Advancing the Research and Development of Enzyme Replacement Therapies for Lysosomal Storage Diseases

Abstract

With the increasing interest in developing gene therapies for rare diseases, it is easy to overlook that there are numerous rare lysosomal storage diseases (LSDs) with treatments that have been approved by regulatory agencies in the United States and Europe. These primarily consist of enzyme replacement therapies (ERTs)—recombinant human proteins that are delivered for the life of the patient through different routes and may offer distinct safety and distribution advantages over gene therapies. The research and development of ERT is a lengthy and expensive process, usually performed in academic laboratories before transfer to pharmaceutical companies and hence a process ripe for disruption. To realize fully the considerable scientific and investment potential for ERT, we need to develop a pipeline of proteins analogous to what has been created in some open science efforts, as well as apply technologies to decrease manufacturing costs. In this perspective, we illustrate the opportunity to fill the rare LSD treatment gap with ERTs while gene therapies are in development for these life-shortening diseases.

Introduction

Among the estimated 7,000 known rare genetic diseases, the lysosomal storage diseases (LSDs) are a collection of >70 clinical syndromes caused by mutations in ∼50 genes. These genes encode proteins that are required for lysosomal function and are in most cases enzymes that are important for the cellular degradation and trafficking of lipids and other macromolecules.¹ As these are of fundamental importance to the cell, the absence or mutation in any one of these genes has catastrophic downstream consequences impacting development. These mutations dramatically shorten patients' lives, create hardship for families, and cause significant costs for the health care system.

For most LSDs, correlations have been observed between residual enzyme activity and disease severity.² LSDs not only affect children; depending on residual enzyme activity, symptoms and the disease phenotype can appear later in life. For example, higher activities (6%) of β-galactosidase enzyme have been reported in late-onset GM1 gangliosidosis (adult-onset type).³ Each LSD is individually rare but collectively they affect many individuals, with an estimated incidence between 1:5,000 and 1:7,000 live births.⁴ Approximately 70% of these LSD cases represent unique highly progressive neurodegenerative diseases, which also highlights the sensitivity of the central nervous system (CNS) to lysosomal dysfunction and the long-term effects.⁵

In most LSDs, the mutated protein is a soluble acidic hydrolase; in others it is caused by deficiency of an integral membrane, activator, transporter, or nonlysosomal proteins that are necessary for lysosomal function.^1,6 The impact of these LSD mutations on lysosomal function results in the accumulation of complex substrates, which may include various oligosaccharides, cholesterol, peptides, glycosphingolipids, glycosaminoglycans, glycogen, and/or glycoproteins,⁷ as well as the secondary impairment of lysosome-related pathways.¹ The dramatic impact on the storage of these substrates in multiple organs results in the variable association of hematological, skeletal, visceral, ocular, and neurological manifestations that are ultimately characteristic of particular LSDs.^1,8

Virtually every cell in the body has a lysosomal system, and that system is defective in each LSD.⁸ The cell types and organs affected by a specific LSD may be influenced by a variety of factors. These include differential biochemistry of various cell types, substrate turnover rate, enzyme redundancy, changes to counteract the primary defect, cell differentiation/regeneration, and the fetal development stage when substrate storage reaches a critical threshold triggering cell dysfunction.⁹ Therefore, to treat an LSD effectively, this high level of knowledge is vital, as the therapies in question need to be able to access the key anatomical sites and affected cell types.

As the underlying cause of most LSDs is a specific enzymatic deficiency, many of the attempts to treat some of these diseases have focused on enzyme replacement therapies (ERTs).¹⁰ An ERT needs to deliver the enzyme to the appropriate organelle (e.g., lysosome) to correct the defect. This can be achieved by delivering either purified or recombinant enzymes, which represents a well-tolerated approach with minimal systemic adverse reactions for the patient.

This ERT concept was first developed in 1964 by Christian de Duve¹¹ and Roscoe Brady for sphingolipid disorders in 1966.¹² Typically, the enzyme may also be modified to allow for a longer half-life; more potent activity; resistance to degradation; or targeting to a specific organ, tissue, or cell type. As most LSDs are manifested by neurological symptoms, several strategies have been developed to ensure the ERT reaches the CNS, such as cerebrospinal fluid and nanovesicle-based delivery.¹³

Lysosomal enzymes are glycoproteins synthesized in the endoplasmic reticulum (ER) with a specific N-terminal signal sequence. Several glycosylation events occur in the ER to enhance protein folding with the help of multiple ER-resident enzymes and molecular chaperones. Newly synthesized lysosomal enzymes are then specifically modified with mannose 6-phosphate (M6P) residues by the sequential action of two enzymes localized in the Golgi apparatus. A small portion of the ER-synthesized enzymes are also directly secreted extracellularly but are recaptured and internalized through the membranous M6P receptors to be delivered to lysosomes through the secretory pathway.¹⁴

Lysosomal acidity activates the trafficked enzyme through multiple proteolytic and/or folding processes to achieve the fully functional active enzyme conformation.⁶ Based on this lysosomal cell biology, enzyme-deficient cells can then take up exogenous enzyme through the M6P recapture mechanism (cross-correction), whereupon the enzyme enters the endocytic system and is delivered to the lysosome where it can then fulfill its required function.⁸ Therefore, any therapies such as ERT, gene therapy, and stem cell transplantation may restore some of the lost enzyme using this principle.⁵

In practice, ERT is based on the periodic administration of a manufactured enzyme that can be taken up into cells, be delivered to lysosomes, and ultimately reduces substrate storage.¹⁵ It is the primary therapeutic approach for LSD that has been the most therapeutically and commercially successful to date.^15,16 Improvements to ERT are also being investigated to reduce dosage, frequency of delivery, and adverse effects while increasing therapeutic efficacy, and ERT affordability.^13,16 In contrast, small-molecule approaches such as substrate reduction therapy or chaperone therapy are only effective in patients with some residual functional enzyme.

An example is migalastat (Galafold), a small-molecule chaperone therapy indicated for Fabry disease in patients with mutant forms of α-galactosidase A that retains their catalytic activity despite abnormal protein folding.¹⁷ One of the major challenges is the treatment of CNS manifestations in LSDs. Gene therapy may address this as there are various clinical trials, using both in vivo and ex vivo approaches, to evaluate the safety and efficacy to treat neuropathic LSDs.¹⁸ Initial findings for late infantile neuronal ceroid lipofuscinosis (CLN2) clinical trials (NCT01414985; NCT01161576) suggest the treatment lowered the progression of disease, but to halt disease progression, improvements in vector design and delivery strategies will be necessary.¹⁹

Intraparenchymal gene therapy in children is well tolerated, however, the clinical efficacy is variable as a lack of efficacy was observed in metachromatic leukodystrophy (MLD), with a trend toward positive results shown in trials for CLN2, mucopolysaccharidosis (MPS) type IIIB (MPS IIIB), and MPS IIIA.¹⁸ In contrast, very encouraging results have been obtained using ex vivo gene therapy in early onset forms of MLD, which preserved cognitive function and motor development in most patients and slowing demyelination and brain atrophy.²⁰

Advantages and Disadvantages of ERTs

There is considerable activity around gene therapies for rare LSDs, to such an extent that ERT may be viewed as a less exciting technology, less likely to receive investor funding. However, ERT may have significant safety advantages over cell and gene therapy, which could be an important factor in redressing this imbalance (Fig. 1). In gene therapy and some cell therapies, foreign DNA is introduced and may integrate into the host genome (unless using appropriate vectors), potentially causing neoplasia and other safety concerns such as hepatotoxicity²¹ and dorsal root ganglion toxicity.²² Although these toxicities should be considered in context of the shortened life span and neurodegeneration facing some LSD patients, they represent a significant hurdle in clinical trials.

FIG. 1.

Summary of opportunities and challenges for ERT.

ERT, enzyme replacement therapy.

ERT does not have this limitation, lowering the hurdles to success in the clinic. ERT also has a proven track record and can be administered at a controlled dose and frequency¹⁶ (unlike gene therapy, which depends on the extent and duration of transduction) and is also not hindered by the potential problem of anticapsid antibodies.²³ ERTs that are approved by the U.S. Food and Drug Administration (FDA) and European Medicines Agency may require infusions every 2 weeks, such as imiglucerase (Cerezyme™), taliglucerase (Elelyso™), agalsidase β (Fabrazyme™), or every week for laronidase (Aldurazyme™), galsuflase (Naglazyme™), idursulfase (Elaprase^®), elosulfase alfa (Vimzim^®), and sebelipase alfa (Kanuma^®)¹⁶ (Table 1).

Table 1.

Enzyme replacement therapies and research development timeline

Protein	Company and target/disease	Initial protein and cell work	Mouse/rat studies	Cat/dog/monkey studies	Clinical trial	FDA approval
Imiglucerase	Genzyme/Gaucher disease	1974,43 199344	—	—	199545	1995
Laronidase	BioMarin and Genzyme/MPS I	1991,46 199447	—	199448	200149	2003
Agalsidase beta	Genzyme/Fabry disease	199250	200151 200352	—	200353	2003
Galsulfase	BioMarin/MPS VI	199454	—	199655	200456	2005
Idursulfase	Shire/MPS II	199357	200258	—	200659	2006
Alglucosidase alfa	Genzyme/Pompe disease	197360	199161	—	200562	2006
Elosulfase alfa	BioMarin/MPS IVA	2007,63 201064	200763 201064	—	201465	2014
Sebelipase alfa	Alexion/Acid lipase	200566	201167	—	201368	2015
Cerliponase alfa	BioMarin/CLN2	200169	200870	2014,71 201572	201826	2017
Vestronidase alfa	Ultragenyx/MPS VII	199373	199374 199475	—	201876	2018
Velmanase alfaa	Chiesi/α-mannosidosis	200177	200478 201579		201880–82	2018

Table shows dates of key milestones for protein production, animal studies, clinical trial, and FDA approval.

^a European Medicines Agency approval. In addition to the following, pabinafusp alfa was approved in 2021 by the Ministry of Health, Labour and Welfare in Japan for MPS II.⁴²

CLN2, ceroid lipofuscinosis; FDA, Food and Drug Administration; MPS, mucopolysaccharidosis.

ERT also has some disadvantages (besides continuing lifelong infusions). It may produce antidrug antibody responses that can result in hypersensitivity reactions, whereas the long-term impact on efficacy is unknown.²⁴ There are numerous examples of successfully commercialized enzymes for LSDs (Table 1), which have stabilized disease outcomes, alleviated suffering, and extended and improved quality of life in patients. Recently many of these treatments have been considered for in utero treatment, such as Aldurazume (laronidase; NCT04532047).

Investors and the FDA are, therefore, very familiar with ERTs and their potential commercialization pathway, leading to considerable commercial success.²⁵ For example, intraventricular cerliponase alfa ERT (Brineura^®; BioMarin Pharmaceutical) was approved by the FDA in 2017 and became the first globally approved treatment for CLN2 (Batten disease). Replacement of this dysfunctional enzyme through intracerebroventricular infusion of a functional enzyme has been shown to effectively attenuate the progression of the disease in patients. These CLN2 patients appear to have less deterioration in motor-language symptoms compared with historical untreated control groups that may be clinically meaningful.²⁶ ERT clearly improves health-related quality-of-life declines as the disease progresses.^27,28

Changing “One and Done” to a Pipeline of ERTs

Although it is difficult to gauge how long the development of each ERT took to bring from the laboratory to patients (Table 1), estimates suggest it is well over a decade. Examples include agalsidase beta (11 years), sebelipase alfa (11 years) imiglucerase (21 years), idursulfase (23 years), vestronidase alfa (25 years), and alglucosidase alfa (33 years). There has been ∼50 years of effort invested to develop treatments for Sanfilippo syndrome (MPS III); early work on MPS IIIB purified the enzyme N-acetyl-α-glucosaminidase, which was used to show metabolic correction in patient fibroblasts.²⁹ However, subsequent ERT clinical trials³⁰ have not so far resulted in FDA approvals. This suggests we need to find ways to shorten the development time as this is not scalable or sustainable.

Despite more decades of development for LSDs, there are no currently approved treatments for almost two-thirds of LSDs³¹ that could be addressed by ERT (Supplementary Table S1). This trajectory is simply insufficient to impact current patients in their lifetime. These challenges also hamper incumbent companies from developing new ERTs because of the long period of research and development investment required, as well as the high costs associated with GMP manufacturing and clinical trials (or other factors). This in turn is reflected in the very high prices these chronic treatments command once approved, which also puts pressure on the health care and insurance systems as well as on the patients and their families who pay for these treatments in some countries.

Efforts to develop ERT for rare diseases may also fail at the first hurdle. The approach for many rare disorders consists of individual research groups working on a single therapeutic approach (a “single protein-single lab” paradigm) (Table 1 and Supplementary Table S1). If they are lucky once they have proof-of-concept data—and if their disease population is sufficiently large—a biotech might come in to license it. In turn, these small companies may also work on a “one-and-done” approach, developing a single treatment for a single disease before exiting, often because they are funded by venture capital and this is a requirement for funding.

We believe that there is a need to change both these academic and commercial approaches by developing simultaneously a pipeline of many proteins for rare diseases to increase ongoing research (Fig. 2). An analogy is in gene therapy, wherein many companies are working on multiple rare diseases with their “plug-and-play” technology using a single viral vector or platform technology to deliver different genes for many unique diseases.³²

FIG. 2.

Generalized ERT development process.

Blue—steps that are automatable, yellow—steps that are more challenging to automate.

We could do well to learn from other areas of science that have moved toward higher throughput and collaborative approaches. One example is the Structural Genomics Consortium,³³ which has efficiently produced thousands of proteins by testing several constructs and cell lines in parallel. These types of protein production consortia have had an impact on structural genomics,³⁴ directly benefiting drug discovery³⁵ and in many cases utilizing automation.³⁶ The early aspects of ERT development could be similarly automated, but the later stages involving manufacturing would be likely more difficult to scale up (Fig. 2).

As stated earlier, there are still LSDs lacking a suitable ERT, even at the research level (Supplementary Table S1). Our literature analysis and first-hand experience with collaborators developing ERTs illustrate the limitations of the current ERT research and development paradigm. As academic laboratories are focused on a single LSD, there are few opportunities for economies of scale, and they also need to use or develop novel technologies that will differentiate themselves from other laboratories if they are to obtain grant funding. This also means that they are completely invested in making a single treatment; if that fails, they have nothing to show for their effort.

Each of these development efforts is often highly dependent on the graduate students and postdocs employed in the academic laboratory. In addition, if the principal investigator is unable to obtain further funding, the project will probably be terminated or shelved, which could represent the only research ongoing for this rare disease. Alternative approaches to developing proteins could assist in creating an ERT pipeline—for example, nascent technologies such as “3D-printing of proteins.”³⁷ Automation of large-scale manufacturing requires bioreactor technologies with online feedback control enabled by monitoring of secreted biomolecular critical quality attributes,³⁸ such that costs for developing gene therapies and ERT may be comparable.

Effective ERT Delivery Remains a Challenge

There are other hurdles for therapeutic development that should be mentioned. For example, there may be challenges for proteins that need to be targeted to the ER, although several proteins such as Ricin can enter mammalian cells by endocytosis and then undergo retrograde transport through the Golgi complex to reach the ER.³⁹ In addition bacterial toxins such as cholera toxin or pseudomonas exotoxin A both carry KDEL or KDEL-like C-terminal tetrapeptides, which can enable their efficient delivery to the ER.⁴⁰ Similarly, fusion proteins may need to be developed for targeting other organelles.

Another hurdle is high cost because the focus has been on a one-drug/therapy-for-one-disease approach.³¹ Researchers are trying to define common mechanisms for LSDs, including inflammation and autophagy, that might help identifying treatments useful for several LSDs³¹ and expand the potential market, which could also drive down costs.

The commercial activity around LSDs has driven improvements in the rates of diagnosis and the earliest possible identification of patients to initiate treatment. The ERTs developed to date (Table 1) have targeted the more prevalent LSDs, leaving behind those diseases involving the brain. Strategies to deliver ERT to the brain across the blood–brain barrier are being actively explored. Several companies are taking this approach and using these technologies to address LSDs (Table 2).

Table 2.

New enzyme replacement therapies in development

Disease	Treatment	References
MPS I, MPS II, MPS IIIA, MPS IIIB, MPS VII, Pompe	J-Brain cargo system, a recombinant fusion protein consisting of a humanized antitransferrin receptor antibody and the enzyme to be delivered.	83–85
Farber disease	ACG-801 is an investigational enzyme replacement therapy for the treatment of Farber disease.	86,87
MPS I, MPS II, MPS IIIA, MPS IIIB, metachromatic leukodystrophy, Pompe	Large molecule transport vehicle platform technology engineers BBB receptor binding into an Fc domain and uses the transferrin receptor enzyme replacement therapy engineered to cross the BBB.	88
Fabry disease	Pegunigalsidase alfa is a plant cell culture-expressed and chemically modified recombinant alpha-galactosidase-A where protein subunits are covalently bound using PEG chains to increase activity and stability.	89
Gaucher disease	A systemic, noninvasive, and central nervous system-selective delivery system based on nanovesicles of saposin C and dioleoylphos-phatidylserine to deliver to acid β-glucosidase.	90
Krabbe disease	Enzyme delivery system based on the encapsulation of cross-linked enzyme aggregates into poly-(lactide-coglycolide) nanoparticles functionalized with brain targeting peptides (Ang2, g7, or Tf2).	91

BBB, blood brain barrier; PEG, polyethylene glycol.

Even very effective ERTs are unable to reach all tissues and organs to a similar extent; this leads to differential efficacy toward various aspects of the overall disease pathology. A prominent example is the ERT for Gaucher disease, which is unable to completely manage bone disease.⁴¹ Thus, there is a need to develop future therapies that treat all compartments of the body effectively for a given LSD.

The challenges and successes of therapeutic development for LSDs may also serve to inform the treatment of other rare diseases. There can be little doubt that this research will also shed light on common diseases of aging, which illustrates why the study and treatment of additional rare diseases are also important for society at large. ERTs still have considerable therapeutic and financial potential, even as gene therapies and other biologics are developed for LSDs. Although gene therapies may seem cheaper in the long run, they may not prove to be the panacea for all diseases for some of the mentioned reasons discussed.

Conclusion

We propose that we should be focusing on technologies and strategies that will allow companies to create a broader pipeline for developing many proteins in parallel for treating LSDs. In a best-case scenario, one could proceed from preclinical research to manufacture of a human recombinant protein in 2–3 years. Accelerating this may be difficult and hampered by commercial concerns investing in gene therapy manufacturing resources rather than recombinant ERT production. Bringing multiple ERTs for LSDs to fruition would be disruptive for the biotech industry, leading to an increase in new therapeutics for the many rare diseases currently without treatments.

This would also address an urgent clinical need for LSD patients. This “older” therapeutic approach still has considerable potential in a world that increasingly values the new technologies over the tried and tested. We advocate for increased investment in ERT research and development on a par with that for gene therapies by governmental agencies, venture capital, and biotechnology companies.

Author Disclosure Statement

S.E. is owner and both authors are employees of Collaborations Pharmaceuticals, Inc.

Funding Information

We acknowledge the support of 1R43NS107079-01, 3R43NS107079-01S1, and 3R43NS107079-01S2 from NIH/NINDS.

Supplementary Material

Supplementary Table S1