The landscape of CRISPR/Cas9 for inborn errors of metabolism

Abstract

Since its discovery as a genome editing tool, the clustered regularly interspaced short palindromic repeats and CRISPR-associated protein 9 (CRISPR/Cas9) system has opened new horizons in the diagnosis, research, and treatment of genetic diseases. CRISPR/Cas9 can rewrite the genome at any region with outstanding precision to modify it and further instructions for gene expression. Inborn Errors of Metabolism (IEM) are a group of more than 1500 diseases produced by mutations in genes encoding for proteins that participate in metabolic pathways. IEM involves small molecules, energetic deficits, or complex molecules diseases, which may be susceptible to be treated with this novel tool. In recent years, potential therapeutic approaches have been attempted, and new models have been developed using CRISPR/Cas9. In this review, we summarize the most relevant findings in the scientific literature about the implementation of CRISPR/Cas9 in IEM and discuss the future use of CRISPR/Cas9 to modify epigenetic markers, which seem to play a critical role in the context of IEM. The current delivery strategies of CRISPR/Cas9 are also discussed.

1. Introduction

Inborn errors of metabolism (IEM) are a group of rare genetic disorders caused by mutations in genes coding for proteins involved in the intermediated metabolism [1]. First described by Dr. A. Garrod a century ago [2], over 1500 conditions have been classified as IEM [3]. As a group, the IEM may have a birth prevalence estimate of 50.9 per 100,000 live births, which varies depending on the studied region [4, 5]. Nevertheless, this prevalence might be underestimated due to several factors, such as the nonspecific clinical presentation of most disorders, the need, in some cases, for specialized centers for diagnosis, and the limited access to clinical training and academic formation in basic, applied, and clinical aspects of IEM [6, 7].

Several treatment strategies have been developed for the management and treatment of patients, which depend on the type of IEM (e.g., type 1: small molecule, type 2: energetic deficit, type 3: complex molecule), the characteristics and properties of the affected protein (e.g., intracellular, secreted, tissue expression, organelle distribution), and the age of diagnosis [5, 6, 8]. Management may involve supportive therapy, nutritional therapy, substrate/metabolite reduction, metabolite/cofactor supplementation, enzyme replacement, protein enhancement (e.g., pharmacological chaperones), hematopoietic stem cell transplantation, and gene therapy (GT) [5, 6, 8–13].

Since all IEM are monogenic diseases, GT presents the opportunity to correct the genetic disorder. Due to the potential of GT to cure and significantly improve the natural history of these disorders, it is necessary to accelerate the development of GT-based strategies by increasing the efficacy and safety of the current ones. In this sense, genome editing may play a key role in designing and developing novel treatment strategies for IEM [14]. Genome editing uses intracellular double-strand breaks (DSB) to generate genes knock-out or knock-in, mainly using the Non-Homologous End Joining (NHEJ) or Homology Direct Repair (HDR) mechanisms [10]. Genome editing tools mainly include Zinc Finger Nucleases (ZFN), Transcription Activator-Like Effector Nucleases (TALENs), or Clustered Regularly Interspaced Short Palindromic Repetitions (CRISPR)/CRISPR-associated nuclease 9 (CRISPR/Cas9) [15, 16]. These strategies have been evaluated in several IEM, showing their potential to correct the disease phenotype [17, 18]. The first clinical trial of a genome editing strategy was carried out on mucopolysaccharidosis type II patients using ZFN. However, the enzyme activity in blood was not well achieved, and no clear clinical improvement was reported [19]. Since CRISPR/Cas9 is much simpler to design and produce than ZFN and TALENs [10], it was established as the leading genome editing technology and has accelerated and improved the use of genome editing as a tool for the treatment of several diseases [17, 18, 20].

In this review, we summarize the status of CRISPR/Cas9 applications in the treatment and study of IEM based on scientific data available on PubMed, Embase, and SpringerLink for the last five years.

2. Inborn Errors of metabolism: An updated overview

IEM constitute about 1500 diseases (1462, according to http://www.icimd.org. Accessed May 1st, 2022) [3]. The classic definition proved by Dr. Garrod in 1902 has begun to fall short in the face of the discoveries of new entities facilitated by the latest technologies available [1, 3, 21, 22]. Morava et al. (2015) describe an IEM as: “any condition in which primary alteration of a biochemical pathway is intrinsic to specific biochemical, clinical, and/or pathophysiological features.” Nevertheless, unlike other metabolic disorders, IEM are monogenic diseases in which a mutation directly affects the protein function.

A detailed IEM classification is provided by Ferreira et al., 2021 [21], who established and classified different types of IEM based on protein defects and their biochemical characteristics into 24 categories with 124 groups (http://www.icimd.org/ Ferreira 2021) [21].

On the other hand, Saudubray et al., 2019 [3] propose a simplified classification with three types focused on clinical presentation and biochemical consequences of IEM; type 1: small molecule disorders, type 2: energy deficit disorders, and type 3: complex molecule disorders [6, 23]. Regardless of the classification, IEM research aims to develop new and better methods and strategies for diagnosis and treatment. Regarding the diagnosis, Mordaunt et al. 2020 [29] reviewed the trajectory of diagnosis for IEM. They highlighted the use of novel approaches based on omics techniques and the need to integrate them with genetic, biochemical, and clinical information for an efficient and timely diagnosis [6, 24, 25].

In addition, it is estimated that about only 25% of IEM have manifestations during the neonatal period, which may have the most common presentation of the diseases [8]; while the late-onset (infant, juvenile, or adult) may represent additional diagnosis challenges since they have more attenuated clinical findings [9]. Regarding the treatment of these disorders, more than 120 are susceptible to some therapy [26]. For instance, https://www.treatable-id.org/ is a database that summarizes the different therapeutic alternatives and the level of available evidence. The treatment concept generally is oriented to four basic strategies, as outlined by Balakrishnan 2021 [6]. Beyond traditional pharmacological treatment and the implementation of transplants and gene or cell therapies, nutritional management has been and continues to be the fundamental strategy for IEM, especially in groups 1 and 2, since it can be oriented to control the biochemical consequences of IEM; accumulation, depletion, intoxication, and alternative pathways activation [6, 27]. Figure 1 summarizes the current classification of the IEM, as well as their biochemical consequences.

Figure 1.

Overall view of the IEM and their biochemical consequences. A. Since the presence of gene mutations affects proteins related to metabolic pathways, several biochemical consequences can occur as follows: accumulation of substrates (metabolite A) which can be toxic or activate alternative metabolic pathways, as well as depletion of its upstream products (Metabolite B). B. Current IEM classification includes type I affecting the metabolism of small molecules such as phenylalanine (Phe) because of the phenylalanine hydroxylase (PAH) impaired activity, type II, where the cellular energy production is compromised (e.g., Leigh syndrome caused by SURF1 mutations), and type III which involves complex molecules (e.g., GAGs accumulation into lysosome in the MPS IVA where N-acetylgalactosamine-6-sulfate sulfatase (GALNS) activity is affected). This figure was created with Biorender.com.

Gene therapy has been extensively evaluated for IEM [5, 13, 17, 28, 29], with over 7000 publications available on Pubmed. The first attempt at gene therapy in humans was carried out on a group of patients with hyperargininemia at the end of the 60s decade [30, 31]. Even though this clinical trial failed to show any clinical or biochemical improvement, in addition to the ethical issues discussed at that time, this study showed, in T. Friedmann’s words, “…the potential use of viruses as vectors to add new genetic information into human cells for therapy” [30]. In addition, the first death of a patient associated with a gene transfer vector was of a young man with a partial deficiency of ornithine transcarbamylase (OTC), an urea cycle disorder, due to an immune reaction to the adenoviral vector [32]. Thus, IEM played an essential role during the growth and development of the gene therapy field. Nevertheless, only two gene therapies for IEM have been approved by FDA: Glybera^® (alipogene tiparvovec) for the treatment of familial lipoprotein lipase deficiency and Libmeldy^® (atidarsagene autotemcel) for the treatment of metachromatic leukodystrophy [14, 15, 33]. Recently, the European Medicines Agency gave a positive opinion for Upstaza (eladocagene exuparvovec), a potential GT for the treatment of aromatic L-amino acid decarboxylase (AADC) deficiency (https://www.ema.europa.eu/en/documents/smop-initial/chmp-summary-positive-opinion-upstaza_en.pdf. Accessed Jun 24, 2022).

3. Molecular bases of CRISPR/Cas9

The CRISPR/Cas9 system was first identified as an RNA-mediated immune system in prokaryotes for protection against bacteriophages and horizontal plasmid transmission [34]. The prokaryotes develop acquired immunity toward exogenous invasion due to the memory and immune function obtained through the adaptive CRISPR/Cas9 system [35]. This system is present in 50% bacterial and 90% archaeal genomes [36]. The CRISPR/Cas systems are classified into two major classes, depending on effector module organization, Class 1 and 2, with six types (from I-VI) and 18 subtypes [34, 37]. Class 1 systems implement multi-protein effector complexes, which makes the usage of this class challenging compared to Class 2 systems. Class 2 systems utilize single-effector complexes, making them simpler to work with [37]. Thus, Class 2 systems, which make up 10% of CRISPR/Cas loci of the archaeal and bacterial genome, have demonstrated more potential in gene editing applications and screening, as evidenced by multifaceted studies done with Cas9, Cas12a, and Cas13 systems [37]. Moreover, the CRISPR/Cas systems also comprise various editing agents based on the desired type of editing required and have been implemented using nucleases, base editors, transposases/recombinases, and prime editors [38]. The CRISPR/Cas9 system of Class 2 has been most extensively used for gene editing applications. Due to its potential, the research has even progressed towards the engineering of new variants of Cas9, such as Cas9 nickase (nCas9) [39], from Staphylococcus aureus (SaCas9), Streptococcus thermophilus (StCas9), as well catalytically inactive (“dead” dCas9) [40].

3.1. Components of the CRISPR-Cas9 system.

The major components of the CRISPR/Cas9 system are the Cas9 protein and a guide RNA (gRNA) (Fig. 2). The functional CRISPR/Cas9 loci consist of an operon of respective cas genes and two specific RNA components: a programmable CRISPR RNA (crRNA, 18–20 base pair) and a trans-activating crRNA (consists of 3 stem-loops, 20-nt gRNA, and 12-nt repeat region) [41]. Distal to the crRNA targeted sequence exists the protospacer-adjacent motif (PAM, 2–5 bp), which plays a pivotal role in gene editing [41].

Figure 2.

Molecular mechanism of CRISPR-Cas9 mediated DNA target recognition and cleavage. Cas9 structure comprises the REC lobe and a NUC lobe, consisting of an HNH, Ruv C, and PAM-interacting domain (PID). Once sgRNA loads, Cas9 confers a conformational rearrangement forming a pre-target recognition structure, where PAM is structured for sampling. On PAM recognition, target DNA unwinds, and R-loop is formed. Conformational change in HNH and RuvC domains leads to the cleavage and formation of DSB of target DNA, which is further repaired by endogenous processes. NUC: nuclease lobe; REC: recognition lobe; PAM: protospacer-adjacent motif; sgRNA: single-guide RNA; DSB: double-stranded breaks. This figure was created with Biorender.com.

Cas9 protein, known as the “genetic scissor,” is a large multi-domain DNA endonuclease (1368 amino acids) responsible for its multi-functional nature, which cleaves the target DNA, forming DSB [40]. The Cas9 protein belongs to Class 2 of the CRISPR/Cas systems, the first to be utilized for gene editing, extracted from Streptococcus pyogenes (SpCas-9). The protein comprises two function domains, namely, the NUC (nuclease) lobe, consisting of HNH (analogous to phage T4 endonuclease VII), RuvC (like the Escherichia coli RuvC), and PAM interacting domain (PID). The second domain is the REC (recognition) lobe, made up of REC1 and REC2 domains, with a long helix bound to repeat: anti-repeat duplex [38, 41]. The HNH and RuvC domains utilize a single mechanism to cut single-stranded DNA, while PID provides specificity and initiates DNA binding. Guide RNA is formed by the combination of tracrRNA (which acts as a binding scaffold for nuclease) and crRNA (which pairs with the target sequence providing specificity) [41]. One of the significant advantages of having a gRNA is that it can be programmed synthetically to target and edit any gene of interest, making CRISPR/Cas9 systems confer wide applications.

3.2. CRISPR/Cas9 gene-editing molecular mechanism.

The underlying mechanism of action involves three consecutive steps:

• Recognition: The gRNA is designed towards a specific target sequence, where the 5′crRNA complementary base pair domain further directs the Cas9 nuclease to recognize its target in the respective gene of interest [41]. In more depth, the step begins with the formation of target: gRNA heteroduplex and repeat: anti-repeat duplex, which leads to the formation of stem-loops. The Cas9 domains NUC and REC then recognizes the two duplexes, stem-loops and the linker, which leads to the formation of the gRNA: Cas9 binary complex that will ultimately identify the target sequence [41].

• Cleavage: Once the binary complex identifies the correct PAM, the PID initiates the binding of the binary complex to the target DNA, and Cas9 nuclease domains cut the complimentary and non-complimentary strands and form blunt DSBs [38, 41].

• Repair: The DSBs are repaired by the host cellular machinery through non-homologous end-joining or homology-directed repair pathways [42, 43]. NHEJ is error-prone due to the indels that the enzymatic joining of the DNA can introduce. At the same time, HDR is more precise and utilizes a homologous DNA template, leading to the accurate insertion of genes with sequence homology at the DSB [44].

Recent gene editing methods that utilize synthetically engineered variants of Cas9 work differently depending on the respective modifications. Some examples involve dCas9 or CRISPRi (CRISPR interference), which has inactive HNH and RuvC domains. They function by direct transcriptional modification of the target sequence, not affecting the DNA sequence [45], and the Cas9n, which is formed by a loss-of-function mutation in nuclease cleavage domains, allowing high specificity and efficiency as it combines sgRNAs with Cas9n [46] (Fig. 3).

Figure 3.

Applications of CRISPR-Cas9 gene-editing system and Cas9 variants. A. The wild-type Cas9 nuclease cleaves the target DNA by introducing DSBs, which either NHEJ or HDR repairs. NHEJ can introduce indels disrupting the target DNA, while HDR results in specific knock-ins. B. Cas9 nickase results from a mutation in either NUC lobe nucleases, e.g., HNH mutation, H840A, and RuvC mutation, D10A. This strategy requires two sgRNA to form DSBs. C. Dead or dCas9 is the result of two mutations, e.g., H840A in the HNH lobe and D10A in the RuvC lobe, which results in loss of cleavage but not binding, due to which this system has been paired with diverse effector domains that allow specific localization, such as transcriptional activators and repressors. Abbreviations: NUC: nuclease lobe; REC: recognition lobe; DSBs: double-stranded breaks; NHEJ: non-homologous end joining; HDR: homology-directed repair, PID: PAM-interacting domain. This figure was created with Biorender.com.

The CRISPR/Cas9 systems require gRNA construction, delivery, target sequence selection, and system efficiency for developing more precise and effective treatment. The sgRNA determines the specificity of the nuclease, which is why one needs to consider the 5′-NGG-3′ PAM sequence to facilitate cleavage [47]. Thus, it is crucial to determine PAM preference for Cas nucleases. Various methods like in silico, bacterial depletion assays, mammalian cell-based, and, more recently, high-throughput PAM determination assay (HT-PAMDA) have been used to determine PAM preferences, with which a near PAMless Cas9 variant was generated that provided potential for new therapeutic targets and cellular model development [48]. Various modifications of gRNA, like GC content percentage, length, and chemical changes, have been evidenced to reduce off-target effects [49]. A specific sgRNA is key to minimum off-target effects, one of the crucial cellular impacts of the CRISPR/Cas9 system. The construction and delivery of sgRNA depend on the desired application [44, 49]. The off-target effects can be determined by biased and unbiased methods, such as biased tools involving Cas-OFFinder, and CRISTA, while unbiased methods involve Chip-seq and GUIDE-seq [50]. Even though the cellular impact and immune response seem like a hurdle, the success of CRISPR-Cas9 system gene editing is due to its efficiency, cost-effectiveness, and easy use [10].

Although more advances to the CRISPR-Cas9 system are necessary for endorsing specificity, efficacy, and safety, leading to improved treatments for IEM, several in vitro and in vivo approaches have shown encouraging outcomes. In the following section, we will review the most recent data about the use of CRISPR/Cas9 in the context of IEM following the updated classification of these groups of diseases as follows: Type I: Small molecule, Type II: Energy deficiency, and Type III: Complex molecule.

4. CRISPR/Cas9 and IEM type I: Small molecule

IEM type I includes disorders with an accumulation of toxic metabolites, such as amino acid disorders, organic acidemia, and urea cycle disorders. In this section, the reader will find a review of research articles in which CRISPR has been used with therapeutic effect or as a molecular tool to generate animal models for studying this group of IEM.

4.1. Urea cycle disorders.

The urea cycle is the primary metabolic mechanism through which ammonia is converted into urea and excreted in the urine in a complex and multi-step process [51]. The total or partial impairment in an enzyme or transporter biological function from the urea cycle may lead to hyperammonemia with potential severe neurologic damage [52]. Those inherited metabolic disorders have been named according to the defective enzyme and belong to a group of diseases known as Urea Cycle Disorders (UCDs), which are estimated to have an overall incidence of 1:35,000 to 52,000 births [53, 54]. OTC is the most frequent UCD. Current treatments for these disorders focus on diet management, medications, or liver transplantation.

4.1.1. OTC deficiency.

OTC deficiency is caused by impaired OTC activity, which catalyzes the condensation between carbamoylphosphate and ornithine to synthesize citrulline as an intermediated step to the breakdown and removal of nitrogen in the body [55]. OTC deficiency is an X-linked recessive disorder affecting males, although females can also develop some disease characteristics. Among all the urea cycle defects, OTC deficiency is the most common [55]. Classical clinical onset is characterized by hyperammonemia, leading to cognitive and neurological impairment. Untreated hyperammonemia can result in early patient death.

Yang et al. (2016) reported an HDR-based editing approach using the CRISPR/Cas9 system and the hepatotropic adeno-associated virus 8 (AAV8) as a viral delivery strategy to correct a punctual G-to-A on the exon 4 at the OTC locus on an in vivo model of OTC deficiency [56]. Upon intravenous administration in newborn mice, they observed a correction of ~10% of the evaluated hepatic cells, with a consistent recovery of ~20% and ~16% of OTC enzyme activity after 3 and 8 weeks of treatment, respectively. The consequent decrease in blood ammonia levels in treated animals reached the wild-type levels. Despite these positive results, several NHEJ events were identified in adult animals (8 to 10 weeks), which exacerbated the phenotype. In a similar approach, Wang et al. (2020) reported the successful implementation of CRISPR/Cas9 to correct the same transition G-to-A [57]. Upon a single co-injection of two AAV8 vectors, carrying out the CRISPR/Cas9 system and a donor vector containing a codon-optimized human OTC as well as the liver-specific thyroxine-binding globulin (TBG) promoter upstream, the authors reported a recovery ~24% of functional hepatocytes after eight weeks post-treatment with almost the normalization of ammonium and enzyme activity to wild-type levels [57]. Nevertheless, the authors reported the survival of all the animals after seven days post-treatment, even with a high-protein diet.

4.2. Phenylketonuria (PKU).

PKU is a disorder of phenylalanine (Phe) metabolism, resulting in severe impairment in the brain [58]. Current treatment is mainly based on the management of a low Phe diet. Although pharmacological therapies have been evaluated, no positive outcomes have been achieved [59]. For CRISPR/Cas9, it has been assessed to correct a single missense mutation (c.835T>C) on the phenylalanine hydroxylase (PAH) gene in mice models with hyperphenylalaninemia phenotype [60]. After 16 and 24 weeks post-treatment with AAV2/8 carrying out the CRISPR/Cas9 system, an increase in the HDR frequency from ~1% to ~13% was reached by concomitant use of an inhibitor of the NHEJ repair pathway [60]. Consequently, a significant reduction of 4-fold blood Phe concentration was achieved in treated mice compared to the untreated ones. Likewise, liver PAH activity increased up to ~10% of wild-type levels in treated mice. These findings remained stable after 65 weeks of treatment.

Moreover, CRISPR/Cas9 genome editing was used to create a pig model of classical PKU by knocking out exons 6 and 7 [61]. This novel model exhibited high blood and urine Phe concentrations and hypopigmentation, although they did not exhibit any neurocognitive impairment, which is a classical manifestation in PKU patients. Although this PKU pig could be used to assess potential therapeutic alternatives, more accurate models expressing the complete phenotype should be developed.

4.3. Methylmalonic acidemia (MMA).

MMA is a group of disorders caused by defects in the metabolism of branched-chain amino acids, characterized by the accumulation of methylmalonic acid (MAc) [62]. Recently, Schneller et al. (2021) used CRISPR-Cas9 genome editing to develop a knock-in mouse model of MMA that recapitulates severe and partial phenotypes seen with corresponding mutations for severe (p.R106C) and partial (p.G715V) ones on methyl malonyl-CoA mutase (MMUT) gene in patients [63]. For the classical genome edition using Cas9 cut and sgRNA addressed to intron 1 or the starting codon in the exon 1 of albumin locus, the authors found a significant MMUT gene expression increase with the consequent decrease in the MAc plasma concentration throughout six months of treatment [63]. Given that the homologous left arm from the donor used during CRISPR/Cas9-based genome editing for starting codon strategy remains the albumin promoter, the authors found a sustained expression of the MMUT gene and MAc decrease when the donor vector was injected without the presence of Cas9, suggesting that the therapeutical effect was because of the episomal state of donor vector [63]. These exciting findings support a new approach for targeting the start codon since the donor can be knocked out or remains episome into the nucleus.

5. CRISPR/Cas9 and IEM type II: Energy deficiency

IEM type II includes disorders affecting the metabolism of fatty acids, carbohydrates, and mitochondria which affect energy cell homeostasis. CRISPR/Cas9 has been used for the generation of new models, and they will be summarized in the current section.

5.1. Fatty acids oxidation.

LentiCRISPRv2 carrying the CRISPR/Cas9 system was used to edit into 143B human osteosarcoma cells the ACADM gene, which encodes for the medium-chain acyl-CoA dehydrogenase (MCAD) enzyme [64]. 143B MCAD knock-out cells showed a decrease in almost all the OXPHOS complexes, except complex V and outer mitochondrial membrane (TOM) complex, resulting in lower oxygen consumption in the presence of galactose than in control cells. 143B knock-out cells were more sensitive to superoxide mitochondrial-dependent production in the presence of antimycin A, while this effect was not noticed with rotenone. The results suggested that loss of MCAD increases the susceptibility to oxidative stress through complex III [64].

5.2. Carbohydrates metabolism.

Glycogen storage disease type Ia (GSD-Ia) is caused by the deficiency of glucose-6-phosphatase-α (G6PC), characterized by severe hypoglycemia in patients [65]. Recently, Arnaoutova et al. (2021) used CRISPR/Cas9 on a newborn mouse model of the disease, which carried out the p.R83C mutation, a common mutation in GSD-Ia. After a single AAV8 injection containing the CRISPR/Cas9 system (sgRNA and donor), the animals showed an increase in the hepatic G6PC activity (~3% of normal activity) [66]. This slight increase was enough to regulate glucose homeostasis and related metabolites such as triglycerides, cholesterol, lactic acid, and uric acid up to 16 weeks post-treatment.

5.3. Mitochondria-related energy.

As previously described, IEM type II includes all the metabolic processes related to energy production, by which the mitochondrion becomes an essential intracellular organelle. Mitochondria is an endomembranous organelle containing a compact DNA-based genome (mtDNA) [67], which encodes for 13 proteins necessary for the proper oxidative phosphorylation chain function and the subsequent ATP generation [67, 68]. For many years, the mitochondria were considered CRISPR/Cas9 inaccessible organelles because of their biological structure and the limited nucleic acids importation to the mitochondrial matrix [69]. However, Jo et al. (2015), using a FLAG-tagged spCas9, demonstrated the presence of Cas9 in the nucleus, cytoplasm, and mitochondria. After transfection of sgRNAs against COX1 and COX3, more than 80% of the protein reduction was achieved [70]. These results suggest the feasibility of CRISPR/Cas9-mediated mitochondrial genome editing without additional modification of the canonical system. A new version of Cas9, named mtCas9, was created by incorporating a mitochondrial targeting sequence of cytochrome C (MTS-Cox), improving the organelle specificity of the Cas9 with non-compromise in mitochondrial genome editing [70]. Nevertheless, the biological process underlying the mitochondrial Cas9 transport continues to be elucidated. One related approach using MTS-Cox was later used on zebrafish embryos with similar outcomes [71].

Since the mtDNA encodes for only 1% of the total mitochondrial proteins [72], exploring the CRISPR/Cas9-based genome editing in the nucleus remains crucial. SURF1 is a nuclear gene located in chromosome 9 implicated in Leigh syndrome (LS), a mitochondrial disease with central nervous compromise. Recently, Inak et al., 2021 corrected the transition c.769G>A (p.G257R) on the SURF1 gene in fibroblasts LS-derived iPSCs, and they introduced the same mutation on normal fibroblasts- iPSCs using lipotransfection of CRISPR/Cas9, to study the impact of the mutation beyond the genetic background. Using organoid cultures from iPSCs cells, the authors showed impaired proliferation and maturation profiles when c.769G>A was present, as well as an arrest in the glycolytic proliferative state of neuronal progenitor cells [73], providing novel insights into the physiopathology of the LS syndrome.

Although CRISPR/Cas9 has not been implemented as a therapeutic alternative for pyruvate dehydrogenase (PDHD) or carboxylase deficiencies, Liu et al., 2019 reported its use for PDHA1 knock-out. PDHA1 gene encodes for pyruvate dehydrogenase E1 alpha subunit, which is pivotal for converting pyruvate into acetyl coenzyme A and flux OXPHOS [74]. Using esophageal squamous cancer cell line KYSE450, authors obtained a stable PDHA1 non-expressing cell line, which resulted in a metabolic reprogramming to Warburg status instead of the normal oxidative profile, with an increased glucose and glutamine uptake to KYSE450^PDHA1+/+ [74]. Decreased OXPHOS is a common finding in patients suffering from PDHD deficiency [75, 76].

6. CRISPR/Cas9 and IEM type III: Complex molecule

IEM type III represent the inherited metabolic disease involving all the organelles except mitochondria. Consequently, complex molecules are involved mainly in the lysosome, peroxisome, and Golgi apparatus.

6.1. Lysosomal storage disorders-LSDs.

LSDs are a heterogeneous group of diseases, all characterized by the accumulation of substrates partially or not degraded into the lysosome because of the impaired activity of proteins related to the lysosomal function [10, 77]. According to the accumulated substrate, LSDs can be grouped as glycogenosis, lipoproteinoses, oligosaccharidosis, mucolipidoses, mucopolysaccharidoses, neuronal ceroid lipofuscinoses, and sphingolipidoses, among others [77].

6.1.1. Mucopolysaccharidoses.

Mucopolysaccharidoses (MPS) are a group of LSDs caused by the total or partial absence of a specific lysosomal enzyme required for the degradation of glycosaminoglycans (GAGs), heparan sulfate (HS), dermatan sulfate (DS), chondroitin sulfate (CS), keratan sulfate (KS) or hyaluronic acid (HA). The lack of a specific lysosomal enzyme and the resulting GAG accumulation lead to twelve MPS, which may involve systemic and central nervous system (CNS) manifestations [78].

6.1.1.1. MPS I or Hurler syndrome.

MPS I is caused due to the deficiency of alpha-L-iduronidase (IDUA), leading to the accumulation of GAGs (DS and HS) [78, 79]. A non-viral vector was used to deliver a plasmid encoding for the gRNA and Cas9 protein and a donor template to correct the p.Trp402Ter mutation [80]. IDUA activity increased after nanoemulsions or liposomal co-complexed with the CRISPR-Cas9 system and donor fragment in MPS I fibroblasts [80, 81]. In vivo evaluation in neonatal MPS I mice showed an increased enzyme activity in serum and several tissues, including the brain [82, 83]. Although GAGs are significantly reduced in several tissues such as the kidney, liver, lung, and spleen in treated animals respect untreated ones, the brain remains unaffected by the CRISPR/Cas9-based GT in GAGs and the neuroinflammation profile, suggesting the need to improve the current strategies [82–84].

On the other hand, ex vivo CRISPR/Cas9 gene therapy using the CCR5 locus as a safe harbor in human CD34⁺ cells allowed the expression of IDUA and correction of musculoskeletal, neurological, and biochemical manifestations in a MPS I immunocompromised mice. The gene-editing efficacy in Hematopoietic Stem Progenitor Cells (HSPCs) was improved by chemically modifying gRNA with 2′-O-methyl 3′phosphorothioate and was delivered into cells with Cas-9 through electroporation. In addition, AAV6 was used as a viral vector to deliver a donor template, after which IDUA activity was demonstrated to be reconstituted in serum and organs with a reduction in GAGs except for the brain [85]. The mouse model transplanted showed improved neurological and bone pathology, demonstrating a safe and effective approach to treating MPS.

6.1.1.2. MPS IVA or Morquio syndrome A.

MPS IVA is caused by a mutation in the gene coding for the enzyme N-acetylglucosamine-6-sulfatase leading to the accumulation of GAGs, keratan, and chondroitin 6-sulfate [86]. Bone pathology is the hallmark clinical manifestation of MPS IVA. Recently, a modification of the Cas9 enzyme, Cas9 nickase, was utilized for targeting the AAVS1 safe harbor locus for knock-in of N-acetylgalactosamine-6-sulfate sulfatase (GALNS) cDNA in MPS IVA patients’ fibroblasts with mutation (p.R386C/p.F285del). The study demonstrated successful knock-in using the CRISPR-Cas9 system, evident with an increase in GALNS enzyme activity up to 40% of WT and a reduction in GAGs, lysosomal mass, and mitochondrial-dependent oxidative stress [87]. Also, this CRISPR/nCas9-based system was assessed on human MPS IVA fibroblasts carrying different mutations [88]. The results showed an apparent mutation-dependent phenotype recovery, suggesting that potential epigenetic features could influence the gene editing response [88]. Epigenetics findings on IEM will be discussed later.

6.1.2. Neuronal ceroid lipofuscinoses-NCLs.

NCLs are a group of severe LSD affecting the CNS. CRISPR/Cas9 has been used chiefly for creating new models for type I [89], II [90], III [91], and V [92] by targeting PPT1, TPP1, CLN3, and CLN5 genes, respectively. Although NCL II, III, and IV have been developed using in vitro approaches, NCL I CRISPR/Cas9-based model was done in sheep. Animals showed reduced PPT1 enzyme activity and typical symptoms such as brain atrophy accompanied by brain mass reduction. These animals constitute novel in vivo models for evaluating potential therapeutic strategies, including CRISPR/Cas9 [89].

6.1.3. Sphingolipidoses.

Accumulating sphingolipids, a group of complex lipids containing a backbone of sphingosine [42], cause the sphingolipids, which comprise the following diseases: gangliosidoses, Fabry (FD), Metachromatic leukodystrophy, Krabbe, Gaucher (GD), Niemann-Pick, and Farber. Concerning CRISPR/Cas9-based genome editing, a previous paper by Santos & Amaral, 2019 reviewed its potential therapeutical use [93]. Herein, we will include the newly available data ranging from 2019 to the date.

6.1.3.1. Gaucher.

Gaucher disease is caused by mutations in the GBA gene, resulting in deficient glucocerebrosidase (GCase) activity, affecting primarily macrophage lineage [94]. Using an ex vivo strategy, Scharenberg et al., 2020 showed the modification of HSPCs by introducing the normal sequence of GCase to be inserted into locus CCR5 using CRISPR/Cas9. Upon AAV6 transduction containing the CRISPR/Cas9 system, HSPCs were differentiated to monocyte/macrophage lineage, mediating the phagocytosis process and expressing GCase [95]. Later, engineered-HSPCs were transplanted into immunodeficient mice. Serial transplantation resulted in the preservation of chimerism of edited HSPCs and up to ~2-fold enzyme activity on macrophages isolated from bone marrow [95]. Further assays using mice models of Gaucher should be performed to know the real therapeutic potential of this novel approach.

6.1.3.2. Gangliosidoses.

Gangliosidoses comprise GM1 and GM2 gangliosidoses. For GM1 gangliosidoses, a complete review was published [96]. In the case of GM2 gangliosidoses, these include three different diseases Sandhoff, Tay-Sachs, and the AB variant [42].

In an in vivo approach using a Sandhoff mice model, Ou et al. (2020) evaluated the hydrodynamic administration of AAV8 containing a CRISPR/Cas9 system to knock in a coding sequence of HEXM in the albumin locus [97]. HEXM is a chimeric protein composed of the alfa α-subunit active site and the stable β -subunit interface [98]. The long-term edition showed supraphysiological levels of HEXM activity in plasma, leading to a significant increase in the brain compared to untreated animals. Although a substantial reduction in gangliosides was observed in the liver, heart, and spleen close to wild-type levels, that profile was not achieved in the brain. Interestingly, the rotarod analysis showed an improvement in treated animals compared to untreated ones [97], suggesting a positive outcome despite the persistent brain GM2 gangliosides accumulation.

Recently, a new strategy using a variant of Cas9, denominated nickase Cas9, in which the RuvC motif has been mutated [39], was used as proof of concept using human fibroblasts. In this study, the authors used an expression cassette containing the functional version of HEXA, HEXB cDNA, and two sgRNA to address nCas9 at the AAVS1 locus as a safe harbor [99]. The results showed that 30 days post-treatment, classical biomarkers such as lysosomal mass, mitochondrial-dependent oxidative stress, and secondary glycosaminoglycan accumulation were normalized [99].

6.1.3.3. Metachromatic leukodystrophy-MLD.

MLD is caused by mutations in the ARSA gene, which encodes for the arylsulfatase-A. MLD patients present impairment of the central and peripheral nervous systems [33]. In a study using CD34⁺ HSPCs isolated from MLD patients with infantile and juvenile-onset, CRISPR/Cas9 was evaluated as a therapeutic approach to correct point mutations. As proof of concept, the authors reported an increase in ARSA activity between 20- and 32-fold compared with untreated cells. A significant unwanted effect was observed on the proliferation/viability after electroporation of AAV6 carrying the CRISPR/Cas9 system, although any impact on lineage differentiation was not reported [100].

6.2. Peroxisomal diseases (PDs).

PDs are caused by defects in biogenesis, peroxisomal enzyme deficiency, or defective substrate transport to the peroxisome [101]. X-linked adrenoleukodystrophy (X-ALD) is the most frequent PD. It is associated with mutations in the ABCD1 gene, which encodes for an ATP-binding cassette transporter responsible for transporting very long-chain fatty acids [102]. CRISPR/Cas9 has been used for the development of X-ALD models using microglia [103], embryonic stem cells [104], and fibroblast patient-derived iPSCs [105]. Also, CRISPR/Cas9 was recently used to correct the disease in an X-ALD mouse model through homology-independent targeted integration (HITI) [106]. After intravenous administration of AAV9, the authors observed an increase in the ABCD1 mRNA compared to untreated animals and a reduction in C24:0-LysoPC and C26:0-LysoPC in plasma after three weeks of treatment [106].

6.3. Glycosylation defects (GDs).

GDs are a heterogeneous group of disorders characterized by abnormal protein and lipids glycosylation patterns [107]. In a recent paper published by Ng et al., 2021 the authors identified the mutation c.1267C>T (p.Arg423*) on the SLC37A4 gene in 7 patients, all of them with coagulopathies, liver dysfunction, and abnormal N-glycans on plasma. [108]. SLC37A4 encodes for the glucose 6-phosphate translocase, which is located on the endoplasmic reticulum (ER). Given the liver-specific impact of this mutation, in this study, CRISPR/Cas9 was used to include that mutation on Huh7 cells. A miss localization between the ER/Golgi apparatus compartments was identified. This phenomenon was accompanied by a decreased Golgi pH and an altered architecture in this organelle [108], suggesting an impaired activity of the N- and O-glycan modifying enzymes.

7. CRISPR/Cas9 and Epigenetics

Epigenetics involves studying environmental factors affecting DNA expression without changes in the nucleotide sequences [109]. Epigenetic changes seem to impact metabolic processes more than previously thought considerably. The role of this regulation has been explored in different metabolic disorders such as diabetes, obesity, and cardiovascular diseases [110]. The epigenetic mechanisms implicated in metabolic disorders are DNA methylation, histone modifications, and micro-RNAs (Fig. 4) [109, 111]. Epigenetic mechanisms and metabolism are closely related. In fact, this interaction is demonstrated by the regulation of important metabolites that are essential in the signaling pathways of epigenetic mechanisms; for example, DNA methyltransferases (DNMTs) and histone methyltransferases (HMTs) require S-adenosyl methionine (SAM). Likewise, the histone acetyltransferases (HATs) require acetyl-CoA for the addition of the acetyl group to the residues of histone tail proteins [16, 112], while histone deacetylation needs nicotinamide adenine dinucleotide (NAD⁺) [16]. In this sense, metabolic alterations are characterized by dysregulation of the intracellular and extracellular metabolite levels that could alter signaling pathways of different epigenetic mechanisms [16].

Figure 4.

Overall epigenetic mechanisms. A. In general, three epigenetic mechanisms can activate (Green arrow) or repress (Blocking red circle) the gene expression. Histone modifications can regulate gene expression through activation or repression [16, 115, 116]. DNA packs histone protein groups into a chromatin structure; depending on the kind of residue and type of modification, the architecture of the chromatin structure can be changed [16]. Some modifications of the histone proteins include methylation (Me), acetylation (Ac), phosphorylation (Ph), ubiquitination (Ub), and Sumoylation (Su) [110, 115–117]. Several modifications can have a dual role. For example, the histone three lysine 4 methylation and the Histone 2A ubiquitination activate/allow the transcription, whereas the histone 3 lysine 9 dimethylation and Histone 2B ubiquitination reprime it. DNA methylation is based on the transfer of a methyl group to the fifth carbon of a cytosine residue located in Cytosine-phosphate-Guanine (CpG) dinucleotides of a gene promoter region or regulatory elements [116, 118]. The methylation of cytosine residues in CpGs controls the gene expression in two ways: 1) Impeding the binding of transcription factors. 2) Recruiting regulatory proteins such as the methyl CpG binding domain (MBD), which is a member of a family of proteins that can modify and regulate chromatin structures by acetylation or deacetylation of histones proteins [116, 119, 120]. Finally, MicroRNAs (miRNAs) can modulate protein levels through post-transcriptional regulation without affecting the genome. B. Potential use of CRISPR/Cas9 includes dead Cas9 (dCas9), which can be coupled to several modifying proteins like Tte1 to remove or include epigenetic marks altered on IEM. This figure was created with Biorender.com.

Although epigenetic regulation has not been widely investigated in IEM, some studies have shown the relevance of epigenetic mechanisms in metabolic regulation and rare diseases. Given the impact of epigenetic regulation on gene expression and the reversibility of the epigenetic marks, it could be a promising opportunity to explore new treatments that can correct or compensate for the abnormal protein in IEM and other rare diseases [16].

7.1. CRISPR/dCas system and epigenetics.

Recently, a technology evolving CRISPR/dCas design and epigenetic editing (EGE) has been explored. This technique is known as CRISPR-based epigenetic engineering, which is based on the ability of a nuclease-deactivated Cas (dCas) protein to bind to a specific locus by a gRNA-sequence and an epigenetic enzyme, with can edit or reprogram exiting epigenetics signature derived from DNA methylation and histone modification [16, 113, 114]. Epigenetic enzymes used in EGE, also call as epigenetic writers or erasers, depend on their function; writers are DNA methyltransferases, histone, and lysine/methyltransferases. On the other hand, erasers are histone deacetylases/demethylases or can remove DNA methylation (Fig. 4) [16, 114].

7.1.1. Epigenetics alterations in MPS.

Some reporters have shown modifications in DNA methylation patterns in MPS. Analysis of the GALNS gene in MPS IVA patients showed an interesting correlation between the distribution of transitional mutation and levels of DNA methylation in CpG sites [112, 121]. Increased DNA methylation in the coding region of the GALNS gene elevates the mutability and deficient expression of the GALNS enzyme [116, 121]. Moreover, MPS IVA fibroblasts containing different mutations were recently characterized for histone acetylation (H3K14ac) and histone methylation (H3K9me3) [122]. An increased pattern for both epigenetic markers was reported, which was restored upon treatment with Trichostatin A (TSA), a histone deacetylases inhibitor, supporting the crucial role of epigenetic changes in the MPS IVA pathology [122]. A similar increased methylation pattern was reported early for MPS I, using bisulfite genomic sequencing analysis [116, 123].

7.1.2. Epigenetic alterations in Sphingolipidoses.

Epigenetic alterations have been detected for FD, GD, and Niemann Pick type C [124–128]. For FD, Hübner et al. (2015) showed the presence of methylated CpG islands at position 78504 in the promoter of the calcitonin receptor gene on blood leukocytes from FD patients under enzyme replacement therapy (ERT) [124]. In GD, an increase in the histone H4 acetylation [125] and miRNA deregulation [126] have been reported. MiR127–5p, which targets LIMP2, a protein involved in sorting the GCase from ER to the lysosome, down-regulated the GCase activity on Gaucher cells [126, 127]. Interestingly, miR16–5p and miR195–5p have shown the opposite effect to miR127–5p, leading to the upregulation of the GCase [126]. The potential modulation of miRNA could be explored to develop novel therapeutical strategies for GD. Finally, for Niemann-Pick type C (NPC), early studies showed the therapeutic effect of histone deacetylase inhibitors (HDACi) [128, 129]. Some HDACi, such as panobinostat (LBH589) or TSA at sub-micromolar concentrations, normalize the cholesterol homeostasis in NPC fibroblasts.

As observed, new evidence supports the relationship between IEM and epigenetic changes; however, no preclinical or clinical trials are undergoing. Consequently, treatments with CRISPR/dCas-based epigenetic approaches could offer novel therapeutic strategies for these monogenic diseases [16].

8. CRISPR/Cas9 delivery approaches

CRISPR/Cas9 can be delivered to several cells or tissues using different platforms: Ribonucleoprotein complex (RNP), mRNA, or DNA [48]. Depending on the platform chosen, both viral and non-viral vectors can be implemented [44, 130]. In figure 5, we present the most used delivery strategies for IEM.

Figure 5.

CRISPR/Cas9 platforms and current delivery strategies for IEM. A. CRISPR/Cas9 system can be used as a ribonucleoprotein complex (RNP), RNA, or DNA. B. To date, 30 classical GT-based clinical trials are ongoing for IEM type I (n=3), II (n=5), and III (n=22) (https://clinicaltrials.gov). C. Schematic representation of AAVs. Note that the maximum cargo is around 4.5Kbs. Promoters (P), several genes of interest (GOI), or regulatory (R) sequences (e.g., Poly A, etc.) can be inserted. ITR: Inverted terminal repeats. D. Despite viral vectors remaining the most popular carriers for CRISPR/Cas9, novel non-viral vectors have been tested in IEM. Primarily, cationic liposomes with encapsulated or attached DNA are used. Also, liposomes containing magnetic nanoparticles (MNP), susceptible to being manipulated through magnetic fields, were recently tested. Both general strategies, viral and non-viral vectors, can be modified with several molecules (See the square, circle, and diamond on MNPs) acting as ligands to increase their specificity to relevant IEM tissues. This figure was created with Biorender.com.

8.1. Viral vectors.

Viral vectors used for gene therapy remain limited to nucleic acid transport [131–133]. Viral vectors continue to be utilized to deliver gene information to cells because of their natural capacity to infect cells [134]. Clinically relevant viral vectors include adenovirus (Ad), AAVs, herpes simplex virus (HSV), retrovirus, and lentivirus [134, 135]. A complete review of molecular structure and mechanism was recently published by Bulcha et al., 2021 [135]. Regarding CRISPR/Cas9 and IEM, AAVs are the commonest and have shown positive outcomes in both in vitro and in vivo models. AAVs are small, single-strand DNA viruses with low pathogenicity, a packing capacity of around ~4.5 Kb, and broad tissue tropism depending on the serotype [134, 136]. Moreover, capsid modification with specific ligands could increase tissue targeting. For example, AAV2 with an aspartic acid octapeptide on the N-terminal region of the VP2 capsid displayed a significant biodistribution in bone after intravenous administration in MPS IVA animals, possibly by the interactions between positively charged amino acids and hydroxyapatite [137]. Like these interesting approaches, several modifications in the capsid can be made to improve the natural tropism of viral vectors carrying the CRISPR/Cas9 system. According to Clinicaltrials.gov (https://www.clinicaltrials.gov, Accessed by November 1, 2022), no clinical trials involving the CRISPR/Cas9 system for treating IEM have started; however, 30 classical gene therapies by using viral vectors as carriers are ongoing: IEM type I - aromatic L-amino acid decarboxylase deficiency (1), methylmalonic acidemia (n=1), and OTC (n=1), type 2 - glycogen storage disease type Ia (n=3), Wilson disease (n=2), and type 3 - Fabry (n=3), Krabbe (n=1), Pompe (n=2), leukodystrophy (n=4), neuronal ceroid lipofuscinosis (n=1), GM1 Gangliosidosis (n=3), GM2 Gangliosidosis (n=1), MPS II (n=2), MPS IIIA (n=4), and MPS VI (n=1). These clinical trials are significant advances in GT-based strategies; however, IEM remain understudied since only 15 out of over 1,500 disorders are under clinical trials.

8.2. Non-viral vectors.

Non-viral vectors use several materials to transport and deliver molecules, such as RNA, DNA, and proteins [130, 138]. These include polymers, lipids, peptides, inorganic structures, and hybrid systems [139]. For IEM, some studies, mainly focused on MPS I, have used cationic liposomes as carriers of the CRISPR/Cas9 system because of their simple synthesis and biocompatibility with biological membranes [81–83]. More recently, magnetoliposomes (MLPs) were reported as a potential carrier for CRISPR/nCas9 with encouraging results on several models of LSD [99]. MLPs are hybrid systems composed of magnetite (Fe₃O₄) nanoparticles (MNPs) and a covering liposome structure (Fig. 5) [140]. Similar to viral and non-viral vectors, MLPs can be modified with several proteins acting as ligands to confer specific properties [140]. For example, conjugation of several proteins, such as outer membrane protein A from Escherichia coli or buforine II, has shown endosomal escaping of MNPs upon their uptake by phagocytic and non-phagocytic cells [141, 142]. Since they are ferromagnetic materials [143], the external magnetic field could be another mechanism for the controlled delivery of the CRISPR/Cas9 system in relevant IEM-related targeting tissues. This assumption has been tested for cancer-related disorders confirming its suitability [144].

9. Current challenges for CRISPR/Cas9 implementation and future perspectives

CRISPR is the foundation of a revolutionary gene-editing system. Over the past decade, the CRISPR/Cas9 system has been widely used for several approaches, including model disease development and therapeutics strategies for IEM. Despite the above, the CRISPR/Cas9 system has not reached clinical practice yet due, at least in part, to several concerns behind its use. In addition to the previously exposed challenges related to delivery systems, unwanted effects due to off-target cuts, immunogenic toxicity, bioethical concerns, and limited regulation in some countries remain major challenges that the CRISPR/Cas9 system should overcome [10]. Bioethical concerns and regulations were recently reviewed and can be consulted elsewhere [145, 146]. Table 1 summarizes the advantages and challenges of the CRISPR/Cas9 system.

Table 1.

Advantages and challenges of vectors in combination with the CRISPR/Cas9 system

Vector	Molecule	Advantages	Challenges	Ref.
Adeno-associated virus	DNA	High transduction efficiency Transduces dividing and non-dividing cells Low immunogenicity	High cost Limited packaging capacity (~4kb) Preexisting neutralizing antibodies Constant expression of Cas9 increases the risk for off-target effects	[134, 136, 147]
Lentivirus	RNA	High transduction efficiency Large packaging capacity (~10kb) Transduces dividing and non-dividing cells Low immunogenicity	High cost Constant expression of Cas9 increases the risk for off-targeting effects	[132, 147]
Non-viral vectors	DNA, RNA, RNP complex	*High biocompatibility Low immunogenicity Limited Cas9 expression when RNA and RNP decrease the risk for off-targeting effects.	**Relative toxicity Limited delivery efficiency Cargo degradation	[148, 149]

*Compatibility and **toxicity depends on the chemical composition of non-viral vectors. A comprehensive review was recently published in this regard by Duan et al., 2021 [148].

Nevertheless, the CRISPR/Cas9 system could make it possible to develop cures for most IEM in the near future. An ex vivo CRISPR-based treatment to treat beta-thalassemia or sickle cell disease conducted by Frangoul et al., 2021 has shown remarkable recoveries [150]. Likewise, in 2020, the first in vivo clinical trials started by injecting CRISPR directly into living humans (subretinal injection in a single eye), aiming to repair a genetic mutation that causes blindness [151]; Leber congenital amaurosis [152, 153]. Various ongoing in vitro and in vivo studies on IEM will shed light on future clinical trials for IEM patients.

Many researchers, clinicians, and families hope CRISPR-based therapies could eventually cure genetic diseases, including IEM, with promising results in animal studies and initial human clinical trials. However, more studies are necessary before the widespread use of the CRISPR/Cas9 system in medicine to avoid unintended consequences.