Synthetic genistein derivatives as modulators of glycosaminoglycan storage

Background Mucopolysaccharidoses (MPS) are severe metabolic disorders caused by accumulation of undegraded glycosaminoglycans (GAGs) in lysosomes due to defects in certain lysosomal hydrolases. Substrate reduction therapy (SRT) has been proposed as one of potential treatment procedures of MPS. Importantly, small molecules used in such a therapy might potentially cross the blood–brain barrier (BBB) and improve neurological status of patients, as reported for a natural isoflavone, 5, 7-dihydroxy-3- (4-hydroxyphenyl)-4 H-1-benzopyran-4-one, also known as genistein. Although genistein is able to cross BBB to some extent, its delivery to the central nervous system is still relatively poor (below 10% efficiency). Thus, we aimed to develop a set of synthetically modified genistein molecules and characterize physicochemical as well as biological properties of these compounds. Methods Following parameters were determined for the tested synthetic derivatives of genistein: cytotoxicity, effects on cell proliferation, kinetics of GAG synthesis, effects on epidermal growth factor (EGF) receptor’s tyrosine kinase activity, effects on lysosomal storage, potential ability to cross BBB. Results We observed that some synthetic derivatives inhibited GAG synthesis similarly to, or more efficiently than, genistein and were able to reduce lysosomal storage in MPS III fibroblasts. The tested compounds were generally of low cytotoxicity and had minor effects on cell proliferation. Moreover, synthetic derivatives of genistein revealed higher lipophilicity (assessed in silico) than the natural isoflavone. Conclusion Some compounds tested in this study might be promising candidates for further studies on therapeutic agents in MPS types with neurological symptoms.


Background
Mucopolysaccharidoses (MPS) are rare lysosomal storage disorders caused by deficiencies in activities of several different lysosomal hydrolases. Mutations in genes coding for these enzymes lead to defects in degradation of glycosaminoglycans (GAGs) [1,2]. Excessive accumulation of undegraded GAGs in lysosomes causes severe problems in most tissues and organs and usually leads to death in childhood [2]. Currently, only two therapeutic procedures are available for treatment of some of MPS types: bone marrow (or hematopoietic cell) transplantation and enzyme replacement therapy (ERT) [3]. The former procedure is the therapy of choice in MPS I, as it can halt neurocognitive decline when performed early, preferably before the age of 2.5 years [4]. However, efficacy has only been demonstrated in MPS I and MPS VI, and not in MPS III [3,4]. The latter treatment is based on administration of the lacking enzyme, and it is currently available for MPS I, MPS II and MPS VI [5]. Although this treatment is to some extent effective in management of somatic symptoms of the disease, in many MPS types (MPS IH, MPS II, all MPS III subtypes, MPS VII) central nervous system (CNS) is also affected, and ERT seems to be of low efficacy in treatment of neurological symptoms because of the poor delivery of enzyme molecules to CNS across the blood-brain barrier (BBB) [4,6].
Substrate reduction therapy (SRT) is one of putative alternative methods for MPS treatment [6]. A specific kind of SRT for MPS, based on administration of genistein (5, 7-dihydroxy-3-(4-hydroxyphenyl)-4 H-1-benzopyran-4-one also known as 4', 5, 7trihydroxyisoflavone), has been proposed [7]. In vitro, genistein significantly inhibits GAG synthesis and results in a decrease in lysosomal storage in MPS cells [7]. Expression of genes coding for enzymes involved in GAG synthesis might be controlled by signalling pathways dependent on a tyrosine kinase activity of epidermal growth factor receptor (EGFR) [8,9], and genistein has been reported to inhibit this enzymatic activity [10]. In fact, genistein-mediated SRT was reported to act due to inhibition of phosphorylation of EGFR [11] and subsequent putative modulation of gene expression. Therefore, this specific kind of SRT has been named 'gene expression-targeted isoflavone therapy' or GET IT [12][13][14].
Since genistein was reported to cross the BBB to some extent [15], it has been suggested that GET IT might be effective in treatment of neurological symptoms of MPS. In fact, in vivo studies performed with MPS IIIB mouse model revealed a significant reduction of lysosomal storage in liver of MPS IIIB mice treated with genistein for 8 weeks [16], and correction of the abnormal behavior in a long-term (9 month) experiment with high genistein dose (160 mg/kg/day) [17]. Additionally, in both open label and placebo-controlled studies with MPS III patients treated with a genistein-rich soy extract at relatively low doses (5-15 mg/kg/day), somethough only limitedpositive effects on urinary and plasma GAG levels, hair morphology, cognitive functions and behavior were reported [18][19][20][21][22]. This low efficacy of GET IT in clinical trials, which is in contrast to promising results of experiments performed in vitro and with mice, has recently been suggested to be due to low genistein doses in former studies (5-15 mg/kg/day in clinical studies vs. 160 mg/kg/day in animal-based experiments) [14]. Nevertheless, other mechanisms, like limited effects of genistein in human body and/or low efficiency of crossing BBB by this isoflavone (this efficiency was estimated to be below 10% in rats [15]), could not be excluded.
Simultaneously to clinical trials, further laboratory experiments on GET IT have been performed and it was demonstrated that some other natural isoflavones, or even flavonoids, may also cause an inhibition of GAG synthesis and reduction of their accumulation in MPS cells [23,24]. Therefore, one might speculate that chemical modification(s) of genistein might improve either its efficiency in GAG synthesis inhibition or efficiency in crossing BBB. If so, GET IT could be of higher efficacy in MPS patients. In this study, we aimed to test a series of synthetic derivatives of genistein in terms of efficiency of GAG synthesis inhibition and potential ability to cross BBB.

Cytotoxicity and proliferation assay
Cytotoxicity and cell proliferation was assessed using MTT assay. Cells were seed in 96-well plates in a number of 6 x 10 3 cells per well (cytotoxicity assay) or 10 3 cells per well (proliferation assay). After an overnight incubation, growth medium was substituted with medium supplemented with appropriate concentrations of genistein synthetic derivatives or 0.05% DMF as a control and cells were incubated for 24-or 48-hours (cytotoxicity assay) or for 7-days (proliferation assay). Then, medium was substituted with MTT solution (1 mg/ml in RPMI-1640 medium) and following 2-hour incubation at 37°C the amount of a purple formazan product dissolved in DMSO was quantified by measuring the absorbance at 550 nm. LC 50 (cytotoxicity assay) or IC 50 (proliferation assay) index values were determined relative to nontreated cultures (incubated with DMF only).

Measurement of kinetics of GAG synthesis
Cells seed in a number of 2 x 10 4 cells per well (48-well plate) were incubated in growth medium overnight. Then, the medium was substituted with another one, containing appropriate concentrations of genistein synthetic derivatives or 0.05% DMF as a control, and cells were grown for 48 hours. In the next step, the medium was substituted with growth medium without inorganic sulfates (MEM, Joklik's modified) mixed with standard DMEM medium (1:1) supplemented with FBS. GAGs were labeled with 20 μCi/ml of H 2 [ 35 S]O 4 (Hartmann Analytic) for 24 hours. Cells washed with PBS were digested for 3 hours with 0.03% papain (prepared in 100 mM sodium acetate with 5 mM L-cysteine, pH 7.0) (Merck KGaA, Darmstadt, Germany). 35 S incorporation was measured in a scintillation counter and calculated per DNA amount, which was determined in samples with Quant-iT™ PicoGreen W dsDNA Reagent (Molecular Probes, Inc.) according to the manufacturer's protocol.

Cytotoxicity and antiproliferative activity of genistein synthetic derivatives
We have tested 32 synthetic derivatives of genistein in order to establish cytotoxicity and antiproliferative activity in cultured fibroblasts after 24 h exposition to these compounds in a concentration range 1-30 μM. We found toxicity of several synthetic derivatives as low as that of genistein (compounds: IFG-032, IFG-034, IFG-036, IFG-038, IFG-053, IFG-054, IFG-066, IFG-070, IFG-071, IFG-072), while others exhibited higher cytotoxicity (see LC 50 values, Table 2). Additionally, most of the compounds with low cytotoxicity (excluding IFG-053 and IFG-070) exhibited antiproliferative activity similar to or lower than that of genistein (see IC 50 values, Table 2).

Decrease of lysosomal storage of GAGs by some genistein synthetic derivatives
For further studies, we have selected the derivatives which revealed: (i) low cytotoxicity, (ii) antiproliferative activity similar to genistein, and (iii) efficient inhibition of GAG synthesis. Assuming that reduction in GAG synthesis may lead to a decrease in lysosomal storage, as concluded previously [13,14], we assessed the storage in MPS cells by using electron microscopic techniques. The number of different lysosomal structures were counted and calculated per 100 μm 2 of cell crosssection. Examples of abnormal storage structures, observed in MPS IIIA and MPS IIIB cells, are presented in Figure 1.
We have observed a statistically significant decrease in the number of different abnormal lysosomal structures in MPS IIIA or MPS IIIB fibroblasts after 6-day exposure to all selected genistein synthetic derivatives at 30 μM, relative to untreated cells (Table 3). Average size of these structures did not change significantly (data not shown). Interestingly, unusual large vacuolelike structures of unknown function or content, were observed in cells treated with compound IFG-066 ( Figure 1D).

Phosphorylation of EGF receptor in the presence of genistein synthetic derivatives
Reduction of GAG synthesis is assumed to be the result of impaired expression of genes coding for enzymes n/a denotes experiments in which no antiproliferative activity was observed (relative proliferation was over 95% at concentrations ranging from 1 to 30 μM).
-denotes experiments in which IC 50 for antiproliferative activities could not be determined due to cytotoxic effects of the tested compounds.
required for GAG production, and involves an intracellular signaling pathway depending of tyrosine kinase activity of EGFR [13,29]. Since genistein was reported to inhibit the activity of tyrosine kinase of EGFR, and thus to impair production of GAGs, we have tested the ability of some genistein synthetic derivatives (able to inhibit GAG synthesis) to impair phsophorylation of EGFR. We have observed that none of the tested synthetic derivatives  of genistein affects the tyrosine kinase activity of EGF R. Comparing to genistein or a potent tyrosine kinase inhibitor PD168390, no decrease of phosphorylation of EGFR was observed (Figure 2), suggesting that the synthesis of GAGs is reduced by investigated compounds by some other, unknown, mechanism(s).

Computational prediction of BBB penetration
Because genistein was reported to cross BBB only to some extent (several percent efficiency) we aimed to develop new derivatives of genistein with more lipophilic properties that could be able to cross this barrier more readily. Chemical modifications of the genistein molecule resulted, in most of the cases, in an increase of cLogP values, suggesting an increase in lipophilicity (Table 4). Moreover, some of genistein derivatives revealed tPSA values optimal for compounds able to cross BBB (tPSA < 90 Å 2 ) ( Table 4). These results suggest a possible improvement in physicochemical properties of some modified genistein molecules compared to unmodified ones, in the light of efficiency of crossing BBB.

Discussion
An important issue in development of therapeutic approaches for MPS types with neurological symptoms is the ability of potential therapeutic agents to cross BBB. Enzyme replacement therapy, a treatment based on systematic, intravenous administration of the lacking enzyme, although effective -to some extent -in treatment of visceral organs, is of low efficacy in treatment of cases where the central nervous system is affected [1,6]. Alternative therapeutic approaches, such as substrate reduction therapies, are based on assumptions that lowmolecular-weight molecules might be able to cross BBB and penetrate the brain readily [29].
The results presented in this report indicate that some synthetic derivatives of genistein, particularly, IFG-060 and IFG-066, are potent inhibitors of GAG synthesis. Impairment of GAG synthesis by IFG-032, IFG-034, IFG-036, IFG-038, IFG-066, IFG-071 and IFG-072 was also an effective method for reduction of lysosomal storage in MPS IIIA and/or MPS IIIB cell cultures, as it was previously reported for genistein [7]. Studies on MPS IIIB mice suggested that GET IT may be a promising treatment [16,17]. Thus, according to results obtained in this study, we suggest that artificial genistein derivatives listed in Table 3 might be considered as potential drugs to be used in treatment of MPS.
In the development of new therapies, it is crucial for a potential drug to be safe for humans. In this study, some synthetic derivatives of genistein (including the efficient reducers of GAG storage, listed in Table 3) revealed low cytotoxicity and minor effects on cell proliferation. This appears important in the light of safety problems with another effective inhibitor of GAG synthesis, rhodamine B [30,31]. Therefore, it seems that some derivatives of genistein (e.g. IFG-032, IFG-034, IFG-036, IFG-038, IFG-066, IFG-071, IFG-072) possess desirable biological properties for a potentially safe and effective drug. Moreover, predicted changes of physicochemical properties of some synthetic derivatives, relative to genistein (as assessed in silico), might result in improvement of ability to cross BBB (see Table 4). On the other hand, it is necessary to stress that such an improved crossing of BBB was only calculated in silico by using algorithms based on putative physicochemical properties of compounds, predicted from their formulas, according to previously described models [28]. One has to consider that such an approach, although based on solid physical and chemical assumptions, cannot reflect all biological processes, among which a possible active transport of tested compounds may be especially important. Therefore, it should be noted that for determination of actual abilities of penetration of BBB by all compounds described in this report, it will be necessary to perform experiments with either BBB models or (preferably) laboratory animals. Synthesis of labeled isoflavones should be the first step in the way to assess the real (not only calculated or predicted) efficiency of BBB penetration by tested genistein derivatives.
Interestingly, the mechanism of action by which selected synthetic derivatives of genistein inhibit GAG production seems to be different from that described previously for genistein. Namely, contrary to this natural isoflavone, its artificial derivatives did not affect the EGF-dependent pathway, as they were not able to inhibit the EGFR kinase activity. It is worth noting that similar phenomenon was observed for various natural flavonoids causing GAG synthesis inhibition [24]. It is, therefore, tempting to speculate that various chemical modifications of the genistein molecule destroy its activity of the EGFR kinase inhibitor, while either retaining/enhancing or gaining a new function of GAG synthesis inhibitor by influencing another, as yet unidentified, biochemical pathway.
Finally, one should note that apart from genistein derivatives that decreased GAG synthesis, there were also compounds significantly enhancing the efficiency of this process, like IFG-062. Therefore, we assume that the set of artificial genistein derivatives described in this report might be a useful tool in further studies on molecular mechanisms of regulation of GAG synthesis.

Conclusions
Some synthetic derivaties of genistein revealed low cytotoxicity and small (if any) effects on cell proliferation, while slowing down GAG synthesis (though by a pathway other than inhibition of EGF receptor's tyrosine kinase activity) and decreasing lysosomal storage. These compounds had higher potential abilities to cross BBB than genistein. Thus, we suggest they are promising candidates for further studies on therapeutic agents in MPS types with neurological symptoms.

Competing interest
Genistein and its derivatives listed in Table 1