mtDNA Deletion Diseases Cure Roadmap

We're thrilled to share with the world the Mitochondrial DNA Deletion Diseases Cure Roadmap prepared for Dave and Loretta McGovern and the United Mitochondrial Disease Foundation/UMDF

May 27, 2022

United Mitochondrial Disease Foundation (UMDF) - Ciitizen

Perlara PBC was founded in 2014 as the world’s first biotech public benefit corporation. Today we are a fully decentralized biotech company and globally distributed cure consultancy. We guide families and foundations on their cure odyssey at every step from diagnosis to treatment. The very first step of the journey is to make a plan of action — a Cure Roadmap.

Financial support for the creation of this Cure Roadmap was provided by the Cavan McGovern Family Research Fund at the United Mitochondrial Disease Foundation (UMDF). If you benefit from this document, please consider making a donation to support additional research of mitochondrial disease.

Authors

Justin Donnelly (PhD candidate, Stanford University, Cure Guide, Perlara PBC)
Shiri Zakin, PhD (Cure Guide, Perlara PBC)
Ethan O. Perlstein, PhD (CEO of Perlara PBC & Maggie’s Pearl LLC)

Executive summary

Towards a future where people with mitochondrial DNA deletion disorders (MDDDs) will lead healthy, interactive, and fulfilled lives. This multi-year, multi-modality strategic plan describes the scientific and commercial landscape for finding a treatment for MDDD patients. Based on the current knowledge of the pathogenesis of these disorders, we recommend moving forward with a drug repurposing approach. In parallel, the document further discusses other exploratory therapeutic modalities. We explore the clinical development landscape and the current status of endpoint development and biobanking. Finally, we encourage the establishment of a drug repurposing consortium that will bring together academic and industrial researchers and clinicians to facilitate transfer of knowledge, patients’ information, and biological samples to accelerate the process of drug discovery and improve the lives of MDDDs patients.

The goal of the MDDD Cure Roadmap is to accelerate the identification of a treatment for Cavan McGovern, our pioneer patient whose family is a partner in the development of this Cure Roadmap. We also hope and expect that developments that arise from this work will produce new treatment options across the space of MDDDs, and perhaps even mitochondrial disease more generally.

Our key recommendations for the next 12-24 months are:

In 4-6 months, the preliminary experimentation will be performed to assess the likelihood for success of the drug repurposing. A drug repurposing screen consisting of 1,500-2,000 compounds will be tested on cardiomyocytes derived from Cavan’s samples. The screen will also test a few compounds that according to our experts' interviews and literature overview show promise for treatment of MDDDs.
In 6-12 months, validation of drugs that had positive effect in the cardiomyocytes screen will be tested in a model organism (C. elegans and zebrafish).
In 12-24 months, we hope to generate preclinical data packages based on
our drug repurposing results that will form the basis for tackling fundraising and partnership opportunities with biotech and pharmaceutical companies.

Roadmap Reviewers

Philip Yeske, PhD (Science and Alliance Officer, UMDF)
Tamas Kozicz, MD, PhD (Mayo Clinic)
Michio Hirano, MD (Columbia University)
Ana Andreazza, PhD (University of Toronto)
Jim Stewart, PhD (Newcastle University)

Abstract

While most genetic information in a mammalian cell is stored in nuclear DNA (nDNA), mitochondria are unique among our organelles in that they also carry some of their own genetic information. Mitochondrial DNA deletion disorders consist of three rare diseases - Kearns-Sayre syndrome (KSS), Pearson syndrome, and progressive external ophthalmoplegia (PEO) - that are all characterized by kilobase-scale deletions of mitochondrial DNA (mtDNA). Those deletions impair mitochondrial function. Organs with high energetic requirements such as the brain, heart and muscles are usually most affected. Currently, there is no cure for mitochondrial DNA deletion disorders, and the condition is managed with supportive therapy. In January 2022 a family of a young boy diagnosed with KSS partnered with Perlara and the United Mitochondrial Disease Foundation (UMDF) seeking therapeutic options for their son as well as others living with mitochondrial DNA deletion disorders. Following a comprehensive literature review and interviews with experts from academia and industry, Perlara has produced a document considering potential therapeutic strategies. Our key recommendation is to do a drug repurposing screen in an aim to accelerate the identification of a treatment for mitochondrial DNA deletion disorders. To this end, we propose to establish a drug repurposing consortium- a collaborative network of researchers, clinicians and knowledgeable experts working together to find a cure for those disorders.

Introduction

Disease Overview: Alterations in Mitochondrial Genetics

Unlike most other genetic diseases, MDDDs generally arise sporadically: only about 4% of KSS cases are inherited from maternal mtDNA, for example [2]. As compared to nDNA, mtDNA is significantly more susceptible to mutation. Its proximity to the mitochondrial respiratory chain exposes mtDNA to much higher concentrations of reactive oxygen species (ROS) than nDNA; surprisingly though, it has since been demonstrated that the primary source of mtDNA mutation is a result of the fact that its polymerase is significantly more error-prone [3]. The vast majority of MDDDs are believed to arise early in development, and when and where that lesion occurs can have a major impact on the clinical outcomes of the disease. While not all patients share the same specific deletion, 60% of deleted sequences in KSS are flanked by a 13 bp direct repeat, and an additional 30% are flanked by a variation on this repeat [4]. Interestingly, while this suggests a potential recombination-dependent mechanism for gene deletion, it is worth noting that mammalian mitochondria do not exhibit significant HDR activity - how these apparently contradictory findings fit together remains an open question [5]. It should also be noted that - due to the specific mechanisms of sequencing technology - in some cases, patients may exhibit mitochondrial gene duplications that can be hard to distinguish from deletions. This further complicates the picture of the mitochondrial genetics underlying MDDDs [6].

Most mitochondrial proteins are encoded by nDNA, but thirteen, including key components of the respiratory chain, are encoded in mtDNA. While proteins can be imported into the mitochondria by the TIM/TOM complex, no analogous endogenous mitochondrial RNA importation pathway has been definitively identified in humans. As a result, in order to effectively express mtDNA-encoded mitochondrial proteins, mitochondria also need to internally express their own tRNAs and rRNAs, all of which are also encoded in mtDNA. As many of these key genes are deleted in MDDDs, mutant mitochondria fail to produce functional protein products [7], and this can result in the buildup of paracrystalline deposits and osmiophilic inclusion bodies in affected mitochondria. Tissue-level effects from compromised mitochondrial function are especially notable in muscle, where excessive proliferation of damaged mitochondria results in muscle fibers that appear ragged and red [7]. These myopathies are key drivers of MDDDs’ clinical pathology.

Disease Overview: Heteroplasmy, Purifying Selection, and the Mitochondrial Fitness Landscape

Unlike the nuclear genome, which is usually present in just one or two copies in healthy adult somatic cells and whose replication is strictly regulated as part of the cell cycle, the average mtDNA copy number per cell in an adult human is ~1000 and can vary widely. Moreover, all human mitochondrial genomes, including healthy ones, exhibit some degree of heterogeneity in the sequences of different mtDNA copies - even within a single cell [3]. This is referred to as heteroplasmy. In KSS, the fraction of mutant mitochondria (also referred to herein as the heteroplasmy percentage or heteroplasmy rate) has been found to range between 27% and 85%, and this value can vary significantly between biopsies of the same patient [7]. Cavan McGovern, whose family has partnered with Perlara, PBC and the United Mitochondrial Disease Foundation (UMDF) to produce this Roadmap, was found to have a heteroplasmy percentage of 67% in a blood sample.

Additionally, mtDNA copy number can vary between individual mitochondria and is usually somewhere between two and ten. Mitochondrial copy number is controlled by mtDNA replication as well as the dynamics of three additional processes key to mitochondrial quality control: mitochondrial fusion, fission, and mitophagy [8]. Mitochondrial fusion is controlled by the activity of three GTPases (Mfn1, Mfn2, and Opa1) [9], and fusion of genetically damaged mitochondria with healthy ones allows missing or non-functional mutant genes to be complemented [8]. Conversely, mitochondrial fission (mediated by a single GTPase, Drp1 [9]) can segregate mutant mtDNA copies from healthy ones, and the mutant mitochondria can be eliminated via mitophagy, which can be activated by either a ubiquitin- or receptor-mediated pathway. While fission has been proposed to be necessary for mitophagy, the specific mechanisms of mtDNA segregation are not fully understood - however, it is believed that faults in the dynamics of heteroplasmy management may be key to the molecular pathogenesis of MDDDs [8].

Emerging from the heteroplasmic genetics of the mitochondrial population, the fusion and division of individual mitochondria, and the continuous elimination of damaged mitochondria via mitophagy is a mitochondrial fitness landscape that is tightly regulated and key to mammalian cells’ ability to generate energy, produce key metabolites, and regulate redox balance. Even mitochondrial haplotypes considered to be “wild type” display biased selection, which can vary based on cell type and nuclear genetic background [3]. Additionally, certain cells, including naïve T cells [8], adult stem cells [10], and female germ cells [3], exhibit significantly reduced mitochondrial heteroplasmy due to a process in their development known as purifying selection, whereby the mammalian host cell regulates the fitness landscape and population size of its mitochondria to reduce their genetic variability. Even in patients with MDDDs, mitochondrial heteroplasmy is significantly reduced in actively dividing cells, indicating some mechanistic linkage between cell proliferation and purifying selection [6]. While the existence of endogenous mechanisms to maintain mitochondrial genetic integrity points to potential opportunities to treat MDDDs, the exact mechanism of this process is not fully understood.

Three main non-mutually exclusive hypotheses exist to explain mitochondrial purifying selection: selective replication, selective degradation, and selective transport. Selective replication posits that mutant mtDNA replicates less efficiently than wild-type mtDNA. This hypothesis is supported by evidence that mtDNA replication is upregulated at stages where selection occurs and that Pink1, a kinase associated with mitophagy, accumulates on damaged mitochondria and inhibits translation of the mitochondrial polymerase. Interestingly, a reduction in mitochondrial polymerase levels has also been shown to promote elimination of mutant mtDNA. Selective degradation proposes that mutant mtDNA or mitochondria are selectively eliminated by host cells, while wild-type mtDNA persists. Evidence in support of this hypothesis includes data demonstrating that knock-down of genes associated with mitophagy interfere with purifying selection and that mitophagy processes are upregulated during oocyte differentiation. Selective transport suggests that during the course of differentiation, wild-type mitochondria are selectively transported into the maturing cell. Knockdown of mitochondrial transport genes has been shown to increase mutant heteroplasmy in developing oocytes, but it is unclear whether this is a major driver of purifying selection compared to selective replication and degradation [3].

Disease Overview: Clinical Features of MDDDs

Interestingly, while the three MDDDs - KSS, Pearson syndrome, and PEO - are all characterized by similar large-scale deletions in the mitochondrial genome, the clinical features these diseases present can be quite divergent. This is thought to be driven by variability in the level of mtDNA deletion lesions in different tissues among patients with different MDDDs, though the mechanism behind this variability is unknown [11].

Two MDDDs - KSS and PEO - are primarily neuromuscular diseases [12]. While the effects of PEO are more limited to the neuromuscular deficiency, particularly in the eye [13], KSS is characterized by ophthalmoplegia, cardiac branch block, and pigmentary retinopathy [12] and can affect the entire system, including the kidneys and gut. Fatality from these diseases usually arises due to cardiac conduction failure [14], so these patients are generally given prophylactic pacemakers [13]. As with overall clinical severity, the age at which cardiac failure occurs can be variable. Most patients survive to adulthood and many survive to old age, though they may experience muscle weakness, progressive hearing loss, and other comorbidities [14].

Pearson syndrome, on the other hand, manifests primarily as a hematologic disease [13]. Clinical features of Pearson syndrome include sideroblastic anemia (failure to produce sufficient red blood cells) and pancreatic dysfunction. Without appropriate hematologic management, this disease can be fatal in infancy [13]. However, if managed properly through transfusion therapy, Pearson syndrome can resolve in childhood. This is presumed to be due to selective pressure on hematopoietic stem cells (HSCs), either at the cellular or mitochondrial level, that eliminate damaged mtDNA [6]. Interestingly, both mice [15] and humans [16] with mutations in the mitochondrial DNA polymerase also exhibit this anemic phenotype, potentially suggesting a similar pathological mechanism. Furthermore, given their shared genetic background, it is not surprising that a majority of infants who suffer from Pearson syndrome - 64% in one study [13] - will exhibit KSS later in life.

Cavan McGovern, a partner of Perlara in the development of this Roadmap, is living with severe KSS. He suffers from hearing loss, CPEO, ptosis, urinary and fecal incontinence, and complains regularly of gastrointestinal pain. His heart conduction issues are severe enough to warrant a pacemaker, which was implanted in September of 2020. He has had one seizure, in November 2021, which led to a full week in the hospital. When he regained consciousness he did not recognize his parents, and was severely exhausted when his memory returned. Cavan fatigues easily, and has balance and coordination issues which have resulted in falls and concussions. Stunted growth is being addressed with daily growth hormone injections.

Cavan’s brain MRIs demonstrate progressive changes to the white and grey matter in his brain, and the MRI of his spine also demonstrates disease progression and signal changes. Cavan complains of poor memory, can become confused, and has developed debilitating obsessive-compulsive disorder, for which he recently began taking sertraline (Zoloft). His family is most deeply concerned about the brain and spinal effects of his disease, as well as the conduction issues in his heart. These concerns offer irreplaceable insight into what constitutes a meaningful therapeutic effect in Cavan’s case as well as for KSS patients more generally [17].

The current model that explains how the presence of mtDNA lesions - in the case of MDDDs, large-scale genetic deletions - results in the symptoms observed in the clinic proposes that affected cells exhibit a bioenergetic defect [5]. In other words, in patients with MDDDs, a significant proportion of the mitochondria in some cells are damaged due to their inability to synthesize functional mtDNA-encoded proteins. If these cells rely heavily on the oxidative phosphorylation (OXPHOS) metabolism that mitochondria perform, as muscles or neurons do, they will find themselves unable to meet their energetic requirements. The greater this deficit in a given tissue, the more severe the clinical presentation will be for that system. This model posits that any intervention that corrects the bioenergetic defect in affected tissues will result in significant (if not total) correction of the disorders’ effects [5,18]. As such, there are many therapeutic approaches with the potential to positively impact the progression of MDDDs (see Therapeutic Readiness).

While many questions regarding the relationships between the molecular pathogenesis of MDDDs and their clinical presentation remain to be answered, some key clinical relationships are reported in the literature, and these can be very enlightening with respect to potential therapeutic opportunities. One key modulator of clinical severity in MDDDs is aerobic exercise. Regular exercise was found to increase mitochondrial function by multiple metrics and improve patient quality of life. Additionally, there was some evidence of a reduction in mutant heteroplasmy percentage in patients who engaged in exercise. It is thought that this effect is driven by the lower heteroplasmy percentage in muscle stem cells, known as satellite cells, which are activated during exercise and are responsible for the generation of new muscle cells and tissue [19].

A retrospective natural history study found two key indicators with a strong relationship to disease severity. The first is the age of diagnosis. In general, patients who are diagnosed earlier will suffer from more severe disease, and in particular are more likely to display neurological manifestations of their disease. Patients diagnosed before age 9 are at the highest neurological risk, while those diagnosed after 20 have lower risk. Additionally, this study also found that the presence of mutant mtDNA in blood was more strongly correlated with clinical severity - defined as both a neurological phenotype as well as reduced life expectancy - than the heteroplasmy rate in muscle tissue. Why this is the case is not completely clear, but it has been suggested that the presence of mutant mtDNA in the blood could be related to when the mutation arose in embryogenesis or differences in the ability of HSCs to segregate damaged mitochondria in these different patients [20].

Despite these studies, there are many significant gaps in our clinical understanding of MDDDs. It is important, for example, to acknowledge that while the field has coalesced around an understanding of bioenergetic defects resulting from genetic faults in the mitochondrial network as the key driver of disease pathology in mitochondrial disease, including MDDDs, a possible role for other elements of mitochondrial biology, such as Fe-S cluster biogenesis, has not been fully ruled out [18]. Further, metabolic and protein biomarkers with definitive linkages to clinical outcomes in MDDDs (and in mitochondrial disease more generally) are sorely needed [6]. These will have value in research, but will also be a key component of an effective regulatory strategy (see Clinical Development). Additionally, the effect of the nuclear genetic background on MDDD severity and progression is of great interest, but very challenging to study due to the rare nature of the disease [20]. There are conflicting clinical data surrounding the roles of oxidative and reductive stress in MDDDs - oxidative stress is known to increase DNA damage, but supplementation of the precursors for NAD+, a cellular oxidant, is has been shown to promote mitophagy and improve mitochondrial function [6,8]. Lastly, investigations into the role of diet, particularly ketogenic diet, produced mixed results, unexpectedly putting patients’ safety at risk [21,22]. Greater clarity on these linkages between molecular mechanism and clinical severity will undoubtedly facilitate the development of new MDDD therapeutics.

Disease Models: Cellular Models

A key step in the cure development process is the establishment of disease models that can be used to investigate mechanistic hypotheses or test new potential therapies. Evidence of safety and efficacy in a trustworthy disease model is a key prerequisite to clinical investigation of any new therapeutic. The first of these models are cellular models, which can be cultured in a dish, making the process of generating and handling them relatively simple. Cellular disease models are key for performing large-scale screens for new therapeutics as well as answering important questions about their activity and mechanism of action. Perlara PBC is currently working with the McGovern Family and Dr. Michio Hirano at Columbia University to establish patient-derived cellular models of KSS which may also be applicable to the study of other MDDDs.

There are two central challenges in developing useful patient-derived cellular models for mitochondrial disease. Given that the key driver of the disease is thought to be a bioenergetic defect in affected tissues, in order to be maximally representative of the clinical condition, cells should meet two key criteria: they should contain defective mitochondria (i.e. they should have a sufficiently high mutant heteroplasmy rate), and they should rely on OXPHOS to meet their energy requirements [23]. It turns out that each of these criteria represents a challenge for cultured cells. The most technically straightforward mammalian cells to culture are those that are actively proliferating. However, actively proliferating cells rely more heavily on glycolytic metabolism, which produces the carbon-based building blocks and reducing equivalents necessary to build new cells. They also tend to have a lower heteroplasmy percentage, presumably due to purifying selection [6,23]. Thanks to efforts in academic discovery science, however, a number of options exist to overcome these challenges (see Drug Repurposing).

The simplest patient-derived model to produce and culture is a patient-derived fibroblast model. Fibroblasts are cells that are responsible for the production of collagen and other fibrous proteins that constitute the extracellular matrix. They are resident in nearly all solid tissues, and so can be obtained by a number of different biopsy strategies. They are easily cultured and can be maintained for a significant amount of time. Unfortunately, in Cavan’s case, these cells did not exhibit a significant heteroplasmy rate once they were maintained in culture, creating concerns that they may not be able to accurately represent the bioenergetic defect that drives Cavan’s disease pathology [6].

Another cellular model currently being explored is a patient-derived induced pluripotent stem cell (iPSC) model. iPSCs can be generated from cells obtained from blood samples known as peripheral blood mononuclear cells (PBMCs), a subset of white blood cells. By activating certain genetic programs in these cells, they can be de-differentiated to a pluripotent, stem cell-like state. By controlling certain inputs, these cells can then be re-differentiated into a number of different cell types, including myocytes (muscle cells), neurons, immune cells, and fibroblasts. While iPSCs are actively dividing glycolytic cells, it is possible to convert them to oxidative metabolism. Further, differentiation of these cells can result in a restoration of mutant heteroplasmy (though this mutation may differ from the lesion found in other patient biopsies), and their mitochondrial defects match those observed in the original patient-derived tissue [23, 24, 25].

A final option involves the generation of quiescent (non-dividing), primary cell models. It is likely that, among cellular models, quiescent primary cells are the most representative of the patient’s actual disease, especially if they are derived from tissues that are severely affected. However, they are also challenging to work with. As these cells are not actively dividing, multiple biopsies may be required and the number of experiments that can be performed using them is limited. Many of these cell types are not robust and die easily under culture conditions. Further, non-dividing cells are extremely difficult to genetically engineer. While these limitations make primary, quiescent models incompatible with many experimental approaches including drug screening, they may be important for mechanistic investigation or to validate the mechanism of therapeutic candidates in a lower-throughput context.

Disease Models: Whole-Organism Models

Cell-based and animal models of mitochondrial diseases allow for the experiments necessary to identify potential therapies. A key challenge in the production of mitochondrial disease models is the relative difficulty of creating stable genetic alterations in mtDNA compared to nDNA. Nevertheless, many in vivo models for mitochondrial diseases have been developed by the research community and characterized through basic research.

Invertebrate Models:

Much of the current understanding of mitochondrial diseases originally came from studies in S. cerevisiae. Yeast have a unique ability to survive mutations that inactivate OXPHOS in the presence of fermentable sugars. Inhibition of the glycolytic pathway allows for the identification of mechanisms that restore OXPHOS [26]. Another advantage of using yeast for studying mitochondrial deletions is that, because yeast are unable to stably maintain heteroplasmic mtDNA [27], it is relatively easy to obtain homoplasmic yeast populations in which all mtDNA molecules carry a mutation of interest.

Caenorhabditis elegans (C. elegans) and Caenorhabditis briggsae (C. briggsae) are excellent multicellular model organisms for studying human mitochondrial diseases. Many genes in those model systems have orthologs in humans. Studies in C. elegans have shown that mitochondrial respiratory chain mutations can mimic symptoms of human primary mitochondrial disorders (neuromuscular deficits, developmental delay, altered anesthetic sensitivity, and increased lactate levels) [28,29]. Drosophila melanogaster (fruit fly) is another whole-animal model for studying the mechanisms underlying disorders of mitochondrial function. Human and Drosophila mtDNAs encode the same set of products, and the nucleus-encoded genes required for mitochondrial function are also conserved. In addition, Drosophila contain sufficiently complex organ systems to effectively recapitulate many basic symptoms of mitochondrial diseases [30,31].

Currently, however, C. elegans is the only whole animal model developed that can recapitulate large-scale mtDNA deletion [18,32]. This can be overcome in other species by establishing a pharmacological model - using a combination of compounds inhibiting different pathways of cellular respiration. All of these models allow for HTS of a large number of compounds for their ability to restore mitochondrial function, and can be used for first line screening or for hit validation following a drug repurposing screen in patient derived cells.

Vertebrate Models:

Owing to their genetic similarity to humans, mouse models are indispensable for studying the pathological mechanism of mitochondrial disorders, as well as for searching new treatments. Numerous mouse models of mitochondrial disorders (targeting both mitochondrial and nuclear genes) have been established, including a mouse with a single mtDNA deletion at high mutation load (Mito-mouse) [33]. The deletion region of mtDNA almost completely matches the common deletion region reported in human patients with chronic PEO and KSS. Mito-mice could be considered as a mouse model for KSS [34].

Pigs: Pigs are excellent models of human physiology and pathophysiology. Moreover, embryology and development processes in pigs are very similar to that of the human. Compared to mice, pigs have a longer gestation period. This allows mtDNA mass to accumulate, and thus more variants accumulate during the embryonic development [35]. For this reason, while not all mtDNA lesions found in human mitochondrial disease produce a phenotype in mice, they can produce this effect in pigs, despite the fact that humans and mice have a more similar genetic background [5]. As a result, pigs should be considered as an important animal model in any preclinical studies submitted as part of an investigational new drug (IND) application to FDA.

Zebrafish: A high degree of conservation of molecules and processes between zebrafish and humans makes this model an attractive one for human diseases. Zebrafish mtDNA has approximately 70% identity to human mtDNA and the same complement of 37 genes. Moreover, A zebrafish model of a single large deletion spanning from nd5 to atp8, mimicking the common deletion found in Kearns-Sayre syndrome and Pearson syndrome, has been developed [36]. A clear advantage of the zebrafish model is its accessibility for in-vivo imaging experiments. Mitochondria can be imaged in zebrafish using fluorescent proteins or chemical probes [37].

Therapeutic Readiness

There are several therapeutic modalities with potential applicability to intervene in the currently-understood mechanisms of MDDDs to benefit patients. There is significant variability of development strategy, precedence, timeline, likelihood of success for these different modalities. In this section, we outline these modalities and situate them in the context of an overall treatment strategy for Cavan as well as other patients suffering from MDDDs. The first of these is drug screening, which Perlara is currently implementing in collaboration with academic partners (see Biotechnology Readiness). The second modality, mitochondrial transplantation, is in clinical development by a biotechnology company that could potentially be accessed by compassionate use. Finally, we explore two additional modalities that are currently under development - while their development timelines are longer, they are useful starting points to understand how the MDDD treatment arsenal will expand in the next decade. The end of each subsection highlights key challenges for each approach and, when possible, potential short-term opportunities to gain greater insight into their feasibility.

Modalities: Drug Screening

The fastest route to a clinical therapeutic that could make a meaningful impact on the course of Cavan’s disease (and of others who suffer from MDDDs) is to screen drugs that are already clinically approved for activities that could potentially be therapeutically useful. Given that the specific mechanisms of both disease pathology and drug activity are not fully understood, in many cases unbiased drug repurposing screens can yield valuable and promising results.

There are two central choices to sort out in the design of a drug screening campaign: a readout of drug activity and a model system in which to test it. Readouts can be either direct or functional. Direct readouts report directly on the activity of a protein or the expression of a gene, while functional readouts assess some phenotypic change indicative of therapeutic activity. While direct readouts are easier to interpret, functional readouts have the advantage of being mechanistically agnostic - it does not matter how they produce the phenotypic effect that we observe, as long as they produce it. As previously discussed, the most well accepted model of MDDDs posits that clinical phenotypes are observed as a result of a bioenergetic defect arising from large-scale mtDNA deletions (see Clinical Features). Thus, a functional readout reporting on mitochondrial bioenergetic function is likely to be the best option for a repurposing screen.

One strategy to investigate mitochondrial energy production is to assess ATP levels or the ATP/ADP ratio. ATP level is an excellent readout of OXPHOS activity because OXPHOS produces far more ATP than other forms of carbon utilization such as glycolysis. On the other hand, by restricting cells to oxidative metabolism we can more easily interrogate this energy deficit using a more functional readout. One way to accomplish this is to replace the high concentration of glucose in cell culture media with galactose - as a result, this approach is known as a galactose challenge assay. Most cells in culture rely heavily on glycolysis for their metabolism (see Cellular Models). Removing glucose and replacing it with another sugar, galactose, forces cells to rely almost entirely on OXPHOS, performed by mitochondria, to meet their energy demands [38]. Cells with a more functional mitochondrial population will be healthier, survive more, or proliferate more rapidly under these conditions, and these readouts can be easily monitored by a variety of strategies. By investigating how different drugs affect this readout in a high-throughput format, we can screen for drugs that can correct or supplement malfunctioning mitochondria.

Selection of an appropriate model system will also be important to promote success of a repurposing campaign. To maximize the chances that the results of this screen will be applicable to Cavan’s specific disease biology, efforts are ongoing in collaboration with the Hirano lab to generate cultured iPSC lines from Cavan’s samples that exhibit his mtDNA deletion at a sufficiently high heteroplasmy rate. Additionally, these cells should effectively represent a tissue that is affected in Cavan’s disease [18,23,39] - such as brain, cardiac muscle, retinal epithelia, or gut. In the past, Perlara PBC has explored partnering with Dr. Tamas Kozicz, a physician-scientist at the Mayo Clinic, to perform drug repurposing screens in cardiomyocytes derived from iPSCs. If the process of generating and validating iPSCs derived from Cavan’s blood is successful, these cells could enable such an approach to be repeated in the context of his disease.

Additionally, the question of which compound library to screen is not trivial. It is feasible to screen up to approximately 2,000 compounds, and several off-the-shelf drug-repurposing libraries are available - however, it will also be advantageous to ensure that key mechanistic pathways and compound classes are included in this library based on our existing knowledge of the disease mechanism (see Disease Overview). This can be achieved either by combining or supplementing compound libraries, or by building a new one.

GTPases that control mitochondrial fusion and fission events, such as Opa1 and Drp1 respectively, could be targets of activators or inhibitors that could be identified through such a screening approach [40], so known GTPase modulators should be well-represented in the library. Interestingly, it has been suggested that an upregulation of both events could achieve therapeutic benefit - by shuffling mtDNA genomes throughout the mitochondrial network through increased mitochondrial fusion and fission, the likelihood that mutant mtDNA will be complemented by a functional wild type chromosome can be elevated [18]. Additionally, the compound 2-deoxyglucose (2DG) has been shown to decrease the heteroplasmy percentage of mutant mtDNA in cells by inhibiting mtDNA synthesis via restriction of glycolytic and glutaminolytic metabolism [41], demonstrating the key importance of these metabolic fluxes in regulating heteroplasmy. Another approach to inhibiting mtDNA synthesis, the inhibition of the mitochondrial polymerase PolG1, has been shown to induce mitochondrial hyperfusion [18] - together, these data indicate that perturbing mtDNA synthesis offers an opportunity to modulate mitochondrial quality control. Alternatively, redox balance has been shown to be a key player in mtDNA damage and MDDD progression [8] - modulators of key redox regulators, such as aldehyde dehydrogenases or precursors to redox-regulators such as NAD+ [8] may prove fruitful additions to the library.

If we are able to obtain hit compounds from our screen that putatively restore mitochondrial function in patient-derived cardiomyocytes, the subsequent challenge will be to validate these hits’ therapeutic potential. Not all screening hits will validate - for example, a compound may affect assay results through a disease-irrelevant mechanism. In this case, an example of such a compound could be a mitogen (a compound that increases cell proliferation in a general fashion). It will be necessary to follow up with a more detailed analysis of each hit compound’s mechanism of action, which may be similar or different to the therapeutic mechanism for its primary indication. One key follow-up strategy will be to perform a Seahorse assay, which measures oxygen consumption and CO2 production by sampling a cell’s growth medium. These data directly report on mitochondrial function and glycolytic activity, respectively [42]. COX staining could be an additional follow-up to measure heteroplasmy, or mitochondrial ATP production could be monitored to look for correction of a bioenergetic defect. Further follow ups in animal models, such as C. elegans and zebrafish, will provide in vivo validation of the compounds’ potential efficacy (see Animal Models).

Lastly, a key challenge for this approach is the question of generalizability. While there is reason to believe that KSS (Cavan’s disease) and other MDDDs share common underlying molecular mechanisms, it is not guaranteed that the most promising hits in Cavan’s cells will also be applicable to other patients - even other patients with KSS [43]. On the other hand, it is possible that a compound that corrects the bioenergetic defect observed in Cavan’s disease could be applicable to a wide array of mitochondrial diseases, not just MDDDs. Validating the generalizability of a screening hit will require access to a broader array of patient-derived cell lines (see Biobanking).

Modalities: Mitochondrial Transplantation Therapy

Another exciting therapeutic approach involves mitochondrial transplantation. The principle behind this approach is that, rather than degrading damaged DNA, promoting clearance of abnormal mitochondria, or compensating for impaired mitochondrial function, new healthy mitochondria can simply be added to the system. This approach is being pioneered by Minovia Therapeutics, which is a biotechnology company developing a strategy to produce healthy mitochondrial populations and administer them to patients. In a preliminary exchange, Minovia expressed openness to treating Cavan under compassionate use (see Biotechnology Readiness).

In Minovia’s approach, mitochondria isolated from maternal peripheral blood mononuclear cells (PBMCs) or healthy donor term placenta are added to patient-derived human CD34+ hematopoietic stem and progenitor cells (HSPCs) isolated from the patients (either from the bone marrow aspiration, umbilical cord blood, or mobilized peripheral blood). The mitochondria are taken up by the patients HSPCs (a process called mitochondria augmentation therapy or MAT). The HSPCs enriched with healthy mitochondria are then transplanted back into the patient. Pre-clinical results from a PolG mutant mouse model with a higher mtDNA mutation rate indicate transfer of donor mitochondria into recipient host hematopoietic cells. Donor mitochondria were also present in myeloid cells, as well as B cells in the peripheral blood [41]. A clinical trial evaluating the safety and therapeutic effects of transplantation of mnv-bm-bld in pediatric patients with Pearson Syndrome was recently completed, however the results have not been published yet [44].

Another clinical trial aimed evaluating the safety of Infusion of MNV-BM-PLC (autologous CD34+ cells enriched with placenta-derived allogeneic mitochondria) in patients with primary mitochondrial diseases associated with mtDNA mutation or deletion is expected to start in the near future [45]. Additionally, it should be noted that while Minovia is the most well-developed company in this space, others, such as Mitrix Biosciences, are also developing mitochondrial transplantation approaches, offering further evidence in favor of the feasibility of and interest in this approach.

A key risk of mitochondrial transplantation worth highlighting is the open question of whether the treatment can have a durable effect in patients with MDDDs. There is considerable evidence that the mitochondrial fitness landscape plays a key role in MDDD pathogenesis (see Disease Overview). If the elevated mutant heteroplasmy percentage in MDDD patients is the result of an ongoing selective pressure, which certainly seems to be the case, it is conceivable that the effect of such treatments will be temporary and brief. Nevertheless, experts we interviewed alluded to the potential for engineering solutions to this challenge. Ongoing academic research, potentially in partnership with one of the companies developing mitochondrial transplantation therapy, should seek to address this question, as it will offer valuable insight into the feasibility of this approach in MDDDs specifically.

Modalities: Nuclease Therapy (NTx)

The apparent relationship between mitochondrial heteroplasmy rate and the development of MDDDs raises the question of whether a viable treatment for MDDDs may be to selectively eliminate mutant mtDNA rather than supplement healthy mitochondria. Sequence-specific nucleases are a class of proteins that cut DNA selectively at a given nucleotide sequence, and a significant degree of discovery science in recent decades (including research that was awarded the 2020 Nobel Prize in Chemistry, for CRISPR/Cas9) has gone into identifying and, where possible, reprogramming these nucleases to cut new sequences. These tools are fundamental to current life sciences research and paved the way for the first bona fide gene therapies, first approved by the FDA in 2017.

There are three major classes of programmable nucleases that are of interest for their potential clinical utility: zinc-finger nucleases (ZFNs), TAL-effector nucleases (TALENs), and the CRISPR/Cas9 system. The CRISPR/Cas9 system demonstrates superior programmability thanks to its single-guide RNA (sgRNA), which targets the Cas9 nuclease to a complementary strand of DNA and can thus be easily engineered [46]. However, any NTx targeting mtDNA will have to be able to physically access the mitochondrial matrix, where mtDNA is located. For Cas9, this includes its sgRNA, which cannot be imported to the mitochondria (see Disease Overview). Thus, ZFNs and TALENs are more viable options for mtDNA-targeted NTx. We will focus on ZFNs in this section as a greater degree of proof-of-concept work has been done on this class of nucleases [47,48], though the underlying principles for TALENs are analogous.

ZFNs are engineered proteins composed of a zinc-finger DNA binding domain, a small, zinc ion-dependent, easily engineered protein that binds a specific DNA sequence, fused to a Fok1 nuclease. After the target mtDNA is cut, the mutation can be eliminated either via degradation of the cut mtDNA [47] or homologous recombination with an additional copy of wild-type mtDNA present in the mitochondrion [49]. However, because the DNA-binding sequence itself does not cut DNA, another enzyme, such as a deaminase, can also be utilized [50]. Key to the design of this therapeutic approach, as well as how much of the patient population could potentially benefit, is the question of targetable sequence. 90% of patients with KSS exhibit deletions flanked by some variation of a 13 bp direct repeat [4] (see Disease Overview). Specific cutting at this sequence could enable mutant mtDNA to be distinguished from wild-type copies and could even potentially induce reparative mtDNA recombination. The remaining 10% patients could still potentially benefit from NTx, though likely via a differently-programmed nuclease.

A key advantage of nuclease therapy is the relative straightforwardness of its design compared to small molecules, which are almost always identified via some sort of screening approach. However, given the relative novelty of this modality - zero current approvals with a leading candidate currently in a Phase III clinical trial - preclinical and clinical development of this modality may potentially pose a greater challenge than for more traditional approaches. We simply have less understanding of how these molecules behave in a human body than small molecules, and thus their optimization can be riskier and more challenging. Further, a better understanding of the complex relationship between mtDNA heteroplasmy and the clinical features of MDDDs will shed more light on the potential for NTx to provide clinical benefit.

Modalities: Mitochondrial Autophagy-Targeting Chimeras (AUTACs)

Mitochondrial discovery science has generated significant interest in mitophagy as a key regulator of the mitochondrial fitness landscape and a critical mechanism for cells to clear damaged or dysfunctional mitochondria. While drugs that generally increase mitophagy or even autophagy may offer clinical benefit, a new modality known as an autophagy-targeting chimera (AUTAC) may offer the opportunity to specifically target mutant mitochondria for autophagic degradation.

AUTACs are part of a larger class of molecules known as heterobifunctional molecules. These molecules contain two binding domains and function by recruiting their respective targets into close proximity with each other [51]. The first of these molecules to enter the clinic are known as proteolysis-targeting chimeras (PROTACs). PROTACs work by binding both a target protein and an E3 ubiquitin ligase, an enzyme that attaches ubiquitin modifications to other proteins. Because ubiquitination of a protein acts as a signal for its degradation, PROTACs have been described as “targeted protein degraders”.

It has been known for some time now that protein modification by cGMP, a process found in many bacteria, can promote autophagy. In a 2019 publication, Japanese researchers demonstrated that a small molecule mimetic of this S-guanylation moiety, known as p-fluorobenzylguanine (FBnG), can also recruit and activate cellular autophagy machinery. Moreover, they showed that, when elaborated with a targeting warhead, FBnG can selectively induce autophagic degradation of the target protein. In a nod to PROTAC technology, these researchers termed this technology AUTACs [52]. AUTACs offer an exciting potential strategy to treat MDDDs, given the importance of mitochondrial autophagy in MDDD progression (see Disease Overview). For example, the protein Pink1 has been shown to accumulate on the surface of mutant mitochondria [3] - this makes it an excellent potential target for an AUTAC that could selectively clear mutant mitochondria while sparing wild-type, functional mitochondria.

The authors of this study also tested the ability of the FBnG ligand to induce mitophagy by targeting it to a protein in the outer mitochondrial membrane. Interestingly, this ligand alone was not sufficient to induce mitophagy. Induction of the mitophagy response also required knockdown of the protein Opa1, one of the GTPases regulating mitochondrial fusion (see Disease Overview). Opa1 knockdown alone, it should be noted, also did not promote mitophagy [52]. This study highlights both the interdependence of mitochondrial fusion, fission, and mitophagy processes as well as the potential value of developing pharmacological modulators of each of these processes, as intervention in just one pathway may not be sufficient to induce the desired response.

Unlike the small molecule screening approach previously described, AUTACs as a class have not been clinically validated - in other words, they have not been successfully applied to treat a disease in a living human. While this inherently carries additional risk, the technology has the potential to produce a more significant and more targeted therapeutic effect in patients and is thus worth investigating further. Additionally, the entry of PROTACs into the clinic with leading candidates in Phase II validates the clinical feasibility of directing cell processes like proteolysis using heterobifunctional molecules. Another key challenge is that the medicinal chemistry optimization required for heterobifunctional development is at least as challenging, if not more so, than a traditional inhibitor - even if potent, specific binders to each of the desired target proteins already exist. As a result, the cost and time required to develop this strategy will likely be greater than for traditional small molecules.

Clinical Research and Development Readiness

The development of new therapeutics, particularly in rare disease, requires hand-in-glove coordination between clinicians and researchers - even at the earliest stage. Having reviewed the scientific landscape, laboratory model systems, and promising technologies that define the space of possible therapeutic candidates, here we explore the readiness of the clinical healthcare system to support development of these candidates as well as their evaluation in human patients - the final, costliest, and riskiest stage of the drug development process.

Clinical Research: Patient Identification and Community Engagement

Engagement of a patient community in the drug development process is crucial. Enrollment of a sufficient number of patients in clinical trials for FDA approval is a massive bottleneck in the drug development pipeline. This bottleneck is a major contributor to the reality that the clinical development phase is often the most time-consuming in the entire drug development pipeline. As a result, early efforts to diagnose, identify, and register patients can pay significant dividends, both financially and for the health of the patient community.

Patients with KSS, the most common MDDD, are diagnosed via a clinical triad of PEO with ptosis, cardiac conduction abnormality, and pigmentary retinopathy. PEO itself is diagnosed based on muscle weakness in the eyes and limbs and is sometimes referred to as a KSS spectrum disorder. Pearson syndrome in infants is diagnosed based on the presence of sideroblastic anemia [6,13].

Despite such seemingly clear guidelines, however, patients with mitochondrial disease often undergo a significant “diagnostic odyssey” before receiving their accurate diagnosis - the median number of clinicians required to secure this diagnosis is five [53] - as a result, many patients are likely to go unidentified. KSS can often be mistaken for other disorders of the eye including myasthenia gravis, for example, complicating the diagnostic process [13]. Increased awareness of mitochondrial DNA tests on the part of doctors and access to them on the part of payers will certainly help, but clinicians also face the challenge of heteroplasmy mosaicism, sometimes needing multiple biopsies in order to collect cells where mtDNA deletion can be detected. Less invasive diagnostic strategies or biomarkers to suggest whether a biopsy to look for mtDNA lesions is necessary could also facilitate more accurate diagnoses (see Biomarkers).

Despite these challenges, the North American Mitochondrial Disease Consortium (NAMDC), an organization funded by the National Institutes of Health (NIH), has established a registry of approximately 1,800 MDDD patients at the time of this writing [6]. Additionally, the UMDF, a privately funded non-profit and a partner in the development of this Roadmap, engages patients by helping connect them to clinical, scientific, and community resources. Since many factors influence the specific number of patients necessary to power a given clinical trial (see Clinical Development), these ongoing efforts to identify patients and keep them informed of and invested in the advancement of therapeutic candidates through the drug development pipeline must be continued. Strong patient engagement will be a key advantage with the potential to cut several years from these candidates’ clinical development timelines.

Clinical Research: Natural History

A detailed understanding of the natural history of a disease is fundamentally important to a successful drug development campaign, particularly in the case of a rare disease. There are two major reasons for this: first, the natural progression of a disease offers clues as to how a drug that corrects it might work; second, a robust natural history study has significant regulatory implications. As part of a modernization of its handling of the regulation of new medicines for rare diseases, FDA now accepts natural history as a control group on a case-by-case basis. Much as a strong patient network can accelerate a clinical development timeline by facilitating enrollment, a natural history control group accelerates timelines by reducing the total number of patients required to produce a signal of efficacy that FDA requires in order for a drug to be marketed. This also has the advantage of reducing trial costs and addresses ethical concerns associated with withholding a potentially effective drug from a patient, often a child, with severe disease in order to recapitulate a natural history that we already understand (see Regulatory Landscape and Strategy).

Fortunately, longitudinal natural history studies of KSS/PEO patients are available, the first of which was published in 2007. This study offered clear prognostic indicators of disease: mtDNA in the blood and age of onset. It also offered a key clinical indicator of disease severity, a neurological phenotype, that was strongly associated with an effect on a given patient’s life expectancy [20]. The understandings from this study and others like it will facilitate the establishment of inclusion criteria and clinical endpoints - however, a key gap in our understanding of MDDD natural history remains that warrants attention: molecular biomarkers of disease progression.

Alongside their utility as controls in clinical evaluation, natural history studies have significant implications for regulatory strategy in another respect by furnishing surrogate endpoints (see Regulatory Strategy, Biomarkers). Ongoing natural history investigations should focus on the application of modern laboratory techniques like single-cell genetic analysis, proteomics, and metabolomics to observe longitudinal changes that are associated with increasing disease severity or progression. Additionally, the nuclear genetic background of MDDD patients is often overlooked [6]. While there are currently too few patients to do large-scale genome-wide association studies that are often used to identify molecular determinants of disease in humans, the role of the nuclear genome in regulating the mitochondrial fitness landscape has been established in cells [3] and humans [54], and therefore may offer clues toward useful biomarkers.

Clinical Research: Biobanking

Reliable patient samples are fundamental components to multiple ongoing efforts necessary to develop an MDDD therapeutic (see Models, Natural History). The core of patient sample collection, cataloging, and distribution to researchers is the biobank or biorepository. Fortunately, NAMDC has also begun the process of building a centralized MDDD biobank. They currently have samples, primarily blood samples, from approximately 338 US patients at the time of this writing [6]. Some smaller additional biobanks exist, but NAMDC is likely to constitute the primary resource for drug developers working in this space.

In service of the production of useful cellular models for research and screening, the samples most likely to be useful to biobank include blood, skin, and muscle biopsies. While skin and muscle biopsies are significantly more invasive and painful than blood samples, these samples contain fibroblasts, which are superior screening lines than iPSCs due to their robustness in culture. Additionally, muscle biopsies contain myocytes, which are quiescent and therefore generally have a higher heteroplasmy percentage and reliance on oxidative phosphorylation as their primary metabolic strategy. Such cells are challenging to acquire and culture, but also are the most representative of the physiologic condition possible in a cellular model. Overall, in order to maximize the representativeness of a patient-derived model for a given individual’s disease, cellular models should be generated from tissues most highly affected by the disease whenever possible [18,23].

Clinical Development: Regulatory Landscape and Strategy

The FDA plays a key role in the healthcare system by serving as gatekeeper to the pharmaceutical marketplace - for access, any new therapeutic must demonstrate both safety and efficacy in human clinical data. FDA is also tasked with ethically evaluating and approving the clinical trials necessary to generate these safety and efficacy data based on the risks and potential benefits to the patients involved in the study. As such, a well-developed regulatory strategy is an indispensable component to a successful drug development campaign.

Traditionally, FDA has operated with a degree of aloofness toward drug developers in service of its role as a neutral gatekeeper. However, driven in particular by the need for innovative regulatory approaches to ensure drug approvals for patients suffering from rare diseases, FDA can now be engaged at a much earlier stage in the process and in many cases can operate as a partner and resource. Additionally, there are a series of regulatory designations that FDA now offers to facilitate the development of drugs in therapeutic areas that are traditionally underserved by the biopharmaceutical industry. Fortunately, therapeutic candidates for MDDDs are likely to be eligible for multiple such designations - their specific details, as well as their implications for the overall development process, are explored here.

The Orphan Drug Act (ODA), passed by Congress in 1983, grants FDA the authority to provide specific incentives for the development of drugs for diseases affecting fewer than 200,000 patients per year in the United States. All three MDDDs fall into this category. The most significant outcome of ODA designation is a seven year period of market exclusivity which is awarded to the entity that submits the drug for FDA approval. As market exclusivity, traditionally secured by patents, is the key incentive driving private investment in the high upfront development cost of drugs, the introduction of the ODA designation has radically shifted the calculus around whether a drug development program has the potential to be profitable, particularly if the underlying asset is difficult to protect through patents (often the case with repurposed drugs). Because of this, the certainty of an ODA designation from FDA will be a major advantage in securing private, for-profit funding to finance clinical development of any new therapeutic candidates identified through a repurposing screen.

Any new therapeutic candidate for an MDDD will also be eligible for Fast Track designation, as there are no current therapeutics for MDDDs - this is known as an unmet need. Fast track designees have the advantage of more frequent meetings and communication with FDA, which increases the probability of success of a clinical trial. A similar designation known as Priority Review can accelerate the timeline on which FDA reviews the final application for approval - any new MDDD therapeutic with significant efficacy has a strong likelihood of receiving this designation. Again, because they shorten the clinical development timeline, these designations can both reduce costs and help patients benefit from new treatments faster.

Two other FDA designations exist that warrant attention at this stage: Accelerated Approval and the Breakthrough Designation. Accelerated approval allows for a drug to be approved earlier on the basis of a surrogate endpoint - essentially, this is a biomarker that is not itself a measure of clinical benefit but is thought to be a strong predictor of it. Clinical endpoints often require long observation times and have more variability due to their more subjective nature than biomarkers measured in a laboratory. As such, the ability to conditionally approve a drug based on a surrogate endpoint can also significantly reduce the clinical development timeline. Evaluation on the basis of clinical endpoints then continues while the drug is conditionally approved. If these surrogate endpoint data suggest that a drug is in fact achieving a substantial benefit in a serious condition, that drug will be eligible for Breakthrough designation, which offers all Fast Track features as well as intensive guidance from senior FDA managers - drugs that receive this designation receive the full support of FDA to ensure that a drug’s safety and efficacy is evaluated quickly, efficiently, and reliably. In order to maximize the probability of accessing these two valuable FDA designations, the establishment of a validated surrogate endpoint will be critical.

Clinical Development: Biomarkers and Candidate Surrogate Endpoints

To date, there are no biomarkers for MDDDs that have been approved by the FDA for use as primary surrogate endpoints. In terms of impact on clinical development timeline and cost, this is a significant bottleneck to clinical readiness for an MDDD therapeutic. FDA will require significant evidence that any surrogate biomarker serving as a primary endpoint in a clinical trial is highly representative of the desired clinical outcome. As such, research toward the identification and validation of potential surrogate biomarker candidates and their specific relevance to MDDD disease progression should be prioritized and supported.

Existing research into biomarkers for mitochondrial disease yields some interesting starting points. Serum lactate levels have been proposed as a biomarker for mitochondrial function, as lactate is produced via anaerobic ATP production [55], but in practice its measurement is far too variable to be useful as a surrogate biomarker. This is probably due to some combination of the fact that it varies between individuals and that metabolite levels can continue to change in blood after it is drawn, introducing technical variability. Other metabolic biomarkers that face similar technical challenges include pyruvate, creatine, and amino acids. While none of these biomarkers offer a strong candidate for a surrogate biomarker themselves, they do point to mitochondrial function as a key phenotype associated with the severity and progression of mitochondrial disease.

Following this logic, two protein-based biomarkers have been identified to be upregulated in skeletal muscle tissue in patients exhibiting mitochondrial insufficiency. These are the growth factors FGF-21 and GDF-15. As protein biomarkers, these have the advantage of being detectable via ELISA (rather than mass spectrometry, which is far more expensive and requires specialized operators) and they are less likely to break down after being sampled [55]. Considerable evidence has accumulated to support the predictive value of GDF-15, in particular, for the severity of a different rare mitochondrial disease, TK2-deficient myopathy [56]. GDF-15 is currently being evaluated in a clinical trial for this disease [6]. A key question that remains unanswered is whether this predictive utility is generalizable to MDDDs or to mitochondrial disease in general.

Another approach to a molecular biomarker that has been considered is the heteroplasmy rate. Indeed, heteroplasmy rate has been shown to correlate to some degree with disease outcome, and the existence of mutant mtDNA in the blood in KSS is considered an important prognostic indicator [20]. Though the existence of a relationship between heteroplasmy rate and clinical outcomes is known, much as with metabolites, variability in the measurement of heteroplasmy rate - between patients as well as between individual biopsies of the same patient [6] - complicates its use as a surrogate biomarker. Nevertheless it is likely that a highly effective therapeutic for KSS and other MDDDs will exhibit an effect on the mtDNA mutational load, so improvements to the biomarker’s reproducibility would be beneficial. These refinements can potentially be achieved via prospective longitudinal studies, which have reduced inter-patient variability, and the identification of specific tissues and cell types whose mutational load correlates with MDDD severity, which can reduce the variability between biopsies of the same patient.

Clinical Development: Clinical Endpoints

In nearly all cases, the primary relevant outcome to FDA (and, generally, to patients) is whether the drug produces some kind of clinical benefit that is meaningful. Even if surrogate endpoints are used to approve a drug, ongoing investigation to demonstrate such clinical benefit will continue, and if there is no evidence of clinical benefit the drug can be removed from the marketplace. As such, even though defining strong surrogate endpoints is key at this stage, it is important to consider clinical endpoints as well.

Clinical endpoints are generally more challenging, expensive, and time-consuming to measure than biomarker endpoints. Clinical endpoints are more subjective, measured over longer periods of time, and capture more inherent biological variability than biomarker endpoints do. Thus, reaching a clinical endpoint is a high bar that most therapeutic candidates do not achieve - but to offer a patient anything less than this high standard would be unacceptable.

Several clinical endpoints have been defined for mitochondrial diseases in general and their use is precedented in clinical trials for MDDDs. These include metrics like the Newcastle Mitochondrial Disease Adult Scale (NMDAS), which scores a patient on the basis of a series of clinical indicators of mitochondrial disease, and fatigue [57]. The role of neurological phenotypes as a severity differentiator [20] suggests that clinical endpoints related to neurological and cognitive function are worth consideration. Lastly, the importance of heart conduction defects as a key mechanism of reduced life expectancy for patients with MDDDs [14] indicates that they may be representative of clinical outcomes.

Biotechnology Ecosystem Readiness

Having explored the scientific landscape and considerations for clinical development of a therapeutic candidate, we turn our attention in this final section of the Roadmap to the biotechnology ecosystem, encompassing academia, industrial biotechnology, and public and private funding sources. The successful development of new therapies nearly always requires engagement of multiple players in this ecosystem. In this section, we explore potential partners who can facilitate development of the various modalities previously discussed (see Therapeutic Readiness) as well as patient access to those therapies.

To facilitate effective coordination across this ecosystem, Perlara and the UMDF are partnering to establish a drug repurposing research consortium. This consortium will consist of academic labs, core facilities and screening centers, CROs, and other research facilities engaged in mitochondrial repurposing research. Perlara is available to provide coordination and project management services to this consortium. Objectives, progress, and results will be tracked and managed through an online platform to which consortium members will have access. In fact, it was highlighted that such an online research platform would be of great service to the mitochondrial research community more broadly [23].

The Hirano lab at Columbia will be a key founding member of this consortium and provide necessary patient samples from Cavan, our pioneer patient, including blood-derived iPSCs [6]. We have also added the Kozicz lab at the Mayo Clinic as an additional founding member to develop drug screening assays for mitochondrial disease in iPSC-derived cardiomyocytes. Additionally, we identified several other potential consortium members worth considering. Ana Andreazza at the University of Toronto expressed interest in research collaboration during her interview, and she is an expert in iPSC models of mitochondrial disease. Her lab is developing improved methodologies for culturing and differentiating them, making her a strong potential addition [23]. Samantha Lewis of UC Berkeley has developed C. elegans models of mitochondrial disease, including mtDNA deletion [16] - as part of the consortium, her lab could help develop methodologies for validating screening hits in an animal model.

To facilitate scaling of this repurposing approach to a larger cohort of patients with mitochondrial disease, we are exploring engagement of academic screening cores and CROs that can perform repeated screens on a steady influx of new samples. The potential for collaboration with screening cores at Stanford and UCSF will be investigated initially. In the past, Perlara has worked with Charles River for contract research, but we will accept bids from multiple contractors if we go this direction. Additionally, the MITO2i academic-industry conference happening in April 2022, run by Prof. Andreazza, will serve as an excellent opportunity to identify and engage potential consortium partners.

In parallel to this drug repurposing consortium, there are two key partners we are currently exploring that are developing novel therapeutic modalities. The more advanced of these potential partners is Minovia Therapeutics, developing mitochondrial transplantation therapy (see Modalities). While their therapy is in clinical development, it is not yet available for compassionate use. This is likely due to process development issues involved in producing high-quality transplantable mitochondria at a large scale. The other potential partner is Pretzel Therapeutics, a self-described mitochondrial disease platform company developing mitochondrially-targeted nuclease therapies. While they have not yet reached clinical development, they are less likely to encounter the same degree of process challenges as Minovia due to the fact that large scale production of therapeutic-grade protein is a more well-solved problem than mitochondria. Together, these potential partners offer additional treatment strategies that could complement or augment a drug repurposing approach.

References

Taanman, J. W. (1999). The mitochondrial genome: Structure, transcription, translation and replication. Biochimica et Biophysica Acta - Bioenergetics, 1410(2), 103–123. https://doi.org/10.1016/S0005-2728(98)00161-3
Finsterer, J., & Zarrouk-Mahjoub, S. (2018). Kearns-sayre syndrome is genetically and phenotypically heterogeneous. Pediatria Medica e Chirurgica, 40(1), 31. https://doi.org/10.4081/pmc.2018.193
Jeedigunta, S. P., Minenkova, A. V., Palozzi, J. M., & Hurd, T. R. (2021). Avoiding Extinction: Recent Advances in Understanding Mechanisms of Mitochondrial DNA Purifying Selection in the Germline. Annual Review of Genomics and Human Genetics, 22, 55–80. https://doi.org/10.1146/annurev-genom-121420-081805
Saldaña-Martínez, A., Muñoz, M. de L., Pérez-Ramírez, G., Montiel-Sosa, J. F., Montoya, J., Emperador, S., Ruiz-Pesini, E., Cuevas-Covarrubias, S., López-Valdez, J., & Ramírez, R. G. (2018). Whole sequence of the mitochondrial DNA genome of Kearns Sayre Syndrome patients: Identification of deletions and variants. In Gene (Vol. 688). Elsevier B.V. https://doi.org/10.1016/j.gene.2018.11.085
Stewart, Jim. Expert Interview. 3 Mar 2022.
Hirano, Michio. Expert Interview. 28 Feb 2022.
Bohlega, S., Al-kawi, M. Z., Chavis, P., Haider, A., & Omer, S. (1993). MITOCHONDRIAL DNA DELETION IN KEARNS-SAYRE SYNDROME : Annals of Saudi Medicine, 13(2), 197–202.
Rong, Z., Tu, P., Xu, P., Sun, Y., Yu, F., Tu, N., Guo, L., & Yang, Y. (2021). The Mitochondrial Response to DNA Damage. Frontiers in Cell and Developmental Biology, 9(May), 1–10. https://doi.org/10.3389/fcell.2021.669379
Detmer, S. A., & Chan, D. C. (2007). Functions and dysfunctions of mitochondrial dynamics. Nature Reviews Molecular Cell Biology, 8(11), 870–879. https://doi.org/10.1038/nrm2275
Shoubridge, E. A., Johns, T., & Karpati, G. (1997). Complete restoration of a wild-type mtDNA genotype in regenerating muscle fibres in a patient with a tRNA point mutation and mitochondrial encephalomyopathy. Human Molecular Genetics, 6(13), 1–4. papers://395e1851-bc1d-4e5f-8c70-4fdc5dc2d371/Paper/p18
Jeppesen, T. D., Duno, M., & Vissing, J. (2020). Mutation Load of Single, Large-Scale Deletions of mtDNA in Mitotic and Postmitotic Tissues. Frontiers in Genetics, 11(October), 1–6. https://doi.org/10.3389/fgene.2020.547638
Nelson, I., Degoul, F., Obermaier-Kusser, B., Romero, N., Borrone, C., Marsac, C., Vayssiere, J.-L., Gerbitz, K., Fardeau, M., Ponsot, G., & Lestienne, P. (1989). Mapping of heteroplasmic mitochondrial DNA deletions in Kearns-Sayre syndrome. October, 17(20), 8117–8124.
Goldstein A, Falk MJ. Mitochondrial DNA Deletion Syndromes. 2003 Dec 17 [Updated 2019 Jan 31]. In: Adam MP, Ardinger HH, Pagon RA, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2022.
Tsang, S. H., Alicia, R. P., & Sharma, T. (n.d.). CHAPTER 30 Mitochondrial Disorder : Kearns-Sayre Syndrome. https://doi.org/10.1007/978-3-319-95046-4
Chen, M. L., Logan, T. D., Hochberg, M. L., Shelat, S. G., Yu, X., Wilding, G. E., Tan, W., Kujoth, G. C., Prolla, T. A., Selak, M. A., Kundu, M., Carroll, M., & Thompson, J. E. (2009). Erythroid dysplasia, megaloblastic anemia, and impaired lymphopoiesis arising from mitochondrial dysfunction. Blood, 114(19), 4045–4053. https://doi.org/10.1182/blood-2008-08-169474
Hikmat, O., Charalampos, T., Klingenberg, C., Rasmussen, M., Tallaksen, C. M. E., Brodtkorb, E., Fiskerstrand, T., McFarland, R., Rahman, S., & Bindoff, L. A. (2017). The presence of anaemia negatively influences survival in patients with POLG disease. Journal of Inherited Metabolic Disease, 40(6), 861–866. https://doi.org/10.1007/s10545-017-0084-9
McGovern, Loretta. Personal Interview. 11 Mar 2022.
Lewis, Samantha. Expert Interview. 7 Mar 2022.
Tarnopolsky, M. A. (2014). Exercise as a therapeutic strategy for primary mitochondrial cytopathies. Journal of Child Neurology, 29(9), 1225–1234. https://doi.org/10.1177/0883073814538512
Auré, K., Ogier De Baulny, H., Laforêt, P., Jardel, C., Eymard, B., & Lombès, A. (2007). Chronic progressive ophthalmoplegia with large-scale mtDNA rearrangement: Can we predict progression? Brain, 130(6), 1516–1524. https://doi.org/10.1093/brain/awm067
Ahola, S., Auranen, M., Isohanni, P., Niemisalo, S., Urho, N., Buzkova, J., Velagapudi, V., Lundbom, N., Hakkarainen, A., Muurinen, T., Piirilä, P., Pietiläinen, K. H., & Suomalainen, A. (2016). Modified Atkins diet induces subacute selective ragged‐red‐fiber lysis in mitochondrial myopathy patients. EMBO Molecular Medicine, 8(11), 1234–1247. https://doi.org/10.15252/emmm.201606592
Pitceathly, R. D., & Viscomi, C. (2016). Effects of ketosis in mitochondrial myopathy: potential benefits of a mitotoxic diet. EMBO Molecular Medicine, 8(11), 1231–1233. https://doi.org/10.15252/emmm.201606933
Andreazza, Ana. Expert Interview. 2 Mar 2022.
Duong, A., Evstratova, A., Sivitilli, A., Hernandez, J. J., Gosio, J., Wahedi, A., Sondheimer, N., Wrana, J. L., Beaulieu, J. M., Attisano, L., & Andreazza, A. C. (2021). Characterization of mitochondrial health from human peripheral blood mononuclear cells to cerebral organoids derived from induced pluripotent stem cells. Scientific Reports, 11(1), 1–16. https://doi.org/10.1038/s41598-021-84071-6
Lorenz, C., Lesimple, P., Bukowiecki, R., Zink, A., Inak, G., Mlody, B., Singh, M., Semtner, M., Mah, N., Auré, K., Leong, M., Zabiegalov, O., Lyras, E. M., Pfiffer, V., Fauler, B., Eichhorst, J., Wiesner, B., Huebner, N., Priller, J., … Prigione, A. (2017). Human iPSC-Derived Neural Progenitors Are an Effective Drug Discovery Model for Neurological mtDNA Disorders. Cell Stem Cell, 20(5), 659-674.e9. https://doi.org/10.1016/j.stem.2016.12.013
Rinaldi, T., Dallabona, C., Ferrero, I., Frontali, L., & Bolotin-Fukuhara, M. (2010). Mitochondrial diseases and the role of the yeast models. FEMS Yeast Research, 10(8), 1006–1022. https://doi.org/10.1111/j.1567-1364.2010.00685.x
Okamoto, K., Perlman, P. S., & Butow, R. A. (1998). The sorting of mitochondrial DNA and mitochondrial proteins in zygotes: Preferential transmission of mitochondrial DNA to the medial bud. Journal of Cell Biology, 142(3), 613–623. https://doi.org/10.1083/jcb.142.3.613
Dancy, B. M., Sedensky, M. M., & Morgan, P. G. (2015). Mitochondrial bioenergetics and disease in Caenorhabditis elegans. Frontiers in Bioscience - Landmark, 20(2), 198–228. https://doi.org/10.2741/4305
Wagner, J. T., Howe, D. K., Estes, S., & Denver, D. R. (2020). Mitochondrial DNA variation and selfish propagation following experimental bottlenecking in two distantly related Caenorhabditis briggsae isolates. Genes, 11(1). https://doi.org/10.3390/genes11010077
Chen, Z., Zhang, F., & Xu, H. (2019). Human mitochondrial DNA diseases and Drosophila models. Journal of Genetics and Genomics, 46(4), 201–212. https://doi.org/10.1016/j.jgg.2019.03.009
Sen, A., & Cox, R. T. (2017). Fly Models of Human Diseases: Drosophila as a Model for Understanding Human Mitochondrial Mutations and Disease. In Current Topics in Developmental Biology (1st ed., Vol. 121). Elsevier Inc. https://doi.org/10.1016/bs.ctdb.2016.07.001
Liau, W. S., Gonzalez-Serricchio, A. S., Deshommes, C., Chin, K., & LaMunyon, C. W. (2007). A persistent mitochondrial deletion reduces fitness and sperm performance in heteroplasmic populations of C. elegans. BMC Genetics, 8, 1–11. https://doi.org/10.1186/1471-2156-8-8

Inoue, K., Nakada, K., Ogura, A., Isobe, K., Goto, Y. I., Nonaka, I., & Hayashi, J. I. (2000). Generation of mice with mitochondrial dysfunction by introducing mouse mtDNA carrying a deletion into zygotes. Nature Genetics, 26(2), 176–181. https://doi.org/10.1038/82826
Ishikawa, K., & Nakada, K. (2021). Attempts to understand the mechanisms of mitochondrial diseases: The reverse genetics of mouse models for mitochondrial disease. Biochimica et Biophysica Acta - General Subjects, 1865(3), 129835. https://doi.org/10.1016/j.bbagen.2020.129835
Cagnone, G., Tsai, T. S., Srirattana, K., Rossello, F., Powell, D. R., Rohrer, G., Cree, L., Trounce, I. A., & St. John, J. C. (2016). Segregation of naturally occurring mitochondrial DNA variants in a mini-pig model. Genetics, 202(3), 931–944. https://doi.org/10.1534/genetics.115.181321
Campbell, J. M., Perales-Clemente, E., Ata, H., Vidal-Folch, N., Liu, W., Clark, K. J., Xu, X., Oglesbee, D., Nelson, T. J., & Ekker, S. C. (2018). Engineering targeted deletions in the mitochondrial genome. BioRxiv, 1–23. https://doi.org/10.1101/287342
Steele, S. L., Prykhozhij, S. V., & Berman, J. N. (2014). Zebrafish as a model system for mitochondrial biology and diseases. Translational Research, 163(2), 79–98. https://doi.org/10.1016/j.trsl.2013.08.008
Aguer, C., Gambarotta, D., Mailloux, R. J., Moffat, C., Dent, R., McPherson, R., & Harper, M. E. (2011). Galactose enhances oxidative metabolism and reveals mitochondrial dysfunction in human primary muscle cells. PLoS ONE, 6(12). https://doi.org/10.1371/journal.pone.0028536
Mochly-Rosen, Daria. Expert Interview. 25 Feb 2022.
Rosdah, A. A., K. Holien, J., Delbridge, L. M. D., Dusting, G. J., & Lim, S. Y. (2016). Mitochondrial fission - a drug target for cytoprotection or cytodestruction? Pharmacology Research and Perspectives, 4(3), 1–19. https://doi.org/10.1002/prp2.235
Pantic, B., Ives, D., Mennuni, M., Perez-Rodriguez, D., Fernandez-Pelayo, U., Lopez de Arbina, A., Muñoz-Oreja, M., Villar-Fernandez, M., Dang, T. mai J., Vergani, L., Johnston, I. G., Pitceathly, R. D. S., McFarland, R., Hanna, M. G., Taylor, R. W., Holt, I. J., & Spinazzola, A. (2021). 2-Deoxy-D-glucose couples mitochondrial DNA replication with mitochondrial fitness and promotes the selection of wild-type over mutant mitochondrial DNA. Nature Communications, 12(1). https://doi.org/10.1038/s41467-021-26829-0
Agilent. (2022). Seahorse XF Pro Platform Solution. https://www.agilent.com/en/product/cell-analysis/real-time-cell-metabolic-analysis/xf-analyzers/seahorse-xf-pro-analyzer-1980223/transforming-bioenergetic-measurements
Jacoby, E., Ben Yakir-Blumkin, M., Blumenfeld-Kan, S., Brody, Y., Meir, A., Melamed-Book, N., Napso, T., Pozner, G., Saadi, E., Shabtay-Orbach, A., Yivgi-Ohana, N., Sher, N., & Toren, A. (2021). Mitochondrial augmentation of CD34+ cells from healthy donors and patients with mitochondrial DNA disorders confers functional benefit. Npj Regenerative Medicine, 6(1), 1–12. https://doi.org/10.1038/s41536-021-00167-7
Minovia Therapeutics (27 Dec 2017 - 31 Aug 2021). A Study to Evaluate the Safety and Therapeutic Effects of Transplantation of MNV-BM-BLD in Pediatric Patients With Pearson Syndrome. Identifier NCT03384420. https://clinicaltrials.gov/ct2/show/NCT03384420
U.S. National Library of Medicine. (16 Sep 2020 - ). A First in Human Study to Evaluate the Safety of Infusion of MNV-BM-PLC (Autologous CD34+ Cells Enriched With Placenta Derived Allogeneic Mitochondria) in Patients With Primary Mitochondrial Diseases Associated With Mitochondrial DNA Mutation or Deletion. Identifier NCT04548843. https://clinicaltrials.gov/ct2/show/NCT04548843
Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. A., & Zhang, F. (2013). Genome engineering using the CRISPR-Cas9 system. Nature Protocols, 8(11), 2281–2308. https://doi.org/10.1038/nprot.2013.143
Minczuk, M., Papworth, M. A., Miller, J. C., Murphy, M. P., & Klug, A. (2008). Development of a single-chain, quasi-dimeric zinc-finger nuclease for the selective degradation of mutated human mitochondrial DNA. Nucleic Acids Research, 36(12), 3926–3938. https://doi.org/10.1093/nar/gkn313
Gammage, P. A., Rorbach, J., Vincent, A. I., Rebar, E. J., & Minczuk, M. (2014). Mitochondrially targeted ZFNs for selective degradation of pathogenic mitochondrial genomes bearing large-scale deletions or point mutations. EMBO Molecular Medicine, 6(4), 458–466. https://doi.org/10.1002/emmm.201303672
Ramirez, C. L., Certo, M. T., Mussolino, C., Goodwin, M. J., Cradick, T. J., McCaffrey, A. P., Cathomen, T., Scharenberg, A. M., & Joung, J. K. (2012). Engineered zinc finger nickases induce homology-directed repair with reduced mutagenic effects. Nucleic Acids Research, 40(12), 5560–5568. https://doi.org/10.1093/nar/gks179
Lim, K., Cho, S. I., & Kim, J. S. (2022). Nuclear and mitochondrial DNA editing in human cells with zinc finger deaminases. Nature Communications, 13(1), 1–10. https://doi.org/10.1038/s41467-022-27962-0
Conway, S. J. (2020). Bifunctional Molecules beyond PROTACs. Journal of Medicinal Chemistry, 63(6), 2802–2806. https://doi.org/10.1021/acs.jmedchem.0c00293
Takahashi, D., Moriyama, J., Nakamura, T., Miki, E., Takahashi, E., Sato, A., Akaike, T., Itto-Nakama, K., & Arimoto, H. (2019). AUTACs: Cargo-Specific Degraders Using Selective Autophagy. Molecular Cell, 76(5), 797-810.e10. https://doi.org/10.1016/j.molcel.2019.09.009
Grier, J., Hirano, M., Karaa, A., Shepard, E., & Thompson, J. L. P. (2018). Diagnostic odyssey of patients with mitochondrial disease Results of a survey. Neurology: Genetics, 4(2), 0–6. https://doi.org/10.1212/NXG.0000000000000230
Gustafson, M. A., McCormick, E. M., Perera, L., Longley, M. J., Bai, R., Kong, J., Dulik, M., Shen, L., Goldstein, A. C., McCormack, S. E., Laskin, B. L., Leroy, B. P., Ortiz-Gonzalez, X. R., Ellington, M. G., Copeland, W. C., & Falk, M. J. (2019). Mitochondrial single-stranded DNA binding protein novel de novo SSBP1 mutation in a child with single large-scale mtDNA deletion SLSMD) clinically manifesting as Pearson, Kearns-Sayre, and Leigh syndromes. PLoS ONE, 14(9), 1–24. https://doi.org/10.1371/journal.pone.0221829
Hubens, W. H. G., Vallbona-Garcia, A., de Coo, I. F. M., van Tienen, F. H. J., Webers, C. A. B., Smeets, H. J. M., & Gorgels, T. G. M. F. (2022). Blood biomarkers for assessment of mitochondrial dysfunction: An expert review. Mitochondrion, 62(October 2021), 187–204. https://doi.org/10.1016/j.mito.2021.10.008
Dominguez-Gonzalez, C., Badosa, C., Madruga-Garrido, M., Martí, I., Paradas, C., Ortez, C., Diaz-Manera, J., Berardo, A., Alonso-Pérez, J., Trifunov, S., Cuadras, D., Kalko, S. G., Blázquez-Bermejo, C., Cámara, Y., Martí, R., Mavillard, F., Martin, M. A., Montoya, J., Ruiz-Pesini, E., … Jimenez-Mallebrera, C. (2020). Growth Differentiation Factor 15 is a potential biomarker of therapeutic response for TK2 deficient myopathy. Scientific Reports, 10(1), 1–13. https://doi.org/10.1038/s41598-020-66940-8
Hugarth, Matilda, Abliva, AB. (25 Mar 2019 - 20 Oct 2021). A Phase Ia/Ib, SAD and MAD Study of of KL1333 in Healthy Subjects and Patients With Primary Mitochondrial Disease. Identifier NCT03888716. https://clinicaltrials.gov/ct2/show/NCT03888716

Perlara’s Cure Odysseys

Discussion about this post