A 2.5-kilobase deletion containing a cluster of nine microRNAs in the latency-associated-transcript locus of the pseudorabies virus affects the host response of porcine trigeminal ganglia during established latency

Abstract

The alphaherpesvirus pseudorabies virus (PrV) establishes latency primarily in neurons of trigeminal ganglia when only the transcription of the latency-associated transcript (LAT) locus is detected. Eleven microRNAs (miRNAs) cluster within the LAT, suggesting a role in establishment and/or maintenance of latency. We generated a mutant (M) PrV deleted of nine miRNA genes which displayed properties that were almost identical to those of the parental PrV wild type (WT) during propagation in vitro. Fifteen pigs were experimentally infected with either WT or M virus or were mock infected. Similar levels of virus excretion and host antibody response were observed in all infected animals. At 62 days postinfection, trigeminal ganglia were excised and profiled by deep sequencing and quantitative RT-PCR. Latency was established in all infected animals without evidence of viral reactivation, demonstrating that miRNAs are not essential for this process. Lower levels of the large latency transcript (LLT) were found in ganglia infected by M PrV than in those infected by WT PrV. All PrV miRNAs were expressed, with highest expression observed for prv-miR-LLT1, prv-miR-LLT2 (in WT ganglia), and prv-miR-LLT10 (in both WT and M ganglia). No evidence of differentially expressed porcine miRNAs was found. Fifty-four porcine genes were differentially expressed between WT, M, and control ganglia. Both viruses triggered a strong host immune response, but in M ganglia gene upregulation was prevalent. Pathway analyses indicated that several biofunctions, including those related to cell-mediated immune response and the migration of dendritic cells, were impaired in M ganglia. These findings are consistent with a function of the LAT locus in the modulation of host response for maintaining a latent state.

Importance: This study provides a thorough reference on the establishment of latency by PrV in its natural host, the pig. Our results corroborate the evidence obtained from the study of several LAT mutants of other alphaherpesviruses encoding miRNAs from their LAT regions. Neither PrV miRNA expression nor high LLT expression levels are essential to achieve latency in trigeminal ganglia. Once latency is established by PrV, the only remarkable differences are found in the pattern of host response. This indicates that, as in herpes simplex virus, LAT functions as an immune evasion locus.

FIG 1

(A) Physical map of the PrV-Ka genome containing unique (U_L and U_S) and inverted repeat (IR and TR) sequences. BamHI restriction sites and fragments, as well as the insertion of a bacterial vector and of an EGFP reporter gene cassette at the gG gene locus in pPrV-ΔgGG (24), are indicated. (B) An enlarged section shows the boundary between U_L and I_R with the open reading frames of the regulatory proteins EP0 and IE180. Viral mRNAs and the spliced large latency transcript (LLT) are indicated by dotted arrows. Identified miRNAs (22, 23) are shown as arrowheads numbered from 1 to 11 (corresponding to miRNA genes from prv-mir-LLT1 to prv-mir-LLT11). In pPrV-ΔmiRN, the majority of the miRNA genes were deleted and replaced by selection markers (RpsL and KanR) used for BAC mutagenesis in E. coli.

FIG 2

Replication of pPrV-ΔgGG and pPrV-ΔmiRN in PK15 (A) and RK13 (B) cells. Progeny virus titers were determined between 4 and 24 h after infection at a multiplicity of infection (MOI) of 10 (PK15) or 5 (RK13). Titers represent mean values from three independent experiments with bars showing standard deviations.

FIG 3

RT-qPCR expression kinetics of EP0 (A and D [after primer-specific RT]), IE180 (B), and the LLT exon1/exon2 junction (C) during PrV infection in vitro. PK15 cells were infected with pPrV-ΔmiRN (light gray) and pPrV-ΔgGG (dark gray) at an MOI of 10. Values are provided as mean C_T values and are the averages from three biological replicates (higher C_T values mean decreased gene expression levels). The qPCRs were normalized to the input amount of total RNA.

FIG 4

Establishment of latency in vivo. Pigs were infected with either pPrV-ΔgGG (WT 54 to 58) or pPrV-ΔmiRN (M 49 to 53) or were mock infected (C 22 to 25). (A and B) DNAs from nasal swabs of animals infected by WT PrV (A) or M PrV (B) were analyzed by RT-qPCR of the PrV gB gene log₂ (RQ), log₂ of relative quantity. (C and D) The host antibody response was analyzed by ELISA using PrV gB as the antigen. The threshold value of the assay (0.7) is indicated as a horizontal line.

FIG 5

Relative amounts of PrV genomes in latent trigeminal ganglia. (A) The PrV genome copy value per 100 ng of genomic DNA was quantified by qPCR using a GFP amplicon. (B) PA-GFP-coilin C2 plasmid DNA standard curve. The x axis represents the input copies of plasmid DNA, and the y axis represents the mean threshold cycle (C_T mean).

FIG 6

RT-qPCR profiles of prv-miR-LLT1, prv-miR-LLT2, and prv-miR-LLT10 in trigeminal ganglia latent for the WT or M PrV (A) and in PK15 cells at 12 h p.i. with the WT PrV (B). Values are normalized against the background and are indicated as 2^−ΔCT(± standard deviations).

FIG 7

Pattern of transcription of three regions of LLT (exon 1, ex1/ex2 junction, and exon 2) in trigeminal ganglia latent for WT or M PrV. RT-qPCR values were calibrated versus the relative amounts of PrV genomes. Data are 2^−ΔCT (± standard deviations) values calculated from three technical replicates.

FIG 8

Visualization of the distribution of RNA-seq reads obtained by RNA-seq profiling of trigeminal ganglia latent for the mutant (M) or parental (WT) PrV on the PrV genome.

FIG 9

IL-6, gamma interferon, and TNF were identified by IPA as the most significant upstream regulators (z scores >2) to explain the pattern of transcription of 20 DE genes, 15 of which belong to the top IPA network, termed cell-mediated immune response, cellular movement, hematological system development, and function (17 DE genes). Left, WT versus C; right, M versus C. Numbers are the logFC values of each comparison. Red, upregulated; green, downregulated; orange, leads to activation; blue, leads to inhibition; yellow, finding inconsistent with state of downstream molecules; gray, effect not predicted.