Amyotrophic Lateral Sclerosis

Super Memory Formula

Instant Remedies for Dementia

Get Instant Access

Amyotrophic lateral sclerosis (ALS) is an age-dependent degenerative disorder of motor neurons with sporadic and inherited forms. The disease is characterized by a progressive loss of motor neurons in the spinal cord and brain. The cause of ALS is not known. The incidence of the disease worldwide is 1-2 cases per 100,000. Pathologically, the degeneration of lower motor neurons of the spinal cord and brainstem and loss of large motor neurons in the cerebral cortex are prominent features causing weakness and spasticity (117). The selective vulnerability of motor neurons is relative because peripheral nerves reveal reduced numbers of large myelinated axons, axonal degeneration, and distal axonal atrophy (117).

4.1. Genetics/FALS

Multiple mechanisms have been postulated to be the cause of the sporadic form of the disease, including excitotoxicity (118,119), oxidative injury (120), cytoskeletal abnormalities with aggregates containing SOD1 (121), and autoimmunity (122,123). Up to 70% of sporadic cases have varying loss of the glutatmate transporter, EAAT2 (118,124). Aberrant EAAT2 mRNA species seem to account for the regional selective loss of EAAT2, most likely the result of RNA processing errors (125).

Amyotrophic lateral sclerosis is inherited in about 10% of cases, and of those cases, about 20% have been linked to mutations in sod1, the gene that encodes human copper-zinc superoxide dismutase (Cu,Zn-SOD) (126,127). To date, more than 60 different dominantly inherited point mutations have been found.

No correlation between SOD activity and the frequency or severity of the disease has been demonstrated so far, suggesting that mutant SOD does not cause FALS because of a deficiency in SOD activity. In addition, an elevation of wild-type Cu,Zn-SOD in Down's syndrome causing the symptoms of the disease could be excluded, a lack of correlation of symptoms with overdosage was found in partial trisomy 21 (128,129).

Familial ALS mutations in Cu,Zn SOD are dominant and are currently believed to exert their effects because of a gain of function instead of a loss of function. There is increasing evidence for a gain of toxic function. First, the expression of FALS mutant human Cu,Zn-SOD in transgenic mice causes motor neuron disease, whereas expression of wild-type human Cu,Zn-SOD does not (130132). Second, Cu,Zn-SOD knockout mice do not develop a FALS-like syndrome (133), although female homozygous knockout mice showed a markedly reduced fertility compared with that of wildtype and heterozygous knock-out mice. Further studies revealed that these mice ovulated and conceived normally, but exhibited a marked increase in embryonic lethality (134). Third, human Cu,Zn-SOD is antiapoptotic, whereas the expression of FALS mutant protein can be proapoptotic in cultured cells (135). Fourth, some FALS mutations retain near-normal levels of enzyme activity or stability (136) and mutant SOD1 subunits do not alter the metabolism or activities of wild-type SOD1 in a dominant negative fashion (137). Fifth, poorly or unstably folded mutants are sufficient to cause the formation of SOD1-containing aggregates that are toxic to motor neurons (138,139). Recently, the gain of function has been best exemplified by Estevez and colleagues who showed that an altered Cu(II) coordination by Zn(II)-deficient SOD was responsible for the conversion of the antioxidant activity into a pro-oxidant activity of SOD (140).

4.2. Superoxide Dismutase

The lack of a changed phenotype in SOD1 null mice indicates that the enzyme is probably not necessary for normal development and life, but rather may be involved with other antioxidant enzymes in the response to oxidative stress. This might also be taken as another argument for the appealing hypothesis that SODs are not only enzymes that rapidly dismutate O2- radicals, but that they can also act as metal storage proteins and are otherwise involved in metal-ion homeostasis (141).

A role for Cu,Zn-SOD in the free-radical theory of senescence was provided by the shortened lifespan of Drosophila with a mutational defect in Cu,Zn-SOD, and by Candida elegans, which had only half the normal complement of SOD (142,143). Mice lacking SOD exhibited a shortened life-span (144) but appeared normal while young and were less able to recover from axonal injury (133). A lack of Mn-SOD imposed more serious consequences by shortening the life-span and exhibiting faulty mitochondrial activities in several tissues, especially the heart (145,146). Thus, Mn-SOD may play a more critical role than Cu,Zn-SOD in antioxidant defense mechanisms under normal physiological conditions (147). This is supported by the findings that mice lacking Mn-SOD die at very young age (145,146).

Homodimeric Cu,Zn-SOD acts to disproportionate two molecules of superoxide anion to hydrogen peroxide and water in a reaction mediated through cyclic reduction and oxidation of copper bound to the active site [(148); i.e., reactions (1) and (2)].

O2- + Cu(I)Zn-SOD + 2H+ — H2O2 + Cu(II)Zn-SOD (2)

Cu,Zn-SOD is a cytosolic antioxidant enzyme that lowers concentrations of superoxide by dispropor-tionation to give hydrogen peroxide and dioxygen [i.e., reaction (3)].

The mechanism involves alternate reduction and reoxidation of the copper ion at the active site of the enzyme [i.e., reactions (1) and (2)]. Approximately one-third of the total cellular SOD1 exists as the apoprotein (149). The incorporation of copper into SOD1 in intact cells is absolutely dependent on the presence of the SOD1-specific copper chaperone CCS that belongs to a group of small cytoplasmic proteins involved in copper trafficking (150). A colocalization of CCS and SOD1 in mammalian tissues of the CNS (150) already suggested that CCS can directly insert copper into the enzyme and is active at very low copper concentrations (151). Moreover, there is a physiological requirement for CCS in vivo, and it has been proposed that in the absence of CCS, intracellular SOD1 exists with one Zn(II) bound per SOD1 dimer (152).

Cu,Zn-Superoxide dismutase has also been known for some time to catalyze reactions of hydrogen peroxide with certain substrates in competition with autoinactivation by the enzyme itself [i.e., reaction (6)] (153). The mechanism for this peroxidative activity of Cu,Zn-SOD involves alternate reduction and reoxidation of the copper ion in the enzyme, but this time by hydrogen peroxide [i.e., reactions (4)-(6)].

H2O2 + Cu(II)Zn-SOD — O2- + Cu(I)Zn-SOD + 2H+ (4)

H2O2 + Cu(I)Zn-SOD — OH- + (• OH)Cu(II)Zn-SOD (5)

Spin trap + (• OH)Cu(II)Zn-SOD — Oxidized spin trap +Cu(II)Zn-SOD (6)

Spin-trap reagents in vitro and physiological substrates in vivo are oxidized by hydrogen peroxide in a reaction catalyzed by Cu,Zn-SOD [i.e., reactions (4) and (5)] and are able to inhibit the inactivation of the enzyme (154,155). Three hypotheses have been proposed to account for the possible involvement of copper in the gain of function of the human FALS mutant Cu,Zn-SOD:

• Copper ions in the mutant enzymes could catalyze deleterious oxidation reactions of substrates by hydrogen peroxide (154,155).

• Copper ions in the mutant enzymes could be catalyzing nitration of tyrosine residues by peroxynitrite

• Copper could be leaching out of the mutant proteins and could be inducing toxic effects at another site in the cell.

Most likely, FALS mutations in Cu,Zn-SOD cause a change in the structure of the enzyme that results in an enhancement of the ability of the enzyme-bound copper ion to catalyze destructive oxidative reactions in substrates with peroxides. SOD mutations may destabilize SOD or the dimer and may lead to misfolding (136,159). Misfolding and perturbed dimer formation could lead to aggregation of the protein with a subsequent gain-of-function toxicity (139).

The toxicity of mutant SOD has been proposed to involve an increase in peroxynitrite formation (156), an increase in peroxidase activity (160,161), a loss of shielding of metal ions (162), and aggregation of the enzyme (139). The idea that the FALS mutant Cu,Zn-SOD enzymes may have more accessible active sites is supported by recent crystallographic studies of the G37R mutant (163). In this variant, spinal motor neurons are the most profoundly affected cells, showing axonal and dendritic abnormalities that include SOD1 accumulations in irregularly swollen portions of motor axons, abnormal axonal cytoskeleton architecture, and small vacuoles, derived from damaged mitochondria, in both axons and dendrites (4,162,164-166).

4.3. Deposits

The different mutations can be associated with different types of cellular pathology in mice. The presence of G85R SOD1 had no effect on the level or the activity of wild-type SOD1 (121), but astrocytes contain SOD1 aggregates induced by mutant human G85R SOD1 expression. This is supported by ubiquitin-immunoreactive Lewy-body-like inclusions in astrocytes, which are observed before clinical signs appear (121). At later stages, motor neurons also contain SOD1 and ubiquitin-positive aggregates.

Aggregates of SOD1 are common to disease caused by different mutants, raising the possibility that coaggregation of unidentified components or aberrant catalysis by misfolded mutants may underly mutant-mediated toxicity. Similar SOD1 aggregates have also been reported for some human sporadic and familial ALS cases (167-169). Wild-type protein was a component of aggregates in human disease, but the wild-type protein was not required for aggregation, neither for stabilizing the mutant nor for contributing to aggregates nucleated by mutant SOD1.

4.4. Role of Copper

Although the absence of influence of SOD1 activity on mutant toxicity was supposed to challenge the idea that toxicity derives from oxidative stress arising from superoxide (121), one has to take into consideration that copper may play an important role in the as-yet-unidentified mechanism leading to misfolded aggregated mutants. Administration of the antioxidant vitamin E and selenium (which raises concentrations of the antioxidant enzyme glutathione peroxidase) modestly delays both the onset and progression of disease without affecting survival, although riluzole and gabapentin (two putative inhibitors of the glutamatergic system) do not influence the onset or progression of disease but prolong survival (170). Oral administration of D-penicillamine (a copper chelator) delays the onset of disease (171). Coexpression of the antiapoptotic protein Bcl-2 and even mutant SOD1 extends survival but has no influence on disease progression (172). Expression of two FALS-related mutant

SODs (A4V and V148G) caused cell death in differentiated PC12 cells, superior cervical ganglion neurons, and hippocampal pyramidal neurons. Death could be prevented by Cu chelators, Bcl-2, glutathione, vitamin E, and inhibitors of caspases.

4.5. Conclusion

The extraordinary metabolic activity of motor neurons at least partly resulting from their large surface area and dependence on intracellular transport proteins (173) puts them at a heightened risk of damage from free-radical species that are generated as a result of the mutant SOD expression. Other factors present in the CNS may confer a selective vulnerability to motor neurons.


Age is by far the most important risk factor for dementia, because susceptibility to dementia and Alzheimer's disease increases exponentially with age. However, Alzheimer's disease is clearly distinguishable from normal age-related changes, indicating that it is a specific disease process. The proportion of individuals with AD increases by approx 30-40% with each decade of life to a peak after the age of 80.

5.1. Genetic Background

Three identified genes are involved in the autosomal dominant forms of early-onset AD. Mutations in the gene encoding the amyloid precursor protein (APP) and two homologous genes, presenilin 1 (PS-1) and presenilin 2 (PS-2), result in a dominantly inherited form of the disease that usually begins prior to the age of 60 yr and has been found to alter amyloid p (Ap) protein metabolism. Screening a population-based sample of patients with early-onset AD (<65 yr) revealed APP mutations in 0.5% of the patients, PS-1 mutations in 6%, and PS-2 mutations in 1% (174).

Alzheimer's disease is a multifactorial disease in which several genetic and environmental factors have been implicated. The disease can often result from an interplay of different genetic and environmental risk factors. Familial and sporadic forms of Alzheimer's disease (Fig. 2) occurring later in life have been associated with the e4 polymorphism in the gene encoding apolipoprotein E (APOE). The APOE gene is an important genetic determinant for early-onset AD and for the predominant late-onset form of the disease, which affects 90-95% of all patients. APOE-e4 lowers the typical age of late-onset AD (175) and APOE-e4 also decreases the age at onset of familial AD with an APP mutation (176). Each allele lowers the age of onset by 7-9 years (177).

5.2. Brain Lesions

Alzheimer's disease is characterized pathologically by neuronal degeneration and the deposition of amyloid plaques and neurofibrillary tangles in the brains of affected individuals. Amyloid deposits are primarily composed of the Ap amyloid protein of which the Ap1-40 isoform is the predominant soluble species in biological fluids and the Ap1-42 isoform is the predominant species in developing plaque deposits (178-186). When transgenic mouse models, which overexpress mutant human APP, were immunized with Abeta42, either before the onset of AD-type neuropathologies (at 6 wk of age) or at an older age (11 mo), immunization of the young animals essentially prevented the development of Ap plaque formation (187).

Neuritic plaques and neurofibrillary tangles are the major neuropathological lesions that allow a definitive diagnosis of AD postmortem. Neuritic plaques are spherical multicellular lesions that are usually found in moderate or large numbers in limbic structures and in the association neocortex. They are composed of extracellular deposits of the amyloid Ap protein with degenerating axons and dendrites within and intimately surrounding the amyloid deposit. In regions that are less affected, such as the cerebellum and the thalamus, Ap deposits are of a primarily diffuse type with an amor familial disease genetic factors (mutation)

familial disease genetic factors (mutation)

• sporadic disease environmental toxins disease-modifying genes loss-/ gain-of-f unction

• sporadic disease environmental toxins disease-modifying genes altered protein conformation/ clearance loss-/ gain-of-f unction toxic intermediates active meta I-ions/ reactive oxygen species antioxidant activity/ proteosomal degradation detoxification antioxidant activity/ proteosomal degradation detoxification synaptic dysfunction/ neurotoxicity:

cell death synaptic dysfunction/ neurotoxicity:

cell death i i

Fig. 2. Accumulation of amyloid Ap is invariably associated with the pathology of AD. Overexpression of APP, or disease-modifying genes, shorten the time of onset. Early accumulations of amyloid Ap constitute the first abnormality, possibly representing toxic forms of Ap. If there is no detoxification, extracellular deposition occurs that is central to the pathogenesis of the disease. Vaccination with amyloid Ap can dramatically reduce amyloid deposition in a mouse model of AD.

Fig. 2. Accumulation of amyloid Ap is invariably associated with the pathology of AD. Overexpression of APP, or disease-modifying genes, shorten the time of onset. Early accumulations of amyloid Ap constitute the first abnormality, possibly representing toxic forms of Ap. If there is no detoxification, extracellular deposition occurs that is central to the pathogenesis of the disease. Vaccination with amyloid Ap can dramatically reduce amyloid deposition in a mouse model of AD.

phous and nonfibrillar form of Ap, as found in brains of aged normal humans that often contain Ap deposits.

Neurofibrillary tangle is the other classical lesion that occurs independently in AD. The tangles occur in large numbers in the AD brain, the entorhinal cortex, hippocampus, amygdala, and association cortices of lobes. The subunit protein of the paired helical filaments (PHF) is the tau protein, present in tangles as bundles of paired, helically wound tau aggregates. These intraneuronal proteina-ceous inclusions are often ubiquitinated deposits, similar to those found in other neurodegenerative disorders, such as Parkinson's and Lewy-body disease. Although the tau gene has not been found to be the site of mutations in familial AD, mutations in tau have been discovered in families with fron-totemporal dementia with parkinsonism linked to chromosme 17 (FTDP-17) (17,188,189).

5.3. The Amyloid Precursor Protein

The amyloid precursor protein is a transmembrane glycoprotein that undergoes extensive alternative splicing (190). APP belongs to a multigene family that contains at least two other homologs known as amyloid precursor-like proteins (APLP1 and APLP2) (191-194). APP and APLPs share most of the domains and motifs of APP, including a hydrophobic membrane-spanning domain, N- and O-glycosylation sites, metal-ion-binding domains, and the Kunitz-type protease inhibitor (KPI) domain (not found in APLP1). Only APP contains the Ap region and can be cleaved by p- and y-secretase to generate Ap. Thus, APLPs cannot contribute to Ap deposition in Alzheimer's disease, but they may compensate for the function of APP.

The normal functions of APP and APLPs are not well understood. There exists at least some evidence for neuritic and protective roles (195-198). APP binds Zn(II) at higher nanomolar concentrations (199-

202) and an altered APP metabolism or expression level is believed to result in neurotoxic processes (108,196,203-207). APP can reduce Cu(II) to Cu(I) in a cell-free system, potentially leading to increased oxidative stress in neurons (108). The domain that contributes to such activities is the copper-binding domain (208) residing between residues 135 and 158 of APP, a region that shows strong homolgy to APLP2 but not to APLP1. Five years after the initial identification of the amino acid residues involved in the redox reaction, it has been confirmed by Ruiz et al. that cysteine 144 is a key residue in the reduction of Cu(II) to Cu(I) by soluble APP (108,208,209).

Potentially, APP-Cu(I) complexes may be involved to reduce hydrogen peroxide to form an APP-Cu(II)-hydroxyl radical intermediate (204,205). Recently, it has been discovered that APP residues 135-158 consisting of cysteine and Cu-coordinating histidine residues can modulate copper-mediated lipid peroxidation and neurotoxicity in culture of APP knockout (APP0/0) and wild-type (wt) neurons (206). The wt neurons were found to be more susceptible than the APP0/0 neurons to physiological concentrations of copper but not to other metals. The increased levels of lipid peroxidation products in wt neurons were most likely the result of copper-mediated oxidative stress. Similar effects were obtained with the APP142-166 peptide containing the APP copper-binding domain. The specificity for the increased copper-mediated toxicity by the APP142-166 peptide was shown by the failure of a mutant peptide to bind copper and to potentiate toxicity and the failure of the wt peptide to mediate toxicity from zinc or iron (207). Because an inhibition of copper toxicity was observed in the presence of the copper(I) chelator bathocuproine (BC) in wt cultures of neurons, the increased toxicity in wt neurons could be related to Cu(I) production of APP or secondary reactions of Cu(I), leading to an inbalance of cellular antioxidants.

The amyloid precursor protein and APLP2 are most likely involved in maintaining copper levels in tissues of adult mice (210). APP0/0 mice have significantly increased copper but not zinc or iron levels in the cerebral cortex and liver, compared to age and genetically matched wt mice. APLP20/0 mice also revealed increases in copper in the cerebral cortex and liver. These findings suggest that the APP family can modulate copper homeostasis and that APP/APLP2 expression may be involved in copper efflux from the liver and cerebral cortex (108,210). The fact that APP0/0 but not APLP20/0 neurons revealed increased resistance to copper toxicity in vitro may be explained by the observation that APLP2 has only 60% of the copper redox activity of APP (204,211). The redox activity may be directly linked to toxicity and could be critical for the potential ability of APP and APLP2 to transport copper. This point of view is supported by the finding that Cu(II) reduction modulates copper uptake in eucaryotic cells (212). Most importantly, copper was found to influence APP processing in a cell culture model system when copper was observed to greatly reduce the levels of amyloid Ap peptide and copper also caused an increase in the secretion of the APP ectodomain (213). An increase in intracellular APP levels that paralleled the decrease in Ap generation suggested that additional copper was acting on two distinct regulating mechanisms (i.e., Ap production and the other on APP synthesis) (213).

5.4. APP in Antioxidant Responses

Neurotrophic factors and neuronal injury upregulate APP expression and induce secretion of APP (197,214,215). The addition of sAPP to culture medium protects cortical and hippocampal neurons from neurotoxic insults induced by hypoglycemia and excitatoxic amino acids. APP was reported to act by stabilizing intracellular Ca2+ levels and reducing oxidative stress (216-218). Also overexpression of human APP in cell lines and transgenic mice can result in protection against oxidative stress and resistance to excitotoxicity (215,219). In contrast to these experiments with exogenous sAPP or transfected cell lines, no differences in cell survival were observed in APP0/0 compared with APP+/+ neurons when both were exposed to various oxidative insults (207).

5.5. Metal-Ion Homeostasis in AD

Metal-ion homeostasis is severely dysregulated in AD (220-225). Increased concentrations of copper, iron, and zinc are detected in the neuropil of the AD-affected brain where they are highly

Neuron Copper Transporter

Fig. 3. Possible involvement of copper in Alzheimer's disease. First, APP transport is disturbed by intracellular accumulations of amyloid Ap. Second, the transport of APP to synaptic sites is inhibited followed by a loss of cell surface APP. Third, reinternalization of APP-Cu complexes is impaired, leading to increased levels of copper in the neuropil. Fourth, increased extracellular copper may trigger amyloid Ap aggregation extracel-lularly.

Fig. 3. Possible involvement of copper in Alzheimer's disease. First, APP transport is disturbed by intracellular accumulations of amyloid Ap. Second, the transport of APP to synaptic sites is inhibited followed by a loss of cell surface APP. Third, reinternalization of APP-Cu complexes is impaired, leading to increased levels of copper in the neuropil. Fourth, increased extracellular copper may trigger amyloid Ap aggregation extracel-lularly.

concentrated within amyloid plaques and reach concentrations of up to 0.4 mM (Cu) and 1 mM (Fe and Zn) (223,226,227). A likely reason is that Ap binds equimolar amounts of Cu(II) and Zn(II) at pH 7.4 (199,200,228-230). Human Ap can directly produce hydrogen peroxide by a mechanism that involves the reduction of the metal ions Fe(III) or Cu(II), setting up conditions for Fenton-type chemistry (231).

In the presence of submicromolar concentrations of zinc and copper, synthetic peptides of Ap aggregate into amyloid (230). Evidence that these metal ions may play a role in cerebral amyloid assembly is provided by the observation that Cu/Zn selective chelators enhance solubility of Ap from postmortem brain tissue of AD patients and transgenic mouse brains (232). In addition, by histologi-cal fluorescent techniques, it has been shown that Zn(II) is present in amyloid deposits in human brain (233) and in plaques of transgenic mice (234). Most likely, an abnormal zinc and/or copper homeostasis mediated by APP initiates the Ap deposition followed by the formation and accumulation of Ap that acts as a sink, drawing in metal ions into its mass and away from APP (Fig. 3). This might be further facilitated by the mildly acidic conditions that have been reported in AD brains and that allows copper to completely displace zinc from synthetic Ap aggregates (230,235), reflecting the fact that copper and zinc are biologically antagonistic (236). Under acidic conditions, Api-40 strongly favors copper binding over zinc in the presence of an equimolar amount of both metal ions (237). Also, the affinity of copper for Api-42 is seven orders of magnitude higher than for Api-40, most likely the result of a higher p-sheet content of Ap42 because p-sheet or p-barrel conformations frequently mediate the high-affinity copper binding of cuproproteins (238). Although it remains to be determined whether Ap is at all metal bound in vivo, Ap-mediated cytotoxicity could be shown to be potentiated by Cu(II) and was greatest for the variant Api-42 in vitro (239). Thus, the early unifying hypothesis for free-radical-based neurotoxicity of Ap (240) was recently attributed to metal-catalyzed redox reactions (231,241).

A marked enrichment of copper, iron, and zinc was observed in AD neuropil and amyloid plaques (223). In addition, a 2.2-fold increase of the concentration of copper in the CSF of AD patients was found accompanied by an increase of the copper-transport protein ceruloplasmin (242). A marked accumulation of ceruloplasmin has been described within neurons, astrocytes, and neuritic plaques in AD hippocampus, where it may reflect a response to increased copper levels (243).

5.6. Oxidative Metabolism

Energy metabolism studies showed that glucose-6-phosphate dehydrogenase activities and heme oxygenase-1 (HO-1) levels were increased in AD brains (244,245). The increased G6PDH activity might be the result of the fact that glycolytic enzyme activity is decreased and more NADPH is demanded by detoxifying systems. A considerable number of oxidative stress markers are found in AD brains, such as increased protein carbonyl formation in the inferior parietal lobule and hippocampus (246,247), protein nitration (226,248), lipid peroxidation (240,249-252), DNA oxidation (253,254), and advanced Maillard reaction products (255). Indirect evidence that the AD brain is under oxidative stress can be concluded from reports that treatment of individuals with antioxidants such as vitamin E and selegeline (256) and with anti-inflammatory medications (257-259) delay the progression of the disease.

5.7. Copper and AD Pathology

The possibility that copper may contribute to AD pathology is suggested by the perturbed cerulo-plasmin and copper levels in AD patients (223,243) and the production of free radicals by APP-Cu complexes and increased Aß aggregation in the presence of zinc and/or copper (203,230). Aß can generate ROS from transition metals (260) and deplete neuronal glutathione levels (261,262). On the other hand, redox activities of APP and APLP2 (with APLP2 having only 60% of the copper redox activity) may not only be critical to the ability of the proteins to transport copper but may also be involved in radical production during copper uptake that requires Cu(II) reduction (108).

Was this article helpful?

0 0
All About Alzheimers

All About Alzheimers

The comprehensive new ebook All About Alzheimers puts everything into perspective. Youll gain insight and awareness into the disease. Learn how to maintain the patients emotional health. Discover tactics you can use to deal with constant life changes. Find out how counselors can help, and when they should intervene. Learn safety precautions that can protect you, your family and your loved one. All About Alzheimers will truly empower you.

Get My Free Ebook

Post a comment