Human U90926 orthologous long non-coding RNA as a novel biomarker for visual prognosis in herpes simplex virus type-1 induced acute retinal necrosis

Acute retinal necrosis (ARN) is a form of infectious uveitis caused by alpha herpesviruses, including herpes simplex virus type 1 (HSV-1). We previously found that the long non-coding RNA (lncRNA) U90926 is upregulated in murine retinal photoreceptor cells following HSV-1 infection, leading to host cell death. However, to date, an orthologous transcript has not been identified in humans. We investigated U90926 orthologous transcript in humans and examined its utility as a prognostic marker for visual acuity in patients with ARN. We identified two human orthologous transcripts (1955 and 592 bases) of lncRNA U90926. The amount of the longer human U90926 transcript was approximately 30- and 40-fold higher in the vitreous fluid of patients with ARN than in those with sarcoidosis and intraocular lymphoma, respectively. Furthermore, the expression of the longer human U90926 transcript in the vitreous fluid was highly correlated with the final best-corrected logarithm of the minimum angle of resolution visual acuity in patients with ARN (r = 0.7671, p = 0.0079). This suggests higher expression of the longer human U90926 transcript in the vitreous fluid results in worse visual prognosis; therefore, expression of the longer human U90926 transcript is a potential negative prognostic marker for visual acuity in patients with ARN.

Acute retinal necrosis (ARN) is one of the leading causes of blindness in developed countries. It is a type of infectious retinitis caused by members of the alpha herpesvirus subfamily 1 . In a previous study, the main causative viruses of ARN were varicella zoster virus (VZV) (76.1%) and herpes simplex virus (HSV) (15.2%), and ARN caused by HSV was about one-fifth of all cases of ARN caused by VZV 2 . ARN is chiefly diagnosed through viral DNA detection in the intraocular fluid using a high-sensitivity polymerase chain reaction (PCR) assay 1,3 . Although the accurate prediction of visual prognosis is critical to selecting an optimised treatment, biomarkers to predict visual outcomes in patients with ARN are not known. Therefore, it is critical to identify objectively measurable biomarkers that can predict visual prognosis.
Long non-coding RNAs (lncRNAs) are a group of functional RNAs that have diverse biological functions and pathogenic roles, such as immune response 4,5 , pathogenic infection 6,7 , and cancer development [8][9][10][11] . In ARN, the neuronal cells of the retina, such as the retinal photoreceptor cells, have been shown to be the primary infected cells 12 . We have previously shown the upregulation of lncRNA U90926 in murine retinal photoreceptor cells post herpes simplex virus type 1 (HSV-1) infection; this upregulation plays a crucial role in HSV-1 proliferation 13 . However, an orthologous U90926 transcript has not been identified in humans to date. Therefore, in this study, we investigated the presence of U90926 orthologous transcripts in humans and examined whether it could be a prognostic marker for visual acuity in patients with ARN caused by HSV-1. Here, we identified two human

Results
Identification of human U90926 orthologous transcript. To identify the human ortholog of U90926, we searched the Ensembl genome browser (www. ensem bl. org). The mouse U90926 gene is located between Uso1 and Ppef2 on chromosome 5; this gene is evolutionarily conserved between humans and mice ( Fig. 1, top). In the human genome, we found that the AC110615.1 gene was annotated on the region located between USO1 and PPEF2 on chromosome 4 ( Fig. 1, middle). We also found that AC110615.1 encodes short and long variants,   14 , and it is divided into the following three main categories based on aetiology: infectious uveitis, noninfectious uveitis, and masquerade syndrome 15 . Within these categories, ARN is classified as infectious uveitis. We examined the expression levels of the long and short human U90926 transcripts in the vitreous fluid of patients with ARN caused by HSV-1 (a type of infectious uveitis), sarcoidosis (a type of non-infectious uveitis), and intraocular lymphoma (a type of masquerade syndrome). Reverse transcription-quantitative real-time PCR (qRT-PCR) analysis revealed that the expression of the long human U90926 transcript was markedly higher in the vitreous fluid from patients with ARN caused by HSV-1 than from those with sarcoidosis (approximately 30-fold, p = 0.0005) or intraocular lymphoma (approximately 40-fold, p = 0.0005) (Fig. 2, left). The expression of the short human U90926 transcript was almost constant among these three patient groups (Fig. 2, right).  Table 2). In contrast, the viral load in the vitreous fluid was not correlated with the final best-corrected logMAR visual acuity (correlation analysis using Spearman's rank-correlation coefficient; Table 2). These results indicate that expression of the long human U90926 transcript in the vitreous fluid is an excellent negative biomarker for the visual outcomes of patients with ARN caused by HSV-1, although the viral load is not useful as a predictor of visual acuity.  Table 2). This result suggests that expression of the long human U90926 transcript reflects the viral load in the vitreous fluid.

Discussion
In this study, we identified long and short human U90926 transcripts based on high sequence similarities to the mouse U90926 transcript and found that the long human U90926 transcript was dramatically upregulated in the vitreous fluid of patients with ARN caused by HSV-1. Furthermore, in patients with ARN caused by HSV-1, we demonstrated that there was a high correlation between the expression of the human long U90926 transcript in the vitreous fluid and the final best-corrected logMAR visual acuity as well as a moderate correlation between the expression of the long human U90926 transcript and the viral load in the vitreous fluid.
Retinal photoreceptor cells are divided into two major types: cone and rod cells. In humans, cone cells are responsible for colour vision and high visual acuity. Consequently, the death of cone cells leads to significant vision loss 16 . We previously showed that the mouse U90926 transcript was dramatically upregulated in 661W cells derived from cone cells post HSV-1 infection 13 . In 661W cells, lncRNA U90926 acted as a major contributor to host cell death by promoting HSV-1 proliferation 13 . Considering these results, we presumed that upregulated In this study, all patients with ARN caused by HSV-1 were treated with systemic acyclovir therapy before the vitreous fluid was collected. Therefore, we had to consider the possibility that the viral load may have decreased at the time of the vitreous fluid collection due to the acyclovir therapy. However, there was a moderate correlation between the expression of the long human U90926 transcript and the viral load in the vitreous fluid. If the vitreous fluid prior to systemic acyclovir treatment had been used for analysis, it is possible that there would have been an even stronger correlation between the expression of the long human U90926 transcript and the viral load in the vitreous fluid.

Methods
Sequence alignment. Pairwise sequence alignment was performed using Clustal Omega 17 .  6)]. Written informed consent was obtained from all patients. All procedures were performed in accordance with the principles of the Declaration of Helsinki. This study included a total of 21 eyes of 21 patients (12 males and 9 females; average age = 59.0 ± 18.1 years) who underwent pars plana vitrectomy for the diagnosis and/or treatment of ARN, sarcoidosis, and intraocular lymphoma in the University of Tokyo Hospital and National Defense Medical College Hospital between December 2013 and December 2018. There were 8 patients (11 eyes) diagnosed with ARN caused by HSV-1 (7 males and 4 females; average age = 53.5 ± 16.7 years), 5 patients (5 eyes) with sarcoidosis (2 males and 3 females; average age = 57.0 ± 20.0 years), and 5 patients (5 eyes) with intraocular lymphoma (3 males and 2 females; average age = 73.2 ± 14.0 years). The diagnoses of ARN, sarcoidosis, and intraocular lymphoma were made according to certain diagnostic criteria. A diagnosis of ARN was based on the previously reported criteria for ARN in Japan 3 . For sarcoidosis, we used the diagnostic criteria established by the International Workshop on Ocular Sarcoidosis 18 . An intraocular lymphoma diagnosis was made if at least two of the following four criteria were met: (1) the result of cytological diagnosis based on Papanicolaou's classification using Papanicolaou-stained cytological preparations was greater than class III 19 ; (2) interleukin (IL)-10/IL-6 ratio > 1, or IL-10 > 50 pg/mL in the intraocular fluid 20,21 ; (3) strong deviation in the free light chain κ/λ ratio on fluorescence-activated cell sorting analysis 22 ; and (4) positive PCR results for immunoglobulin heavy-chain gene rearrangement [23][24][25] .

Collection of vitreous fluid and measurement of viral loads in the vitreous fluid. A standard
3-port 23-gauge or 25-gauge vitrectomy was performed, and vitreous fluid was collected aseptically without dilution. The collected vitreous fluid was stored immediately in a pre-sterilised cryotube at − 80 °C. DNA was extracted from each vitreous fluid, and then multiplex quantitative PCR assay for viral DNA (HSV-1, HSV-2, varicella zoster virus, Epstein-Barr virus, cytomegalovirus, human herpesvirus 6, human herpesvirus 7, and human herpesvirus 8) was performed as previously described 26 .  Measurement of visual acuity. Decimal visual acuity was measured using a Landolt ring, and then converted to logMAR visual acuity using the following formula: logMAR visual acuity = log (1/decimal visual acuity). As previously described 27 , logMAR visual acuity of finger counting, motus manus, and light sense was defined as 2.4, 2.7, and 3.0, respectively. A higher logMAR value indicates lower visual activity.

Statistical analysis.
Values were presented as median with interquartile range. Comparison of continuous variables between the two groups was performed using Mann-Whitney U-test. Correlation between two variables was analysed using Spearman's rank-correlation coefficient. A p value < 0.05 was considered statistically significant. Both Mann-Whitney U-test and correlation analysis were performed using GraphPad Prism version 7 (GraphPad Software, San Diego, CA).

Data availability
The materials and datasets used and analysed during the current study can be obtained from the corresponding author on reasonable request.