NEUROPROTECTION AGAINST MITOCHONDRIAL DYSFUNCTION WITH NEAR INFRA-RED LIGHT

A. SPECIFIC AIMS

There is an overriding need to develop effective therapies and preventive interventions for neurodegenerative diseases. Mitochondrial dysfunction and the concomitant decrease in energy production and free radical damage are believed to play a central role in neurodegeneration. Near infra-red light (NIL) has been demonstrated to have a wide range of biological effects including enhancement of ATP production, induction of gene expression, and decrease in cell death.

The overarching hypothesis of the present proposal is that NIL therapy, by enhancing ATP production and protein expression, prevents neurodegeneration induced by mitochondrial inhibition in the rat retina in vivo. The long-term goal of this work is contributing to the development of a non-invasive strategy to prevent neuronal toxicity, and the concomitant biochemical, morphological and behavioral derangements induced by mitochondrial inhibition. The results could support further research aimed at developing effective and efficient therapies for neurodegenerative disorders.

I propose a series of experiments to study the effect of near-infra-red light (NIL) therapy as an in vivo neuroprotective intervention against rotenone-induce neurodegeration in the retina. These experiments will be divided in the following specific aims:

Specific Aim No. 1 is to characterize the neuroprotective effects of NIL therapy on the visual pathway of the rat using functional parameters. A battery of behavioral tests of visual acuity including 1) the visual cliff for depth perception, and 2) a visible platform modality of the water-maze with light intensity threshold discrimination will be used. These tests will be used to determine the efficacy of different energy doses of NIL therapy in preserving visual function after an intravitreal injection of rotenone, a specific inhibitor of mitochondrial function. 3) The functional impact of NIL therapy on retinal neuronal activity will be measured by estimating transynaptic cellular activity in the central visual pathway in brain sections and stained histochemically for cytochrome c oxidase activity. It is expected that subjects exposed to NIL therapy after rotenone intravitreal injections will improve their performances in the behavioral tests of visual function and will show increased levels of neuronal activity in the central visual pathway as compared to non-exposed subjects.

Specific Aim No. 2 is to determine the effects of NIL as a neuroprotective intervention by means of morphological parameters. 1) The retinas of rotenone/NIL radiation-treated subjects will be analyzed morphometrically to estimate ganglion cell layer (GCL) and nerve fiber layer thickness (RNFL), retinal volume, ganglion cell count, ganglion cell density and 2) apoptotic cell count using unbiased stereological tools. It is expected that NIL therapy will prevent rotenone-induced decrease in GCL + RNFL thickness and cell counts at 48 hrs, and reduce the number of apoptotic cells.

Specific Aim No. 3 is to characterize the mechanisms of action of NIL therapy. The in vivo effects of NIL therapy on 1) cellular oxygen consumption and 2) ATP production will be measured in brains of animals treated with NIL therapy by means of biochemical assays. 2) The impact of NIL therapy on expression of proteins related to neuronal damage and defense mechanisms will be determined. The expression of the phosphorylated form of the cAMP response element binding protein (p-CREB), superoxide dismutase (SOD), and the inducible neuronal nitric oxide synthase (nNOS), will be determined in rotenone-treated eyes, using an immunohistochemical approach. It is hypothesized that NIL therapy will increase oxygen consumption, ATP production, p-CREB and SOD expression and will decrease the levels of nNOS expression.


B. BACKGROUND AND SIGNIFICANCE

The specific approach selected for attaining the long-term goal of this proposal is based on the following key observations: 1) NIL therapy constitutes a novel and largely unexplored noninvasive intervention that is known to increase neuronal survival after cell damage (Eells JT, 2004). Such an intervention might represent a potential preventive strategy to avoid an otherwise irreversible process: neurodegeneration. 2) Using a model of neurodegeneration caused by mitochondrial dysfunction for testing NIL therapy is relevant because mitochondrial dysfunction is being recognized as a crucial event underlying neurodegeneration (Rego and Oliveira, 2003). Neuronal bioenergetics is directly related to many cellular functions and many cellular derangements observed in neurodegeneration could be understood and approached in relation to neuronal energetic metabolism. 3) Limited knowledge about the role of environmental toxins in the pathogenesis of neurodegenerative diseases is available. Hence, it is of interest to use rotenone as a neurotoxin in the proposed experiments because it is a mitochondrial inhibitor normally found in nature. As such, rotenone may play a role as an environmental risk factor for neurodegeneration (Sherer et al., 2003). 4) The retina represents a simplified model of neural organization that can be efficiently used to analyze the in vivo responses of neural tissue to neurotoxins and candidate neuroprotective agents (Zhang et al., 2002).

Near-infra-red light: biological effects and mechanism of action

Near-infra-red light (NIL) therapy is a strategy based on exposure of a target tissue to a low power and high fluency source of directional and monochromatic light in order to administer energy doses that are too low to cause heating, but high enough to modulate cell functions (Sommer et al., 2001). NIL therapy delivers low energy photon irradiation in the far red to near infra-red extreme of the visible spectrum (630-1000 nm).This strategy has been found to modulate various biological processes in cell culture, animal models, and clinical conditions (Whelan et al., 2003; Whelan et al., 2001; Yu et al., 1994).

Current evidence indicates that NIL photobiomodulation stimulates cellular energy metabolism and energy production by activating the mitochondrial respiratory chain components, which results in initiation of a signaling cascade that promotes cytoprotection (Wong-Riley MTT, 2005). The initial biological response to photoradiation is the interaction of photon energy with specific molecules in the target tissue that act as primary photoacceptors absorbing the incident photon energy. Biological tissue is characterized by the presence of photoacceptor molecules that represent an adaptation to maximize the assimilation of non-coherent electromagnetic radiation from the environment (e.g. sun light) (Alberts, 2002). This adaptation reaches its highest degree of efficiency in the thylakoid membrane of chloroplasts in plants, but is conserved in animal tissue as well. In mammalian tissue, the known photoacceptors are mainly heme containing metalloproteins. The three most important metalloproteins are hemoglobin, myoglobin and cytochrome c oxidase but others such as nitric oxide synthase, cytochrome b, cytochrome c, catalase, and the cryptochromes may also play a role as photoacceptors. All these molecules may be involved in the photobiomodulation exerted by NIL (Wong-Riley MTT, 2005).

In neural tissue, cytochrome c oxidase is regarded as the primary photoacceptor of light in the red to near-infra-red region of the visible spectrum (Fan et al., 2006; Yamanaka et al., 1988). Cytochome c oxidase is the terminal member of the electron transport chain of mitochondria. It catalyzes the reduction of more than 95% of the oxygen taken up by aerobic organisms and constitutes an efficient energy-transducing device acting as a redox-linked proton pump that creates a transmembrane electrochemical gradient to eventually drive ATP synthesis (Hatefi, 1985). NIL photoradiation is absorbed by cytochorme c oxidase, and enhances the chemiosmotically coupled activity of this enzymatic complex promoting an increase in energy production. It has been shown that peaks in the absorption spectrum of cytochrome c oxidase (670 and 830 nm) highly correlate with its peaks in catalytic activity and ATP production in vitro (Figure 1). Irreversible inhibition of cytochrome c oxidase by potassium cyanide prevents NIL therapy-induced increases in neuronal ATP content (Wong-Riley MTT, 2005).

After enhancement of cytocrome c oxidase activity and the concomitant stimulation of neuronal energy metabolism, NIL also affects gene expression and cell survival, through unknown mechanisms. NIL effects reported up until now in vitro include: 1) increased expression of the anti-apoptotic protein Bcl-2 and reduced expression of the pro-apoptotic protein Bax (Shefer et al., 2002), 2) decrease in the number of apoptotic cells after exposure to a neurotoxic fragment of the amyloid beta protein (Duan R, 2003; Wong-Riley MTT, 2005), 3) functional improvement of cortical neurons inactivated by toxins (Wong-Riley MTT, 2005), 4) promotion of neurite outgrowth (Wollman and Rochkind, 1998; Wollman et al., 1996), 5) regulation of genes encoding for DNA repair proteins, antioxidant defense enzymes, molecular chaperones, protein biosynthesis enzymes, and trafficking and degradation of proteins involved in energy metabolism and cell growth (Eells JT, 2004), and 5) increased proliferation of olfactory ensheating stem cells (Byrnes et al., 2005), Schwann cells (Van Breugel and Bar, 1993), astrocytes and oligodendrocytes (Rochkind et al., 1990). In vivo, NIL therapy has been shown to induce peripheral and central nerve regeneration after trauma (Anders et al., 1993; Byrnes KR, 2005), reduce neuroinflammation (Byrnes KR, 2005; Yu et al., 1997), and prevent methanol-induced photoreceptor degeneration (Eells JT, 2003). No clinical applications of NIL therapy in human neurological conditions exist so far, but NIL photobiomodulation has been applied clinically with success in the prevention of oral mucositis in immunocompromised patients (Whelan et al., 2002), and in stimulation of wound closure and skin graft healing (Conlan et al., 1996; Sherer et al., 2003).

Several considerations about the biological effects of NIL photoradiation on neurons are relevant for its applicability in neuroprotective interventions. 1) NIL photomodulation depends on the cell type, the viability of the cell and the specific wavelength, dosage, intervals and source of radiation. In particular, photostimulatory or photoinhibitory effects are obtained with high (> 10 J/cm2) and low (0.001 – 10 J/cm2) energy fluencies, respectively (Brondon et al., 2005). 2) Light tissue penetration depends on both the type of target tissue and the source of NIL. Light penetration in the eye is maximal, where absorbance by the cornea and lens is negligible (<>NIL can be produced by light-emitting diodes (LED) arrays and lasers. Laser sources allow better penetration than LED sources, but they have limitations in beam width, wavelength capabilities, and target tissue size that can be treated. In addition, lasers area associated with heat production that can induce tissue damage (Eells JT, 2004). In contrast, LED arrays generate negligible amount of heat reducing the risk of thermal injury. Directionality of the electromagnetic radiation produced by LED arrays is lower than that of laser sources, but spectroscopic measures have shown that most photons at wavelengths between 630-800 nm are able to travel approximately 23 cm even in tissues with relatively low transparencies such as skin and muscle (Chance et al., 1988). LED arrays are compact, portable, and have achieved non-significant risk status for human trials by the FDA (Whelan et al., 2001). 4) NIL may be superior to other neuroprotective agents in regards to their range of effects. The vast majority of neuroprotective compounds studied so far act as antioxidants (Ezaki et al., 2005), metabolic substrates (Andres et al., 2005), membrane stabilizers (Hurtado et al., 2005), channel blockers (Stieg et al., 1999), or growth factors (Fan et al., 2006). The neuroprotective effects of these compounds are usually limited by constrains related to bioavailability, drug-drug interactions, pharmacogenetics, and therapeutic index. In contrast, given its effect on energy metabolism and gene expression and the probable effects on a wide range of metalloproteins, NIL therapy may have a wider range of cytoprotective effects.

Neurodegeneration, mitochondrial dysfunction and rotenone

Neurodegeneration is the progressive damage or death of neurons leading to a gradual deterioration of the functions controlled by the affected neural system. The cause of neurodegeneration in most instances is not known, but it is probably the result of the interplay of genetic and environmental factors. Environmental neurotoxins have received a great deal of attention during recent years. Their effects on the neuronal physiology have been linked to mitochondrial dysfunction (Beal, 2005; Bolanos et al., 1997). Inhibition of the electron transfer chain in mitochondria has been associated with cell dysfunction and neuronal death. In particular, inhibition of complex I (NADH Coenzyme Q reductase) has been linked to production of reactive oxygen species, which can in turn affect diverse cellular systems, including the respiratory chain complexes themselves, lipid membranes, channels, and lead to activation of proteases, endonucleases, and impairment of gene expression and protein degradation (Koopman et al., 2005; Smeitink et al., 2004; Triepels et al., 2001). These events lead to further activation of the apoptotic cascade and necrosis. Complex I deficiency may be relevant in the pathogenesis of neurodegenerative disorders and it has been identified in the substantia nigra of Parkinson’s disease brains (Schapira et al., 1990). Several complex I inhibitors have been demonstrated to be linked to neurodegeneration and are used in models to study the toxicological mechanisms of degeneration. Rotenone is a naturally occurring ketone that is commonly used as a pesticide and to kill fish that are perceived as pests in lakes and reservoirs (Lindahl and Oberg, 1961). Rotenone is a high affinity inhibitor of complex I of lipophilic nature that freely crosses cellular membranes independently of any transporters. It easily crosses the blood–brain barrier, enters into the brain rapidly, and accumulates in subcellular organelles, such as mitochondria, where it impairs the generation of the mitochondrial transmembrane potential needed to maintain oxidative phosphorylation (Sherer et al., 2003). Besides resulting in reduced production of ATP, rotenone also increases the generation of reactive oxygen species. Complex I inhibition is the main source of superoxide in rotenone-treated cells and rotenone produces an increase in nitric oxide production, lipid peroxidation and a decrease in oxygen consumption in brain tissue (Koopman et al., 2005; Zhang et al., 2006). Apoptosis has been induced by rotenone after 20 min in a liver cell line, and it is not prevented with the removal of rotenone (Isenberg and Klaunig, 2000). This indicates that mitochondrial dysfunction induced by acute rotenone administration may trigger a cascade of further damage in the absence of rotenone, eventually leading to cell death. For these reasons rotenone has been proposed to be one of the etiological components of neurodegenerative disorders. It is possible that complex I inhibition contributes to the deleterious effects of excitotoxicity through an increase in oxidative stress which in turn would lead to macromolecule modification, protein misfolding and apoptosis activation. Additionally, rotenone would reduce the capacity to stabilize cell membrane potential due to inhibition of oxidative phosphorylation and absence of energy, leading to further excitotoxic damage. Because of its impairment of energy metabolism at the mitochondrial level and its probable role as an environmental toxin in neurodegenerative diseases it is relevant to use rotenone in the proposed experiments to ascertain the neuroprotective effects of NIL photoradiation.

An in vivo model for neuroprotective intervention screening

Zhang et al. (2002) previously reported that a single intravitreal injection of rotenone causes retinal degeneration which could constitute an ideal model to address the efficacy of neuroprotective interventions. In this model, the eye is used as “an accessible part of the brain” since it may allow efficient assessment of neuronal damage at three levels of organization: behavioral (visual function), morphological (retinal thickness and neuronal number), and biochemical (protein expression). This combinatorial level of evidence is relevant for the study of neuroprotective interventions due to two main reasons. First, it allows determining whether the intervention is effective (through interpretation of behavioral and morphological data). Second, it facilitates insights into the mechanism of action of the intervention (through analysis of the biochemical and morphological data).The use of this multilevel approach in the study of the effects of NIL therapy is more amenable to contributing in the establishment of a solid basis for translational research on NIL as a therapy in neurological disorders.

The main areas of opportunity in this study include: 1) No previous studies have addressed the question of whether NIL therapy can maintain the function of the visual system as determined by neuropsychological tests. When behavioral tests are used, studies on neuroprotection usually rely on observation of effects in motor systems which are grossly evident. For these reasons, evaluation of the visual function in studies of neuroprotection, if ever done, usually consist on electrophysiological recordings in the form of visual evoked potentials, neuroretinography or single unit recording. Electrophysiological recordings are complicated by the use of sophisticated apparatuses to perform the recordings, usually with a restrained or anesthetized animal that has undergone an additional surgical intervention for electrode implantation. 2) No previous studies have described NIL effects in rotenone-treated systems in vivo and elucidation of the mechanism of action of NIL in neural tissue under rotenone-induced mitochondrial dysfunction is also required. Of particular interests are the pathways that translate cytochrome c oxidase activation into nuclear and mitochondrial gene activation as well as the effect of NIL on metalloproteins that may also be play a role in the neuroprotective effects of this intervention.


C. RESEARCH DESIGN AND METHODS

The proposed experiments are divided in three specific aims each one characterizing the neuroprotective effects of NIL therapy in vivo at one of three different levels: functional, morphological and biochemical. These experiments are expected to determine whether NIL therapy is an effective neuroprotective intervention against rotenone-mediated toxicity and to improve the understanding about the mechanisms mediating the effects of NIL radiation in neural tissue.

1. Description of individual experiments and expected results

General outline of the experimental approach

Subjects used in specific aim 1 (n = 40, experiments 1-3) will be divided into four groups: 1) control, 2) rotenone, 3) NIL dose 1, and 4) NIL dose 2. 1) Subjects in control group (n = 10) will receive bilateral intravitreal injections of the vehicle dimethylsulfoxide (DMSO) and will be further subdivided in a group irradiated with a control white LED light (n = 5) and a non-radiated group (n = 5). 2) Subjects in rotenone group (n = 10) will receive bilateral intravitreal injections of 0.2 mg/kg rotenone and no light treatment. 3) Subjects in NIL dose 1 group (n = 10) will receive bilateral intravitreal injections of rotenone and will be irradiated with 0.5 J/cm2 NIL. 4) Subjects in NIL dose 2 group (n = 10) will receive bilateral intravitreal injections of rotenone and will be irradiated with a fluency ten times higher than NIL group 1 (5 J/cm2 NIL). The NIL doses chosen are based on previous studies showing stimulatory effects at fluencies between 0.001 J/cm2 and 10 J/cm2. The total NIL doses administered will divided in three sessions each one occurring at 1 hr, 24 hr and 48 hr after rotenone injections respectively. This dose fractionation will be implemented based on previous studies reporting that multiple NIL treatment sessions over days are more beneficial than the strategy of delivering a single treatment (Figure 2). Visual tests will be performed in sessions before and after bilateral intravitreal injections of rotenone to allow for within-group paired comparisons in performance. After the last visual test session, animals will be decapitated and eyeballs and brains will be extracted for further histological processing.

Subjects used in specific aim 2 (n = 40, experiments 4-5) will be divided into four groups similar to specific aim 1 (control, rotenone, NIL dose 1 and NIL dose 2) and will be treated accordingly. Animals will be decapitated and eyeballs and brains will be extracted for further histological processing. This tissue will also serve to conduct experiment 8 in specific aim 3.

Brains of subjects in specific aim 3 (n = 30, experiments 6-7) will be irradiated with three different NIL doses through a skull window exposing the frontal cortex. Biopsies of the irradiated areas will be taken and homogenized.

Possible pitfalls and alternate approaches. 1) Intravitreal injections of rotenone might not impair vision. Previous studies have shown that intravitreal administration of 0.2 mg/kg of rotenone causes a reduction in thickness of the ganglion cell layer and nerve fiber layer of the retina by 80%, and a 20% reduction in the number of ganglion cells at 24 hr. However, these morphological changes might not be enough to impair visual function in a way that is detectable by the proposed behavioral tests. In this case, a higher dose of rotenone (2 mg/kg) will be used in each intravitreal injection. 2) The effects of rotenone on visual function may also be time-dependent and further damage and impairment of vision could be detected only several days after injection instead of 24 hr. Hence, an alternate experiment could include evaluation of vision and morphology after 7, 14 or 21 days after the injection. 3) In addition, NIL treatments may not be effective in impairing the neurotoxic effects of rotenone. In this case, the energy doses would be administered reducing the power intensity in every session and increasing the number or duration of sessions. Also, NIL therapy sessions could be given before the intravitreal injection of rotenone, instead of after. This is based on previous evidence showing that NIL pre-treatments are able to reverse the detrimental effects of potassium cyanide on cell survival in vitro. Alternatively NIL treatments could be administered over a broader time lapse including sessions before and after rotenone injections.

Specific Aim No. 1. The first aim is to characterize the neuroprotective effects of NIL therapy on the visual pathway of the rat using functional parameters.

Experiment 1. The visual cliff. The visual cliff will be used to provide a measure of gross visual ability. This apparatus evaluates the ability of the subjects to see the vertical drop-off at the edge of a horizontal surface. A sheet of clear Plexiglas will be placed across the top horizontal plane, extending across the cliff. Thus, the subject will have the visual appearance of a cliff but in fact the Plexiglass will provide a solid horizontal surface. Each subject will be placed onto the center of a ridge at the edge of the cliff at the beginning of a trial. If the subject sees the cliff, it will stop at the edge of the vertical drop-off, while a blind animal may move forward across the “cliff” without a pause. Each subject will be tested in 10 trials and the time taken for the subject to move off the ridge will be recorded as well as whether the subject chooses to step down onto the horizontal (safe) surface or onto the vertical-appearing (cliff) surface. A low latency of response or a 50/50 (0.5) ratio of responses is expected to be found in animals with deficient visual function (e.g. rotenone-treated group), whereas high latencies, prolonged exploratory behavior on the ridge, and high safe-cliff ratio (0.8-1) is expected from subjects with normal or almost normal visual function (e.g. control and NIL therapy groups) (Figure 3).

Possible pitfalls and alternative approaches. 1) This test only provides a gross and qualitative assessment of visual function and it may not be sensitive to detect changes in visual function induced by the different treatments. The sensitivity of this test could be increased however, by introducing a light intensity threshold sensitivity component, similar to that described in experiment 2. In this instance, the visual cliff will be performed under different light intensities. Low intensities could synergize with the neurodegenerative effect of rotenone, decreasing the performance in the visual cliff. At the same time, this alternate approach would give an idea of the robustness of the NIL treatments and would introduce a quantifying variable to identify the threshold of illumination in which the subject is able to perform at levels different from chance. 2) In addition, the visual cliff task may be confounded by the ability of rats to use nonvisual senses for navigation. If a subject is blind it may successfully navigate using feedback from the olfactory system or from whiskers and paws of the tactile feel of the continuous Plexiglas floor. The olfactory cues could be eliminated by rotating the Plexiglas and the ridge and by cleaning all surfaces before every trial. Also it may be possible to eliminate the sensory information from the whiskers by shaving off the whiskers before the visual cliff test. In this case, the visual cliff will be performed only after the treatments and will be done at the end of all behavioral training and testing, since whisker removal will affect performance in other behavioral tasks.

Experiment 2. Light intensity threshold discrimination in the water-maze. Light threshold represents the minimal amount of luminous energy that is needed to elicit an action potential in a neuron. Under normal conditions, neural activity dictates the rate of energy metabolism. However, if energy metabolism is impaired, neural activity is also reduced and this is associated with an elevation of the light threshold. This means that more luminance is needed to excite the neuron and trigger an action potential. Thus, a sensitive, quantitative and fine evaluation of the visual function can be obtained by determining the absolute dark-adapted threshold to light. A behavioral test for light intensity threshold is proposed using a visible platform version of the Morris water maze. In this test, an intact, freely moving subject will be presented with a relatively normal situation involving identification of an object in a dark environment. The subject will be trained to climb out of the water using a dimly illuminated escape platform, visible above the water level. Illumination of the platform could be attenuated and latencies to escape will be compared within and between groups. A subject with intact visual function should be able to identify the escape platform and reach it in a short period of time, even with low levels of illumination. A subject with impaired vision should show increased latencies to escape for a given level of illumination, especially when this level is low. The minimum illumination level at which the subject is able to identify the escape platform before the end of a trial can be regarded as the light intensity threshold. The advantages of the use of this behavioral test of visual function in the evaluation of NIL therapy effects include: a) the possibility of describing dose-response relationships, and b) elimination of somatosensory and olfactory cues during performance of the task. A threshold elevation of at least one Log luminance unit (cd/m2) is expected in subjects treated with rotenone. No changes in intensity threshold are expected in the control group and no changes in threshold or intermediate thresholds are expected in the NIL therapy groups (Figure 4).

Possible pitfalls and alternative approaches. 1) Light sensitivity may not be affected by the treatments. In previous toxicological models of retinal degeneration, cell damage has also associated with loss of spatial and temporal resolution and only minor impairments in light intensity perception. For this reason, a water Maze apparatus with spatial frequency and orientation discrimination could be implemented. In this apparatus, the water maze would include a Y maze in which only one arm contains an escape platform. This arm will be associated with a visual cue consisting of an oblique (10-80°) grating pattern with low spatial frequency that should be identified and used for navigation by the subject. The other arm of the Y maze will be associated with a horizontal grating pattern with high spatial frequency and will not contain an escape platform. The subject should be trained to find the escape platform using the visual cues and the latency to escape as well as the number of correct choices (e.g. choosing the arm with an escape platform) will be computed. An animal with intact vision is expected to show lower latencies to escape and a higher number of correct choices than animals with impaired vision.

Experiment 3. Metabolic activity in the central visual pathway. An indirect functional measurement of retinal neural function can be determined by analyzing the activity of the energy-deriving mitochondrial enzyme cytochrome c oxidase in the central visual pathway. Upon perception of visual stimuli, components of the visual pathway are expected to show and increase in neural activity. In the brain neural activity is supported by energy derived from oxidative metabolism. Because there is tight coupling between energy metabolism and neuronal activity, it is reasonable to expect that altered activity within the central visual pathway will be reflected by regional changes in levels of energy metabolism. These adjustments involve the level of activity of cytochrome c oxidase. It has been demonstrated that sensory deprivation of the auditory, somatosensory and visual systems results in a histochemically detectable lowering of the cytochrome c oxidase activity within relay centers one to several synapses away. Optical density of histochemically label sections closely correlates with the amount of cytochrome c oxidase activity in nervous tissue. Hence, in this experiment, brains from the subjects in the four above-mentioned treatment groups (control, rotenone, NIL dose 1 and NIL dose 2) will be sectioned and stained for cytochrome c oxidase activity for analysis of metabolic activity in the central visual pathway. Densitometric measurements will be taken from the superficial gray layer of the superior colliculus, dorsal lateral geniculate nucleus and primary visual cortex. Sensory deprivation elicited by rotenone-induced retinal neurodegeneration should lead to a decrease in neural activity identifiable as a decrease in cytochrome c oxidase activity in the central visual pathway after two days. It is also expected that the NIL treatment groups will show no changes in visual pathway metabolic activity compared to the control group.

Possible pitfalls and alternative approaches. 1) No changes in cytochrome c oxidase activity may be detected in the rotenone-treated group. Even a dose of rotenone effective for inducing visual impairment and retinal degenerative changes determined by histological examination may not be associated with a change in the metabolic activity of the visual pathway. This may occur if deafferentiation is not prolonged enough to allow changes in levels of protein expression to take place. Cytochrome c oxidase activity changes are regulated at the level of protein synthesis and they are based on long-term metabolic demands of the tissue. Thus, in order to enhance the sensibility of cytochrome c oxidase as a marker of neuronal activity, survival of subjects could be increased to 7, 14 or 21 days after rotenone injection.

In summary, experiments 1-3 will allow determining the efficacy of NIL therapy to prevent rotenone-induced neurodegeneration in the retina using functional parameters. This will be done by evaluating gross and fine visual abilities with behavioral tests and quantifying metabolic activity in central connections of retinofugal neurons with a histochemical approach.

Specific Aim 2. The second aim is to determine the effects of NIL radiation as a neuroprotective intervention by means of morphological parameters.

Experiment 4. Retinal morphometry. The histopathological changes in the retina will be evaluated in the four treatment groups (control, rotenone, NIL dose 1 and NIL dose 2) using unbiased stereological techniques. This experiment will allow identifying structural changes induced by rotenone and will help to demonstrated whether NIL therapy is able to prevent these structural changes. The major neuronal type in the retina is the ganglion cell, which directly sends projections to the central nervous system. The innermost layer of the retina, the retinal nerve fiber layer (RNFL), contains the axons of the ganglion cells. The ganglion cell layer (GCL) is contiguous to RNFL and contains mainly the cell bodies of the ganglion cells. Morphometric measurements will include retinal volume, ganglion cell count, ganglion cell density, and GCL + RNFL thickness. Cell counts will be obtained using the optical disector principle in cresyl violet-stained sections. Retinal volume will be estimated using the Cavalieri principle in sections stained for mitochondrial complex I activity. Cell density will be estimated using cell counts with the optical disector. GCL + RNFL thickness will be estimated using linear analysis in complex I-stained sections. Complex I staining is ideal for thickness and volumetric analyses, since it allows identification of both cell bodies and neuropil, whereas cresyl violet staining only delineates cell bodies. It is expected that after the treatments, subjects in the rotenone group will show a decrease in ganglion cell counts, retinal volume, ganglion cell density and GCL + RNFL thickness 48 hr after the intravitral injection. It is also anticipated that NIL therapy groups will show no changes in this morphological parameters, compared to controls.

Possible pitfalls and alternative approaches. 1) Higher doses of rotenone or NIL may be needed to induce morphological changes in the retina. Doses of 0.4, 0.8 or 2 mg/kg rotenone may be used to induce morphologically detectable neurodegeneration in the retina. In addition, NIL dose fractionation could be attempted as indicated in the general outline. 2) Efficiency of stereological estimates can be maximized by increasing the precision of estimates (e.g. reducing the coefficient of error), by analyzing at least 5 or 6 systematic slices through the object interest (e.g. eyeball) (15) and by performing cell counts in the same sections used to estimate retinal volume, thus sampling a larger fraction of the organ (16). Bias in the optical disector technique can be reduced by obtaining thick sections (40-100 mm) and adjusting for lost caps and discarding the top and bottom planes of the sections. Finally, having the same experienced experimenter performing all measurements will reduce the margin of error in stereological estimates.

Experiment 5. Induction of programmed cell death. In situ visualization of programmed cell death at the individual cell level will be accomplished by staining ganglion cells in retinal sections from subjects in the four treatment groups (control, rotenone, NIL dose 1 and NIL dose 2) for nuclear DNA fragmentation using terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL). Cells undergoing programmed cell dead are characterized by chromatin condensation and fragmentation, events that can be histochemically detected by TUNEL even in the absence of the defining morphological features of apoptosis. The number of cells undergoing programmed cell death will be calculated by means of stereological particle counting with the optical disector 48 hours after rotenone administration. It is expected that NIL therapy will prevent rotenone-induced increase in positive TUNEL cells at 48 hrs.

Possible pitfalls and alternative approaches. 1) The time point chosen to analyze program cell death may not be the most informative. It has been reported that apoptotic cells can be identified from 20 min to 14 days after toxin exposure. Preliminary findings demonstrate an increase in TUNEL positive cells in retinas of rotenone-treated subjects compared to control subjects, as soon as only one hour after injection with no difference at 24 hrs. Thus the effect of NIL therapy on program cell death could be biased by the absence of late rotenone effects as detected by TUNEL. To avoid this bias, effects of treatments could be evaluated after one hour of injection and after only one NIL session. Alternatively, immunostainning for proteins expressed early in the apoptotic cascade, such as caspase 3 or caspase 8 may be done at one hour after rotenone injection.

It is expected that experiments 4-5 will help to demonstrate that ganglion cells are vulnerable to rotenone toxicity and that NIL therapy can prevent cell loss induced by rotenone. These experiments will also help in assessing whether NIL therapy prevents activation of programmed cell death mechanisms.

Specific Aim 3. The third specific aim is to characterize the mechanism of action of NIL therapy using biochemical markers.

Experiment 6. Effect of NIL therapy on oxygen consumption. This experiment will test the hypothesis that NIL radiation increases oxygen consumption in vivo. In situ oxygen concentrations will be measured during irradiation with a control white light and during irradiation with NIL with an oxygen sensor inserted in the frontal cortex. A skull window exposing the frontal lobe will be used to administer irradiation to the frontal cortex during one hour divided in two sessions. During the first 30 min session, the brain will be irradiated with a source of white light at fluencies of 0.1, 1 and 10 J/cm2 as a control. During the second 30 min session the same brain region will be exposed to NIL radiation at the same fluencies. Oxygen consumption will be expressed as the ratio of oxygen concentration in the second interval over oxygen concentration during the first interval. It is expected that NIL will enhance oxygen consumption in a dose-dependent manner.

Possible pitfalls and alternative approaches. 1) No changes in oxygen consumption may be detected in vivo. To reduce the variability of the oxygen concentration measurements, they can be carried out in an in vitro system, following a setting similar to the one presented in experiment 6. Brain biopsies from irradiated areas will be obtained (one hemisphere use as a control), homogenates will be created and oxygen concentration measures will be done every 10 sec, during a 5 min period. Oxygen consumption will be determined by percent changes in oxygen concentration compared to the controls.

Experiment 7. Effect of NIL therapy on ATP production. This experiment will evaluate the effects of NIL radiation at fluencies of 0.1, 1 and 10 J/cm2 on in situ ATP production. The experiment will help to determine whether NIL therapy is able to maintain the chemiosmotic coupling in the internal mitochondrial membrane. The brain of anesthetized subjects will be exposed to light through a skull window exposing the frontal lobe. Irradiation will occur during one hour divided in two sessions. During the first 30 min session, the brain will be irradiated with a source of white light at fluencies of 0.1, 1 and 10 J/cm2 as a control. During the second 30 min session the same brain region will be exposed to NIL radiation. Punch out biopsies of the irradiated areas will be obtained after each session and will be used in homogenates preparation and mitochondria isolation. ATP concentrations will be determined by means of the luciferin-luciferase assay in an illuminometer. It is expected that NIL will enhance ATP production and no change in oxygen consumption will be observed with radiation with white light.

Possible pitfalls and alternative approaches. 1) No differences in ATP content may be detected after the radiation sessions. To increase the sensitivity of the test isolation of mitochondria could be avoided, working with whole brain homogenates. Alternatively, mitochondrial isolates from naïve animals could be radiated in an in vitro system and ATP levels could be determined using the luciferase assay.

Experiment 8. Effect of NIL therapy on p-CREB, SOD and nNOS expression. Eyeballs from subjects divided into four groups as described in the general approach section (control, rotenone, NIL dose 1 and NIL dose 2) will be collected after the last NIL radiation session (e.g. 2 days after rotenone injection), sectioned and immunohistochemically stained for the expression of the phosphorylated form of the cAMP response element binding protein (p-CREB), superoxide dismutase (SOD), and the inducible neuronal nitric oxide synthase (nNOS). This experiment will allow determining: 1) whether there is a change in protein expression following NIL therapy, specifically of proteins involved in antioxidant defenses (SOD) and enhancement of mitochondrial respiratory chain inhibition and cell damage (nNOS) and 2) whether changes in protein expression are mediated by the activation of the transcription factor CREB. It is expected that NIL treated subjects will show an increase in p-CREB and SOD levels, and a decrease in nNOS expression.

Possible pitfalls and alternative approaches. 1) Changes in protein expression may not be detectable at the proposed time point. In this case, tissue analysis could be done at shorter survival times (e.g. 1, 12 or 24 hr after rotenone injection and 1 NIL radiation session). Additional proteins could be analyzed, including catalase and cytochrome c oxidase. If immunohistochemical analysis of retinal tissue is not informative, an additional approach could be followed to further elucidate the in vivo impact of NIL therapy on nitric oxide-related pathways. This would consist on measuring in situ levels of NO using a selective microelectrode assay of NO3- and NO2- ions, NO derivatives, in brains radiated with NIL. A third option will include measurement of nNOS activity through histochemical detection of NADPH-diaphorase activity.

In summary, experiments 6-8 will help to understand the mechanism of action of the neuroprotective effects of NIL in vivo, determining whether NIL enhances chemiosmotically coupled oxygen consumption and regulates protein expression associated with neuroprotection via a CREB-mediated pathway.

2. Methods

Subjects and tissue processing. Male Long-Evans rats (n = 110, weight 250-350 g) will be used in all experiments. Subjects will be decapitated; brains and eyeballs will be extracted and frozen in -40°C isopentane. Eyeballs and brains will be sectioned at 4 or 40 mm and mounted in glass slides to create 3 series per specimen.

Rotenone intravitreal injections. Intravitreal microinjections will be performed in the superior and temporal quadrant of the eye with a 30-G dental injection needle connected by polyethylene tubing to a 10-ml Hamilton microsyringe. Injections will be delivered over 2 min using a microinjection pump (Monoject, Sherwood Medical Company, Norfolk, NE) (0.5 mL, final volume).

NIL therapy. A self-made GaAlAs LED array (40 cm x 40 cm) of 670 nm wavelength will be placed at 2.5 cm of the rostrum. Power intensities and times of radiation will vary according to the target energy dose.

Visual cliff. A transparent Plexiglas surface (100 cm x 50 cm) will be placed at the edge of a horizontal surface that will continue with a vertical drop-off (37.5 cm) accentuated by a black and white checkerboard pattern. The Plexiglas surface will extend across the cliff. A rectangular metallic ridge (80 cm x 8 cm x 3 cm) will be placed at the edge of the cliff. Each subject will be placed onto the center of the ridge at the beginning of a trial, making sure that it perceives both sides of the ridge (horizontal vs. cliff). Subjects will be tested in 10 consecutive trials and the time taken for the subject to move off the ridge will be recorded as well as whether the subject chooses to step down onto the horizontal surface or onto the vertical-appearing (cliff) surface.

Water maze. A circular pool (diameter 1.7 m; height 0.6 m) will be filled to within 20 cm of the top with cool tap water (18°C). An ramp (13 x 13 cm) with its top surface 1 cm above the water surface will be used as an escape platform. No intramaze cues will be available except for the escape platform. Ambient luminances will be attenuated with neutral density filters. Subjects will be trained to escape in four consecutive trials each day under full illumination (70 cd/cm2) for ten days. Performances will be recorded according to latency to escape. After training, probe trials with decreasing luminances will be done until a significant decrease in performance is observed (e.g. a performance that differed from the initial latencies after training). The light level prior to the decrease in performance will be defined as the subject’s dark-adapted threshold.

Complex I and Nissl staining. Histochemical staining for mitochondrial complex I will be done following a previously described procedure using tetrazolium salts (Jung et al., 2002).
Cytochrome c oxidase staining. Histochemical staining for cytochrome oxidase activity of fresh-frozen brain sections will be done with the principle of diaminobenzidine conversion to an indamine polymer as described previously. (Gonzalez-Lima and Cada, 1994, 1998; Gonzalez-Lima and Jones, 1994).

Apoptosis staining. A series of retinal fresh frozen-sections will be stained for the TUNEL-based detection of apoptosis, using the FD NeuroApop Kit (FD Neurotechnologies, Inc, Ellicot MD).
p-CREB, nNOS and SOD immunostaining. Indirect immunohistochemistry will be used to demonstrate protein expression in fresh frozen retinal sections.

Stereological measurements. Total retinal volume will be estimated using the Cavalieri principle. Ganglion cells density (cells/volume) will be estimated using the optical disector methodology. Total ganglion cell number will be estimated from the ganglion cell density and the retinal volume. An estimate of ganglion cell layer (GCL) + retinal nerve fiber layer (RNFL) thickness will be obtained by means of a systematic linear analysis.

Densitometric measurements of enzymatic activity of cytochrome oxidase. Using an image-processing system (JAVA, Jandel Scientific, Corte Madera, CA), optical density will be sampled from the regions of interest. Optical density values will be then converted to cytochrome oxidase activity units, which will be determined by spectrophotometry of cytochrome c oxidase standards as described by Gonzalez-Lima & Cada (1994).

Oxygen consumption. Oxygen concentrations will be measured at intervals of 5-minutes with a fast response fiber optic fluorescence oxygen sensor (Ocean Optics, Tampa, FL) inserted in the frontal cortex. Changes in dissolved oxygen concentration in the two sessions will be compared and oxygen consumption will be expressed as the ratio of oxygen concentration in the second interval over oxygen concentration during the first interval.

ATP production. ATP content will be determined by means of the luciferin-luciferase method using the components of the bioluminiscent somatic cell assay kit (Stratagene, La Jolla, CA).

Statistical Analyses. 1) Differences in performances in the visual cliff and water maze and oxygen consumption and ATP production w
ill be determined using Bonferroni-corrected repeated-measures ANOVA and paired student t tests. 2) Differences in densitometric measures of cytochrome c oxidase activity, retinal volume, total ganglion cell number, GCL + RNFL thickness, apoptotic cells, SOD, nNOS and p-CREB expression between treatment groups will be determined with Bonferroni-corrected independent samples one-way ANOVA. 3) Tests for normality will be conducted in stereological data. Monte Carlo simulations will be used to run non-parametric tests if data is found to violate the normality assumption. All tests will use a two-tailed p <>LITERATURE CITED
Alberts, B. (2002). Molecular biology of the cell, 4th edn (New York: Garland Science).
Anders, J. J., Borke, R. C., Woolery, S. K., and Van de Merwe, W. P. (1993). Low power laser irradiation alters the rate of regeneration of the rat facial nerve. Lasers Surg Med 13, 72-82.
Andres, R. H., Huber, A. W., Schlattner, U., Perez-Bouza, A., Krebs, S. H., Seiler, R. W., Wallimann, T., and Widmer, H. R. (2005). Effects of creatine treatment on the survival of dopaminergic neurons in cultured fetal ventral mesencephalic tissue. Neuroscience 133, 701-713.
Beal, M. F. (2005). Mitochondria take center stage in aging and neurodegeneration. Ann Neurol 58, 495-505.
Bolanos, J. P., Almeida, A., Stewart, V., Peuchen, S., Land, J. M., Clark, J. B., and Heales, S. J. (1997). Nitric oxide-mediated mitochondrial damage in the brain: mechanisms and implications for neurodegenerative diseases. J Neurochem 68, 2227-2240.
Brondon, P., Stadler, I., and Lanzafame, R. J. (2005). A study of the effects of phototherapy dose interval on photobiomodulation of cell cultures. Lasers Surg Med 36, 409-413.
Byrnes KR, W. R., Ilev IK, Wu X, Barna L, Smith K, Heckert R, Gerst H, Anders JJ. (2005). Light promotes regeneration and functional recovery and alters the immune response after spinal cord injury. Lasers Surg Med 36, 171-185.
Byrnes, K. R., Wu, X., Waynant, R. W., Ilev, I. K., and Anders, J. J. (2005). Low power laser irradiation alters gene expression of olfactory ensheathing cells in vitro. Lasers Surg Med 37, 161-171.
Chance, B., Nioka, S., Kent, J., McCully, K., Fountain, M., Greenfeld, R., and Holtom, G. (1988). Time-resolved spectroscopy of hemoglobin and myoglobin in resting and ischemic muscle. Anal Biochem 174, 698-707.
Conlan, M. J., Rapley, J. W., and Cobb, C. M. (1996). Biostimulation of wound healing by low-energy laser irradiation. A review. J Clin Periodontol 23, 492-496.
Duan R, Z. L., Cheng-Yi Liu T, Li Y, Liu J, Jiao J, Xu X, Yao L, Liu S. (2003). Light emitting diode irradiation protect against the amyloid beta 25-35 induced apoptosis of PC12 cell in vitro. Lasers Surg Med 33, 199-203.
Eells JT, H. M., Summerfelt P, Wong-Riley MTT, Buchmann EV, Kane M, Whelan NT. (2003). Therapeutic photobiomodulation for methanol-induced retinal toxicity. PNAS 100, 3439-3444.
Eells JT, W.-R. M., VerHoeve J, Henry M, Buchmann EV, Kane MP, Gould LJ, Das R, Jett M, Hodgson BD, Margolis D, Whelan HT. (2004). Mitochondrial signal transduction in accelarated wound and retinal healing by near-infrared light therapy. Mitochondrion 4, 559-567.
Ezaki, Y., Nishihara, E., Shibata, Y., Matsuo, T., Kitagawa, N., Nagata, I., and Shinohara, K. (2005). Vitamin E prevents the neuronal cell death by repressing cyclooxygenase-2 activity. Neuroreport 16, 1163-1167.
Fan, W., Agarwal, N., and Cooper, N. G. (2006). The role of CaMKII in BDNF-mediated neuroprotection of retinal ganglion cells (RGC-5). Brain Res 1067, 48-57.
Hatefi, Y. (1985). The mitochondrial electron transport and oxidative phosphorylation system. Annu Rev Biochem 54, 1015-1069.
Hurtado, O., Moro, M. A., Cardenas, A., Sanchez, V., Fernandez-Tome, P., Leza, J. C., Lorenzo, P., Secades, J. J., Lozano, R., Davalos, A., et al. (2005). Neuroprotection afforded by prior citicoline administration in experimental brain ischemia: effects on glutamate transport. Neurobiol Dis 18, 336-345.
Isenberg, J. S., and Klaunig, J. E. (2000). Role of the mitochondrial membrane permeability transition (MPT) in rotenone-induced apoptosis in liver cells. Toxicol Sci 53, 340-351.
Jester, J. V., Moller-Pedersen, T., Huang, J., Sax, C. M., Kays, W. T., Cavangh, H. D., Petroll, W. M., and Piatigorsky, J. (1999). The cellular basis of corneal transparency: evidence for 'corneal crystallins'. J Cell Sci 112 ( Pt 5), 613-622.
Koopman, W. J., Visch, H. J., Verkaart, S., van den Heuvel, L. W., Smeitink, J. A., and Willems, P. H. (2005). Mitochondrial network complexity and pathological decrease in complex I activity are tightly correlated in isolated human complex I deficiency. Am J Physiol Cell Physiol 289, C881-890.
Lindahl, P. E., and Oberg, K. E. (1961). The effect of rotenone on respiration and its point of attack. Exp Cell Res 23, 228-237.
Rego, A. C., and Oliveira, C. R. (2003). Mitochondrial dysfunction and reactive oxygen species in excitotoxicity and apoptosis: implications for the pathogenesis of neurodegenerative diseases. Neurochem Res 28, 1563-1574.
Rochkind, S., Vogler, I., and Barr-Nea, L. (1990). Spinal cord response to laser treatment of injured peripheral nerve. Spine 15, 6-10.
Schapira, A. H., Cooper, J. M., Dexter, D., Clark, J. B., Jenner, P., and Marsden, C. D. (1990). Mitochondrial complex I deficiency in Parkinson's disease. J Neurochem 54, 823-827.
Shefer, G., Partridge, T. A., Heslop, L., Gross, J. G., Oron, U., and Halevy, O. (2002). Low-energy laser irradiation promotes the survival and cell cycle entry of skeletal muscle satellite cells. J Cell Sci 115, 1461-1469.
Sherer, T. B., Betarbet, R., Testa, C. M., Seo, B. B., Richardson, J. R., Kim, J. H., Miller, G. W., Yagi, T., Matsuno-Yagi, A., and Greenamyre, J. T. (2003). Mechanism of toxicity in rotenone models of Parkinson's disease. J Neurosci 23, 10756-10764.
Smeitink, J. A., van den Heuvel, L. W., Koopman, W. J., Nijtmans, L. G., Ugalde, C., and Willems, P. H. (2004). Cell biological consequences of mitochondrial NADH: ubiquinone oxidoreductase deficiency. Curr Neurovasc Res 1, 29-40.
Sommer, A. P., Pinheiro, A. L., Mester, A. R., Franke, R. P., and Whelan, H. T. (2001). Biostimulatory windows in low-intensity laser activation: lasers, scanners, and NASA's light-emitting diode array system. J Clin Laser Med Surg 19, 29-33.
Stieg, P. E., Sathi, S., Warach, S., Le, D. A., and Lipton, S. A. (1999). Neuroprotection by the NMDA receptor-associated open-channel blocker memantine in a photothrombotic model of cerebral focal ischemia in neonatal rat. Eur J Pharmacol 375, 115-120.
Triepels, R. H., Van Den Heuvel, L. P., Trijbels, J. M., and Smeitink, J. A. (2001). Respiratory chain complex I deficiency. Am J Med Genet 106, 37-45.
Van Breugel, H. H., and Bar, P. R. (1993). He-Ne laser irradiation affects proliferation of cultured rat Schwann cells in a dose-dependent manner. J Neurocytol 22, 185-190.
Whelan, H. T., Buchmann, E. V., Dhokalia, A., Kane, M. P., Whelan, N. T., Wong-Riley, M. T., Eells, J. T., Gould, L. J., Hammamieh, R., Das, R., and Jett, M. (2003). Effect of NASA light-emitting diode irradiation on molecular changes for wound healing in diabetic mice. J Clin Laser Med Surg 21, 67-74.
Whelan, H. T., Connelly, J. F., Hodgson, B. D., Barbeau, L., Post, A. C., Bullard, G., Buchmann, E. V., Kane, M., Whelan, N. T., Warwick, A., and Margolis, D. (2002). NASA light-emitting diodes for the prevention of oral mucositis in pediatric bone marrow transplant patients. J Clin Laser Med Surg 20, 319-324.
Whelan, H. T., Smits, R. L., Jr., Buchman, E. V., Whelan, N. T., Turner, S. G., Margolis, D. A., Cevenini, V., Stinson, H., Ignatius, R., Martin, T., et al. (2001). Effect of NASA light-emitting diode irradiation on wound healing. J Clin Laser Med Surg 19, 305-314.
Wollman, Y., and Rochkind, S. (1998). In vitro cellular processes sprouting in cortex microexplants of adult rat brains induced by low power laser irradiation. Neurol Res 20, 470-472.
Wollman, Y., Rochkind, S., and Simantov, R. (1996). Low power laser irradiation enhances migration and neurite sprouting of cultured rat embryonal brain cells. Neurol Res 18, 467-470.
Wong-Riley MTT, L. H., Eells JT, Chance B, Henry MM, Buchmann E, Kane M, Whelan HT. (2005). Photobiomodulation directly benefits primary neurons functionally inactivated by toxins. J Biol Chem 280, 4761-4771.
Yamanaka, T., Fukumori, Y., Numata, M., and Yamazaki, T. (1988). The variety of molecular properties of bacterial cytochromes containing heme a. Ann N Y Acad Sci 550, 39-46.
Yu, W., Chi, L. H., Naim, J. O., and Lanzafame, R. J. (1997). Improvement of host response to sepsis by photobiomodulation. Lasers Surg Med 21, 262-268.
Yu, W., Naim, J. O., and Lanzafame, R. J. (1994). The effect of laser irradiation on the release of bFGF from 3T3 fibroblasts. Photochem Photobiol 59, 167-170.
Zhang, X., Jones, D., and Gonzalez-Lima, F. (2002). Mouse model of optic neuropathy caused by mitochondrial complex I dysfunction. Neurosci Lett 326, 97-100.
Zhang, X., Rojas, J. C., and Gonzalez-Lima, F. (2006). Methylene blue prevents neurodegeneration caused by rotenone in the retina. Neurotox Res 9, 47-57.