Probes for INS

Herpes simplex virus 1 (HSV-1) establishes latency in neurons and expresses long noncoding RNAs termed the latency-associated transcripts (LATs). Two previous studies using HSV-1 recombinants containing deletions in the LAT promoter revealed opposing effects of the promoter deletion regarding the heterochromatinization of latent viral genomes in mice ganglia. Confounding variables in these studies include viral strains utilized (17syn+ versus KOS), anatomical infection site (footpad versus eye) and infectious virus dose (500 versus 1 × 105 PFU), and to date the basis for the differences between the two studies remains unresolved. We recently reported that 17syn+ and KOS display distinct differences in heterochromatin levels during latency in human neurons. This raised the possibility that the discrepancy seen between the two previous studies could be explained by strain-specific differences within the LAT region. Here, we examine two recombinants containing orthologous 202 bp LAT promoter deletions, 17ΔPst and KOSΔPst, in a human neuronal model of latency and reactivation (LUHMES). We found that LUHMES neurons recapitulate previous observations in mice where deletion of the LAT promoter results in an increase in H3K27me3 deposition on the viral genome compared to the parental strain 17syn+ but a decrease compared to the parental strain KOS. We also found distinct strain-specific differences in the production of viral transcripts and proteins during latency. These results indicate that the function and/or regulation of the LATs differs between HSV-1 strains and may shed light on some discrepancies found in the literature when examining the function of the LATs. IMPORTANCE Herpes simplex virus 1 (HSV-1) establishes a lifelong infection in neuronal cells. Periodically, the virus reactivates from this latent state and causes recurrent disease. Mechanisms that control entry into and maintenance of latency are not well understood, though epigenetic posttranslational modification of histones associated with the viral genome are known to play an important role. During latency, the latency-associated transcript (LAT) is known to impact epigenetic marks, but the ultimate effect has been a point of controversy. Here, we utilize a human neuronal cell line model of HSV latency and reactivation (LUHMES) to characterize latency for two HSV-1 wild-type strains and their respective LAT promoter deletion viruses. We find that the LAT acts in a strain-specific manner to influence levels of chromatin marks, viral transcription, and viral protein production. This work highlights the need to account for strain-specific differences when characterizing the LAT's function and the dynamics of reactivation.

To understand the subcellular localization of RUNX2 and two lncRNAs, LINC02035 and LOC100130207, immunocytochemistry (for RUNX2 protein) and RNA _in situ_ hybridization assays (for both lncRNAs) were performed using human primary chondrocytes isolated from knee cartilage of OA patients. We confirmed that the RUNX2 protein was strongly detected in the nucleus of chondrocytes isolated from damaged cartilage (Figure 4A). The fractionated western blot results also showed that the RUNX2 protein was detected only in the nucleus of chondrocytes isolated from damaged cartilage (Figure 4B). To further understand the molecular mechanisms of the lncRNAs LINC02035 and LOC100130207, we performed an _in situ_ assay using primary chondrocytes derived from patients, because primary chondrocytes are a valuable model for studying OA pathogenesis. The results showed that both LINC02035 and LOC100130207 were highly expressed in chondrocytes isolated from the knee cartilage of patients with OA (Figure 4C). We then evaluated the mRNA levels and subcellular localization of both lncRNAs to elucidate their site of action using a commercially available kits in primary chondrocytes isolated from intact or damaged cartilage tissues. The results showed that both lncRNAs were more upregulated in primary chondrocytes isolated from damaged cartilage tissue than in intact cartilage tissue (Figure 4D). In primary chondrocytes, LINC02035 and LOC100130207 were merely detected in the cytoplasm of human primary chondrocytes and both lncRNAs were localized to nucleus (Figure 4E). Likewise, we also studied the subcellular localization of both lncRNAs in TC28a2 cells. The results showed that LINC02035 and LOC100130207 were evenly distributed in the nucleus and cytoplasm of normal chondrocytes (Figure 4F, left). However, both lncRNAs were preferentially localized to the nucleus and to a lesser extent to the cytoplasm after TC28a2 cells were treated with hypertrophic medium or TNF-α (Figure 4F, middle and right). To investigate whether RUNX2 is regulated at the post-translational level during hypertrophic changes in chondrocytes, human primary chondrocytes or TC28a2 cells were treated with the proteasome inhibitor MG132. The results showed that the protein level of RUNX2 was dose-dependently increased by MG132 treatment (Figure 4G-H), indicating that the upregulation of RUNX2 in osteoarthritic or hypertrophic chondrocytes occurs at the post-translational level. To examine whether both lncRNAs are involved in the stabilization of RUNX2 protein during hypertrophic differentiation and the inflammatory response in chondrocytes, IP was conducted to confirm the ubiquitination of RUNX2 protein. First, we investigated how the ubiquitination of RUNX2 protein is regulated during hypertrophic differentiation or the inflammatory response of chondrocytes, and as a result, it was confirmed that ubiquitination of RUNX2 was reduced by hypertrophic medium or TNF-α treatment (Figure 4I). However, ubiquitination of RUNX2 protein was clearly increased in TC28a2 cells transfected with siRNAs targeting LINC02035 or LOC100130207, even though the cells were treated with hypertrophic medium or TNF-α (Figure 4J-K). These results suggest that both lncRNAs upregulated during hypertrophic differentiation and the inflammatory response in chondrocytes contribute to the stabilization of the RUNX2 protein.