Cell type-dependent RNA recombination frequency in the Japanese encephalitis virus

Abstract

Japanese encephalitis virus (JEV) is one of approximately 70 flaviviruses, frequently causing symptoms involving the central nervous system. Mutations of its genomic RNA frequently occur during viral replication, which is believed to be a force contributing to viral evolution. Nevertheless, accumulating evidences show that some JEV strains may have actually arisen from RNA recombination between genetically different populations of the virus. We have demonstrated that RNA recombination in JEV occurs unequally in different cell types. In the present study, viral RNA fragments transfected into as well as viral RNAs synthesized in mosquito cells were shown not to be stable, especially in the early phase of infection possibly via cleavage by exoribonuclease. Such cleaved small RNA fragments may be further degraded through an RNA interference pathway triggered by viral double-stranded RNA during replication in mosquito cells, resulting in a lower frequency of RNA recombination in mosquito cells compared to that which occurs in mammalian cells. In fact, adjustment of viral RNA to an appropriately lower level in mosquito cells prevents overgrowth of the virus and is beneficial for cells to survive the infection. Our findings may also account for the slower evolution of arboviruses as reported previously.

Figure 1

The schematic sketch designed to identify RNA recombination between viral strains. A fragment (868 bp) comprised of the C/preM junction (nt 10~877) of viral RNA extracted from coinfected BHK-21 or C6/36 cells was amplified, cloned, and then used for an RFLP analysis with SmaI or Alw44I. Two and one recombinant form(s) were, respectively, identified in selected samples from BHK-21 and C6/36 cells, when they were coinfected with the T1P1-S1 and CJN-S1 strains of the Japanese encephalitis virus.

Figure 2

RNA recombination between genomic RNA and a transfected RNA sequence. No fragment was seen in tests with RNA extracted from cells following mock treatment (with neither infection nor transfection), virus infection only, or transfection only. Although amplification of a 564 bp fragment showing RNA recombination was present in the control group which contained a mixture of RNAs extracted from infected and transfected cells, RNA recombination was significantly elevated in BHK-21 and C6/36 cells infected by the Japanese encephalitis virus (Nakayama strain) following transfection with the (+)5′3′-UTR-I plasmid RNA. According to the image-density analysis, it seems that RNA recombination occurred less frequently in mosquito cells. A specific fragment of viral RNA (529 bp) was used as an internal control in all groups with viral infection.

Figure 3

Status of RNA recombination after inhibition by exoribonuclease with PAP (3′-phosphoadenosine-5′-phosphate, an inhibitor of exoribonuclease). (a) The RNA recombination rate increased to a higher level in BHK-21 cells after treatment with PAP, compared to that of untreated cells. In contrast, no effect of PAP on increasing RNA recombination of the virus was shown in C6/36 cells despite a very low level of RNA recombination still being observed. Viral RNA was not affected after treatment with PAP, suggesting exoribonuclease-mediated degradation of transfected RNA fragments might increase RNA recombination of the virus strains, especially in mammalian cells. (b) Treatment with PAP in C6/36 cells did not cause degradation of the transfected (+) RNA fragment at 3 h until 6 h after transfection at which a partial effect appeared.

Figure 4

Degradation of double-stranded (ds) RNA fragments transfected into mosquito cells. A fragment (807 bp) of viral RNA extracted from C6/36 cells was detected through an RT-PCR at 0 h after transfection (hpt) with dsRNA derived from (+) or (−) 5′3′-UTR RNA. The transfected dsRNA had faded at 3 and 6 h after transfection, suggesting that dsRNAs may have been cleaved, and thus generated undetectable short interfering RNAs.

Figure 5

Viral RNA, either T1P1-S1 or CJN-S1, accumulated in C6/36 cells more slowly than in BHK-21 cells. (a) The RNA amount of T1P1-S1 remained at the baseline level until 12 h after infection (hpi) (3.81-fold change), which increased to 169.72-fold at 24 hpi in C6/36 cells. In contrast, T1P1-S1 RNA, respectively, increased to 3.09-, 28.99-, 429.05-, 4396.07-, and 5487.75-fold, at 3, 6, 9, 12, and 24 hpi in BHK-21 cells. The RNA amount of CJN-S1 also accumulated more slowly in C6/36 cells than BHK-21 cells, which remained at the baseline level until 12 hpi (2.36-fold increase) and had increased to 152.32-fold by 24 hpi in C6/36 cells. Although the amount of CJN-S1 RNA did not evidently increase until 6 hpi (1.35-fold change), it increased to 16.64-, 111.43-, and 554.87-fold at 9, 12, and 24 hpi, respectively, in BHK-21 cells. (b) Stability of viral RNA was evaluated after a fragment of (+)5′3′-UTR-II RNA was transfected into BHK-21 or C6/36 cells. Transfected fragments were insignificantly degraded even at 3 or 6 h after transfection in BHK-21 cells while more obvious degradation appeared in C6/36 cells.