The eye has always been considered a window to the soul. Today this saying is more relevant than ever, given the growing evidence of retinal manifestations in a range of diseases typical of the central nervous system (CNS). Indeed, being an integral part of the CNS, the retina shares many mechanisms with the brain and often mirrors CNS disease processes. Indeed, the brain, spinal cord, and retina share a common embryological origin and appear to be affected by the same neurodegenerative processes (Kavcic et al., 2011; London et al., 2013; Chang et al., 2014; Lee et al., 2014; Javaid et al., 2016; Mancino et al., 2019).

Visual abnormalities and signs of retinal pathology have been reported in patients with Alzheimer’s disease (AD), Parkinson’s disease (PD), and Huntington’s disease, multiple sclerosis and amyotrophic lateral sclerosis (ALS; Gupta et al., 2021). In addition, the eye offers a unique opportunity for direct observation of nerve tissue due to its transparent optical media. For these reasons, the retina is now called the window to the brain (Gupta et al., 2021). This has prompted interest in using retinal imaging to diagnose and monitor neurodegenerative diseases in vivo with noninvasive techniques, bringing with it the need to identify biomarkers to diagnose and monitor neurological conditions. However, to rationally implement retinal imaging biomarkers in clinical practice, it is necessary to understand the cellular and molecular processes occurring in the retina and their relationship to brain pathology. In recent years, progress has been made in understanding neuropathological changes in the retina of AD and PD patients. Alzheimer’s disease patients and animal models show retinal alterations, such as thinning of the ganglion cell layer, the presence of protein aggregates, glial activation, and alterations in the vasculature (den Haan et al., 2017; Koronyo et al., 2017; Grimaldi et al., 2018, 2019); in Parkinson’s disease, in addition to motor symptoms, visual function is also impaired, and postmortem evidence reports reduced retinal dopamine content and accumulation of α-synuclein (Altintaş et al., 2008).

In this context, even for ALS, traditionally considered a disease of the pure motor system, there is a growing body of evidence supporting the involvement of other non-motor systems, including the eyes (Nepal et al., 2022). Indeed, the traditional view that the visual system is spared in ALS (Ringelstein et al., 2014) has recently been challenged: atrophy of retinal layers (Moss et al., 2012; Fawzi et al., 2014; Rohani et al., 2018) has been reported, and retinal thinning may reflect an underlying neurodegenerative process (Volpe et al., 2015).

Amyotrophic lateral sclerosis, the most common form of motor neuron disease, is a progressive and ultimately fatal condition, with an average life expectancy of 2–3 years after diagnosis (Hübers et al., 2016).

Despite increasing knowledge, so far the diagnosis of ALS is still made by exclusion and often delayed, as there is no specific test that can confirm the disease (Marin et al., 2018). Several biomarkers have recently been identified, but none has already been implemented in clinical practice (Bakkar et al., 2015). Thus, there is an urgent unmet need to find effective tools to aid in the accurate diagnosis, stratification, and monitoring of disease progression in ALS patients.

While only a small percentage of ALS cases have a familial origin (fALS), attributable to selective mutations in causative and heritable genetic (SOD1, TARDBP, C9orf72, etc.), the vast majority of cases are believed to have a sporadic origin (sALS), in which genes and environmental factors interact. However, a number of cellular pathways/functions have been associated with the pathogenesis of sALS and fALS, such as RNA processing, dysfunctional protein homeostasis, endoplasmic reticulum stress, dysfunctional mitochondria and/or energy metabolism, increased oxidative stress, altered axonal transport, and dysfunction of RNA-binding proteins (Zhao et al., 2022).

In particular, the presence of protein aggregate inclusions in spinal cord tissue is a hallmark of ALS, present in 97% of familial and sporadic cases. These inclusions may involve different components such as neurofilaments, ubiquitin, RNA-binding proteins such as TDP-43 (coded by TARDBP), and the antioxidant enzyme superoxide dismutase 1 (SOD1). In most cases, these inclusions are positive for ubiquitin/p62, ubiquilin-2, and p-TDP-43 (Arai et al., 2006; Neumann et al., 2006). In familial and sporadic cases, TDP-43 observed in autopsy tissue is found in cytoplasmic aggregates and is aberrantly phosphorylated, ubiquitinated, and cleaved. TDP-43 is normally a primarily nuclear protein, with roles in mRNA translation, splicing and retention of cryptic exons, nuclear export, and axon growth. Cytoplasmic mislocation of TDP-43 is considered to be an important pathogenic mechanism in ALS, likely affecting several biological processes that reduce neuronal viability. TDP-43-positive protein aggregates in ALS often also contain the p62 protein. The consistent presence of p62 in pathological inclusions and the identification of Sequestosome 1 (SQSTM1)/p62 mutations as a rare cause of ALS suggest a role in pathogenesis (Falchetti et al., 2009 and reviewed in Rea et al., 2013).

Retinal involvement in ALS has only been demonstrated in familial forms using SOD1 and C9orf72 mice and in a patient with ALS and expansion of the GGGGCC hexanucleotide repeat in the C9orf72 gene (C9-ALS). Notably, SOD1 mutant mice show retinal cell degeneration and protein aggregates have been reported to parallel spinal cord degeneration (Hashizume et al., 2008; Ringer et al., 2017; Soldatov et al., 2021). In addition, C9orf72 mutant mice show retinal inclusions of p62 in the inner layer (Soldatov et al., 2021). In contrast, the mouse model of ALS FUS revealed no significant structural changes in the retina (Soldatov et al., 2021). Regarding human tissues, histopathological examination of the retina, optic nerve, and CNS of a C9orf72 ALS patient demonstrated p62-positive and pTDP-43-negative perinuclear inclusions in the inner nuclear layer of the retina and CNS. These cytoplasmic inclusions of ALS-related misfolded proteins suggest that the inner layer is the site of ALS-associated neurotoxicity (Rojas et al., 2020). However, data on sporadic ALS and other human tissues are not available.

To better understand and identify new molecular targets for a more comprehensive panel of putative biomarkers for ALS diagnosis, we performed immunostaining on human retinas obtained from sporadic ALS patients and age-matched controls. We report that the retinas of ALS patients show, in the granule cell layer, p62 inclusions, cell mislocalisation of TDP-43, and neuronal apoptosis, together with the increase in microglia density. These results highlight that retinal protein aggregates and cellular markers are targets to be considered for ALS diagnosis.

We report for the first time, the presence of cytoplasmic inclusions of TDP-43 and p62 in retinal cells along with degeneration of neurons in the retina of sporadic ALS patients.

Recently, many studies have reported decreased retinal nerve fiber layer (RNFL) thickness associated with various neurological and neurodegenerative diseases. This has been observed in AD, minimal cognitive impairment (Ono et al., 1988; Howland, 1990; Philippou and Joubert, 2013; Bozzoni et al., 2016), multiple sclerosis (Pugdahl et al., 2007; Sharma et al., 2011), Parkinson’s disease (Moss et al., 2012), diffuse Lewy body disease (DLB; Cerveró et al., 2019), and Wilson’s disease (Yap et al., 2019). Furthermore, in AD, the retina shows the presence of amyloid beta protein and tau aggregates, neurodegeneration, glial reactivity, and altered vascularity (Gupta et al., 2021). Therefore, the possibility of using the eye as a window to the CNS has prompted further research into the presence of protein aggregates, typical of other neurodegenerative disorders, in the retina of patients and animal disease models. In particular, for ALS cases, typical motor neuron inclusions are positive for ubiquitin/p62, ubiquilin-2, SOD1 (Soldatov et al., 2021), TDP-43, and pTDP-43 (Arai et al., 2006; Neumann et al., 2006).

However, retinal involvement of ALS-related protein aggregates has only been reported in one fALS patient with the C9orf72 mutation and in two fALS mouse models to date. Specifically, SOD1 inclusions have been found in hSOD1G93A mice (Soldatov et al., 2021), in which retinal cell degeneration and protein aggregates have been reported to parallel spinal cord degeneration (Hashizume et al., 2008). In addition, retinal inclusions of p62 were observed in the inner layer of C9orf72 mouse retina. In contrast, the ALS FUS mouse model showed no significant structural retinal changes (Soldatov et al., 2021). We observed, for the first time, in the retinas of sALS patients, the presence of TDP-43 and p62 co-localized inclusions at the level of ganglion cells and inner nuclear layers. Notably, while the nuclear expression of TDP-43 was similar in ALS patients compared with age-matched controls, its cytoplasmic localization was present only in ALS patients.

This is consistent with the idea that the inner layer may be the site of ALS-associated neurotoxicity (Rojas et al., 2020). Given that histopathological analysis of the retina from a C9-ALS patient showed p62-positive intracytoplasmic perinuclear inclusions in the IL similar to those observed in the dentate gyrus of ALS patients with the C9orf72 mutation. Notably, in line with what reported in motoneurons (Arai et al., 2003; Nakano et al., 2004; Mizuno et al., 2006; Tanji et al., 2012; Foster et al., 2021), we observed a strong co-localization of p62 and TDP-43 staining in RGC. Indeed, cytoplasmic aggregates of TDP-43-positive proteins also contain the p62 protein. We can hypothesize that, similar to what occurs in motoneurons, these inclusions may comprise different protein components, such as neurofilaments and RNA (Blokhuis et al., 2013).

We also reported the presence of retinal ganglion cell degeneration in sALS patients, evidenced by cleaved Caspase-3 positivity of RGC neurons (TuJ1-positive cells). This is in line with the RNFL thinning in ALS patients recently demonstrated by optical coherence tomography (Rohani et al., 2018).

We also, evaluated the glial reactivity in retinal slices as, among the multifactorial mechanisms underlying ALS-associated neurodegeneration, inflammation is one of the most prominent hallmarks of sporadic and familial forms of ALS (Liu and Wang, 2017), with overall reactivity of astrocytes and microglia and overproduction of inflammatory cytokines (Peric et al., 2017; Clarke and Patani, 2020; Vahsen et al., 2021; Van Harten et al., 2021).

It should be noted that we found no signs of astrocyte reactivity in the retina of ALS patients contrary to expectations. On the other hand, counting microglia cells, we found increased microglia density in ALS retina.

It is possible, however, that the glial phenotype is underestimated in this context. Indeed, the use of coronal retinal slices prevented a morphological analysis of astrocytes and microglia because of the extremely thin thickness of the slices. In addition, inclusion in paraffin prevented a Real Time PCR analysis.

In conclusion, our data support the idea that although ALS has traditionally been considered a disorder purely of the motor system, several non-motor systems, including the eyes, are involved and affected by the disease. Furthermore, the presence of protein aggregate inclusions in the retina could become a new fingerprint panel to characterize ALS, supporting the possibility of using ocular biomarkers for early diagnosis of ALS-associated neurodegeneration.

Indeed, to date, there is a lack of reliable and noninvasive clinical assessment to monitor ALS patients and document response to therapy in ALS clinical trials. Measurement of retinal changes could become an inexpensive and noninvasive method to assess neurodegeneration and, given its correlation with disease severity, can be used to assess disease progression in ALS patients (Soldatov et al., 2021).

However, this will require the development of specific ligands for ALS biomarkers and remote high-resolution imaging techniques to achieve noninvasive and inexpensive diagnosis and monitoring of ALS by scanning the retina.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The use of Human tissue has been approved by the Ethics Committee of Fondazione Santa Lucia I.R.C.C.S. to GR. Written informed consent was not provided because we used tissues obtained from a tissue bank.

NP, YG, SC, and MP designed, carried out, and analyzed the immunofluorescence experiments on protein aggregates and neurodegeneration. LM designed, carried out, and analyzed the immunofluorescence experiments on astrocytes and microglia. NP and YG designed, carried out, and analyzed the confocal acquisitions. SG designed and wrote the Matlab-based code. SDA wrote the manuscript with the help of GR and SG and conceived, supervised, and administered the project. SDA and GR acquired funds for the project. All authors contributed to the article and approved the submitted version.

This research was funded by the REGIONE LAZIO (19036AP000000019 and A0112E0073; to SDA); Fulbright (FSP-P005556; to SDA); Rome Technopole PNRR, FP7 (to SDA); ERC-2019-Synergy Grant (ASTRA, n. 855923; to GR); EIC-2022-PathfinderOpen (ivBM-4PAP, n. 101098989; to GR). This research was partially funded by grants from Sapienza University (to SDA).

YG and MP are employed by D-Tails s.r.l. SDA is a scientific advisor of D-Tails s.r.l.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.