Pharmacologic chaperoning as a strategy to treat Gaucher disease

Authors


J. W. Kelly, Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute, 10550 North Torrey Pines Road, BCC265, La Jolla, CA 92037, USA
Fax: +1 858 784 9610
Tel: +1 858 784 9880
E-mail: jkelly@scripps.edu

Abstract

We briefly introduce the most common lysosomal storage disorder, Gaucher disease, concisely describe the Food and Drug Administration approved strategies to ameliorate Gaucher disease, and then outline the emerging pharmacologic chaperone strategy that offers the promise to remedy this malady.

Abbreviations
CNS

central nervous system

ER

endoplasmic reticulum

ERAD

endoplasmic reticulum-associated degradation

ERT

enzyme replacement therapy

GC

glucocerebrosidase

GD

Gaucher disease

WT

wild type

Introduction

Human lysosomal storage diseases are loss of function disorders, typically caused by a deficient lysosomal glycolipid hydrolysis activity, leading to intralysosomal accumulation of the enzymes substrate(s) [1,2]. Although each lysosomal storage disease has unique characteristics, generally they are progressive in nature and lead to an enlarged liver and spleen, bone and skeletal changes, short stature and respiratory and/or cardiac problems.

Gaucher disease (GD) is caused by deficient lysosomal glucocerebrosidase (GC or acid β-glucosidase) activity [3,4]. Glucocerebrosidase degrades glucosylceramide (Fig. 1) into glucose and ceramide, which are recycled in the cytoplasm. Mutations in both alleles of GC sometimes result in the accumulation of glucosylceramide in the lysosomes of monocyte-macrophage cells, often leading to hepatomegaly, splenomegaly, anemia and thrombocytopenia, bone lesions, and sometimes central nervous system (CNS) involvement [5,6]. Patients not exhibiting CNS symptoms are classified as type 1, whereas the 4% of patients presenting with CNS involvement are classified as either type 2 (acute infantile) or type 3 (juvenile or early adult onset).

Figure 1.

 Chemical structures of glucosylceramide, the substrate for glucocerebrosidase, Zavesca®, the substrate reduction therapy compound approved by the Food and Drug Administration and selected glucocerebrosidase pharmacologic chaperones.

Of the 200 mutations associated with GD, only a few are prominent. For example, over 70% of the variant alleles among the Ashkenazi Jewish subjects are N370S (Fig. 2B) [5,7–9]. The neuropathic L444P allele occurs at a much higher frequency (37.5%) among non-Jewish subjects (Fig. 2B). GD is recessive, meaning that patients require mutations in both GC alleles to present with symptoms and, even then, the penetrance is variable, suggesting that physiological and genetic background differences also influence disease onset.

Figure 2.

 (A) Ribbon diagram representation of the crystal structure of glucocerebrosidase depicting the location of the GD-associated point mutations. (B) Frequency of GC point mutations in human GD patients.

GD is currently treated by enzyme replacement therapy (ERT) [10], wherein a recombinant GC enzyme is administered intravenously. Identification of a mannose receptor on macrophages made it possible to specifically target this cell type by creating recombinant ‘mannose-terminated’ GC that is recognized by mannose receptor, endocytosed and delivered to the lysosome, where it partially restores GC activity. In spite of the fact that lysosomal localization is very inefficient, ERT is currently the treatment of choice for non-neuropathic GD. Unfortunately, GC replacement therapy does not ameliorate the damage to the CNS that exists in type II/III patients because the recombinant enzyme used in ERT does not cross the blood–brain barrier.

Another strategy for treating GD is substrate reduction therapy [11]. The premise behind this strategy is that intralysosomal glucosylceramide accumulation will occur in individuals where the amount of substrate exceeds the capacity of the endogenous mutant GC enzyme to degrade it. Because reducing glucosylceramide influx will restore the balance between substrate synthesis and degradation in the lysosome, inhibition of glucosylceramide biosynthesis may improve the clinical course of disease. Zavesca® (Actelion Pharmaceuticals, South San Francisco, CA, USA) has recently been approved in Europe and the USA for use in patients with mild to moderate type 1 GD, for whom enzyme replacement therapy is not a feasible option. Conditional approval resulted because Gaucher patient response was better with ERT [12]. Yet another possible strategy to treat GD is gene therapy mediated by adeno- and lentiviral vector delivery, although significant hurdles still exist with the implementation of gene therapy as a practical and safe therapeutic strategy [13].

GD is generally caused by GC mutations that compromise folding inside the endoplasmic reticulum (ER). Hence, clinically important variants such as N370S and L444P GC are largely degraded by endoplasmic reticulum-associated degradation (ERAD) mediated by the proteasome, instead of being properly folded in the ER and trafficked to the lysosome. Because of extensive ERAD, there is little mutant GC in the lysosome, and the fraction that does localize properly only has fractional glucosylceramide hydrolase activity. That said, the fractional activity appears to be sufficient to ameliorate disease, when folding and trafficking efficiency is increased, resulting in an increase in the mutant GC concentration in the lysosome.

Permissive growth temperatures (below 37 °C) often enable enhanced folding and lysosomal trafficking of GC variants in patient derived cells, providing hope that one can restore proper cellular folding and trafficking to these endogenous enzymes utilizing a small molecule strategy [14]. Moreover, by growing cells at a temperature that permits enhanced GC ER folding and trafficking to the lysosome, the temperature can then be increased to 37 °C revealing that these mutant enzymes are stable and functional in the lysosomal environment once folded. Biophysical studies using cell-derived mutant GC proteins and recombinant mutant GC proteins reveal that these enzymes often exhibit substantially decreased stability at the neutral pH condition found in the ER, yet these mutant enzymes generally exhibit near wild type (WT) stability at lysosomal pH (approximately pH 5).

That it is possible to correct the folding and trafficking of mutant GCs in cells using a permissive growth temperature motivated us, and subsequently others, to explore whether ER permeable active-site-directed inhibitors of GC could bind to and stabilize these folded mutant enzymes in the ER, enabling their trafficking on to the lysosome (Fig. 3). These so-called ‘pharmacologic chaperones’ are envisioned to assist the macromolecular chaperones by binding to the small fraction of mutant GC that does fold in the ER, stabilizing that folded conformational ensemble and thereby enabling coupling to the secretory apparati. Thus, by LeChatlier's principle, pharmacologic chaperones shift the equilibrium towards folding at the expense of ERAD, enabling folded GC to engage the exocytic pathway that carries it to the lysosome. Once mutated GC is localized to the lysosome, the glucosylceramide substrate is able to displace the inhibitor and allow the enzyme to turn over glucosylceramide, owing to the high lysosomal concentration of glucosylceramide. Thus the cellular GC activity goes up because of an increased lysosomal concentration, despite the fact that the pharmacologic chaperone is actually a GC inhibitor.

Figure 3.

 Mechanism of pharmacologic chaperoning, adapted with permission from Sawkar et al.[5]. Pharmacologic chaperones bind to the folded glucocerebrosidase (GC) enzyme population in the ER, shifting the equilibrium toward the folded conformational ensemble, away from the unfolded ensemble that is subject to ERAD mediated by the proteasome. Pharmacologic chaperoning of GC enables a greater proportion of GC to be folded and thus engage the exocytic pathway that trafficks it to through the Golgi and on to the lysosome, increasing the concentration of the folded enzyme having partial wild-type activity in the lysosome. Glucosylceramide displaces the pharmacologic chaperone in the lysosome, enabling the enzyme to cleave glucosylceramide into glucose and ceramide. Pharmacologic chaperoning of GC increases the folding and trafficking efficiency of GC, increasing its concentration in the lysosome, restoring partial function, and offering the possibility to ameliorate GD.

Unlike nonspecific, low molecular weight osmolytes, such as glycerol, dimethyl sulfoxide and trimethylamine N-oxide that have been shown to increase proper folding and trafficking of variant proteins when included in the cell culture medium at high (mm) concentrations [15], pharmacologic chaperones are typically effective at much lower concentrations (nm to µm) and stabilize just one protein and thus are generally protein and disease selective, if not specific.

Many of the clinically important Fabry disease-associated α-galactosidase A variants (causing another lysosomal storage disease) were shown to be folding and trafficking mutants [16] before this was explored as a possibility in GD. Galactose administration increased Q279E α-galactosidase A residual activity in patient derived cells; thus, galactose was demonstrated to be first active-site-directed pharmacologic chaperone for a lysosomal storage disease. Galactose administration (1 g·kg−1 body weight) every other day proved to be effective therapy for a Fabry disease patient harboring the G328R variant, meaning that a heart transplant was no longer required [17]. An active-site-directed pharmacologic chaperone for α-galactosidase A discovered by Jian-Qiang Fan and developed by Amicus Therapeutics is currently in Phase II clinical trials for Fabry disease [18,19]. A thorough review of α-galactosidase A pharmacologic chaperones for Fabry disease is provided in an accompanying minireview by Fan & Ishii [20].

The GD-associated N370S, G202R and L444P GC mutations reduce lysosomal GC concentration by impairing proper folding and trafficking, apparently by similar, but not identical mechanisms. These GC variants exhibit distinct subcellular localization patterns in patient-derived fibroblasts: N370S GC exhibits weak lysosomal localization, G202R GC is retained in the ER, and L444P is largely degraded with a small fraction making it to the lysosome [14,21]. The N370S, L444P and G202R GC mutations reduce the stability of GC in the ER as an apparent consequence of the neutral pH environment there, resulting in enough ERAD to reduce lysosomal GC concentration and activity [14]. The folding and trafficking of G202R and L444P GC is temperature-sensitive, providing further evidence that these variants are deficient in folding and are recognized by ERAD [14,21].

Several GC variants have been shown to be amenable to pharmacologic chaperoning in patient-derived cell lines [14,22–30]. Moreover, several distinct structural classes of GC pharmacologic chaperones have been discovered [14,22–30]. In 2002, we demonstrated that the active-site-directed GC inhibitor N-(n-nonyl)deoxynorjirimycin acted as pharmacologic chaperone for N370S, but not L444P GC in patient-derived fibroblasts [22], stabilizing GC against thermal denaturation and increasing cellular N370S GC activity two-fold by increasing ER folding efficiency and lysosomal trafficking. Several other nitrogen-containing heterocycles and monosaacharides that also inhibit enzymes that make and break glucosyl bonds were also shown to be N370S GC pharmacologic chaperones, including morpholine and piperazine-based molecules. N-(n-butyl)deoxynorjirimycin (Zavesca®), does not act as a pharmacologic chaperone under comparable conditions in these cell lines [22,25,26]. In 2004, Lin and colleagues reported that application of N-octyl-β-valienamine (Fig. 1) increased the cellular activity of F213I GC six-fold; however, this compound proved to be ineffective in the N370S and L444P cell lines tested [23]. In 2005, we reported that terminating the DNJ N-alkyl chain with an adamantyl group results in very active N370S and G202R GC pharmacologic chaperones [25]. N-octyl-isofagomine and N-octyl-2,5-dideoxy-2,5-imino-d-glucitol were also reported to be pharmacologic chaperones, enhancing cellular N370S and G202R GC activity [25]. Collectively, the data demonstrate that distinct GC mutations exhibit different pharmacologic chaperoning profiles in patient-derived cell lines. In 2005, Pocovi and colleagues reported that increased N370S activity was observed with 10 µm Zavesca® in transfected COS-7 cells, in contrast to our observations in patient-derived cell lines [22,24]. In 2006, Asano and colleagues reported that α-1-C-nonyl-1,5-dideoxy-1,5-imino-d-xylitol was more selective, but less potent as a pharmacologic chaperone than N-(n-nonyl)deoxynorjirimycin [27,29]. In 2006, Fan and colleagues reported that the hydrophilic amino sugar isofagomine (Fig. 1) is a potent inhibitor of GC and serves as a GC pharmacologic chaperone that increased cellular N370S GC activity two-fold by enhancing its cellular folding and trafficking [26]. Kornfield and colleagues reported a very similar result with isofagomine, just a few months later [28]. Isofagomine is now being evaluated in a phase II clinical study for GD by Amicus Therapeutics. In 2007, we reported additional adamantyl terminated N-alkyl isofagomines and 2,5-anhydro-2,5-imino-d-glucitol derativatives that are potent GC pharmacologic chaperones [30]. More than a seven-fold enhancement of cellular G202R GC activity was observed when cells were cultured with N-adamantyl-4-((3R,4R,5R)-3,4-dihydroxy-5-(hydroxymethyl)piperidin-1-yl)-butanamide (Fig. 1) for 5 days (cellular N370S GC is increased by more than 2.5-fold). These structure–activity relationships are now easily rationalized by the 2007 GC structure of Petsko and coworkers, revealing two hydrophobic binding clefts proximal to the active site where the monosaacharide substructure binds [31]. Collectively, these data demonstrate that pharmacologic chaperoning increases mutant GC folding efficiency in the ER enhancing lysosomal trafficking, which increases the lysosomal concentration of partially active GC variants, as demonstrated by increased cellular GC activity, an increased concentration of lysosomal GC glycoforms and increased colocalization of GC with the lysosomal markers based on fluorescence microscopy analysis.

All of the GC variants that are amenable to pharmacologic chaperoning harbor mutations in the active-site domain, whereas the L444P mutation, located in the immunoglobulin-like domain of GC [31,32], does not respond to pharmacologic chaperoning in patient-derived cells when treated identically. Mutations in domains remote from the chemical chaperone binding active-site domain may continue to be subject to misfolding, despite binding-induced stabilization of the active-site domain, especially if the domains are not thermodynamically coupled. In the future, it may be possible to discover a small molecule that binds to and stabilizes the immunoglobulin-like domain, which should correct the folding defect associated with the L444P GC variant. Alternatively, it may be that the L444P GC is actually being partially pharmacologically chaperoned and is more sensitive to inhibition than the other variants because of its lower lysosomal concentration, in which case new dosing and washout regimens may be useful in restoring partial L444P GC activity.

Lastly, in contrast to ERT and like substrate reduction therapy, a pharmacologic chaperone strategy for GD relies on the endogenous activity of the folded mutant GC enzyme. Thus, the pharmacologic chaperoning approach will not be able to increase cellular GC activity in the case of mutations that do not produce a foldable protein or produce a folded product lacking GC activity. In addition, enzymes that are unable to bind the pharmacologic chaperone will not benefit from this approach.

The promise of the pharmacologic chaperone strategy for GD

Pharmacologic chaperones penetrate the plasma membrane and the ER, and by binding to the folded mutant GC enzyme population in the ER, shift the equilibrium towards folding allowing mutant GC to be trafficked to the Golgi and on to the lysosome more efficiently, where the high substrate concentration and low pH environment stabilize the GC fold enabling it to degrade glucosylceramide. The resulting increase in mutant lysosomal GC concentration and cellular activity is thought to be sufficient to ameliorate GD, a hypothesis being tested by an ongoing GD clinical trial utilizing the pharmacologic chaperone isofagomine. Orally available pharmacologic chaperones that cross the blood–brain barrier efficiently have the potential to be useful for treating patients with type II and III GD, for which there are currently no therapeutic options available. In the future, it is likely that pharmacologic chaperones will be used in concert with ERT to ameliorate lysosomal storage diseases, including GD.

Acknowledgements

The authors are grateful to Professor Evan Powers for drawing Fig. 3, and to the National Institutes of Health (DK075295), the National Gaucher Foundation, Gaucher Disease Divot Classic (Grant no. 70), the Skaggs Institute of Chemical Biology and the Lita Annenberg Hazen Foundation for financial support of the studies outlined herein.

Ancillary