Spaceflight Effects and Molecular Responses in the Mouse Eye: Preliminary Observations After Shuttle Mission STS-133

The space environment creates challenges for extended human spaceflight and presents a unique combination of stressors: microgravity, high-energy-particle radiation, nutritional deficiencies, hypobaric hypoxia, intermittent hyperoxia, and psychological stress. Lack of gravity implies reduced physical loading, fluid shift, and incompletely understood cellular responses that are reflected by a number of detrimental changes, such as muscle atrophy and loss of bone mass, immunosuppression, and overall gene expression changes (Pietsch et al., 2011; Sundaresan and Pellis, 2009). Ground models of simulated microgravity, namely hindlimb suspension (HS) and bed rest, induce a fluid shift and concomitant vascular pressure and flow alterations (Hargens and Watenpaugh, 1996; Wilkerson et al., 2002), affecting not only cardiovascular physiology but also inducing genome-wide gene expression changes in the central nervous system (Frigeri et al., 2008).

Ocular changes have been reported related to exposure to the space environment. In humans, the direct effect of radiation in the lens results in cataract formation (Cucinotta et al., 2001), which manifests with a higher incidence and earlier onset in the astronaut population. Light flashes in the eye are an occurrence that has been observed by astronauts since the Apollo program (Sannita et al., 2006) -- a phenomenon not completely understood.

Most importantly, recent medical data from astronaut cohorts have reported the development of optic disc edema, choroidal folds, posterior globe flattening, and a resulting hyperopic shift (Kramer et al., 2012; Mader et al., 2011) in a fraction of the astronaut population upon return from missions longer than 30 days (NASA, 2010). No clear etiology has been established for these cases, but it is hypothesized that microgravity, the ensuing cephalad fluid shift, and venous congestion may play a role. The perturbations observed in some individuals of the astronaut cohort resemble those found in papilledema associated with idiopathic intracranial hypertension (IIH) also known as pseudotumor cerebri (Friedman, 2007; Kramer et al., 2012; Mader et al., 2011). Because the etiology is still a matter of speculation, investigating whether exposure to microgravity represents a source of stress for the eye is an issue of critical occupational health importance. To this aim, this project examines the effects of spaceflight on the rodent eye and the responses that occur when challenged with exposure to microgravity in combination with other stressors during spaceflight.

Previous spaceflight studies performed on rodents found evidence of retinal degeneration in neonatal rats aboard shuttle mission STS-72 (Tombran-Tink and Barnstable, 2006), and of cell swelling and disruption in rats aboard two experiments on Russian Cosmos satellites (Philpott et al., 1980; Philpott et al., 1978). However, these studies were limited to structural histopathologic observations of the eye. In the present work, we expand the immunohistopathologic analysis to investigate the effects of spaceflight and the elicited responses observed in the eyes of mice aboard shuttle mission STS-133, focusing, for the first time, on molecular and cellular processes subjacent to the histopathologic changes.

While the spaceflight results reported herein represent pilot data due to the small sample size, these data offer, for the first time, direct evidence suggesting that oxidative stress, neuronal damage, and mechanical injury take place in the retina, lens, and optic nerve of rodents flown in low-Earth orbit for a period under two weeks. Several previous studies have shown the occurrence of oxidative stress during spaceflight (Stein, 2002), however, our work gives a first insight into the impact of space-associated factors on biological processes like cell death, oxidative stress, and probable mechanical injury in the rodent eye.

Because the BALB mouse strain used in the STS-133 experiment is susceptible to light-induced retinal degeneration (LaVail et al., 1987), we speculate that this particular strain exhibits an enhanced sensitivity to oxidative stress and/or a reduced stress response, making it a suitable strain in which to identify alerting evidence of risks previously unrecognized in the retinal tissue, while impacting its value as a model for the study of the human changes seen in-flight.

8OHdG, a product of deoxyguanosine oxidation, is a marker of oxidative stress-induced DNA damage. This damage has been observed in mouse cornea exposed to dryness (Nakamura et al., 2007), ultraviolet radiation (Tanito et al., 2003), and in mouse retina exposed to intense light (Tanito et al., 2002; Wiegand et al., 1983). In our study, 8OHdG was present in all acanthotic areas of the cornea. Irregular acanthosis with visible edema was only seen in the VIV samples at R+7, and it was only in this group where positivity at the corneal endothelium was observed since day 1, suggesting an impaired ion and water transport in the cornea.

The retinal response to intense light in susceptible mice has been studied before and has been found to be related to lipid peroxidation at the ONL (Tanito et al., 2002; Wiegand et al., 1983). Likewise, radiation-induced retinopathy is an ocular complication in cancer patients that receive radiation therapy (Parsons et al., 1996). The processes involved in the damage by high-energy-particle radiation in these cases may share commonalities (direct DNA damage and oxidative stress) with exposure to radiation present during spaceflight. The present work shows evidence of both oxidative stress-induced DNA damage in the neuronal layers of flight mice retinas and of an oxidative stress response induced at the gene expression level in these mice. Short-term responsiveness to DNA oxidation followed by DNA repair has been studied longitudinally in blood of trauma patients (Oldham et al., 2002), suggesting that the attenuated DNA damage observed after one week of return from flight may be the result of DNA repair.

Of note, the ground controls kept in the vivarium exhibited a comparable level of retinal oxidative stress to the samples from flight, especially at longer exposures (day R+7). This is likely due to the fact that the illumination conditions in a standard vivarium room are approximately 15-fold in light flux compared to the illumination of an AEM, even if both maintain a 12 hour light-12 hour dark cycle.

Caspase-3 is a pro-enzyme that is activated in the intrinsic apoptotic pathway in all mammals (D’Amelio et al., 2010). In this study, all mice showed positivity for caspase-3 at the level of the cornea. This may be explained by the fact that caspase-3 immunoreactivity in the stratified epithelium of the cornea serves as an internal positive control due to the natural differentiation process that the basal cells suffer towards cornification. Apoptosis can be triggered by oxidative stress, brain trauma, or ischemia. In a model of brain ischemia, the area of neuronal apoptosis has been identified not in the infarct region but in the surrounding area, where the oxygen tension is decreased, but not absent (Pulsinelli et al., 1982). The presence of activated caspase-3 is thus related to hypoxic environment and radiation exposure. In our study, the FLT group at R+1 showed higher positivity compared to the rest of the groups. This may be related to radiation and microgravity exposure during spaceflight. It is important to point out that the effect of high-energy-particle radiation may be overall increased in this susceptible mouse strain.

Qualitative examination revealed that VIV and FLT groups showed more caspase-3-positive cells at the retinal layers than AEM retinas. This may suggest that the damage caused by visible light radiation in the albino strain in the vivarium conditions may be comparable to the damage caused by the exposure to spaceflight environmental factors. We also observed positive microglial (astrocytes) but not Muller cell activation in VIV specimens, which may support the notion of visible light radiation effects as the triggering factor in inner layers of the retina only in these mice (Song et al., 2012).

Both mice in the FLT group at R+1 and one mouse at R+5 showed evidence of apoptosis in the RPE. Apotosis in the RPE has been identified in ocular pathologies like age-related macular degeneration (AMD) secondary to exposure to activated monocytes (Yang et al., 2011), or triggered by oxidative stress with H₂O₂, lipofuscin, or light irradiation (Sparrow et al., 2000). This data also suggest oxidative stress may be an important component in the retinal damage in these mice. Of note, in vitro experiments with human RPE cells cultured in simulated microgravity generated by a NASA-bioreactor resulted in DNA damage and inflammatory response in these cells (Roberts et al., 2006). Retinal pigment epithelium attenuation has been related to retinal choroidal folds previously found in astronauts (Mader et al., 2011). It is yet to be determined whether or not increased RPE apoptosis may contribute to the formation of choroidal folds or if it increases the risk for AMD in astronauts.

Several advances in immunohistochemistry have led to the identification of β-amyloid in traumatic brain injury in humans (Iwata et al., 2002), rats, and pigs (Smith et al., 1999), by tracing not only the full-length protein but also small aminoacid peptides. β-amyloid was present in areas of the brain as soon as one day after brain trauma was provoked by pressure injection of saline into the cranium in a rat model (Pierce et al., 1996). Moreover, β-amyloid deposits showed evidence of optic nerve injury in cases of shaken-baby syndrome (Gleckman et al., 2000). Previous studies in animal models have shown distribution of β-amyloid in the mouse retina that suggests its involvement in the pathophysiology of glaucoma (Kipfer-Kauer et al., 2010). We report that β-amyloid deposition was present in the neural retina of mice in all treatment groups and that the VIV mice showed a slightly higher positivity in both RGC and INL compared to the rest of the mice. Interestingly, β-amyloid was present in the optic nerve of both mice in the FLT group at R+7 and had the unique characteristic of being at the level of lamina cribrosa or immediately distal to it. This compares with the findings in traumatic injury in children of shaken-baby syndrome where most of the axonal changes are seen in the postlaminar region (Gleckman et al., 2000). This may be associated to the anatomy of this region where the nerve is anchored by the fibers of the lamina cribrosa but immediately posterior to this or beyond this area the nerve can move freely. Thus, in the event of mechanical trauma the immediate fibers in the postlaminar region may be the ones demonstrating more damage. The trauma may include increased intracranial pressure that is transmitted into the nerve, positional or whiplash (similar, although in a less intense manner to what happens in shaken baby syndrome), or vibration (as the one occurring during launch or landing). However, there is the need to further investigate the nature of the changes through additional experimental work.

GFAP is an intermediate filament protein known to be present in astrocytes, Muller cells, and oligodendrocytes in the post-laminar optic nerve. GFAP is elevated when there is stress in the central nervous system and has been shown in the injured retina mostly present in the activated Muller cells (Lewis and Fisher, 2003). In this paper, we show that the optic nerves of several mice were positive for GFAP and β-amyloid; however, it was only the FLT group at R+7 that showed increased expression of GFAP at the postlaminar optic nerve. These findings suggest that the astrocytes and oligodendrocytes were activated in this region probable secondary to mechanical trauma. The causes of this, either vibration or fluid shift-related, need to be further investigated.

In addition, only FLT mice sacrificed at day 1 (FLT R+1) were immunoreactive in the neuronal layer for all β-amyloid, GFAP, caspase-3, and 8OHdG, suggesting increased oxidative and possibly mechanical damage. This may be explained by the possible correlation of β-amyloid deposition and activation of astrocytic cells, both triggering reactive oxygen species production (Lamoke et al., 2012).

The gene expression profiling results with BALB mice in flight STS-133 support the immunohistopathologic findings and suggest that: a) Oxidative stress-induced DNA damage was higher in the FLT samples compared to controls on R+1, and decreased on R+7. A trend toward higher oxidative and cellular stress response gene expression was also observed on R+1 compared to AEM controls, and these levels decreased on R+7. Several genes coding for key antioxidant enzymes, namely, heme-oxygenase-1, peroxiredoxin, and catalase, were among those elevated after flight. Likewise, the inflammatory response genes Nfkb1and Tgfb1 were elevated after flight. The fact that only two mice flown on STS-133 were genetically analyzed per day of sacrifice creates a major limitation in any statistical analysis. However, this does not preclude the comparisons of samples. b) There is an apparent correlation trend in the stress parameters measured in the different animals and there is certain variability in the stress response among the individual animals. For example, mouse # 13 in the FLT group at R+1 suffered from overall elevated stress, demonstrated by the highest 8OHdG levels, induction of antioxidant enzymes, induction of Nfkb1, and concomitant lower levels of the cytoprotective heat shock protein Hsp1a1. Sirtuin 1 gene expression results were non-conclusive, but further analysis is required to determine if translocation of sirtuin 1 may occur and how this may affect the expression of downstream cellular stress response genes (Jaliffa et al., 2009; Ozawa et al., 2010). c) Spaceflight represents a source of environmental stress that translates into oxidative and cellular stress in the retina, which is partially reversible upon return to Earth. Also, retinas from VIV control mice evidenced higher oxidative stress markers, Nfkb1 and Tgfb1, likely due to the more intense illumination in vivarium cages versus the AEM.

In addition, mice in FLT group at R+7 were positive for both β-amyloid and GFAP, and it was only in these mice that there was increase in GFAP staining adjacent to lamina cribrosa in the optic nerve. We suspect some long term damage in the optic nerve may be seen after spaceflight because this did not resolve after seven days on Earth. Additional quantitative experiments are needed to give a better understanding on this finding.

These preliminary data suggest that spaceflight represents a source of environmental stress that directly translates into oxidative and cellular stress in the retina, which is partially reversible upon return to Earth. Moreover, the optic nerve findings suggest that the lesion may be mechanical in nature and that does not resolve after return to Earth, at least in the animals studied. Further work is needed to dissect the contribution of the various spaceflight factors (microgravity, radiation) and to evaluate the impact of the stress response on retinal and optic nerve health. These preliminary results should inform investigators on the design of future studies utilizing a more suitable mouse strain devoid of photic degeneration predisposition, male animals that better reflect the astronaut population, and statistically powered larger sample sizes.