Correction of Huntington’s Disease Phenotype by Genistein-Induced Autophagy in the Cellular Model

Neurodegenerative diseases will pose an increasing burden on society. Most of them are incurable, and despite extensive work and many efforts, no effective treatment can be proposed to a vast majority of patients (Pritchard et al. 2013). Among these diseases, some are caused by accumulation of protein aggregates and resultant loss of neuronal functions. Huntington’s disease (HD) is an example of such disorders, which can also be considered as a model of them due to its well-defined genetic cause (it is a monogenic disease) and pathomechanism at the cellular level (Morreale 2015).

HD is a monogenic disease inherited in an autosomal dominant manner. The genetic defect consists of expansions of CAG repeats in exon 1 of the HTT gene, coding for the huntingtin protein. This results in appearance of long series of glutamine residues in the huntingtin protein, known as the polyQ tract. American College of Medical Genetics and the American Society of Human Genetics recommend following classification of HD, based on the number of CAG repeats: under 27 repeats, unequivocally normal; 27–35 repeats, normal but may expand in future generations; 36–39 repeats, abnormal but associated with reduced penetrance; 40 or more repeats, abnormal with full penetrance. Mutant huntingtin protein tends to form aggregates that are hardly degradable and became a distinctive feature of Huntington’s disease. Its main clinical symptoms can include movement, cognitive and psychiatric disturbances. Unfortunately, the disease is fatal and most of patients die within 15–20 years from diagnosis (Morreale 2015). Currently, there is no effective treatment of Huntington’s disease, and its management focuses only on relieving the symptoms (Shannon and Fraint 2015). There were various approaches to inhibit expression of the mutated HTT allele, inactivation of the mutated huntingtin, and improvement of cell viability (Shannon and Fraint 2015). Nevertheless, decreasing the level of mutated huntingtin is considered as one of the most promising possibilities to find an efficient therapy for HD (Aronin and DiFiglia 2014). Stimulation of protein degradation appears an obvious option for such an approach.

In human cells, there are two potential degradation pathways of proteins, including huntingtin. One of them consists of the proteasomes, where ubiquitin-labeled proteins are directed to the degradation system (so-called ubiquitin–proteasome system or UPS). The second pathway is based on the autophagy process, in which lysosomal hydrolases degrade damaged organelles and macromolecules sequestered by autophagosomes. It was demonstrated that in HD cells, there are inappropriate interactions between ubiquitin and the proteasome complex which results in impairment of UPS (Jana et al. 2001; Bennett et al. 2007; Finkbeiner and Mitra 2008). Contaminations of proteasomal fragments were found in huntingtin aggregates, suggesting that the latter structures might affect UPS (Gil and Rego 2008). Therefore, putative stimulation of these systems appears unlikely as a method for treatment of HD. Instead, it is postulated that UPS may be involved in degradation of normal huntingtin while the mutant form is directed to the lysosomal pathway (Ravikumar et al. 2002, 2004; Bhutani et al. 2007).

There are many known chemicals acting as autophagy stimulators. However, none of them have been successfully used as a therapeutic molecule since they are either toxic or unable to cross the blood–brain-barrier (Gros and Muller 2014; Yang et al. 2013). On the other hand, recent studies performed in our laboratory indicated that genistein (trihydroxyisoflavone or 5, 7-dihydroxy-3- (4-hydroxyphenyl)-4H-1-benzopyran-4-one), a natural isoflavone, stimulates expression of TFEB, a transcription factor which is a master stimulator of lysosomal biogenesis (Moskot et al. 2014). Moreover, genistein-mediated negative regulation of mTOR caused enhanced dephosphorylation and subsequent translocation of TFEB into cell nuclei. Thus, significant increase in lysosomal content and activity was observed in genistein-treated cells (Moskot et al. 2015a). Since genistein can cross the blood–brain-barrier (Tsai 2005), and it was demonstrated to be safe when used for a long time (over 1 year) at concentration as high as 150 mg/kg/day during a clinical trial (Kim et al. 2013), we aimed to test effects of this isoflavone on mutated huntingtin in HEK-293 cells expressing the HTT gene exon 1 with 74 CAG repeats which is a commonly used cellular model of HD.

HEK-293 cell line transfected with plasmids expressing either no HTT gene (pEGFP), wild-type exon 1 of the HTT gene (pEGFP-Q23; with c.54GCA[23]) or mutant exon 1 of the HTT gene (pEGFP-Q74; with c.54GCA[74]) was employed in this study. We have chosen this model as the aim of this work was to assess effects of genistein on mutated huntingtin that accumulates in cells. The plasmid expressing the c.54GCA[74] variant of HTT ensures high level production of huntingtin with a high number of Q repeats. This allows unambiguous determination of the metabolism of the mutated protein and accurate assessment of effects of tested compounds.

Employing the HEK-293 cell line transfected with either vector or wild-type or the mutated HTT gene fragment, we have found that following relatively short (48 h) treatment with genistein, the level of both aggregates and the soluble form of the mutated gene product were significantly decreased relative to untreated cells (DMSO, used as a control), and the effects were within the dose–response correlation (Fig. 1). No significant differences in the levels of wild-type huntingtin (only soluble form, as no aggregates could be detected) were observed at all tested genistein concentrations. Furthermore, microscopic analyses indicated that number and size of mutated huntingtin aggregates decreased considerably under these conditions (Fig. 2). In control experiments, uniform distribution of wild-type huntingtin in cells was observed (Fig. 2). These results indicate that genistein treatment results in a significant reduction in mutated huntingtin in the HD cell model, while having no considerable effects on the level of the wild-type protein.

To test global effects of genistein on HD cells’ physiology, we assessed cell viability by using the MTT assay. Metabolic activity (being an indicator of cell viability in this assay) of HD cells was significantly decreased relative to control cells transfected with the plasmid expressing wild-type HTT allele (Fig. 3a). However, treatment of HD cells with genistein restored their viability (Fig. 3a). This also demonstrated that genistein is not toxic for cells, even at relatively high concentrations. In control experiments, we aimed to demonstrate deleterious effects of a known cytotoxic compound, and H₂O₂ was highly toxic to the cells, as expected. Nevertheless, it was plausible that lower number of HD cells (Fig. 3a) resulted from slower proliferation of cells rather than from lower viability. Therefore, we have performed the Annexin V assay. No increase in a fraction of apoptotic cells treated with genistein was observed in cells expressing wild-type HTT allele (Fig. 3b), indicating again a lack of cytotoxicity of this compound. In the HD model, the fraction of apoptotic cells was significantly higher, over 30%, but decreased considerably after treatment with genistein (Fig. 3b). These results indicate that genistein is not cytotoxic, but has a positive effect not only on the reduction in number and size of aggregates and the level of mutated huntingtin, but also on the whole cell physiology.

To learn whether genistein can induce lysosomal biogenesis and the autophagy process in HD, we monitored the number and size of lysosomes, and the level of the LC3-II protein (the marker of autophagy) in genistein-treated HD and control cells. Addition of this isoflavone to cell cultures caused a significant increase in the amount of LC3-II irrespective of the HTT allele; however, the induction of autophagy was more pronounced in cells expressing the mutated HTT gene (Fig. 4). Stimulation of lysosomal biogenesis was also evident under these conditions in cells expressing c.54GCA[74] or c.54GCA[23] from plasmid pEGFP-Q74 or pEGFP-Q23, respectively (Fig. 5).

We asked whether the autophagy induction is responsible for the observed changes in the level of mutated huntingtin and number of aggregates. Thus, lysosomal functions were partially blocked by the use of low concentration (10 μM) of chloroquine (higher concentrations were highly deleterious to cells, making the results hardly interpretable), and the effects of genistein were assessed. Inhibition of lysosomes (confirmed by an increase in the level of LC3-II (Fig. 6) which could not be degraded due to chloroquine-mediated increase in intralysosomal pH and resultant impairment of lysosomal enzymes’ activities (see Tasdemir et al. 2008, Mizushima et al. 2010)) resulted in more efficient accumulation of mutated huntingtin and less pronounced effects of the isoflavone on this protein (Figs. 6, 7). Chloroquine caused a strong impairment of genistein-mediated stimulation of mHTT degradation at 30 μM concentration of this isoflavone, while it was significantly less pronounced at higher genistein concentrations (Fig. 6). Therefore, we conclude that genistein treatment causes the correction of the HD phenotype in the cell model, predominantly due to induction of the autophagy process, while at higher concentrations of this isoflavone (60 and 100 μM) other, as yet unidentified, process(es) leading to degradation of mutant huntingting might also be activated. Intriguingly, alleviation of genistein-mediated disappearance of mutated huntingtin by chloroquine was more pronounced in experiments with observation of the aggregates under microscope (Fig. 7) than in determination of soluble protein levels by Western blotting (Fig. 6). This suggests that autophagy may be predominant in destruction of the aggregates, while other processes might be additionally involved in degradation of the soluble form, particularly at higher concentrations of genistein.

The results presented in this report indicate that genistein corrects the HD phenotype in cellular model of this disease (Figs. 1, 2, 3), predominantly by activating the autophagy process (Figs. 4, 5, 6, 7). Autophagy appears to be the major, if not the only, process responsible for mutant huntingtin degradation in genistein-treated HD cells at relatively low (30 μM) concentration of this isoflavone (Fig. 6). At higher genistein concentrations (60 and 100 μM), degradation of mutant huntingtin still occurred, though less efficiently, when autophagy was blocked by chloroquine (Fig. 6). On the other hand, inhibition of destruction of huntingtin aggregates by chloroquine was still very efficient even at high genistein levels (Fig. 7). This may suggest that high genistein concentrations might cause activation of other, as yet unidentified, process(es) leading to degradation of soluble mutant huntingtin. However, it is worth noting that it is unlikely to achieve 60 or 100 μM concentrations of genistein in vivo (in animal or human bodies); thus, results obtained with 30 μM genistein appear more relevant in the light of a possible use of this isoflavone as a therapeutic.

Although autophagy can be induced also by other compounds, apart from genistein, a potential drug for treatment of a neurodegenerative genetic disease must cross the blood–brain-barrier, be non-toxic, and be suitable for a long-term administration. Such features (required to be present all together) are not met by vast majority of these substances (Yang et al. 2013). Particularly, long-term induction of autophagy was suggested as a potentially risky therapeutic option due to a probability of destruction of various cell compounds (Yang et al. 2013). Thus, any putative autophagy-stimulating drug should act gently, rather than strongly, to avoid severe adverse effects, which makes the list of such compounds very short if existing at all. On the other hand, genistein has been proposed previously as a potential drug for treatment of some genetic diseases, including cystic fibrosis and mucopolysaccharidoses, though different properties of this isoflavone (correction of the mutated CFTR protein and impairment of glycosaminoglycan synthesis, respectively) were considered (Węgrzyn et al. 2010). Quite surprisingly, another property of genistein, stimulation of lysosomal biogenesis, appeared to be potentially beneficial in treatment of mucopolysaccharidoses, a group of lysosomal storage diseases (Piotrowska et al. 2006; Jakóbkiewicz-Banecka et al. 2009). In fact, genistein was perhaps the first chemical which caused a complete correction of behavior in studies on an animal model of a neurodegenerative metabolic disease (Malinowska et al. 2010). This isoflavone is able to cross the blood–brain-barrier (Tsai 2005), and it was found to be non-toxic to animals even when used at as high concentration as 160 mg/kg/day (Malinowska et al. 2009). Subsequent clinical studies, performed with children, demonstrated its safety in the long-term (over 1 year) use at the high (150 mg/kg/day) dose (Kim et al. 2013). Currently, a phase III, double-blinded, randomized, placebo-controlled clinical trial of high dose oral genistein aglycone in patients with Sanfilippo syndrome (mucopolysaccharidosis type III) is ongoing (EudraCT number: 2013-001479-18). Thus, we suggest that genistein may be a unique candidate for an effective and safe autophagy-inducing drug.

Demonstration that genistein-mediated stimulation of autophagy causes the disappearance of mutated huntingtin aggregates in HD cells (Figs. 1, 6) may open a new possibility of treatment of this disease. Although our results do not exclude a possibility that other pathway(s) is/are stimulated by this isoflavone, particularly at high concentrations, it appears that autophagy is an effective process responsible for reduction in the protein aggregates in HD cells. In fact, autophagy highly predominates in mutant huntingtin degradation at concentration of genistein (30 μM) which is relevant in the light of its potential use as a therapeutic (Fig. 6). Because aggregates of mutant huntingtin are pathogenic factors, and since genistein fulfills all the requirements for a potential drug to treat patients with metabolic neurodegenerative diseases for a long period, it is reasonable to propose further studies on the use of this isoflavone as an anti-HD drug. Although transcripts containing translated CAG repeats play an auxiliary role in pathogenesis of HD (Galka-Marciniak et al. 2012), mutated huntingtin aggregates are the major cause of neurodegeneration (Zhao et al. 2016). Furthermore, it is worth to remind that protein aggregates are also major factors causing other neurodegenerative disorders, including Alzheimer’s disease and Parkinson’s disease (Goloubinoff 2016). Therefore, it is tempting to suggest that genistein-induced autophagy might be effective in reduction of aggregates of β-amyloid, hyper-phosphorylated τ protein, synuclein and/or parkin, which accumulate in these diseases. If this hypothesis is true, development of effective treatments for a large group of severe, neurodegenerative disorders might be possible. No cure is currently available for vast majority of such diseases, and only alleviation of symptoms and palliative care can be offered to patients.

We are aware that the model HD cells used in this work, HEK-293 cells transfected with a plasmid expressing the 1st exon of the HTT gene with 74 CAG repeats, do not represent neurons which are primarily affected in this disease in humans. Nevertheless, this cell line provides a useful model for investigation of various cellular mechanisms of processes that are characteristic for many cell types, including neurons. In fact, the HEK-293 cell line was successfully used in studies on different neurological diseases (Haas et al. 2014; Lecca et al. 2015; Petrosino et al. 2015), including HD (Pryor et al. 2014; Ratovitski et al. 2015; Bañez-Coronel et al. 2015; Sameni et al. 2016). Such studies can provide molecular bases for further investigations on development of specific treatment procedures and specific drugs.

Although genistein improved significantly the HD phenotype in a cellular model of this disease, there were some side effects observed in our experiments. Namely, at higher genistein concentrations (particularly 100 μM), average size of cellular nuclei increased relative to control cells (without genistein). This was observed in experiments with both pEGFP-Q23 and pEGFP-Q74 (Fig. 2), indicating that such a phenomenon occur irrespective of the form of huntingtin. We speculate that this might arise from the recently reported effects of genistein on the cell cycle (Moskot et al. 2015b). Nevertheless, documented safety of high doses of genistein in both animal studies (Malinowska et al. 2010) and clinical trial (Kim et al. 2013) indicates that this phenomenon, observed in vitro, perhaps does not affect physiology of organisms treated with this isoflavone. In fact, a lack of cytotoxic effects of genistein in the experimental system used in this work has been demonstrated in MTT and Annexin V assays (Fig. 3).

One should also note that it was demonstrated recently that in 3-nitropropionic acid (3-NPA)-treated rats, genistein improves memory impairment (Menze et al. 2015) and sensorimotor gating (Menze et al. 2016). 3-NPA is a mitochondrial toxin that causes lesions in the brain which are similar to that found in HD. Therefore, the 3-NPA-induced animal model reflects only secondary effects of HD (increased level of the brain damage, increased level of reactive oxygen species, energetic disturbances); however, it does not induce the primary cause of this disease, i.e., appearance of mutated huntingtin aggregates. Due to these restrictions, studies on genistein action in this model might only indicate its effects on HD symptoms, with no information about the primary cause. On the contrary, we have demonstrated that genistein corrects the HD phenotype thorough stimulation of the autophagy process and reduction of the primary cause of this disease.

In conclusion, results presented in this report, obtained in studies on the cellular model of HD (transfected HEK-293 cells), indicated that treatment with genistein can correct the storage of mutated huntingtin aggregates (Figs. 1, 2) and improve cell viability (Fig. 3). We recognize these results encouraging, as they provide the basis for further experiments on both molecular mechanisms of genistein-mediated correction of HD and development of an efficient therapy for this inherited neurodegenerative disease. Obtaining such indicative results on cellular models of other genetic diseases has been exemplified in research leading to development of promising therapies, including that based on the use of genistein for treatment of another genetic neurodegenerative disorder, Sanfilippo disease, which after original demonstration of the effects in cell cultures (Piotrowska et al. 2006), reached clinical trials (de Ruijter et al. 2012; Kim et al. 2013) that are now at the level of phase III (EudraCT number: 2013-001479-18).