Bachmann–Bupp syndrome and treatment

Abstract

Bachmann–Bupp syndrome (BABS) is a neurodevelopmental disorder characterized by developmental delay, hypotonia, and varying forms of non-congenital alopecia. The condition is caused by 3′-end mutations of the ornithine decarboxylase 1 (ODC1) gene, which produce carboxy (C)-terminally truncated variants of ODC, a pyridoxal 5′-phosphate-dependent enzyme. C-terminal truncation of ODC prevents its ubiquitin-independent proteasomal degradation and leads to cellular accumulation of ODC enzyme that remains catalytically active. ODC is the first rate-limiting enzyme that converts ornithine to putrescine in the polyamine pathway. Polyamines (putrescine, spermidine, spermine) are aliphatic molecules found in all forms of life and are important during embryogenesis, organogenesis, and tumorigenesis. BABS is an ultra-rare condition with few reported cases, but it serves as a convincing example for drug repurposing therapy. α-Difluoromethylornithine (DFMO, also known as eflornithine) is an ODC inhibitor with a strong safety profile in pediatric use for neuroblastoma and other cancers as well as West African sleeping sickness (trypanosomiasis). Patients with BABS have been treated with DFMO and have shown improvement in hair growth, muscle tone, and development.

The pace of syndrome discovery and causal gene identification has rapidly increased since the turn of the 21st century. Advances in diagnostic technology have allowed conditions previously described clinically to be further explained and clarified, which has resulted in uniting multiple phenotypes, while separating others. The deep repository of animal modeling and bench science can now be mined for clues and confirmations of these syndromes. The metabolic polyamine pathway is an excellent example of this process. Within that pathway, Snyder–Robinson syndrome (SRS) was clinically described in 1969,¹ yet the spermine synthase (SMS) gene was not identified until 2003.² Since that publication, no other syndrome in the pathway had been described until 2018 with the identification of Bachmann–Bupp syndrome (BABS).³ This discovery and ensuing story showcase the promise of genetic diagnoses, novel treatment strategies, and precision medicine.

The origins of BABS exemplify the serendipitous nature that scientific discovery often requires. Dr André Bachmann had long studied the polyamine pathway, first in plants to explore the role of polyamines in foliar senescence^4–6 and later in oncology to study the function of polyamines in cancer, including a phase I trial in pediatric neuroblastoma.^7–17 At the time he was recruited to Michigan State University College of Human Medicine in 2015, the clinical role of polyamines was well established in oncology and parasitic diseases such as West African sleeping sickness (trypanosomiasis) but remained fairly unclear in medical genetics. A study was initiated at Helen DeVos Children’s Hospital in Grand Rapids, MI, USA, to investigate polyamine regulation during pediatric critical illness. This was in part due to the interest in multi-organ dysfunction syndrome by the hospital’s research director and pediatric intensivist physician Dr Surender Rajasekaran, the true matchmaker in bringing together Drs Bachmann and Bupp, which led to the discovery of BABS. During a presentation given at pediatric grand rounds, Bachmann and Rajasekaran described their polyamine research in multi-organ dysfunction syndrome. In that audience was Dr Caleb Bupp, a medical genetics physician at Helen DeVos Children’s Hospital and Corewell Health (formerly Spectrum Health). Unknown to Bachmann and Rajasekaran, Bupp had an undiagnosed patient whose whole-exome sequencing had identified a variant of uncertain significance in ornithine decarboxylase 1 (ODC1), a gene in the polyamine pathway.

The finding of the ODC1 variant warrants some mention as it showcases the diagnostic odyssey that many patients with rare diseases endure. Born after a normal pregnancy the patient spent 5 weeks in the neonatal intensive care unit with low tone, feeding issues, and an unexplained intraventricular hemorrhage. The infant had then shown signs of developmental delay. At the initial evaluation of the patient’s genetics at 10 months of age, a most unusual finding, now thought to be a cardinal sign, was reported: non-congenital alopecia with hair loss shortly after birth. As should be routinely practiced in genetic testing, non-directive counseling about diagnostic workup was provided and genetic testing was offered, which the family declined. This decision was complicated by the cost and insurance coverage barriers to genetic testing that were present in 2017 and persist today. Eventually, when developmental delay did not improve, chromosome microarray was performed with uninformative results. This was then followed by trio whole-exome sequencing, which identified the de novo nonsense variant in ODC1. At that time, this was appropriately classified as a variant of uncertain significance by American College of Medical Genetics and Genomics standards¹⁸ as there was no known syndrome associated with the gene. As often happens with uncertain and unknown results, nothing further was done until the connection between the patient’s ODC1 variant and the work into multi-organ dysfunction syndrome at Helen DeVos Children’s Hospital was made.

Sometime after the grand rounds presentation, as Bupp was reviewing undiagnosed clinical cases, he returned to his patient with the ODC1 variant. That review led to the remembrance of the role of ODC1 in the polyamine pathway. In a meeting shortly following with Rajasekaran, Bupp brought up the possible link, leading to an urgent phone call being made to Bachmann. This telephone call established the connections needed to describe this new syndrome.

BABS was the second polyamine-associated neurodevelopmental disorder to be described after SRS^{2, 19} and we recently proposed to name this new and expanding group of disorders as polyaminopathies (Figure 1). SRS is characterized by intellectual disability, developmental delay, hypotonia, seizures, asthenic body habitus, long hands and great toes, kyphoscoliosis, and a propensity for osteoporosis and bony fractures. As with many rare disorders, the passing of time and accumulating number of affected individuals have helped to clarify the phenotype and natural history. Developmental delay and hypotonia are shared features between SRS and BABS, although this could be said of many other genetic syndromes. Seizures have been seen in one individual with BABS, the oldest known patient at age 24 years, so the potential shared feature of epilepsy in SRS and BABS is yet to be fully determined. The underlying mechanism for bone-related issues in SRS has not been determined; perhaps with more polyaminopathies being discovered, that feature may be seen in other syndromes and the cause may be identified.

FIGURE 1.

Polyamine pathway and associated genetic syndromes. The polyamine pathway is shown, highlighting five gene products (green colored) that are associated with known genetic syndromes and disorders (green boxes). The three principal polyamines putrescine, spermidine, and spermine are synthesized from ornithine by the subsequent catalytic actions of ODC, SRM, and SMS respectively, as well as the catalytic action of AMD1. Both ODC1 and AMD1 genes are direct transcriptional targets of MYC. In addition, polyamines are transported into and out of cells by polyamine transporters, including solute carriers and P5B-ATPases. Polyamines can be acetylated by the action of SSAT-1 and either exported by polyamine transporters or reconverted by PAOX. Alternatively, reconversion may occur directly by SMOX. Polyamines are tightly regulated to maintain cellular homeostasis required to regulate key physiological functions including embryogenesis and organogenesis. If the fine-tuned balance of polyamines is altered, it can result in tumorigenesis. Spermidine is the sole substrate for the post-translational hypusine modification (activation) of eIF5A (at amino-acid residue lysine-50) which requires the consecutive catalytic action of DHPS and DOHH. Therefore, spermidine directly controls eIF5A-dependent protein translation. In addition to SMS, ODC, DHPS, DOHH, and eIF5A, mutations in other polyamine-associated gene products (blue colored) might also be involved in related human genetic disorders, all leading to an imbalance (increase or decrease) of intracellular putrescine, spermidine, and/or spermine pools. Collectively, they present a new group of polyamine-associated neurodevelopmental disorders we have named polyaminopathies. Abbreviations: Amd1, S-Adenosylmethionine Decarboxylase; Arg1, Arginase 1; Az, Antizyme; Azin, Antizyme Inhibitor; Babs, Bachmann–Bupp Syndrome; Dcsam, Decarboxylated S-Adenosylmethionine; Dfmo, α-Difluoromethylornithine; Dhps, Deoxyhypusine Synthase; Dohh, Deoxyhypusine Hydroxylase; Eif5a, Eukaryotic Translation Initiation Factor 5a; Fabas, Faundes–Banka Syndrome; Mat1, Methionine Adenosyltransferase 1; Myc, Myelocytomatosis Oncoprotein; Odc, Ornithine Decarboxylase; Paox, Polyamine Oxidase; Sam, S-Adenosylmethionine; Smox, Spermine Oxidase; Srs, Snyder–Robinson Syndrome; Ssat-1, Spermidine/Spermine N1-Acetyltransferase 1; Srm, Spermidine Synthase; Sms, Spermine Synthase.

Three newer polyaminopathies were reported after BABS, caused by variants in deoxyhypusine synthase (DHPS), eukaryotic translation initiation factor 5A (eIF5A), and deoxyhypusine hydroxylase (DOHH)^20–23 (Figure 1). Patients with DHPS deficiency had mild to severe global developmental delay, abnormal muscle tone (hyper- and hypotonia reported), seizures (multiple subtypes), and most had short stature.²⁰ There was variability among the four patients reported, even between two who were siblings. All individuals reported with eIF5A variants also had developmental delay of varying degrees, and most had microcephaly, micrognathia, and some dysmorphic features.^{21, 22} Again, some consistency is seen with other polyaminopathies, but the presence of craniofacial abnormalities with eIF5A and SRS¹⁹ differs and, so far, does not have a mechanistic explanation. Deficiency in eIF5A is also known as Faundes–Banka syndrome. Lastly, patients with DOHH polyaminopathy show developmental delay and intellectual disability as well as microcephaly.²³ Impairment seems less severe than DHPS and eIF5A, and, for patients with DOHH polyaminopathy, brain magnetic resonance imaging (MRI) showed diffuse atrophy and hypomyelination. MRI results were generally normal for patients with eIF5A polyaminopathy and not commented on in publications about DHPS.^{20, 23} Non-specific abnormalities on brain MRI are a consistent feature of BABS^{24, 25} as well as SRS,¹⁹ although not necessarily atrophy and hypomyelination. Further understanding of the impact on brain development in polyaminopathies will be needed as these disorders are identified in more individuals.

In considering the phenotype of BABS there are additional disorders that can be considered on the basis of shared phenotype. The findings of developmental delay, intellectual disability, and hypotonia have such a broad list of differential diagnoses that they themselves would be highly unlikely to suggest BABS specifically. The ectodermal findings, particularly alopecia, can be seen in many genetic conditions, most commonly the various types of ectodermal dysplasia.^{26, 27} Hair loss in BABS is typically not present at birth, or hair loss has an unusual pattern where hair may be present on the apex of the head and at the bottom with a ring of alopecia between.

ODC1 MUTATIONS ARE THE CAUSE OF BABS

There are 11 patients reported in the medical literature^{3, 24, 25, 28–31} with fewer than 30 known by this team worldwide diagnosed with BABS. Owing to the small number of cases and young age of reported patients, knowledge of the associated clinical symptoms is continuing to evolve. A common feature seen among these patients is the non-congenital alopecia where hair that is typically present at birth, although sometimes sparse and atypical in color, then falls out in the first few weeks of life with future scalp hair growth remaining sparse along with limited to no eyebrow and eyelash growth. As discussed above, this finding is unique and could be considered a cardinal feature of BABS. Additional features seen frequently among affected individuals include hypotonia, moderate to severe global developmental delay (further characterized below), behavior concerns including features of attention-deficit/hyperactivity and autism spectrum disorders, feeding concerns and constipation, macrocephaly with non-specific brain MRI findings, prenatal polyhydramnios, and skin concerns including keratosis pilaris and recurrent follicular cysts.

Other findings seen among patients have been quite variable. As discussed above, the oldest patient known to this team is also the only one known to have a history of seizures which started at 14 years of age. Whether this is part of the condition is hard to say given the young age of all other known patients. Additionally, one individual has unilateral congenital sensorineural hearing loss, two individuals have congenital heart disease (one with mild pulmonary valve stenosis, the other with spontaneously resolved ventricular septal defect), three individuals have nail differences, and one individual has vision differences. Orthopedic conditions such as hip subluxation, osteoporosis, or scoliosis have been noted in this population. Continued follow-up of the known patients along with addition of information as new patients are identified will help to clarify the defining features of BABS. As patients with BABS are treated, the phenotype may evolve and reveal other features, particularly as patients have differing baselines, differing ages at diagnosis, and differing ages when treatment is initiated.

Characterization of the developmental delay in BABS presents some challenges. In particular, there is an overall limited number of patients, of varying ages and stages of development, and a paucity of standardized outcome measures used between patients. The BABS diagnosis was received between 2 weeks old and 22 years old in identified patients. Cognitive, speech/language, and motor deficits have been described for this population of patients.^{32, 33} Onset of independent sitting has occurred within the typical developmental window for some (11 months) or has been delayed in others with onset after 3 years old. Crawling has been documented to have started at 12 to 18 months for some patients and 3 years for others. Onset of walking has been met as early as 4 years old, with some patients beyond this age and not yet independent ambulators. Ambulatory patients present with gait impairments consistent with proximal lower extremity weakness. Some patients communicate with verbal language, others communicate with a variety of sign language or sounds. All patients have been noted to have better receptive language and comprehension skills compared with expressive language (elocution, articulation, word retrieval). Some patients self-feed, and others are orally fed with assistance or are gastrostomy tube dependent.

Collection of data and use of standardized outcome measures at specific ages are current limitations for this group of patients and have been identified as areas for future research. Until this has been established, the International Classification of Functioning, Disability and Health framework is used to describe a bio-psychosocial approach to assessing ability and disability for patients with pathogenic ODC1 variants.³⁴ For patients with BABS, body, structure, and function impairments are known to include hypotonia, hypermobility, muscle weakness, alopecia, macrocephaly, and seizures (rarely reported, severe). Identified activity limitations include static and dynamic postural control, mobility (walking, sitting, standing, transitions), expressive language, personal care, and self-feeding. Participation restrictions occur at home (as a sibling, son, or daughter), school or daycare (as a student), or in the office for one adult patient (as an employee). Environmental and personal factors vary on the basis of each patient and include family dynamics, birth order, geographical location, and proximity and access to resources.

ODC1 FUNCTIONS AND POTENTIAL TREATMENT OF BABS

ODC is a cytosolic protein and part of a complex cellular mechanism that regulates the biosynthesis of aliphatic molecules called polyamines.³⁵ ODC is a rate-limiting enzyme of the polyamine pathway and specifically decarboxylates the non-proteinogenic amino acid ornithine (derived from arginine) into putrescine (a diamine) in the presence of pyridoxal 5′-phosphate (PLP)^{36, 37} (Figure 1). Putrescine is further metabolized into spermidine (a triamine) and spermine (a tetramine) by the catalytic actions of spermidine synthase (SRM) and spermine synthase (SMS) respectively.^{38, 39} Methionine is a precursor to decarboxylated S-adenosylmethionine (dcSAM), which is formed from S-adenosylmethionine (SAM) by the catalytic action of S-adenosylmethionine decarboxylase (AMD1, also known as AdoMetDC or SAMDC) required in the synthesis of spermidine and spermine. In addition to ODC, AMD1 is another rate-limiting enzyme in the polyamine pathway and both ODC1 and AMD1 genes are direct transcriptional targets of the transcriptional regulator MYC.

Polyamines are basic, positively charged small organic molecules that are produced by nearly all eukaryotic and prokaryotic life forms, including plants and bacteria. In humans, they are ubiquitously distributed in tissues and orchestrate diverse physiological processes that control embryonic development, growth, and cell division. Although extensively studied for decades, the precise role(s) of polyamines in cells still needs further exploration. It is well established that polyamines are essential for RNA, DNA, and protein synthesis, angiogenesis, spermatogenesis, and wound healing.^39–41 These molecules interact with negatively charged RNA and DNA, thereby directly contributing to nucleosome, chromatin, and histone biology.^{42, 43} Polyamines also bind to, and regulate, a variety of ion channels.^44–46 In cancer, they play essential roles in cell proliferation, differentiation, apoptosis, tumorigenesis, and immune system function.^47–54

Polyamine biosynthesis is regulated at the transcriptional, translational, and post-translational levels.^{35, 39, 40, 55, 56} MYC proteins are transcription factors that directly bind to, and activate, ODC1 gene transcription by binding to E-box elements.^{38, 57–59} Spermidine serves as the only substrate in the hypusination of eIF5A proteins, a process that requires the enzymes DHPS and DOHH. This unique post-translational modification of the lysine residue at position 50 (K50) to hypusine activates eIF5A, which is critical for eukaryotic protein translation^60–65 (Figure 1). Spermine oxidase (SMOX) is an amine oxidase that directly oxidizes spermine to spermidine. SSAT-1 (also known as SAT1) catalyzes the acetylation of spermidine or spermine, which are then excreted by cells through extracellular transport.^{39, 40, 49, 50} Alternatively, N¹-acetylspermidine or N¹-acetylspermine serve as substrates for the peroxisomal enzyme polyamine oxidase (PAOX), which catalyzes the reconversion of N¹-acetylated polyamines to putrescine and spermidine respectively, depending on the starting substrate (Figure 1). Polyamines are further controlled by antizyme 1–3 proteins (AZ1–3, collectively referred to as AZ), which stimulate ODC degradation by the proteasome through a ubiquitin-independent process^66–71 (Figure 2). Antizyme inhibitor proteins 1–2 (AZIN1–2, collectively referred to as AZIN) in turn bind to and regulate antizymes by competing with ODC for binding to antizymes.^{66, 67, 70–80} Intriguingly, it was found that the final 37 amino-a cid residues of the ODC carboxy (C)-terminal tail constitute an ODC destabilization domain that is required for antizyme-stimulated ODC degradation by the proteasome. If absent/deleted, enzymatically active ODC protein is not properly degraded and consequently accumulates in cells,^66–71 thus representing as a gain-of-function situation. Finally, polyamine levels are controlled in cells by a variety of mammalian polyamine transport systems including the solute carriers (e.g. SLC3A2)^81–83 and P5B-ATPases (e.g. ATP13A2 and ATP13A3),^84–86 which facilitate cellular polyamine import or export to further regulate organelle homeostasis (Figure 1).

FIGURE 2.

Representation of the ubiquitin-independent and antizyme-stimulated degradation of ODC by the 26S proteasome. ODC is a short-lived homodimer protein and regulated by a process that does not require ubiquitination to stimulate its proteasomal degradation. Antizyme binds to transient ODC monomer, which results in the exposure of the ODC carboxy (C)-terminal tail that is subsequently recognized by the 26S proteasome for degradation. Antizyme also binds AZIN, but with greater affinity than ODC, thus sequestering antizyme into a stable complex, which reduces its ability to regulate ODC. If the C-terminal tail of ODC is truncated in the final 37-amino acid residue destabilization region (yellow oval), ODC is not presented to the 26S proteasome, which leads to accumulation of ODC that remains catalytically active, thus producing increased amounts of putrescine, suggesting a gain-of-function situation. The underlying mechanisms by which increased ODC protein and elevated putrescine levels lead to BABS remain to be explored.

Abbreviations: AZ, antizyme; AZIN, antizyme inhibitor; BABS, Bachmann–Bupp syndrome; CT, C-terminal tail; ODC, ornithine decarboxylase.

The complex regulatory mechanisms of polyamines described above ensure a tight control of polyamine titers at multiple levels under physiological conditions. If this sensitive balance and interplay of putrescine, spermidine, and/or spermine is disturbed, it can have detrimental cellular consequences that lead to pathophysiological conditions including cancer and BABS but also SRS, Faundes–Banka syndrome, DHPS-/DOHH-deficiency, and probably other polyaminopathies.

The finding that the C-terminal truncation of ODC leads to elevated ODC activity and high putrescine levels was also confirmed in vivo, using transgenic mice. Before the discovery of BABS, in 1995, Thomas O’Brien and his team designed a transgenic mouse model, which overexpressed a mutated form of ODC.^{87, 88} In these mice a premature stop codon at amino-acid residue position 427 (p.427X) of the Odc1 mouse gene was introduced, resulting in C-terminal truncation. Truncated but active ODC enzyme was constitutively overexpressed and accumulated in the skin of these mice, which induced phenotypic abnormalities such as dermal follicular cysts, excessive skin wrinkling, enhanced nail growth, alopecia, and spontaneous tumor development, thus profoundly affecting skin homeostasis and increasing the risk of neoplastic growth. In 1996, the same team was able to show that treatment of these mice with α-difluoromethylornithine (DFMO) prevented hair loss and partly normalized skin histology, if administered before the onset of ODC overexpression. The transgenic mice that overexpressed and accumulated the truncated ODC in the skin had a normal first hair cycle, but then lost their hair completely 2 to 3 weeks after birth and DFMO reactivated hair growth in mice with complete hair loss.⁸⁹ Remarkably, this ODC skin mouse model that was designed 28 years ago had significant similarities to the skin abnormalities described in the first patient with BABS, including follicular cysts and infants born with thick silverish/grayish hair that fell out in clumps shortly after birth. Additional overexpression and knockout mouse models for ODC1 and other polyamine-related genes (e.g. SSAT-1, AMD1, SMS, AZ, ARG2) have been developed and are summarized in this review.⁹⁰ More recently, transgenic mice deficient in AZIN1 and AZIN2 genes have also been developed.^91–93

The in vitro and in vivo observations described above provided the scientific basis for our own hypothesis that patients with BABS having a premature stop codon within the C-terminal destabilization domain of ODC may not be able to fully clear ODC protein through the ubiquitin-independent proteasomal degradation pathway. Consequently, we expected to find accumulated ODC protein that remains functionally active,⁸⁷ because the C-terminal mutation does not affect the enzyme’s activity center. To investigate this, we analyzed primary red blood cells and primary dermal fibroblasts derived from a skin biopsy of the first patient with BABS reported for ODC protein content, ODC enzymatic activity, and polyamine levels. The results confirmed our hypothesis that the patient with BABS accumulates enzymatically active ODC protein compared with typically developing individuals.³ Moreover, we showed that the polyamine putrescine was significantly elevated as expected. Importantly, treating the patient’s primary dermal fibroblast cells with DFMO demonstrated restoration of putrescine levels to physiological baseline observed in control cells as a result of suppressing the high ODC activity,⁹⁴ providing additional support for consideration of patient treatment.

INNOVATIVE TREATMENT OF BABS TO IMPROVE PATHO PHY SIO LOG ICAL DEFICITS

DFMO is a specific, mechanism-based irreversible (suicide) inhibitor of ODC that was first synthesized by Merrell-Dow in 1978,⁹⁵ following the recognition that ODC, a bona fide oncogene, and its products, the polyamines, are key growth-promoting factors during cell cycle regulation and cellular division. If ODC is overexpressed, the resulting elevated levels of polyamines often contribute to increased cell proliferation and tumorigenesis.^47–52 While effective in laboratory-based studies as a cytostatic antitumor agent, the highly water-soluble DFMO in an oral formulation was only mildly successful as a monotherapy in early clinical trials against human cancers including leukemias and glioblastomas.⁸ However, DFMO was later found to show remarkable chemopreventive properties, for example in combination with sulindac, in a phase III randomized multicenter study of colorectal cancer;⁹⁶ it also showed promise in the relapse prevention of pediatric neuroblastoma.^{97, 98}

Notably, the intravenous and topical formulations of DFMO received approval by the US Food and Drug Administration for the treatment of West African sleeping sickness (trypanosomiasis)^{99, 100} and hirsutism^{101, 102} respectively. DFMO has extraordinary safety and a specific long-term dosing strategy in children with neuroblastoma¹⁴ even if administered daily for several years, making it ideal for clinical applications that require long-term treatments. However, the utility of DFMO is complicated with pharmacokinetic and clinical challenges as it requires continuous high dosing (1.0–6.0 g/m²/day in children) to compensate for rapid renal clearance. The attempt in a neuroblastoma patient-derived xenograft model to delay clearance of DFMO by combination with probenecid showed mild retention improvements.¹⁵ Surprisingly, DFMO is the only ODC inhibitor in clinical use today, and the Bachmann laboratory is in the process of designing novel DFMO analogs with higher potency and superior pharmacokinetic efficacy (ongoing studies, unpublished data) that may become lower-dose treatment alternatives in the future.

Since Bachmann was familiar with the wealth of literature published on DFMO in cell culture experiments, animal models, and clinical trials (cancer, trypanosomiasis, hirsutism, other), combined with the fact that his laboratory had worked with DFMO for more than two decades in the preclinical cancer setting^{7–11, 13, 15, 16, 38, 103–109} and participated in a phase I neuroblastoma clinical trial,¹⁴ he recognized the tantalizing opportunity to repurpose the already existing and safe ODC inhibitor DFMO for BABS.

When a novel rare disease is discovered, the possibility of any treatment, let alone a new treatment, is typically extremely small, particularly if the number of patients affected by a condition seems to be low. A diagnosis can provide answers and reassurance to the patient and family, particularly when there has been a long or complex diagnostic odyssey. Recommendations can sometimes be made for screening or monitoring on the basis of the known phenotype and medical concerns noted in other patients with the condition. In the case of BABS, there were three key factors that played a role in rapidly repurposing DFMO for treatment. First, cultured fibroblast samples from the first patient with BABS treated with DFMO showed biochemical normalization and no apparent concerns in vitro. Second, the previously discussed mouse model presented with both a similar phenotype to patients with BABS as well as clinical improvement and no significant side effects with DFMO treatment in vivo. Lastly, robust pediatric safety data for DFMO were available and able to be referenced when considering BABS treatment. Additional factors of note were access to DFMO supply, which was provided by Sanofi-Aventis. Parental willingness to embark on an ‘n-of-1’ treatment was also incredibly important for BABS. Not all these serendipitous factors will be possible for all rare diseases, and other challenges unique to each condition will exist. In this case, the time from publishing the syndrome discovery to treatment initiation in the first patient was 15 months, a speed that may be unmatched in rare diseases.

RESTORING POLYAMINE HOMEOSTASIS IN BABS

Excited by the potential but facing the challenges associated with treating rare and ultra-rare disorders, with data suggesting that DFMO would be efficacious for patients with BABS, a single-patient Investigational New Drug Application was submitted to the US Food and Drug Administration. Following approval, the Spectrum Health (now Corewell Health) Institutional Review Board reviewed and approved this compassionate-use treatment protocol, which also included approval to collect serial biological research samples from the patient. DFMO was provided by Sanofi-Aventis. Fifteen months after publishing the novel description of BABS,³ the first patient identified with BABS began treatment at age 4 years 8 months,²⁴ using the previously published dosing strategy for DFMO that had been demonstrated to be safe for long-term use in a large pediatric population.¹⁴

After treatment initiation, the patient demonstrated significant clinical improvement, including eyebrow and scalp hair growth 1 and 2 months after treatment initiation respectively. Whereas the patient was previously non-mobile, had limited fine motor skills, and was unable to sit unassisted, 4 months into DFMO treatment she was able to feed herself with a spoon with assistance and sit unassisted,²⁴ skills previously unachieved. Now, just over 3 years into treatment and on maintenance dosing, the patient continues to show significant clinical improvement (Figure 3). These improvements include recently reported resolution of dermatological symptoms in patients with BABS receiving DFMO treatment. Before DFMO treatment, two patients with BABS had a history of recurring follicular cysts, also observed in the transgenic mouse model,^{87, 89} and brittle fingernails. Since initiation of DFMO treatment, follicular cysts have not recurred for either patient.²⁸

FIGURE 3.

Clinical photography. The first patient diagnosed with Bachmann–Bupp syndrome at (a) birth, (b) 2 years 8 months, (c) 4 years 8 months (taken day of α-difluoromethylornithine [DFMO] initiation), (d) 5 years 6 months (11 months into DFMO treatment), and (e) 7 years 11 months. At birth, the patient had a full head of silver/blonde hair which fell out shortly after birth. Before treatment, the patient had complete alopecia with no hair, eyebrows, or eyelashes. After initiation of treatment, hair fully regrew, and the patient had significant improvement in muscle tone and development.

Supporting the improvement of clinical phenotypes seen in our patients, global metabolomic analysis performed on patients’ plasma samples collected before and after dose initiation demonstrated normalization of key metabolites including N¹-acetylputrescine and N-(3-acetamidopropyl) pyrrolidin-2-one (acisoga) immediately after initiation of treatment, and these markers remained unchanged at physiologically normal ranges for all collected timepoints.²⁴ The polyamine metabolite acisoga was recently identified as a novel biomarker for heart failure with reduced left ventricular ejection fraction.¹¹⁰ Despite the promise that DFMO has shown as an effective treatment for patients with BABS, further metabolomic studies of samples collected from untreated patients and serial samples from patients receiving DFMO treatment are warranted to give additional insight into the mechanism of the disease.

CONCLUSIONS

An ever-present challenge for research into rare diseases is the small number of affected patients. Polyaminopathies are no exception, even as multiple conditions in the pathway are discovered and reported. To meet the need for a collaborative resource for polyamine-related research, the International Center for Polyamine Disorders was established in 2020. A sample and data biobank were created to allow collection for study by the Center, but also importantly to serve as a nexus for external collaborators. Patients and family members can give consent for participation, therefore opening opportunities for scientific discovery and clinical advancement modeled on the bedside-to-bench-to-bedside model exemplified by BABS.

The number of distinct yet related polyaminopathies is constantly growing. SRS was the first recognized neurodevelopmental disorder associated with the polyamine pathway, then BABS, followed by DHPS deficiency, Faundes–Banka syndrome, and DOHH-deficiency syndromes (Figure 1). In addition, SSAT-1 gene duplication and putrescine accumulation were described in a patient with keratosis follicularis spinulosa decalvans (or Siemens-1 syndrome) in a case report study,¹¹¹ and overexpression of the SSAT-1 enzyme in a mouse model led to putrescine accumulation and a skin and hair phenotype similar to keratosis follicularis spinulosa decalvans.⁹⁰ Undoubtedly, we will discover additional genes of the polyamine pathway that are connected to polyaminopathies. Such genes may include, but are not limited to, AMD1, MAT1, AZ1–3, AZIN1–2, PAOX, SMOX, and polyamine transporters including solute carriers (e.g. SLC3A2) or those for the P5B-ATPases. Interestingly, ATP13A2 has been linked to neurodegenerative disorders such as Kufor–Rakeb syndrome and early-onset Parkinson disease^{84, 86} or hereditary spastic paraplegia,^{112, 113} and ATP13A3 has been associated with cervical-, head and neck-, and pancreatic cancers as well as pulmonary arterial hypertension.^{84, 114} ATP13A4 has been implicated in a spectrum of neurodevelopmental diseases including autism spectrum disorder, schizophrenia, specific language impairment, and childhood apraxia of speech.⁸⁴ ARG1 and MAT1 lead to arginemia¹¹⁵ and hypermethioninemia¹¹⁶ respectively. All these genes are intrinsically linked to orchestrating polyamine homeostasis, and, if mutated, can lead to loss-of-function or gain-of-function consequences that upset the fine-tuned physiological balance of polyamines in the cell environment. This could potentially lead to new polyaminopathies with distinct phenotypes, possibly dependent also on which of the main three polyamines are either elevated or depleted. Our current understanding of the polyamine pathway and available pharmacological polyamine inhibitors, in addition to DFMO, may aid us in developing new interventional therapies for such related neurodevelopmental disorders.

As more rare disorders are discovered, the pace of developing available treatments for them is not keeping up. With the development of techniques such as clustered regularly interspaced short palindromic repeats (CRISPR), gene therapy, and other novel treatments, there is more hope than ever for the treatment of rare diseases. Challenges include the small number of patients available for treatment for each condition, the high cost of drug discovery and study, and other resource limitations. Collaboration and robust information sharing are key for overcoming these barriers. Patient and family advocacy also plays an important role for navigating the potential barriers. The story of BABS can serve as an encouraging example of how to leverage scarce resources, collaborate with passion and expertise, and achieve a rapid diagnosis to repurposing success for others.

What this paper adds.

• Bachmann–Bupp syndrome and other polyamine-related rare genetic disorders are collectively referred to as polyaminopathies.

• Emerging evidence suggests that polyamine pathway-associated genes play critical roles in various neurodevelopmental disorders.

• Recent advances shed light on the promises of repurposing drugs in rare disease.

• Treatment with α-difluoromethylornithine suppresses ornithine decarboxylase activity and putrescine, resulting in significant overall clinical improvements.

Abbreviations:

AMD1

S-adenosylmethionine decarboxylase

ARG

arginase

AZ

Antizyme

AZIN

antizyme inhibitor

BABS

Bachmann–Bupp syndrome

DFMO

α-difluoromethylornithine

DHPS

deoxyhypusine synthase

DOHH

deoxyhypusine hydroxylase

eIF5A

eukaryotic translation initiation factor 5A

FABAS

Faundes-Banka Syndrome

ODC1

ornithine decarboxylase 1

PAOX

Polyamine Oxidase

SMOX

Spermine Oxidase

SRS

Snyder–Robinson syndrome

SSAT-1

spermidine/spermine N¹-acetyltransferase 1

SRS

Spermidine Synthase

SMS

spermine synthase

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new datasets were generated or analyzed in this study.