*3.7. The Transcriptome Comparison between Untreated (Time Zero) and Treated (3 h* + *6 h) Rpe Cells Revealed the Possible Impairment of Retinal Cells Crosstalk and Synapses, Leading to Rescue or Cell Death*

Many DE and DAS genes only showed significant differences only if analyzed during the whole considered treatment period. Among these DE genes, five (*CCN2*, *ACTG1*, *UTP14C*, *TMSB4XP6* and *TMEM189-UBE2V1*) were linked to extracellular matrix constituent secretion and cellular junctions, as well as to misfolded protein ubiquitination. Furthermore, the number of DAS genes and DTU transcripts changed during treatment was very high. The first ones (*ACADVL*, *GPI*, *HNRNPA1*, *CD81*, *CD63*, *CTSB*, *CTSH*, *LTBP3*, *CAPNS1* and *PAX6*) were enriched in terms related to neuropeptide catabolic processes and extracellular vesicles in the crosstalk of cells, while the second ones (ENST00000586425.2, ENST00000550050.5, ENST00000588991.7, ENST00000527343.5, ENST00000584364.5, ENST00000263645.9, ENST00000615999.4 and ENST00000518154.5) mainly regard dendrite regeneration. Therefore, the overall scenario could reveal a dual response by stressed cells: on one hand, the alteration of retinal cell crosstalk and synapses could lead to various forms of cell death (e.g., autophagy). On the other hand, RPE tries to survive, increasing regeneration of capable parts (e.g., dendrites), by synaptic plasticity. Detailed info on pathways enrichment are available in Table S6 and Table S7).

#### *3.8. The Most Significant DAS Genes Represented the Main Retinal Dystrophy Candidate Genes*

The ToppGene prioritization analysis on known causative retinal dystrophy genes that intersected with most significant DE and DAS genes previously described revealed 19 candidate genes (Bonferroni corrected *p*-value <0.05). Of these, seven showed a strong association with the training genes (Bonferroni corrected *p*-value <0.01). Most of the 19 significant obtained candidate genes, included three of most statistically associated (*PAX6*, *CTSH* and *HNRPA1*) belong to DAS genes, further highlighting the influence of oxidative stress on alternative splicing (Table 1). Details of ToppGene results are available in Table S8.

### *3.9. qRT-PCR Validation*

To validate the authenticity and reproducibility of the RNA-Seq results, the 15 selected mRNAs were validated by qRT-PCR analysis, and obtained expression profiles were similar to the results of transcriptome analysis (Table S9). Moreover, the ANOVA statistics, conducted to compare the means among multiple groups, highlighted high significance (*p*-values < 0.05). The linear regression analysis showed a significantly positive correlation of the relationship between gene expression ratios of qRT-PCR and RNA-Seq for all selected time points (Figure S3), confirming our transcriptomic data validity.

*Antioxidants* **2020**, *9*, 307



#### **4. Discussion**

Retinal dystrophies like age-related macular degeneration and, particularly, retinitis pigmentosa represent a very heterogeneous group of ocular pathologies characterized by a very complex pattern of environmental and genetic causes. One of the most challenging aspects regards the incomplete knowledge of all causative genes and their involved biochemical and molecular pathways, leading to a huge group of orphan forms [37]. Gene mutations or dysfunctional processes not only in the retina but also in RPE could cause inherited retinal degeneration, age-related macular degeneration and other retinal diseases [38]. Such a feature highlights the relevant role of RPE, a high metabolic demand monolayer of pigmented cells that plays fundamental functions for both rods and cones, such as metabolite transport and photoreceptor excitability, regulation of visual cycle, secretion of growth factors, phagocytosis of photoreceptor outer segments (POSs) and oxidative stress defense. Regarding the latter point, oxidative stress represents one of the major lethal mechanisms responsible for age-related RPE damages [39]. Many studies have demonstrated that accumulation of lipid deposit called lipofuscin generates reactive oxygen species through phototoxicity in RPE cells [40–42]. Oxidative stress triggered by photo-oxidation of bis-retinoid A2E, a lipofuscin constituent, is well known to be a progression factor of age-related macular degeneration and also in genetic macular degeneration syndromes such as Stargardt disease [43], but very little is known about A2E involvement in retinitis pigmentosa.

In this study, we treated RPE cells with A2E during a follow-up of two time points (3 h and 6 h) after exposure and compared them to untreated time zero controls. The main purpose of our experiment was the discovery of new pathways potentially involved in retinal dystrophies development, with the further detection of new candidate genes that could be associated or causative of such ocular diseases, emerging from the expression analysis in such altered conditions.

Starting from an initial average value of 16,173 detected genes per sample, about 2432 showed changes in their expression level and 119 were differentially alternative spliced with transiently, late or enduring fluctuations. Selected altered DE and DAS genes were then functionally and statistically analyzed and clustered into final 10 candidate "macro-pathways", showing a very variable time of exposure-related trends. Considering the average fold-change of each constituting genes and their reciprocal connections, we revealed a more detailed functional network. Such connections could help to depict several causative/associative clusters, underlining a more complex pattern of possible retinal dystrophies etiologies.

From analyses of DE genes related pathways, it emerged that, after an early increase of apoptosis processes, the programmed cell death decreased in both considered time points following A2E exposure, probably due to activation of rescue systems and to a limited percentage of survived cells. An opposite trend was shown by steroid receptor and nucleoside transport activities, which evidenced a huge up-regulation of involved genes after 3 h and 6 h. Such results could reflect late alterations in RPE antioxidant and anti-inflammatory abilities [44], as well as inhibition of photoreceptor outer segment (POS) phagocytosis and impairment of ion currents in retinal cells [45,46]. Furthermore, a very interesting result was the sinusoidal trend involving isoprenoid pathway, related to cholesterol-dependent homeostasis of POS [47] and angiogenesis [48]. After 3 h from treatment, it could be possible that disc bulk membranes increased trying to improve phototransduction by residual photoreceptors, despite the decreased choriocapillaris viability due to the reduced vascular endothelial growth factor (VEGF) [49]. The situation reversed at 6 h observation, when the discs turnover was drastically decreased and the VEGF level increased, which could contribute to the subretinal neovascularization already characterized in wet age macular degeneration [50].

Additionally, the extensive alternative splicing information identified from DAS-related pathway analysis highlighted a much higher degree of complexity of regulation in response to A2E-induced oxidative stress, which has been significantly underestimated by analysis of DE genes only. In particular, speed and extent of the oxidative stress-induced AS suggested that AS, together with the transcriptional response, is a major driver of transcriptome reprogramming for RPE cell death and their attempts to

survive. From 3 h and up to 6 h from treatment, an impairment of intracellular traffic related to Rab proteins, already reported in choroideremia [51], was observed along with the alteration of autophagy and accumulation of proteins and damaged organelles. These events are typical of AMD [52] and are also enforced by inactivation of chaperone genes [53,54]. This scenario could reflect a strong reduction of macroautophagy (a catabolic cell survival system) and of a hybrid autophagy–phagocytosis-degradative pathway called LC3-associated phagocytosis (LAP) [55], which plays a critical role in visual pigment regeneration, as well as the complete degradation of phagosomes [56]. The other resulted pathway "big cluster", instead, highlighted a global up-regulation of DAS-involved genes up to the end of the last considered time point. In particular, we demonstrated the dynamic contribution of AS by changes in multiple different mechanisms of transcription and translation. Photosensitization of A2E stimulates oxidative DNA damage, such as the production of 8-oxo-guanines in telomeres, leading to their possible damage [57]. Thus, the resulted alteration of telomerase RNA localization to Cajal body could accelerate the RPE senescence [58], and this process could be further increased by the reduction of FGFR1 signaling and the consequent POS phagocytosis decrease [59]. Moreover, DNA damage response could increase the activity of miRNAs involved in [60,61] and cell death genes transcription by TP53 [62], determining a possible role in retina degeneration. In the meantime, as already discussed, retinal cells could try to fight against induced stress, and one resistance mechanism could be represented by the improved maturation of the large ribosomal unit (LSU) [63] and by the increased polyribosome activity, showing a fundamental role in translation regulation of many retinal genes [64,65]. Detailed analysis of DE and DAS gene-involved pathways, with their possible impact on retinal dystrophies etiopathogenesis, is reported in Table 2.

**Table 2.** Detailed analysis of DE and DAS gene-involved pathways, with their possible impact on retinal dystrophies etiopathogenesis. DE and DAS genes were dysregulated during the whole analysis and showed several fluctuations during observed time points, suggesting that changes in gene-level expression and alternative splicing occurred throughout the whole period, either transiently (occurring after 3 h and returning to initial level after 6 h), occurring later (only after 6 h from treatment) or enduring throughout the whole period.



**Table 2.** *Cont.*

### **5. Conclusions**

We realized a whole RNA-seq experiment on RPE cells treated with A2E, considering two time points (3 h and 6 h) after the basal one. We found 10 different clusters of pathways involving DE and DAS genes, with many highlighted sub-pathways, which could depict a more detailed scenario determined by induced oxidative stress. Regulation and/or alterations of angiogenesis, extracellular matrix integrity, isoprenoid-mediated reactions, physiological or pathological autophagy, cell death induction and retinal cell rescue represented the most dysregulated pathway, probably involved in retinal degeneration. Assembly of splicing and transcriptional networks from analyzed data will further define the contribution of AS, as an extra level of regulation, and the interplay and coordination of the transcriptional and AS responses. However, it is fundamental to highlight several limitations of our study: RPE-cultured cells were not in contact with photoreceptors' outer segments and did not perform any phagocytosis. Moreover, the short-term response (at 3 h and 6 h) detected in vitro do not surely reflect what happens in vivo. Finally, even if it was important to underline the importance to realize an in vivo experiment to confirm what observed in RPE cells, our results could represent an important step towards discovery of unclear molecular mechanisms involved in etiopathogenesis of retinal dystrophies.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-3921/9/4/307/s1, Figure S1: Time-series trend analysis, Figure S2: Gene expression data statistics, Figure S3: qRT-PCR validation of ten most differentially expressed genes, Table S1: Gene expressed in RPE cells transcriptome analysis, Table S2: Gene transcripts expressed in RPE cells transcriptome analysis, Table S3: DE genes profile of RPE Cells, Table S4: DAS genes profile of RPE Cells, Table S5: DTU transcripts profile of RPE Cells, Table S6: Significant DE vs DAS gene enrichment by DAVID, Table S7: Significant DE vs DAS gene enrichment by ClueGO, Table S8: ToppGene prioritization detailed results, Table S9: Expression profiles of 10 among most DE genes.

**Author Contributions:** Conceptualization, L.D.; methodology, L.D.; software, L.D. and C.S.; validation, C.S. and S.A.; formal analysis, L.D.; investigation, L.D.; resources, C.S.; data curation, L.D. and C.S.; writing—original draft preparation, L.D.; writing—review and editing, C.S., R.D. and C.R.; visualization, C.R. and R.D.; supervision, A.S.; project administration, A.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
