Introduction
HSV-1 is a ubiquitous but important human pathogen carried by over half of the world’s population; HSV-1 infection starts with primary infection at the periphery and subsequent lifelong latency in the peripheral nervous system [
1]. In experimental animals such as mouse, acute infection develops following cornea inoculation, the virus replicates in the epithelial cells on the corneal surface and is later transported into trigeminal ganglia, where it establishes latency [
2‐
4]. The acute stage of infection involves lytic, or productive infection of HSV-1 at the site of inoculation, and frequently in the TGs of infected mice; in the tree shrew TGs, however, signs of lytic infected were not seen and no infectious virus could be detected [
5]. During the lytic phase of infection in cultured cells, all viral genes are believed to be expressed in a cascade-dependent manner [
6,
7], but during latency, most viral genes are silenced with the exception of the latency-associated transcript (LAT), multiple miRNAs [
8‐
10] and two small RNAs [
11]. Although latent, the virus reactivates from individual neurons periodically, and could cause more serious diseases including herpes keratitis or herpes encephalitis [
12,
13]. Recently, HSV-1 has been suggested to play a role in Alzheimer’s disease [
14,
15].
Multiple animal species including mice, rabbits and tree shrews [
5,
16,
17] have been used to model human HSV-1 infection, with mice being the most widely used. However, we have recently studied HSV-1 infection in tree shrews, which are more closely related to primates than rodents [
18], and observed differences between tree shrews and mice during acute infection, latency and reactivation [
5,
19]. Here, we performed an in-depth transcriptional profiling of the infected mouse, tree shrew and human TGs, and reveal the differences in viral gene expression patterns. Importantly, we show that latent tree shrew TGs can express many viral genes, including UL6, UL8 and ICP8, which are consistent with spontaneous reactivation. More importantly, during latency tree shrew and human TGs had more similar transcription peaks in the LAT region and possibly more ICP0 transcription, supporting that the tree shrew model better mimics HSV-1 latent infection in human.
Discussion
HSV-1 infection in tree shrew TGs differs from that in mouse, and viral transcriptional level in tree shrews and mice are affected by many factors. In our previous relative HSV-1 genome quantitative data, acutely infected mouse TGs contained higher levels of viral genome than tree shrew TGs [
5], which also contribute to the differences of viral transcription. Another factor that cannot be ignored is whether HSV-1 replicates effectively in the eye of the tree shrew model, because surface replication is the source of the virus transported into the ganglion. In fact, we had tested HSV-1 titer and genome copy number in tree shrew eyes within a period of 46 days after ocular infection (Li et al., in press), and these results support that HSV-1 replication in tree shrew eyes could be comparable with that in mouse eyes.
An important finding from this study is the apparent spontaneous reactivation from latency from two of the tree shrew TGs sequenced, a result supporting our previous study where live virus could be recovered from eye tears in latently infected animals---a feature that mimics human infections seen in tree shrew but less efficient in mouse. From genes detected during spontaneous reactivation, we found immediate early genes ICP0, ICP22 and ICP8, which are important for viral early transcription and genome replication. However, a number of studies support the spontaneous reactivation of infectious HSV-1 in murine sensory ganglia [
31,
32], and ICP0 signal was detected in individual latently infected mice in our RNAseq data, but other lytic viral genes reads are too low to tell if they are positive.
What caused the virus to spontaneously reactivate is currently unknown. LAT is the only known transcript in latently infected tissues [
10,
33,
34], but it is not just a latency marker, it maintains viral latency [
35,
36] by inhibiting apoptosis [
37‐
40], repressing lytic gene expression [
41], etc. There have been many studies supporting LAT is essential for the highly induced reactivation phenotype in the mouse model [
42‐
44], and for the high induced or spontaneous reactivation in the rabbit model [
45‐
53]. If true, then the high level of LAT transcript in tree shrew TGs compared to mouse could partially explain why the virus spontaneously reactivates in tree shrew but not in mouse.
In our analyses comparing latently infected mouse, tree shrew and human TGs, we observed that LAT could be detected from all three, but tree shrew TG was more similar to human TG in that both have high reads in a LAT intron region. Considering that the oligo (dT) method cannot enrich non-polyadenylated LATs, the high level of LAT signal are mainly supported by polyadenylated LATs [
54‐
57], the signal in the LAT intron region is in fact ICP0, but not LAT intron, which is not polyadenylated. This is supported by the strand-specific RT-qPCR experiments shown in Fig.
5a.
A more important similarity between infected tree shrew and human TGs is that both express transcript from the ICP0 region. In contrast, there are no ICP0 transcripts detected in infected mouse TGs, which is inconsistent with some previous reports that the ICP0 signal they detected was located near the first exon at the 5′ end [
58,
59]. In light of this, this comparison could also include infected rabbits TGs in future analysis to examine whether this ICP0 transcripts exist in reactivating samples.
Both in tree shrew and human TGs, most of the ICP0 reads are concentrated at the 3′ end of the ICP0 gene (Fig.
5c & d), while the remaining region of ICP0 has very few reads, raising the possibility that this region is independently transcribed. However, it is also possible that the uneven reads were a result of the high GC content of the HSV genome. Since we could not amplify the full-length ICP0 transcript, we could not rule out the possibility of partial transcripts arising from the 3′ of ICP0. In addition, there is a 0.8 kb LAT insulator and a CTCF binding site within LAT intron region [
60], and these elements are both located up-stream of the LAT/ICP0 transcription peak, making it possible that the CTCF binding site could serve as a promoter to for this transcript, as reporter by others [
61].
Conclusion
In this study, we compared the viral transcriptome of infected mouse and tree shrew TGs during the course of infection, and we compared the pattern of viral transcripts in latently infected mouse and tree shrew TGs with that of human TGs. We found that HSV-1 transcription in acutely infected TGs differs dramatically between mouse and tree shrew, with HSV-1 in mouse TG undergoing productive infection, while that in tree shrew TGs appears to be going through an abortive infectious cycle, missing keys genes needed for viral transcription, replication and maturation. During the latent phase of infection, LAT was detected in mouse, tree shrew and human TGs, but we also detected an ICP0 transcript fragment from tree shrew and human TGs, making tree shrew latent infection more similar to human than mouse. When we analyzed tree shrew TGs individually, we found samples that appeared to be undergoing spontaneous reactivation. Together these analyses support the tree shrew as a better model of human HSV-1 infection in the peripheral nervous system, offering the possibility of a better understand HSV-1 latency and reactivation, and the discovery of potential novel targets for therapeutic interventions. Taken together, the transcriptome data reveals that tree shrews and humans have a more similar transcription pattern in the LAT region during latency than that of mice and humans, supporting tree shrews as a more accurate animal model for research on HSV-1 latency and reactivation.
Materials and methods
Experimental animals
Chinese tree shrews, 6-month-old female, were obtained from the experimental animal core facility of the Kunming Institute of Zoology, Chinese Academy of Sciences. During the experiment, tree shrews were kept in experimental cages of 54 cm × 45 cm × 50 cm, and no more than two animals per cage. The temperature in the room was controlled at 15 ~ 28 °C, the relative humidity was 40 ~ 70%, the daily light was 12 h, and the noise should not exceed 60 dB. BALB/C mice, 6-week-old female, were obtained from the Kunming Medical University. All experimental procedures and animal care were carried out in accordance with the protocols approved by the Institutional Animal Care and Use Committee of the Kunming Institute of Zoology, Chinese Academy of Sciences. The research program was reviewed and approved by the Institutional Animal Care and Use Committee of Kunming Institute of Zoology, Chinese Academy of Sciences.
Virus and cells
HSV-1 strain 17+ was used to infect mouse and tree shrews. Viral culture was performed in Vero cells and titrated by plaque forming assay on RS1 cells, and these two cells were both obtained from Conservation Genetics Academy of Science (CAS) Kunming Cell Bank. The infected cells were cultured in DMEM supplemented with 2% fetal bovine serum (GibcoTM). All cells were maintained at 37 °C with 5% CO2. Experiments involving infectious virus were conducted in a Biosafety level 2 laboratory.
Animal infections and collection of samples
Animal anesthesia, corneal scarification and inoculation with HSV-1 17+ virus was as previously described [
5]. 1 × 10
4 PFU of HSV-1 were used to inoculate each mouse eye, and 1 × 10
6 PFU were used on tree shrews, which was consistent with our previous reports [
5,
19]. The control (mock infected) animals were also scratched and treated with Vero cell supernatant. To generate biological replicates, three groups of animals were infected independently on different days. Samples of infected trigeminal ganglions were collected at 3, 5, 7, 10, 14, 28, 58dpi, and mock infected samples were collected after 24 h of treatment, and then were ground to a fine powder in liquid nitrogen using 50 ml grinding beakers and 20-mm grinding ball for RNA extraction.
Human trigeminal ganglions collection and preparation
Four human trigeminal ganglia were obtained at autopsy and provided by Academy of Forensic Science (Shanghai, China) and Second Affiliated Hospital of Kunming Medical University. According to the record, the material was taken after 2–3 days of death, then stored in liquid nitrogen and transported with dry ice, and demographics are shown in supplemental Table
3. This is about as early as it is possible to legally obtain human tissue for our studies. Tissues were stored at − 80 °C until processing. To disrupt the tissue, frozen ganglia were wrapped in aluminum foil and mechanically broken apart on dry ice using a metal anvil and hammer, both also frozen on dry ice. Small fragments of tissue from a trigeminal ganglion from were used for RNA extraction.
RNA extraction and sequencing
Mouse, tree shrew or human TGs were individually ground to a fine powder in liquid nitrogen before RNA extraction. For each sample, approximately 1 g of powder (from two TGs of one animal) was resuspended in 1 ml TRIzol reagent (Life Technologies) and total RNA was extracted according to the manufacturer’s recommendation. Afterwards, RNA samples were sent to BGI (Wuhan, China) for purification, library preparation and sequencing. The Ambion Trubo DNA-free kit (Life Technologies) was used to eliminate genomic DNA contamination, and an Agilent 2100 Bioanalyzer (Agilent RNA 6000 Nano Kit) to perform the total RNA sample quality control. mRNAs were isolated from total RNA using the oligo (dT) method, and then purified and fragmented using divalent cations under elevated temperature. The oligo (dT) method could not extract, and thus excluded non-polyadenylated transcripts, such as non-polyadenylated LAT introns, and retains only polyadenylated LATs [
54‐
57]. First strand cDNA was synthesized using random primers, and the second strand cDNA was synthesised with Polymerase I and RNase H. cDNA fragments were purified and resolved with EB buffer for end reparation and single nucleotide A (adenine) addition. After that, the cDNA fragments were linked with adapters. Those cDNA fragments with suitable size (300 bp) were selected for PCR amplification. An Agilent 2100 Bioanaylzer and ABI StepOnePlus Real-Time PCR System were used in quantification and qualification of the libraries. Finally, all of the transcriptome libraries were sequenced using an Illumina HiSeq X Ten sequencer with a paired-end protocol.
Read mapping, normalization, and statistical analysis of differential gene expression
According to the BGI’s instructions, the low quality reads (More than 20% of the bases qualities were lower than 10), reads with adaptors and reads with unknown bases (N bases more than 5%) were filtered using an internal software, SOAPnuke, to get the clean reads. After filtering, the remaining reads were stored in FASTQ format. All reads were aligned to the host genome or HSV-1 17+ using HISAT2 software with default parameters [
62]. In viral transcriptome analysis, all reads aligned to host genomes were filtered out, and all unmapped reads were then mapped against HSV-1 17+ genome. The quantification of transcript abundance (raw count) was conducted using featureCounts [
63] software supported by the Subread package [
64], and multi-mapping and multi-overlapping reads were excluded from expression analysis. We chose to normalize our data set using the relative log expression method implemented in the DESeq2 package (v1.22.2) [
20] in R (
www.r-project.org), and only viral genes that had a read count of at least 5 were considered. After DESeq2 analyses, genes with a log2 fold change > 1 and FDR- adjusted p-value <− 0.01 were considered differentially expressed.
Reverse transcription PCR (RT-PCR) and quantitative real-time RT-PCR (RT-qPCR)
In Fig.
5a and b, reverse transcription used strand-specific primers for ICP0 and LAT transcript respectively and was performed with a high fidelity RT-PCR kit (TaKaRa), according to the manufacturer’s instructions. In a 20 μl reaction mixtures contained RNase inhibitor, RTase, PrimeScript buffer, dNTP mixture, 1 μg template RNA and 5pM a strand-specific RT primer. After RT reaction, 1 μL cDNA was used for PCR reaction flowing by 30 cycles of 10 s at 98 °C, then 5 s at 60 °C, and 20s at 72 °C. The RT-PCR product was electrophoresed on a 2% agarose gel for 20 min at 200 V. For gene expression analysis via RT-qPCR, isolated total RNA was reverse-transcribed using a PrimeScriptTM RT reagent Kit with gDNA Eraser (TaKaRa). The RT-qPCR was performed in an Applied Biosystems 7900HT using FastStart Universal SYBR Green Master (ROX, Roche). Per RT-qPCR reaction, cDNA derived from 50 ng RNA was deployed. Cycling conditions were 10 min 95 °C, followed by 45 cycles of 10s at 95 °C, 30s at 60 °C, and 30s at 72 °C. Primers are listed in supplemental Table
2. For the expression of fold changes of RT-qPCR data, the 2 − ΔΔCt method was used [
65].
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.