**1. Introduction**

Diabetes mellitus (DM) epidemic is a global public health problem and the leading cause of preventable blindness in the working-age population [1]. The International Diabetes Federation, in 2019,

stated that DM a ffected an estimated 425 million adults worldwide and this number is likely to increase over the next few years due to urbanization, increased obesity prevalence, and sedentary lifestyles. According to epidemiologic predictions 1 in 10 adults across the globe, will live with diabetes by 2045 [2]. The physiopathology of DR is complex and several interconnecting biochemical pathways have been proposed as contributors in its development. These include the polyol pathway [3], non-enzymatic glycation [4], activation of protein kinase C (PKC) [5], oxidative stress [6–10], and inflammation through proinflammatory cytokines, chemokines, and other inflammatory mediators [11,12]. Among these, oxidative stress and inflammation are major causal factors involved in the endothelial dysfunction of the retina microvasculature that occurs in DR [6–10]. In addition, this chronic low-grade inflammation could finally lead to neovascularization [13]. While improvements in treatment have reduced the macro and microvascular complications of the disease, the increasing number of diabetic patients combined with the extended life expectancy means that more patients will live long enough to develop DR. In fact, it is expected that the number of people with DR will grow from 126.6 million in 2010 to 191.0 million by 2030 [14].

Metabolic abnormalities of diabetes cause mitochondrial superoxide overproduction [15]. This is the central and major mediator of diabetes endothelial dysfunction and tissue damage, with several pathways involved in the pathogenesis: Polyol pathway, increased formation of advanced glycation end-products (AGEs), increased expression of the receptor for AGEs and its activating ligands, activation of protein kinase C (PKC) isoforms, and overactivity of the hexosamine pathway [15].

According to the above, long-chain n-3 polyunsaturated fatty acids (n3-PUFA),including eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) have been studied as an alternative therapy for retinal diseases due to their pleiotropic e ffects including anti-inflammatory, antioxidant, antiproliferative, and antiangiogenic properties. They are essential fatty acids in the human diet that exert anti-inflammatory and antioxidant e ffects by binding cell membrane receptors to a ffect downstream mediators and alter gene expression [16]. DHA is a major structural lipid in the sensory and vascular retina. In fact, the highest body concentrations of DHA per unit weight are found in phospholipids of retinal photoreceptor outer segments. Retinal pigmented epithelium (RPE) is a polarized epithelial monolayer known to synthesize DHA [17]. Moreover, the RPE plays an important role in regulation and delivery of DHA from the plasma to the photoreceptors [18]. EPA is converted intracellularly to DHA at low basal levels, ye<sup>t</sup> exogenous EPA supplementation does not increase DHA levels in human plasma [19]. Rather than rapidly converting to DHA, EPA seems to have a clinically relevant biological activity itself, distinct from that of DHA [20–22].

Several clinical trials have demonstrated the beneficial e ffects of the administration of n3-PUFA in the development of DR. J. Howard-Williams et al. found that poorly controlled patients with low levels of n3-PUFA intake had a significantly greater frequency of retinopathy [23]. Similarly, a reduced severity of DR in well-controlled diabetes patients was observed with increasing n3-PUFA intake [24]. The primary prevention of cardiovascular disease with a Mediterranean diet (PREDIMED) demonstrated that participants taking at least 500 mg/day of long-chain v-3 PUFAs, showed a 48% relatively reduced risk of incident sight-threatening DR compared with those not fulfilling this recommendation [25]. Moreover, the addition of DHA supplement to intravitreal ranibizumab was effective to achieve better sustained improvement of central subfield macular thickness compared with ranibizumab alone [26].

In diabetic animal models, the ability of n3-PUFAs, especially EPA and DHA, to suppress IL-6, TNF-a, ICAM-1, MCP-1, and VEGF production has been demonstrated, as well as the reduction of free radical generation and the restoration of antioxidant homeostasis [27–30]. According to the conclusions of a study conducted by Mahmoudabadi and Rahbar [31], the administration of EPA increases several endogenous antioxidant enzymes, namely superoxide dismutase and glutathione peroxidase, while simultaneously decreasing the levels of malondialdehyde, a classical biomarker of oxidative stress, in type II diabetic patients.

Despite their similarities in their nutritional sources and most of their biological actions, the type of n3-PUFA formulation seems to matter in treating different diseases. DHA is associated with decreased Alzheimer disease risk in humans [32]. Conversely, EPA has a more therapeutic effect in treating depression [19–21] while cardiovascular outcomes have been shown to improve after the intake of combined EPA and DHA supplements [33]. Nevertheless, it should be noted that many of the observational studies published in DR only supplement one type of n3-PUFA (EPA or DHA), while some of them did not measure the type of omega 3 consumed, but rather frequency of fish oil consumption. In addition, the efficacy of the use of different formulations of DHA and EPA have not been studied to date. Currently, the primary dietary source of these fatty acids is fish oil; however, since the global consumption cannot be satisfied due to the increasing demand, alternative sources such as microalgae have emerged [34,35].

Given this lack of evidence, we designed an in vitro study to compare 10 formulations of DHA and EPA supplements from different origins, and assess their safety profile and their ability to rescue retinal pigment epithelium (RPE) cells from the oxidative and inflammatory conditions seen in the DR.

#### **2. Materials and Methods**

## *2.1. Cell Culture*

ARPE-19, obtained from the American Type Culture Collection (CRL-2302, ATCC®, Manassas, VA, USA), were grown to confluence in a standard incubator at 37 ◦C in humidified 5% CO2 condition in a DMEM/F12 medium (1:1) (Sigma-Aldrich, St. Louis, MO, USA) containing 10% fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO, USA), 1% fungizone, and L-glutamine penicillin-streptomycin (Sigma-Aldrich, St. Louis, MO, USA). Cells were passaged every 3–4 days with 0.25% Trypsin-EDTA (Invitrogen, Carlsbad, CA, USA). For all assays, cells were grown to 100% confluence with the exception of BrdU, which requires non-confluent cells grown for 1–2 days.
