Pharmacological chaperones increase residual β-galactocerebrosidase activity in fibroblasts from Krabbe patients

Molecular Genetics and Metabolism 112 (2014) 294–301

Contents lists available at ScienceDirect

Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme

Pharmacological chaperones increase residual β-galactocerebrosidase activity in fibroblasts from Krabbe patients Anna Sara Berardi a,1, Giovanna Pannuzzo b,1, Adriana Graziano b, Elvira Costantino-Ceccarini a, Paola Piomboni a, Alice Luddi a,⁎ a b

Department of Molecular and Developmental Medicine, University of Siena, Siena, Italy Department of Bio-Medical Sciences, Section of Physiology, University of Catania, Catania, Italy

a r t i c l e

i n f o

Article history: Received 25 February 2014 Received in revised form 15 May 2014 Accepted 15 May 2014 Available online 23 May 2014 Keywords: Krabbe disease Pharmacological chaperones Missense mutations β-Galactocerebrosidase Lysosomal storage disease

a b s t r a c t Krabbe disease or globoid cell leukodystrophy is a degenerative, lysosomal storage disease resulting from the deficiency of β-galactocerebrosidase activity. This enzyme catalyzes the lysosomal hydrolysis of galactocerebroside and psychosine. Krabbe disease is inherited as an autosomal recessive trait, and many of the 70 disease-causing mutations identified in the GALC gene are associated with protein misfolding. Recent studies have shown that enzyme inhibitors can sometimes translocate misfolded polypeptides to their appropriate target organelle bypassing the normal cellular quality control machinery and resulting in enhanced activity. In search for pharmacological chaperones that could rescue the β-galactocerebrosidase activity, we investigated the effect of αLobeline or 3′,4′,7-trihydroxyisoflavone on several patient-derived fibroblast cell lines carrying missense mutations, rather than on transduced cell lines. Incubation of these cell lines with α-lobeline or 3′,4′,7trihydroxyisoflavone leads to an increase of β-galacocerebrosidase activity in p.G553R + p.G553R, in p.E130K + p.N295T and in p.G57S + p.G57S mutant forms over the critical threshold. The low but sustained expression of β-galactocerebrosidase induced by these compounds is a promising result; in fact, it is known that residual enzyme activity of only 15–20% is sufficient for clinical efficacy. The molecular interaction of the two chaperones with βgalactocerebrosidase is also supported by in silico analysis. Collectively, our combined in silico–in vitro approach indicate α-lobeline and 3′,4′,7-trihydroxyisoflavone as two potential pharmacological chaperones for the treatment or improvement of quality of life in selected Krabbe disease patients. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Globoid cell leukodystrophy (GLD), also known as Krabbe disease, is a monogenic lysosomal storage disorder (LSD) inherited as an autosomal recessive trait [1,2]. GLD is characterized by a deficiency in galactocerebrosidase (GALC), a lysosomal enzyme essential for normal catabolism of galactolipids, including a major myelin component, galactocerebroside and psychosine [1]. The characteristic biochemical feature of Krabbe disease is the lack of accumulation of the undegraded galactocerebroside in brain, explained by the early degeneration of the myelin forming cells and the block in the synthesis of galactocerebroside [3,4]. However, GALC deficiency results in abnormal accumulation of psychosine, a toxic metabolite which has been demonstrated to induce apoptotic death in oligodendrocytes and Schwann cells throughout Abbreviations: GALC, β-galactocerebrosidase; GLD, globoid cell leukodistrophy; ERT, enzyme replacement therapy; CNS, central nervous system; PNS, peripheral nervous system; LSDs, lysosomal storage diseases. ⁎ Corresponding author at: Department of Molecular and Developmental Medicine, Policlinico Le Scotte, viale Bracci, 53100 Siena, Italy. E-mail address: [email protected] (A. Luddi). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.ymgme.2014.05.009 1096-7192/© 2014 Elsevier Inc. All rights reserved.

respectively the central nervous system (CNS) and peripheral nervous system (PNS) [5–7]. Loss of these myelin-forming cells causes demyelination in both the CNS and PNS during early developmental stages [8,9]. The biochemical disturbances in GLD are manifested by different degrees of brain demyelination, resulting in a broad neurological spectrum of the disease. The rapidly progressive early infantile form is the most common phenotype, which manifests within the first 6 months of life and is characterized by severe neuro-developmental delay, progressive motor dysfunction, and early death [10]. Late onset forms of GLD represent approximately 10% of cases and include the juvenile and adult onsets that predominantly present with motor weakness, sensory and motor neuropathy, cognitive impairment, and psychiatric/behavior disturbances [5,10]. More than 140 mutations have been identified in patients with all clinical types of GLDs, many of which occur in compound heterozygote patterns in patients [2,11–16]. As with many other lysosomal storage diseases, late-onset Krabbe patients with similar or identical genotypes can have varied clinical presentations and course of their disease [13, 17]. In fact, while some mutations clearly result in the infantile form if homozygous or heterozygous with another severe mutation, for most of the mutations it is difficult to establish a genotype–phenotype

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

correlation. A more detailed understanding of individual GALC mutations must be established to develop a tailored therapy for Krabbe disease, an approach never taken into consideration for this pathology. The only available therapy for Krabbe disease is hematopoietic cell transplant using bone marrow or umbilical cord blood cells from healthy donors. This treatment, if done in pre-symptomatic patients, can prevent the rapid neurological course of the disease and the longterm outcome of transplanted infants [18]. Direct intracerebral injection of vectors or transplantation of enzyme-producing cells in the mouse model of the disease has been demonstrated to be effective in enzyme deficiency correction by promoting donor-to-recipient cross correction via enzyme secretionrecapture [19–21]. However, in humans, safety concerns may represent a limit for this approach. The goal of our study was to assess the therapeutic potential of a pharmacological chaperone therapy approach, a novel and emerging therapeutic strategy using small molecules as drugs for the treatment of Krabbe disease. This therapeutic strategy has the potential to treat diseases caused by missense mutations that result in the synthesis of improperly folded lysosomal enzymes. Indeed, pharmacological chaperones are low-molecular-weight molecules designed to selectively work as a folding template for the conformational mutant enzymes, thereby facilitating proper folding and increasing rescuing of misfolded mutant proteins from the endoplasmic reticulum-associated degradation. Due to their small size, pharmacological chaperones have the potential to be orally available with broad biodistribution, including the CNS. The use of small pharmaceutical chaperones for the treatment of Krabbe disease is based not only on the safety problems and unsatisfactory results obtained with gene therapy, but it is also suggested by in vitro studies done on primary oligodendrocytes demonstrating that over expression of the GALC enzyme is detrimental for oligodendrocytes [19]. It is known that GALC, as other lysosomal enzymes, is constitutively expressed at low levels in all tissues. Our attempt to restore low but sustained expression of the enzyme activity is a promising approach, since it is known that in other lysosomal storage diseases a residual enzyme activity of only 15–20% is sufficient for clinical efficacy [22,23]. In the present study, we screened several pharmacological chaperones and tested them on several fibroblast cell lines from infantile, juvenile and adult Krabbe patients bearing diverse missense mutations. We have identified two potential candidates: α-lobeline already known as a weak inhibitor of GALC [24] and 3′,4′,7-trihydroxyisoflavone a novel chaperone, never previously tested on GALC. The approach used for identification and characterization of the two chaperones was based on biochemical studies in cellular systems and on in silico analysis on the interaction between the compounds and the protein. Our studies clearly show that both compounds selectively increase the residual activity in diverse forms of the defective protein. We hope that this approach will be further explored for the development of novel therapeutic treatment of Krabbe patients with responsive mutations, particularly those with severe brain damage. 2. Experimental procedures 2.1. Materials Fibroblasts from patients with Krabbe disease (p.E130K + p.N295T, p.D187V + p.G323R, p.G286D + p.P318R) were kindly provided by the genetic bio-bank of the Gaslini Institute (Genova, Italy). The fibroblast cell lines carrying the p.G553R + p.G553R and the p.G57S + p.G57S mutation were gift of Prof Balestri (University of Siena) and Prof Federico (University of Siena) respectively. The above GALC mutations are numbered according to HGVS nomenclature recommendations, numbering from the first methionine of the complete 42-residue signal sequence. Since the traditional nomenclature uses p.M17 as the first

295

residue, to switching from new to original nomenclature is required to subtract 16 from their numbers. The α-lobeline and 3′,4′,7-trihydroxyisoflavone were purchased from Santa Cruz Biotechnology. Stock solutions were prepared in DMSO (Sigma Aldrich) at a concentration of 10 mM, filtered sterile and stored at 20 °C. Fluorogenic substrates were purchased from Moscerdam Substrates or from Sigma Aldrich. 2.2. GALC enzymatic activity assay Enzyme activity of cerebroside-ß-galactosidase was measured using the fluorogenic substrate 6-hexadecanoylamino-4-methylumbelliferylß-D-galactoside (HMU-ßGal) as originally described [25]. Preliminary studies comparing natural and HMU-ßGal substrate were carried out in order to test the reliability of the water-soluble substrate for this study. No significant differences in the results were obtained; we have therefore decided to use the water-soluble substrate. Duplicates from each sample, containing 10 μg of total protein from normal control fibroblasts and 20 μl of substrate solution, were mixed on a 200 μl PCR tube and incubated at 37 °C for 17 h. After incubation, 200 μl of stop solution (0.5 M NaHCO3/0.5MNa2CO3 buffer pH 10.7 + 0.25% TritonX-100) was added and mixed. The absorbance of the supernatant was measured (excitation at 404 nm and emission at 460 nm) with an F-4500 fluorescence spectrometer (Hitachi, Tokyo, Japan). 2.3. In vitro inhibition assay on GALC The described GALC enzymatic activity assay was used as the screening method to identify GALC inhibitors. Tested drugs or vehicle DMSO (4% v/v) were added to the mixture solution containing 10 μg of total protein from normal control fibroblasts and 20 μl of substrate solution. The final concentration of the two compounds in the sample was 0.4 mM and each sample was run in triplicate. 2.4. In vitro inhibition assay for other lysosomal enzymes Measurements of β-glucosidase, β-galactosidase, α-glucosidase and α-galactosidase enzymatic activity were performed using cultured cell lysates from primary fibroblasts as previously described [26–29]. 2.5. Compound toxicity in cell culture Normal control fibroblast cells were grown in Dulbecco's modified Eagle medium media (Sigma Aldrich) containing 4.5 g/l glucose, 10% fetal calf serum, stable glutamine and 1% antibiotics and maintained at 37 °C in a humidified CO2 incubator. Cells seeded at 2 × 103 cells/well in a 96-well plate were treated with of α-lobeline or 3′,4′,7trihydroxyisoflavone at increasing concentrations (10 to 600 μM) for 72 h. To determine the cell toxicity of the two compounds cell viability was assessed using XTT assay kit (Sigma Aldrich). This assay is based on the reduction of tetrazolium salts (XTT) to formazan by the succinatetetrazolium reductase system in the mitochondria of metabolically active cells. The assay kit was used according to the manufacturer's instructions. The cells were incubated with the tetrazolium salts for 4 h and absorbance was measured at 450-nm using a 550 Ultramark microtiter plate reader (Biorad, Milan, Italy). The experiments were run in triplicates and the results refer to the mean of the three measurements. 2.6. Cell-based chaperone assay for restoring GALC activity Fibroblast cell lines from patients were grown in Dulbecco's modified Eagle's medium (Sigma Aldrich) and 10% fetal bovine serum (Lonza, Basel, Switzerland) supplemented with 1% penicillin/streptomycin (Lonz,a Basel, Switzerland). Cell lines were cultured at 37 °C in a humidified incubator with 5% CO2. Prior to treatment, cells were cultured for at

296

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

least 12 h to reach exponential growth phase. α-Lobeline and 3′,4′,7trihydroxyisoflavone were dissolved in DMSO and added to the culture medium at various concentrations; control cells were treated with the same amount of DMSO for the same time periods and treated in quadruplicates. After treatment, cells were washed, and then harvested by use of trypsin (Sigma Aldrich) and the cell lysate was assayed for GALC activity as described earlier. 2.7. Structure determination In the present study the BLAST program (Basic Local Alignment Search Tool) [30] was used to model human GALC. Mouse GALC (PDB accession code 3zr5A) (residues 40–684) matched to the sequence and predicted fold of the human GALC with the highest percentage of sequence identity (83%) [31]. The structure of galactose was downloaded from the complex with mouse GALC from online Protein Data Bank (3zr6A), the protein was subsequently removed and the structure was subjected to energy minimization. The structures of α-lobeline and trihydroxyflavone were built as pdb files by means of the server PRODRG2 (http://davapc1. bioch.dundee.ac.uk/prodrg/) [32]. The energies of the protein and the molecules were minimized by the steepest descent algorithm. Simulations were performed by using the GROMACS simulation package (4.5.5) [33] and GROMOS 53A6 [34] force field was used to describe atomistically receptor and ligands respectively. 2.8. Structure analysis All docking calculations were carried out using software AutoDock Vina [35]. For GALC preparation, polar hydrogens were added, and then Kollman united atom charges and atomic solvation parameters were assigned. For ligand preparation, Gasteiger partial charges were added, non-polar hydrogen atoms were merged and rotatable bonds were defined. The grid points and spacing were computed using the Autogrid program [36]. The grid must surround the region of interest in the macromolecule. The spacing between grid points was 1 Å for blind docking and 0.37 Å for the active site, respectively. A preliminary blind docking study was performed in order to discriminate the preferential binding sites of the ligand to the receptor. To this aim, for the first simulation, the grid size was properly set up in order to contain the entire receptor structure up (70, 70, and 78 Å along x, y and z, respectively, space between grid points of 1 Å). Visual Molecular Dynamics (VMD) software was used to visualize assay and post-docking analysis. 2.9. Statistical analysis All data are presented as mean values ± SE of at least triplicate samples from representative experiments. Each experiment was independently repeated at least three times. 3. Results 3.1. Selection of pharmacological chaperones Many efforts are being directed towards exploring the efficacy of pharmacological chaperones as therapeutic compounds in lysosomal diseases. In this study we report the results of a pilot enzymatic inhibition assay in which we screened several small molecules that could be potentially useful for the treatment of Krabbe disease. Among several compounds, various iminosugar analogues were initially screened; however, significant inhibitory activity could not be observed for any of them at the concentration used (400 μM). We then focused on two test compounds: α-lobeline, previously reported to be active in restoring GALC on the D528N mutation despite its weak inhibitory activity on GALC [24], and 3′,4′,7-trihydroxyisoflavone, used for the first time, in this study. Both compounds, tested at the concentration of 400 μM

showed a weak inhibitory activity on GALC but were considered suitable for further investigation as pharmacological chaperones. To test the inhibitory activity of α-lobeline and 3′,4′,7trihydroxyisoflavone on GALC, cell extracts from human control fibroblasts were exposed to increasing concentration (from 0,1 μM to 1 mM) of the two compounds. The results indicated an IC50 of 200 μM for α-lobeline (Fig. 1), confirming the weak inhibitory activity observed in our pilot screening. We were able to detect only a very small inflection for 3′,4′,7-trihydroxyisoflavone in the concentration– response at the highest concentration used (Fig. 1) and we were therefore unable to estimate an IC50 for this compound. 3.2. Enzyme specificity To examine the specificity of α-lobeline and 3′,4′,7trihydroxyisoflavone, the inhibitory activity on normal control fibroblasts was tested for other lysosomal enzymes hydrolyzing different glycosidic linkages and/or sugar moieties, namely β-glucosidase, β-galactosidase, α-glucosidase and α-galactosidase. The concentration used was 400 μM for both compounds in all different enzyme assays. All tested enzymes were not inhibited by α-lobeline, confirming the specificity of this compound for GALC. The 3′,4′,7-trihydroxyisoflavone had no inhibitory activity on β-glucosidase and α-glucosidase, but showed very low inhibitory activity (15%) on α-galactosidase and, as expected, higher inhibitory activity (40%) on β-galactosidase; in fact, it is indicated as an inhibitor of β galactosidase [37]. 3.3. Cell toxicity assay To assess cellular toxicity of the two molecules, we next studied the effect of test chaperones on the viability of normal control fibroblasts. The cells were incubated with increasing concentrations of each compound. In all cases, the cells well tolerated concentration of the compounds from 10 to 200 μM, while at higher concentrations (400 and 600 μM) cell viability was significantly reduced (Fig. 2). These data show that these compounds have low cell toxicity, suggesting that they may have a distinct advantage with regard to safety of administration, which makes them good candidates for further drug development. 3.4. In silico analysis of molecular interaction between GALC and α-lobeline or 3′,4′,7-trihydroxyisoflavone The molecular docking analysis of the ligands to their macromolecular receptor (GALC) indicated that α-lobeline and 3′,4′,7trihydroxyisoflavone can bind to GALC in 9 different sites with an energy of binding (ΔG) ranging from − 7.3 to − 6.8 for α-lobeline and from − 7.8 to − 7.0 for 3′,4′,7-trihydroxyisoflavone, while the ΔG of the binding for the natural substrate is −6.4. Interestingly, only one pose of all potential binding sites for both α-lobeline and 3′,4′,7trihydroxyisoflavone overlaps to the binding site of the natural substrate (Fig. 3A–B), but this pose has lower ΔG, so it is disadvantaged in relation to others, as inferred by the weak inhibitory activity measured in the enzyme assay. The bonding interaction between GALC and galactose is shown in panel C (Fig. 3). Our in silico analysis suggests that both compounds interact with GALC and can be, therefore, considered good potential candidates as chaperones. 3.5. Pharmacological chaperoning activity on fibroblast cell cultures Previous studies have used, to test potential chaperones, COS-1 cells over-expressing GALC bearing different human mutations [24]. In this study, being the chaperone activity mutation specific, we aimed at identifying new missense mutant forms of GALC responsive to α-lobeline and/or to 3′,4′,7-trihydroxyisoflavone. The residual enzyme activity was determined in fibroblast cell lines, carrying different missense mutations.

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

297

18 16

GALC activity

14 12 10 8 6 4 2 0 untreated

0.1 µM

1 µM

10 µM

100 µM

200 µM

500 µM

1 mM

Fig. 1. In vitro inhibition of the normal control human GALC. IC50 values were determined by increasing the inhibitor concentrations. Assays were performed with cell extracts in 0.1 M citrate buffer (pH 5.2), using 6-hexadecanoylamino-4-methylumbelliferyl-ß-D-galactoside as substrate. Black bar: untreated; gray bars, α-lobeline treated; open bars: 3′,4′,7trihydroxyisoflavone treated.

The mutations analyzed are the following: p.G553R, p.E130K + p.N295T, p.G57S, p.D187V + p.G323R and p.G286D + p.P318R.

3.5.1. p.G553R + p.G553R The fibroblast cell line we have analyzed was derived from a patient affected by the infantile form of Krabbe disease bearing p.G553R mutation in homozygosis; the residual enzymatic activity reported in association with this mutation is very low [38]. In silico analysis indicates that this mutation induces severe misfolding of GALC [39]. To assess if the selected chaperones could rescue GALC activity, fibroblasts were treated with α-lobeline and 3′,4′,7-trihydroxyisoflavone at the concentration of 50, 100 and 200 μM for 72 h. The treatment with α-lobeline at lower concentrations (50–100 μM) significantly raised (two folds) the enzyme activity, while at the highest concentration used the enzyme activity returned to the level of untreated cells (Fig. 4). In the 3′,4′,7-trihydroxyisoflavone treated cells, a steady increase in the GALC activity was observed with increasing concentration of the test compound (Fig. 4), with the highest activity reached (6 nmol/17 h/mg protein) at 200 μM. This value, corresponding to about 30% of the mean value of normal fibroblasts (control range 16–42, n = 17), is well over the 15–20% needed to prevent accumulation of storage material [22]. These data clearly show the efficacy of 3′,4′,7-trihydroxyisoflavone in restoring GALC activity in this mutation.

3.5.2. p.E130K + p.N295T Treatment with α-lobeline or whith 3′,4′,7-trihydroxyisoflavone was also successful in increasing GALC activity in a fibroblast line bearing the missense mutations p.E130K + p.N295T. When combined these mutations cause the infantile form of Krabbe disease. As for the p.G553R mutation, the p.E130K induces severe misfolding, while the p.N295T induces loss of stabilizing hydrogen bonds [39]; both mutant forms change the structural conformation of GALC. In these fibroblasts the treatment with α-lobeline or 3′,4′,7-trihydroxyisoflavone induced an increase of GALC activity to similar levels (Fig. 5). The concentration range needed to increase by four folds the enzyme activity was between 50 μM and 100 μM of both compounds; when the concentration of both compounds was increased to 200 μM the specific activity decreased significantly, although it remained higher than that of untreated cells. It is not possible, at this time, to determine if one of the two mutations is responsive to both inhibitors or which of the two mutant forms is responsive to which compound. 3.5.3. p.G57S + p.G57S We have focused our efforts at testing the effectiveness of the selected chaperones toward the p.G57S + p.G57S mutation causing a late onset form of Krabbe disease. This mutation is present with high incidence in a restricted geographical area of southern Italy [15]. Finding a pharmacological chaperone active in restoring GALC activity in these patients can be considered an important step forward the

160

Cell viability (% of survival)

Cell viability (% of survival)

160 140 120 100 80 60 40 20 0

10 µM 100 µM 200 µM 400 µM 600 µM untreated

α-Lobeline

140 120 100 80 60 40 20 0 10 µM 100 µM 200 µM 400 µM 600 µM untreated

3’,4’,7-Trihydroxyisoflavone

Fig. 2. Viability of normal control fibroblast cell-line in the presence of vehicle, α-lobeline or 3′,4′,7-trihydroxyisoflavone for 72 h. Cell viability was assessed by XTT assay. The values are the mean of three reactions +SD.

298

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

Fig. 3. Ribbon diagram of GALC complexed with compounds α-lobeline (panel A) and 3′,4′,7-trihydroxyisoflavone (panel B) showing the comparison of galactose binding site relative to α-lobeline and 3′,4′,7-trihydroxyisoflavone binding sites. The atoms of galactose are represented in spheres, α-lobeline and 3′,4′,7-trihydroxyisoflavone are showed as coloured sticks. The amino acids of galactose interacting with the GALC protein moiety are presented in a close-up view. Hydrogen bonds are represented as green cylinders (panel C).

amelioration/treatment of these patients. Consistent with late onset of the disease, fibroblasts from these patients have a residual enzyme activity higher than that measured in the fibroblasts of the infantile forms analyzed above.

A significant increase of GAL activity was measured in fibroblast from these patients with both α-lobeline or 3′,4′,7-trihydroxyisoflavone. The two fold increase measured for both compounds, although attained at different concentrations (100 μM α-lobeline and 200 μM 3′,4′,7-

GALC activity (nmol/17h/mg proteine)

8 7 6 5 4 3 2 1 0 50 µM untreated

100 µM

α-Lobeline

200 µM

50 µM

100 µM

200 µM

3’,4’,7-Trihydroxyisoflavone

Fig. 4. GALC activity in fibroblasts from Krabbe patient, homozygous for the missense mutation p.G553R + p.G553R. GALC enzymatic activity is improved in the presence of the compounds. Fibroblasts were incubated with the increasing concentrations of α-lobeline or 3′,4′,7-trihydroxyisoflavone for 72 h. The values are the mean of three independent experiments ± SD.

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

299

GALC activity (nmol/17h/mg proteine)

0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 50 µM

100 µM

200 µM

α-Lobeline

untreated

50 µM

100 µM

200 µM

3’,4’,7-Trihydroxyisoflavone

Fig. 5. GALC activity in fibroblasts from Krabbe patient, bearing missense mutations p.E130K + p.N295T. Fibroblasts were incubated with increasing concentrations of α-lobeline or 3′,4′,7trihydroxyisoflavone for 72 h. GALC enzymatic activity is improved in the presence of both compounds at the lowest concentration used. At the highest concentration of both compounds the specific activity decreased significantly but remained higher than untreated cells. The values are the mean of three independent experiments ±SD.

trihydroxyisoflavone) is equal to 15% of normal control values (range 16–42, n = 17) (Fig. 6) and in the range expected to prevent accumulation of storage material [22].

treated with α-lobeline or 3′,4′,7-trihydroxyisoflavone, no changes in GALC specific activity were measured at all concentration used. 4. Discussion

3.5.4. p.D187V + p.G323R and p.G286D + p.P318R We also investigated the effect of α-lobeline and 3′,4′,7trihydroxyisoflavone on two fibroblast cell lines from heterozygote compound patients; these mutations do not affect the folding of the protein [39]. The G286D + p.P318R mutation is associated with the adult onset; G286D introduces an acidic side chain, while the P318R introduces a hydrophobic side chain near the substrate binding site. The D187V + p.G323R is associated to the late infantile form of Krabbe disease. It has been reported that D187V mutation causes the substitution of an acidic side chain to a smaller hydrophobic side chain, determining a change of the local surface charge. For the G323R no in silico hypothesis has been published. When the fibroblast of these patients were

The search for new strategies to optimize therapies and to target the different cellular and phenotypic aspects is becoming an important goal of current research due to the limitations of the therapeutic approaches available to treat LSDs. The several attempts and strategies, made to correct the enzyme activity in brain of neuronopathic LSDs, are still a major therapeutic goal. More than 70 GALC mutations causing Krabbe disease have been reported, and many of them cause amino acid substitutions (Human Gene Mutation Database; http://www.hgmd.org). As with several other LSDs, Krabbe patients with similar or identical genotypes can have different clinical presentations and course of their disease [17]; thus, the development of a tailored therapy should represent a successful strategy for the treatment of this patients. Recent studies indicated

GALC activity (nmol/17h/mg proteine)

3.5 3 2.5 2 1.5 1 0.5 0 50 µM untreated

100 µM

α-Lobeline

200 µM

50 µM

100 µM

200 µM

3’,4’,7-Trihydroxyisoflavone

Fig. 6. GALC activity in fibroblasts from Krabbe patient homozygous for the missense mutation p.G57S + p.G57S. Fibroblasts were incubated with the increasing concentrations of αlobeline or 3′,4′,7-trihydroxyisoflavone for 72 h. Both compounds significantly increase the GALC activity. The values are the mean of three independent experiments ±SD.

300

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

the pharmacological or molecular chaperone therapy as a potential targeted therapeutic approach for LSDs caused by missense mutations [40]. The study presented shows the identification of two compounds that selectively rescue GALC activity in fibroblasts derived from patient bearing missense mutations causing infantile, late infantile or adult forms of Krabbe disease. In this study only fibroblast from Krabbe patients were used rather than transduced cell lines as reported in the majority of the studies on chaperones [24,41]. This approach is preferable when measuring the functional activity of lysosomal enzymes involved in lipid degradation because these cell extracts contain all of the necessary co-factors, i.e. saposins, for enzyme activation in the natural environment. Our results show that α-lobeline has a good chaperoning activity on GALC mutations other than those previously reported [24]. Moreover, the 3′,4′,7-trihydroxyisoflavone has been tested for the first time for chaperoning activity and our data demonstrated that this compound can efficiently rescue GALC activity in 4 different missense mutations. In silico analysis suggests that the best-ranked poses of α-lobeline and 3′,4′,7-trihydroxyisoflavone interact with GALC with a binding energy higher than that of the natural substrate. The modeling data are supported by the in vitro results showing that α-lobeline is a weak inhibitor and that 3′,4′,7-trihydroxyisoflavone affects GALC activity only at very high concentrations. Most of the chaperones so far identified for the treatment of LSDs are active site-directed molecules and are reversible competitive inhibitors of the target enzymes. This issue has raised concerns and stimulated the identification of a new generation of chaperones able to protect the enzymes from degradation without interfering with its activity [42–44]. We believe that the compounds we have identified are related to this new generation of chaperones making them suitable for further exploration as pharmacological chaperones. In addition, our data show that, at the concentration used, these compounds are relatively not toxic and do not inhibit other lysosomal enzymes with closely related activities, confirming that these compounds could have a role in the therapy of Krabbe patients. Moreover their potential therapeutic role for diseases with CNS involvement is also supported by the demonstrated capacity of α-lobeline and 3′,4′,7trihydroxyisoflavone to cross the blood brain barrier [45–47], and it can be hypothesized that they penetrate cell membranes and achieve therapeutic concentrations in specific cell compartments. The efficacy of α-lobeline and 3′,4′,7-trihydroxyisoflavone in rescuing GALC activity has been demonstrated in our in vitro studies. In fibroblasts carrying the p.G553R mutations both compounds were helpful, but the 3′,4′,7-trihydroxyisoflavone was found to be more effective than α-lobeline in increasing GALC activity. This cell line was previously used in a different study for the selection of small molecules from a library of pharmacological active compounds, but none of the examined compounds was active in increasing the residual GALC activity [41]. We were particularly interested in analyzing the efficacy of αlobeline and 3′,4′,7-trihydroxyisoflavone on p.G57S because of the high incidence of this mutation in a restricted geographical area of Southern Italy [15]. The development of a tailored chaperone therapy could benefit the patients carrying this mutation. In this case both chaperones were able to rescue GALC activity to the same extent (15% of normal values); however, lower concentrations of α-lobeline were needed to reach this value. It has been demonstrated that a threshold activity of approximately 10% is sufficient to prevent storage in LSDs [23]. Thus, the increase to 15% of normal activity induced by α-lobeline and 3′,4′,7trihydroxyisoflavone is likely to have an impact on disease pathology and to be beneficial for these Krabbe patients. The increase of the GALC activity over the critical threshold is significant from a therapeutic aspect, and in fact, is sufficient to deplete and prevent storage of the substrate. However, even if the increase in enzyme activity remains below the critical threshold, the rescued activity could still be sufficient to induce clinical benefit and ameliorate the patients' quality of life.

5. Conclusion In summary, our combined in silico–in vitro approach allowed for the successful identification of two pharmacological chaperones, α-lobeline and 3′,4′,7-trihydroxyisoflavone, for the treatment of Krabbe disease. Thus, further studies will have to evaluate whether these compounds act similarly on other mutations, and preclinical studies will indicate if these compounds are amenable for development of an individualized therapy for Krabbe disease. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Acknowledgments This study was supported by funds to Alice Luddi from Tuscany Region “Bando salute 2009” n°158. Three of the cell lines used from patients affected by Krabbe disease were obtained from the G. Gaslini Institute–Telethon Genetic Biobank Network (project GTB07001). One of the cell lines used from patients affected by Krabbe disease was obtained from Prof Paolo Balestri, University of Siena. One of the cell lines used from patients affected by Krabbe disease was obtained from Prof Antonio Federico, University of Siena. We are also thankful to Vera Cardile for the kind collaboration. References [1] K. Suzuki, Y. Suzuki, Globoid cell leucodystrophy (Krabbe's disease): deficiency of galactocerebroside beta-galactosidase, Proc. Natl. Acad. Sci. U. S. A. 66 (1970) 302–309. [2] D.A. Wenger, M.A. Rafi, P. Luzi, Molecular genetics of Krabbe disease (globoid cell leukodystrophy): diagnostic and clinical implications, Hum. Mutat. 10 (1997) 268–279, http://dx.doi.org/10.1002/(SICI)1098-1004(1997)10:4b268::AID-HUMU2N3.0.CO;2-D ([pii] 1002/(SICI)1098–1004(1997)10:4b268::AID-HUMU2N3.0.CO;2-D). [3] M.T. Vanier, L. Svennerholm, Chemical pathology of Krabbe's disease. III. Ceramidehexosides and gangliosides of brain, Acta Paediatr. Scand. 64 (1975) 641–648. [4] L. Svennerholm, M.T. Vanier, J.E. Mansson, Krabbe disease: a galactosylsphingosine (psychosine) lipidosis, J. Lipid Res. 21 (1980) 53–64. [5] D.J. Costello, A.F. Eichler, F.S. Eichler, Leukodystrophies: classification, diagnosis, and treatment, Neurologist 15 (2009) 319–328, http://dx.doi.org/10.1097/NRL. 0b013e3181b287c8 (00127893-200911000-00004 [pii]). [6] H. Nagara, H. Ogawa, Y. Sato, T. Kobayashi, K. Suzuki, The twitcher mouse: degeneration of oligodendrocytes in vitro, Brain Res. 391 (1986) 79–84. [7] T. Taketomi, K. Nishimura, Physiological activity of psychosine, Jpn. J. Exp. Med. 34 (1964) 255–265. [8] F. Seitelberger, Demyelination and leukodystrophy at an early age, Bol. Estud. Med. Biol. 31 (1981) 373–382. [9] T. Kobayashi, I. Goto, T. Yamanaka, Y. Suzuki, T. Nakano, K. Suzuki, Infantile and fetal globoid cell leukodystrophy: analysis of galactosylceramide and galactosylsphingosine, Ann. Neurol. 24 (1988) 517–522, http://dx.doi.org/10. 1002/ana.410240407. [10] J. Aicardi, The inherited leukodystrophies: a clinical overview, J. Inherit. Metab. Dis. 16 (1993) 733–743. [11] R. De Gasperi, M.A. Gama Sosa, E.L. Sartorato, S. Battistini, H. MacFarlane, J.F. Gusella, W. Krivit, E.H. Kolodny, Molecular heterogeneity of late-onset forms of globoid-cell leukodystrophy, Am. J. Hum. Genet. 59 (1996) 1233–1242. [12] H. Furuya, Y. Kukita, S. Nagano, Y. Sakai, Y. Yamashita, H. Fukuyama, Y. Inatomi, Y. Saito, R. Koike, S. Tsuji, et al., Adult onset globoid cell leukodystrophy (Krabbe disease): analysis of galactosylceramidase cDNA from four Japanese patients, Hum. Genet. 100 (1997) 450–456. [13] L. Fu, K. Inui, T. Nishigaki, N. Tatsumi, H. Tsukamoto, C. Kokubu, T. Muramatsu, S. Okada, Molecular heterogeneity of Krabbe disease, J. Inherit. Metab. Dis. 22 (1999) 155–162. [14] C. Xu, N. Sakai, M. Taniike, K. Inui, K. Ozono, Six novel mutations detected in the GALC gene in 17 Japanese patients with Krabbe disease, and new genotypephenotype correlation, J. Hum. Genet. 51 (2006) 548–554, http://dx.doi.org/10. 1007/s10038-006-0396-3. [15] W. Lissens, A. Arena, S. Seneca, M. Rafi, G. Sorge, I. Liebaers, D. Wenger, A. Fiumara, A single mutation in the GALC gene is responsible for the majority of late onset Krabbe disease patients in the Catania (Sicily, Italy) region, Hum. Mutat. 28 (2007) 742, http://dx.doi.org/10.1002/humu.9500. [16] D.A. Wenger, M.L. Escolar, P. Luzi, M.A. Rafi, Scriver's The Online Metabolic and Molecular Bases of Inherited Disease (OMMBID). Chapter 147 Krabbe Disease (Globoid Cell Leukodystrophy), Available at: http://www.ommbid.com/OMMBID/the_

A.S. Berardi et al. / Molecular Genetics and Metabolism 112 (2014) 294–301

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26] [27]

[28]

[29] [30]

[31]

[32]

online_metabolic_and_molecular_bases_of_inherited_disease/b/abstract/Part16/ ch147 2013 ([accessed 28 Oct 2013]). A. Fiumara, R. Barone, A. Arena, M. Filocamo, W. Lissens, L. Pavone, G. Sorge, Krabbe leukodystrophy in a selected population with high rate of late onset forms: longer survival linked to c.121GNA (p.Gly41Ser) mutation, Clin. Genet. 80 (2011) 452–458, http://dx.doi.org/10.1111/j.1399-0004.2010.01572.x. P.K. Duffner, V.S. Caviness Jr., R.W. Erbe, M.C. Patterson, K.R. Schultz, D.A. Wenger, C. Whitley, The long-term outcomes of presymptomatic infants transplanted for Krabbe disease: report of the workshop held on July 11 and 12, 2008, Holiday Valley, New York, Genet. Med. 11 (2009) 450–454, http://dx.doi.org/10.1097/GIM. 0b013e3181a16e04. E. Costantino-Ceccarini, A. Luddi, M. Volterrani, M. Strazza, M.A. Rafi, D.A. Wenger, Transduction of cultured oligodendrocytes from normal and twitcher mice by a retroviral vector containing human galactocerebrosidase (GALC) cDNA, Neurochem. Res. 24 (1999) 287–293. A. Luddi, M. Volterrani, M. Strazza, A. Smorlesi, M.A. Rafi, J. Datto, D.A. Wenger, E. Costantino-Ceccarini, Retrovirus-mediated gene transfer and galactocerebrosidase uptake into twitcher glial cells results in appropriate localization and phenotype correction, Neurobiol. Dis. 8 (2001) 600–610, http://dx.doi.org/10.1006/nbdi.2001. 0407 (S0969-9961(01)90407-3 [pii]). M.A. Rafi, H.Z. Rao, P. Luzi, M.T. Curtis, D.A. Wenger, Extended normal life after AAVrh10-mediated gene therapy in the mouse model of Krabbe disease, Mol. Ther. 20 (2012) 2031–2042, http://dx.doi.org/10.1038/mt.2012.153. P. Leinekugel, S. Michel, E. Conzelmann, K. Sandhoff, Quantitative correlation between the residual activity of beta-hexosaminidase A and arylsulfatase A and the severity of the resulting lysosomal storage disease, Hum. Genet. 88 (1992) 513–523. U.H. Schueler, T. Kolter, C.R. Kaneski, G.C. Zirzow, K. Sandhoff, R.O. Brady, Correlation between enzyme activity and substrate storage in a cell culture model system for Gaucher disease, J. Inherit. Metab. Dis. 27 (2004) 649–658. W.C. Lee, D. Kang, E. Causevic, A.R. Herdt, E.A. Eckman, C.B. Eckman, Molecular characterization of mutations that cause globoid cell leukodystrophy and pharmacological rescue using small molecule chemical chaperones, J. Neurosci. 30 (2010) 5489–5497, http://dx.doi.org/10.1523/JNEUROSCI.6383-09.2010 (30/16/5489 [pii]). G. Wiederschain, S. Raghavan, E. Kolodny, Characterization of 6-hexadecanoylamino4-methylumbelliferyl-beta-D- galactopyranoside as fluorogenic substrate of galactocerebrosidase for the diagnosis of Krabbe disease, Clin. Chim. Acta 205 (1992) 87–96. D. Robinson, The fluorimetric determination of beta-glucosidase: its occurrence in the tissues of animals, including insects, Biochem. J. 63 (1956) 39–44. R. Kwapiszewski, J. Szczudlowska, K. Kwapiszewska, M. Chudy, Z. Brzozka, Determination of acid beta-galactosidase activity: methodology and perspectives, Indian J. Clin. Biochem. 29 (2014) 57–62, http://dx.doi.org/10.1007/s12291-013-0318-z (318 [pii]). K. Umapathysivam, J.J. Hopwood, P.J. Meikle, Determination of acid alphaglucosidase activity in blood spots as a diagnostic test for Pompe disease, Clin. Chem. 47 (2001) 1378–1383. N.A. Chamoles, M. Blanco, D. Gaggioli, Fabry disease: enzymatic diagnosis in dried blood spots on filter paper, Clin. Chim. Acta 308 (2001) 195–196. S.F. Altschul, W. Gish, W. Miller, E.W. Myers, D.J. Lipman, Basic local alignment search tool, J. Mol. Biol. 215 (1990) 403–410, http://dx.doi.org/10.1016/S00222836(05)80360-2 (S0022-2836(05)80360-2 [pii]). J. Kopp, T. Schwede, The SWISS-MODEL repository of annotated three-dimensional protein structure homology models, Nucleic Acids Res. 32 (2004) D230–D234, http://dx.doi.org/10.1093/nar/gkh008 (32/suppl_1/D230 [pii]). A.W. Schuttelkopf, D.M. van Aalten, PRODRG: a tool for high-throughput crystallography of protein–ligand complexes, Acta Crystallogr. D Biol. Crystallogr. 60 (2004) 1355–1363, http://dx.doi.org/10.1107/S0907444904011679 (S0907444904011679 [pii]).

301

[33] S. Pronk, S. Pall, R. Schulz, P. Larsson, P. Bjelkmar, R. Apostolov, M.R. Shirts, J.C. Smith, P.M. Kasson, D. van der Spoel, et al., GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit, Bioinformatics 29 (2013) 845–854, http://dx.doi.org/10.1093/bioinformatics/btt055 (btt055 [pii]). [34] C. Oostenbrink, T.A. Soares, N.F. van der Vegt, W.F. van Gunsteren, Validation of the 53A6 GROMOS force field, Eur. Biophys. J. 34 (2005) 273–284, http://dx.doi.org/10. 1007/s00249-004-0448-6. [35] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading, J. Comput. Chem. 31 (2010) 455–461, http://dx.doi.org/10.1002/jcc.21334. [36] G.M. Morris, L.G. Green, Z. Radic, P. Taylor, K.B. Sharpless, A.J. Olson, F. Grynszpan, Automated docking with protein flexibility in the design of femtomolar “click chemistry” inhibitors of acetylcholinesterase, J. Chem. Inf. Model. 53 (2013) 898–906, http://dx.doi.org/10.1021/ci300545a. [37] L. Governini, P. Carrarelli, A.L. Rocha, V.D. Leo, A. Luddi, F. Arcuri, P. Piomboni, C. Chapron, L.M. Bilezikjian, F. Petraglia, FOXL2 in human endometrium: hyperexpressed in endometriosis, Reprod. Sci. (2014), http://dx.doi.org/10.1177/1933719114522549 (1933719114522549 [pii]). [38] B. Tappino, R. Biancheri, M. Mort, S. Regis, F. Corsolini, A. Rossi, M. Stroppiano, S. Lualdi, A. Fiumara, B. Bembi, et al., Identification and characterization of 15 novel GALC gene mutations causing Krabbe disease, Hum. Mutat. 31 (2010) E1894–E1914, http://dx. doi.org/10.1002/humu.21367. [39] J.E. Deane, S.C. Graham, N.N. Kim, P.E. Stein, R. McNair, M.B. Cachon-Gonzalez, T.M. Cox, R.J. Read, Insights into Krabbe disease from structures of galactocerebrosidase, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 15169–15173, http://dx.doi.org/10.1073/ pnas.1105639108 (1105639108 [pii]). [40] K.J. Valenzano, R. Khanna, A.C. Powe, R. Boyd, G. Lee, J.J. Flanagan, E.R. Benjamin, Identification and characterization of pharmacological chaperones to correct enzyme deficiencies in lysosomal storage disorders, Assay Drug Dev. Technol. 9 (2011) 213–235, http://dx.doi.org/10.1089/adt.2011.0370. [41] J. Ribbens, G. Whiteley, H. Furuya, N. Southall, X. Hu, J. Marugan, M. Ferrer, G.H. Maegawa, A high-throughput screening assay using Krabbe disease patient cells, Anal. Biochem. 434 (2013) 15–25, http://dx.doi.org/10.1016/j.ab.2012.10.034 (S0003-2697(12)00553-2 [pii]). [42] G. Parenti, C. Pignata, P. Vajro, M. Salerno, New strategies for the treatment of lysosomal storage diseases (review), Int. J. Mol. Med. 31 (2013) 11–20, http://dx.doi. org/10.3892/ijmm.2012.1187. [43] S. Patnaik, W. Zheng, J.H. Choi, O. Motabar, N. Southall, W. Westbroek, W.A. Lea, A. Velayati, E. Goldin, E. Sidransky, et al., Discovery, structure–activity relationship, and biological evaluation of noninhibitory small molecule chaperones of glucocerebrosidase, J. Med. Chem. 55 (2012) 5734–5748, http://dx.doi.org/10. 1021/jm300063b. [44] J.J. Marugan, W. Zheng, M. Ferrer, O. Motabar, N. Southall, E. Goldin, W. Westbroek, E. Sidransky, Discovery, SAR, and Biological Evaluation of a Non-Inhibitory Chaperone for Acid Alpha Glucosidase, 2010. (doi: NBK153221 [bookaccession]). [45] T.H. Tsai, Concurrent measurement of unbound genistein in the blood, brain and bile of anesthetized rats using microdialysis and its pharmacokinetic application, J. Chromatogr. A 1073 (2005) 317–322. [46] M. Malinowska, F.L. Wilkinson, K.J. Langford-Smith, A. Langford-Smith, J.R. Brown, B. E. Crawford, M.T. Vanier, G. Grynkiewicz, R.F. Wynn, J.E. Wraith, et al., Genistein improves neuropathology and corrects behaviour in a mouse model of neurodegenerative metabolic disease, PLoS One 5 (2010) e14192, http://dx.doi.org/10.1371/ journal.pone.0014192. [47] A. Friso, R. Tomanin, M. Salvalaio, M. Scarpa, Genistein reduces glycosaminoglycan levels in a mouse model of mucopolysaccharidosis type II, Br. J. Pharmacol. 159 (2010) 1082–1091, http://dx.doi.org/10.1111/j.1476-5381.2009.00565.x (BPH565 [pii]).