproteins (Gn and Gc), and nucleocapsid (N) proteins, respectively. Hantaviruses are maintained in nature by cyclical transmission between persistently infection in rodents and incidental infection in humans. They are known to cause two serious and fatal human diseases: hemorrhagic fever with renal syndrome (HFRS) and Hantavirus cardiopulmonary syndrome (HCPS) (Clementetal., 2003; Khaiboullinaetal., 2005; Zeieretal., 2005; Kleinetal., 2007). Seoul virus (SEOV) mainly carried byRattusnorvegicus(Norway rat) was known to be one of crucial causative agents of HFRS in China.

A key genomic characteristic that helps to define Hantavirus as a genus of the familyBunyaviridaeis the presence of distinctive terminal complementary nucleotides that promote the folding of the viral genomic segments into “panhandle” hairpin structures (Osborneetal., 2000; Kaukinenetal., 2005). It was specifically recognized by trimeric hantavirus N protein and was thought to form a double-stranded promoter regulating RNA transcription and replication (Kaukinenetal., 2005; Miretal., 2006), which may be involved in a prime-and-realign mechanism of initiation of RNA synthesis (Garcinetal., 1995). Hantaviruses are enzootic viruses of wild rodents that cause persistent infections (Clementetal., 2003; Khaiboullinaetal., 2005; Zeieretal., 2005; Kleinetal., 2007) and they can also persistently infect cultured mammalian cells, causing little or no cytopathology. Previous work about the long-term Vero-E6 cell cultures by Seoul virus (SEOV) showed that deletions at the 3′ termini of S, M and L virus RNAs accumulated during the infection, which may play a role in Seoul virus persistence (Meyeretal., 2000a; Meyeretal., 2000b). However, because the study was performed in cultured cells without the influence of the host factors, changes of SEOV RNA rodent hosts and patients naturally infected were not clear. Here, we investigated the terminal genomic characteristic of Seoul virus S segment carried by rodent hosts and patients respectively to insight the variations and their influences by different host factors.

1 Materials and Methods

1.1 Materials

Three human blood samples from SEOV positive HFRS patients and four SEOV positive lung tissue samples fromRattusnorvegicuscaptured in field were used in the study (Zuoetal., 2008).

1.2 Reagents

Trizol reagent (Invitrogen, USA); random primers (9 mer) (Takara, Dalian, China); BioDev Gel Extraction kit (Beijing BioDev-Tech., China); pUC19 DNA vector (Takara, Dalian, China); DH5α competentEscherichiacolicells (Takara, Dalian, China).

1.3 Methods

Total RNA was extracted from blood clots of patient or lung tissues of rodent hosts resepctively using Trizol reagent. The exact nucleotide sequences of the 5′ and 3′ termini of the virus segments were determined by the RNA ligation method described previously (Mandletal., 1991; Chizhikovetal., 1995; Kukkonenetal., 1998; Bohlmanetal., 2002; Padulaetal., 2002). cDNA was synthesized with random primers (9 mer) so that both vRNAs (viral RNAs) and cRNAs (complementary RNA) could be mapped simultaneously. PCR was performed using primer pairs that were designed according to SEOV sequence from GenBank containing 5′and 3′ end. The expected PCR products were purified using BioDev Gel Extraction kit and cloned using pUC19 DNA vector and transformed into DH5α competentE.colicells. Positive colonies were confirmed with sequencing on ABI PRISM 3730 Genetic Analyzers (ABI. USA).

In order to study the heterogeneity of the S RNA segment termini, 4-10 clones for each sample were sequenced (Fig.1) and aligned with clustalx1.8 software (Thompsonetal., 1994) and checked manually. The predicted secondary structures of terminal sequence were constructed by RNA structure program (version 5.4) (Reuteretal., 2010) based on the principle of minimizing free energy (Mathewsetal., 1999; Mathewsetal., 2004).

2 Results

Base mutations could be observed frequently at both termini regions of S sequence, no matter from rodent hosts or patients. Among 11 248 nucleotides sequenced, 34 nucleotide changes were observed, giving a mutation frequency of 3.0×10-3for all clones analyzed. There was no significant difference on mutation frequency between clones from rodents and those from patients. Notably, no mutations were observed repeatedly in clones analyzed, suggesting randomness of the events. However, the majority of these changes (25/34) were specific U→C and A→G mutations from both rodents and patients similar to the data reported for other hantaviruses from rodents (Plyusninetal., 1995; Plyusninetal., 1996; Feueretal., 1999) which suggested SEOV exist in the form of quasispecies not only in rodent hosts but also in patients.

In order to get an overall profile for S sequence termini, the clones obtained in this study were aligned with some sequences deposited in GenBank database. Some SEOV cell isolates, such as strain BjHD01 isolated in Beijing, had another three bases CUA (Fig.1) compared with most of clones obtained in this study. Compared with most of clones, only one clone from a rodent had another one bases A belonging to the 3′ end according to previous studies (Kukkonenetal., 1998; Meyeretal., 2000a; Meyeretal., 2000b; Padulaetal., 2002). Nine clones from rodents only had 5′ termini 7 nucleotides missing, but the exact position of deleted bases could not be determined. There were two possibilities for them. The first one was that they had 5′ termini four bases UAGU missing and three internal bases deleted at position nt 10 to nt 12 (Fig.1). The second one was that they had 5′ termini one base U missing and six internal bases deleted at position nt 7 to nt 12 (figure not shown). Five clones seemed to only have 3′ termini missing. One clone from a patient had 3′ termini 22 nucleotides missing, while four clones from rodent hosts had 3′ termini 13, 14 nucleotides missing, respectively. Four clones had both of 5′ end and 3′ end nucleotides deletion. The clones with both termini missing appeared to have longer nucleotides deletion and were all from rodent hosts. One clone with 3′ termini 17 deleted nucleotides also had truncated 5′ termini with 38 nucleotides missing. Another clone with 3′ termini 32 deleted nucleotides also had truncated 5′ termini with 49 nucleotides missing. In addition, two clones had much more deleted nucleotides in that they had 54 deleted nucleotides at 3′ termini and 97 deleted nucleotides at 5′ termini. Clones from different hosts had different deleted nucleotides in this study. But for clones from one rodent, the same deleted nucleotides could be observed once, suggesting fixation of the events in one host.

Several studies have shown that the efficiency of the joining reaction is dependent on the particular 5′ and 3′ terminal nucleotides and the secondary structure of the termini. The secondary structure might alter the interactions between N proteins and panhandles structures of viral genomic RNA. The predicted secondary structure for the RNA of S segment with different termini nucleotides missing varied differently (Fig.2). When the same length of sequence was included in the predicted model, the energy appeared to change with different termini nucleotides missing and become unstable in some structures (Fig.2).

Fig.2 The secondary structures deduced from S segments of SOEVSequence with very large deleted nucleotides missing was not showed. It showed that the models of secondary RNA structures for different termini sequence missing were significantly changed and consequently the energy appeared to change when the same length of sequence were included in the predicated model and it seemed to become unstable for some structures. A: Folding of S viral RNA of some cell isolates in GenBank database, such as strain BjHD01. It showed another three nucleotides CUA, which could only be observed for some S sequence from cell isolates deposited in GenBank database but could not be observed for S sequence from hosts in this study. B: Folding of S viral RNA of most clones. It showed a “panhandle” hairpin structure. C: Folding of S viral RNA of one clone with another one nucleotides A, when compared with most clones. D: Folding of S viral RNA for some clones only with 5′ termini 7 nucleotides missing when compared with most clones. It showed a shorter “panhandle” hairpin structure. All these clones were from rodents. E: Folding of S viral RNA of clones only with 3′ termini 13 nucleotides missing when compared with most clones. It showed a changed structure. F: Folding of S viral RNA of clones only with 3′ termini 14 nucleotides missing from rodent hosts. It showed a changed structure. G: Folding of S viral RNA of clone only with 3′ termini 22 nucleotides missing from a patient. It showed a changed structure. H: Folding of S viral RNA of clones with 3′ termini 17 deleted nucleotides and truncated 5′ termini with 38 nucleotides missing when compared with most clones. It showed a different structure. I: Folding of S viral RNA of clones with 3′ termini 32 deleted nucleotides also had truncated 5′ termini with 49 nucleotides missing. It showed a different structure.

3 Discussions

Prior to this study, the published terminal sequences of SEOV S genome segments had been deduced from PCR primers. In this study, SEOV termini from natural infected rodent hosts and patients, which represent the prototype circulating in nature, were firstly sequenced and analyzed.

Variation of viruses was observed in patients for the first time, demonstrating that SEOV mutate and generate quasispecies in patients, just as in rodent hosts. This might be one of adaptive mechanism for hantaviruses to escape host immunosurveillance and may have implications for viral evolution.

Similar to changes of long-term SEOV in cell cultures (Meyeretal., 2000a), RNA termini deletion of S segment could also be observed in natural infected rodents and patients, which was similar to other hantaviruses (Kukkonenetal., 1998; Padulaetal., 2002). It was reported that Hantavirus N protein specifically recognizes the panhandle structure formed by complementary base sequence of 5′and 3′ ends of viral genomic RNA (Miretal., 2006). As the percentage of terminal deleted virus RNAs increases in the population, they could potentially compete with standard virus and downregulate virus replication and expression to escape detection and clearance by the host immune system, which may have implications for viral persistence (Meyeretal., 2000a; Meyeretal., 2000b). However, RNA deletion at 3′ termini could also be detected in clone from a patient. It suggested that terminal sequence deletion might be one of adaptive mechanism to new host for SEOV. It appeared that the number of deleted nucleotides was different between clones from rodents and patient, consequently, the secondary structure of S segment of SOEV differ significantly (Fig.2), which might have effect on interactions between N proteins and panhandle structures of viral genomic RNA. It suggested that the host factors might be important for viral persistence infection.

Distinct from what was observed in cell cultures, in which only short deletions at the 3′ terminal region that were believed to contain the sequence and/or structural features necessary for initiation of replication and transcription could be found (Meyeretal., 2000b), 5′ end nucleotide deletions and both end sequences missing, especially the typical types of DI RNAs (RNAs with large genome deletions) (Fieldsetal., 1990) could also be detected in rodents in this study. Notably, they were not detected in patients. Diversity of termini sequences were detected in more than one rodent, suggesting the events was not rare. Since it is generally believed that rodent hosts are persistently infected hantaviruses while human are infected temporarily, we presume that these deletions might be influenced by host factors which need to be confirmed in the further study.