Recent advances in our understanding of the pathogenesis of DR have led to the design of new experimental treatments which address early-stage disease. Although the sequence of some pathological events during early DR is not entirely understood, it is established that neurodegeneration plays an important role in a large percentage of patients [30]. Therefore, it could be hypothesized that neuroprotective therapies might be effective, and there is evidence supporting this hypothesis [31,32,33]. Within current experimental approaches aiming to achieve neuroprotection, therapeutic interventions involving GLP-1R activation and, more recently, DPP-4 inhibitors, have demonstrated promising experimental results to address this unmet medical need [5, 6, 9,10,11,12, 34].

In the present study, we tested the effect of topical administration of sitagliptin (a DPP-4i) in 3-month-old Trpv2+/− rats, a new non-diabetic model that replicates the NVU lesions observed in early stages of DR due to impaired retinal blood flow autoregulation [19]. We found that sitagliptin eye drops prevented all the pathological hallmarks shared with DR in Trpv2+/− rats, including retinal thinning, glial activation, inflammation, oxidative stress, and abnormal and sustained vasodilatation. Notably, these beneficial effects were observed without any impact on systemic parameters such as body weight or glycemia, and they cannot be attributed to a possible restoration of the levels of TRPV2 channels mediated by sitagliptin.

We found that Trpv2+/− rats exhibit retinal thinning similar to that seen in DR [34], a progressive event resulting from neurodegenerative mechanisms that affect the retina [35, 36]. Topical administration of sitagliptin prevented this pathological thinning after two weeks of treatment. Since emerging evidence suggests that retinal thinning precedes the onset of DR [37,38,39,40], this result reinforces previous studies indicating that neuroprotective agents can prevent this early hallmark of the disease [41]. This beneficial effect on Trpv2+/− rats could be explained by the neuroprotective properties that we have previously reported in the db/db mouse model, where sitagliptin demonstrated efficacy against oxidative stress [12], inflammation and synaptic damage [10,11,12], thus contributing to restore physiological rates of neurogenesis [9, 12].

Glial activation plays a central role in the NVU dysfunction that occurs in DR, linking the neurodegenerative process with microvascular impairment. In the present study, as previously demonstrated in db/db mice [6], we found that sitagliptin eye drops prevented glial activation in Trpv2+/− rats. This was evidenced by the abrogation of aberrant GFAP overexpression in this non-diabetic model. Consequently, the drug also protected against the downregulation of the Kir4.1 channel and the upregulation of VEGF. In addition, topical sitagliptin prevented the abnormal accumulation of microglial cells, mainly in the IPL layer of the retina, in Trpv2+/− rats. Glial cell activation is crucial in the inflammatory process that occurs in diabetic retina, and we report that sitagliptin prevented the upregulation of several pro-inflammatory cytokines, including IL-6, IL-13, IL-18, ICAM-1, CXCL7 and TNF-α. These results are consistent with our previous study using db/db mice [12] and corroborated by in vitro studies [34, 42].

Both glial activation and inflammation play significant roles in the initial oxidative environment observed in the diabetic retina. This is mainly due to toxic products generated by the saturation of several physiological pathways related to glucose metabolism, including the polyol pathway, the hexosamine biosynthesis pathway, the formation of advanced glycation and lipoxidation end products, and protein kinase C (PKC) activation [1]. Histological sections from the retinas of vehicle-treated Trpv2+/− rats showed higher levels of 8-hydroxyguanosine, indicating greater oxidative damage to DNA compared to wild-type control rats. Additionally, we observed for the first time that the retinas of this new model exhibited lower gene expression of key antioxidant enzymes like Sod1, Sod2, or Gsr, all essential for maintaining the physiological redox state. Furthermore, mRNA levels of the antioxidant transcription factor Nrf2 were also increased, possibly as a defense mechanism to restore physiological levels of these enzymes, among others not studied [43, 44]. Topical administration of sitagliptin inhibited oxidative DNA damage and prevented the downregulation of Sod1 and Sod2 mRNA levels. This corroborates our findings in the db/db mouse model, with the exception that sitagliptin did not prevent the downregulation of Gsr mRNA levels in this case.

Regarding the vascular component of the NVU, we did not observe any visible alterations through fundus imaging among the different experimental groups. The presence of signs of neurodegeneration, such as retinal thinning and glial activation, despite the absence of early visible retinal damage (e.g., microaneurysms, hemorrhages, exudates, or cotton wool spots), confirms that by 3 months of age, the Trpv2+/− rat model successfully recapitulates the early stages of DR observed in most patients. Using ARIA v1.0 software, we detected abnormal vasodilation of the main arterioles and venules of the retina in Trpv2+/− rats, as occurs in DR. This anomaly was prevented with sitagliptin eye drops, with treated heterozygous animals exhibiting vessel diameters similar to those of the wild-type group. Although sitagliptin has been associated with both vasodilatory and vasoconstrictive effects [45, 46], to the best of our knowledge, there is no information on its effect in the retinal microcirculation. Our results suggest sitagliptin exerts a protective effect on the retinal NVU and prevents the abnormal vasodilation that occurs in Trpv2+/− rats. This effect is significant when considering the pathophysiology of DR, as one consequence of diabetes induced NVU impairment is the lack of adaptation to metabolic demands, which could lead to capillary dropout. In fact, we have observed a significant loss of endothelial cells in the capillaries of Trpv2+/− rats, which was prevented by topical administration of sitagliptin. This finding indicates that despite having a more direct effect on neurons, the drug is also capable of protecting the vascular elements of the NVU, thus confirming its dual action: neurotrophic and vasculotrophic.

The precise mechanisms by which Trpv2+/− rats manifest the main features of early stages of DR are not fully understood. However, it should be noted that impaired regulation of blood flow precedes clinically observed lesions in the diabetic retina [37]. Therefore, the Trpv2+/− model replicates the disease from one of its earliest pathological events: the loss of the myogenic response and failed autoregulation within the NVU. Oxidative stress, a major trigger of DR, is increased in this model despite the absence of hyperglycemia. A possible explanation is the initial impairment of neurovascular coupling due to a disrupted myogenic response, reducing the vasculature’s capacity to meet the metabolic demands of the retina, and altering physiological rates of oxygen delivery, consumption and ROS production [47]. Increased oxidative stress could trigger glial activation and inflammation, both of which also produce ROS, further worsening the process [48, 49]. Interestingly, we found that the retinal expression of glutamine synthetase (Glul gene) is not diminished in Trpv2+/− rats. Glutamine synthetase, a key enzyme involved in glial-neuronal transmitter recycling and exclusively expressed by Müller cells within the retina, converts neurotransmitters substances [glutamate and gamma-aminobutyric acid (GABA)] into glutamine to maintain neurotransmitter balance and avoid glutamate excitotoxicity, a major pathogenic pathway of neuronal cell loss in DR [50]. This finding suggests that hyperglycaemia and/or its downstream metabolic pathways are crucial in the downregulation of glutamine synthetase observed in DR and that vascular hemodynamic changes secondary to the abnormal autoregulatory response do not affect this enzyme, at least in the early stages of DR.

Overall, the similarity of retinal findings observed Trpv2+/− rats to those observed in diabetic retinas indicates that the impairment of the autoregulation in the NVU could trigger the neurodegenerative processes present in DR. However, whether Trpv2 is expressed in neurons and glial cells and its potential pathophysiological repercussions remains to be examined. Nevertheless, our results suggest that the Trpv2+/− rat model, at 3 months of age, represents a good model for studying early stages of DR and for testing experimental preventive drugs against these stages of the disease. A graphical summary of the results obtained with sitagliptin eye drops is depicted in Fig. 9.

Primary effects of sitagliptin eye drops in the Trpv2+/− model after two-weeks of twice-daily administration. Topical application of sitagliptin resulted in a reduction of neurodegeneration (preservation of total and inner retinal thickness), glial activation (reduction in GFAP-positive fibers and microglial cells), inflammation (decreased pro-inflammatory cytokines), oxidative stress (diminished DNA/RNA damage and maintenance of antioxidant elements), vasodegeneration (protection against the formation of acellular capillaries) and abnormal vasodilation. These effects were independent of changes in body weight, glycemia, and HbA1c levels

An important aspect of the results obtained is the potential translational value of the beneficial effects of sitagliptin eye drops for other diseases where NVU damage plays a key role in their pathophysiology. These diseases include glaucoma, retinitis pigmentosa, age-related macular degeneration, tauopathies, and even retinal damage associated with neurodegenerative diseases of the central nervous system, such as Alzheimer’s disease and Parkinson’s disease [51,52,53,54].

In conclusion, the beneficial effects of topical administration of sitagliptin, previously demonstrated in the db/db mouse model, are applicable to the non-diabetic Trpv2+/− model, where DR-like lesions are not mediated by hyperglycemia. These findings suggest the potential utility of sitagliptin eye drops for treating early DR in humans and present an opportunity to explore its applicability in addressing other retinal diseases, particularly those characterized by NVU dysfunction and neurodegeneration.