The neuronal ceroid lipofuscinoses (NCL), also known as Batten disease, are a group of inherited neurodegenerative life-limiting diseases that share some common clinical features including epileptic seizures, progressive psychomotor decline, and visual failure, and the accumulation of autofluorescent storage material. NCL usually begins in childhood, and most are inherited in an autosomal recessive manner. More than a dozen genes have been linked to families diagnosed with NCL (Table 1) (1). It is likely that most genes causing NCL have been identified.

This article summarises the genetic basis of NCL and discusses correlations with disease phenotype. All mutation details can be found in the freely accessible NCL mutation database (www.ucl.ac.uk/ncl-disease).

The concept of NCL as a group of inherited diseases first emerged in the 1960s (2), leading to classification into four broad ages of onset: infantile, late infantile, juvenile, and adult. At this time it was assumed that each of these types was caused by mutations in a different gene, named in advance as CLN1, CLN2, CLN3, and CLN4, respectively. The first genes to be identified, in the 1990s (CLN1, CLN2, CLN3), were responsible for the most common paediatric types. However, these first identified genes were not responsible for all childhood onset cases. For example, the genes CLN5, CLN6, CLN7, and CLN8 cause disease with onset in late infancy like CLN2, the first gene identified causing onset at that age (Table 1). CLN4 was not identified until 2011 (3).

A variety of experimental approaches reflecting the available technology were used to identify NCL genes. The first genes were identified using classic and time-consuming genetic linkage approaches requiring large numbers of similarly affected families followed by positional cloning of the genes [CLN1 (4) and CLN3 (5)]. A biochemical approach that detected a missing mannose-6-phosphate tagged lysosomal enzyme in a patient facilitated the identification of CLN2 (6), alongside ongoing genetic linkage studies. Availability of the human genome sequence meant that going forward fewer families were required to provide sufficient power for genetic linkage analysis, facilitating identification of CLN5 (7), CLN6 (8), CLN7 (9), and CLN8 (10). Some genes were identified by recognition of stretches of homozygosity in consanguineous families that narrowed the interval that contained the candidate gene. A gene first identified in an animal model led to identification of CLN10 (11, 12) in human disease. Improvements in sequencing technology later allowed fast and massively parallel sequencing of the whole exome in single families, and facilitated identification of the remaining disease genes [CLN4 (3), CLN11 (13), CLN12 (14), CLN13 (15), CLN14 (16)]. The few families suspected of carrying the putative CLN9 gene were later found to carry mutations in previously identified NCL genes (17, 18).

As monogenic disorders, each NCL is in effect a separate disease entity. All identified NCL genes lie on autosomes. Most cause disease though classic recessive inheritance, where deleterious mutations are present in disease gene alleles inherited from asymptomatic parents. However, adult onset CLN4 disease is dominantly inherited in the few families described with this disease (1, 3). There are three published reports of uniparental disomy in the NCLs, one in which a patient has complete isodisomy of chromosome 8, leading to homozygosity of a maternally-inherited deletion in CLN8 (19), and two patients for CLN1, both with paternal isodisomy of chromosome 1 (19, 20).

The majority of NCL genes encode proteins that reside in the endo/lysosomal pathways (1, 21–23). Most are lysosomal proteins—enzymes and soluble proteins (CLN1/PPT1, CLN2/TPP1, CLN5, CLN10/CTSD, CLN13/CTSF, CLN11/GRN,) or membrane proteins (CLN3, CLN7/MFSD8, CLN12/ATP13A2). Two encode endoplasmic reticulum membrane proteins (CLN6, CLN8). Other NCL proteins are cytoplasmic (CLN4/DNAJC5, CLN14/KCTD7) that peripherally associate with cellular membranes. The in vivo substrates for the lysosomal enzymes are incompletely defined, and much remains to be discovered around the functions of the membrane proteins. Nevertheless, recognition of the genetic basis of the NCLs enables the development of targeted therapies even though the underlying disease mechanism for each NCL is not yet fully delineated. It is unlikely that further NCL genes will be identified unless they cause disease in countries where little genetic analysis has been undertaken.

NCL are considered the most common inherited neurodegenerative disorder of childhood. They occur worldwide, with some forms first recognised in certain geographical regions. Some types are enriched in or absent from certain regions due to historical population (genetic) bottlenecks.

Incidence and prevalence rates are not available world-wide. Incidence rates are probably more robust than estimated prevalence rates, and generally reported between 1 in 14,000 (Iceland) up to 1 in 100,000 (67). The most common NCL in Northern Europe and the UK are juvenile CLN3 disease and late infantile CLN2 disease, but all types are present.

Some NCL genes are conserved in unicellular or simple organisms, indicating their fundamental function within eukaryotic cells (73). For example, yeasts contain homologous genes to CLN1/PPT1, CLN3 CLN10/CTSD, CLN12/ATP13A2. The slime mould Dictyostelium discoideum particularly expresses further NCL gene homologues or members of gene families (e.g., CLN2/TPP1, CLN4/DNAJC5, CLN5, CLN6, CLN7/MFSD8 family). NCL also occurs in animals (e.g., dogs, sheep, cows, monkey). Cell and animal models carrying mutations in genes equivalent to those causing human NCL are well used in research. These range from yeasts, up to rodents and other mammals (for clinical development). Some of these models are naturally occurring (e.g., mouse, dog, sheep), others are engineered models (e.g., mouse, pig). Some animal NCL disease is caused by mutations in genes not reported to cause similar disease in humans [ARSG in dogs (74), CLCN6 engineered in mice (31), CTSB/CTSL engineered in mice as double gene mutations (61)].

The NCL Mutation Database (www.ucl.ac.uk/ncl-disease) lists known disease-causing mutations and sequence variations in NCL genes by gene and by individual. Five hundred and thirty-seventh NCL disease-causing mutations are currently listed (Table 2) across >1,625 patients and >2,160 families. Where possible the age of onset, ethnic background and current location, are listed for each family. Data are gathered from case reports or larger collections in clinical or scientific publications, or referred directly, and updated periodically. These vary in detail according to the report source, e.g., case reports usually have more specifics than reports of large group genetic screens. Mutations are mostly described in single individuals or occasionally siblings from the same family. Some mutations are more common in certain populations due to local founder effects. Several NCL genes have widespread distribution across several continents due to ancient founder effects (Table 1).

An estimate of the proportion of cases caused by each mutation can be made, although there is a considerable under-representation of the occurrence of common mutations since the emphasis is on the collation of novel and rare mutations. The most prevalent mutations are the 1 kb deletion in CLN3 and two mutations in CLN2 (1).

Correlations can be drawn between genotype, phenotype and morphological changes in patients, and have been reviewed previously (47, 75), for example for CLN2 disease (76). These derived correlations can be used to predict the disease course in a newly diagnosed family.

This database is important (1). The severity of mutations has implications for treatment. It may be important to know if residual protein or function remains. Treatments may be developed that do not fully compensate for complete loss of gene function and can reduce but not completely eliminate the disease burden—these may be sufficient to improve health in families carrying so-called mild mutations but not in individuals lacking all gene function (2). The location of mutations in the protein may highlight key residues and functional or regulatory domains, aiding understanding of protein function (3). The data reveals the relative frequency of mutations; as ultra-rare, found only within certain ethnic groups, or widespread (4). The data is freely available and contained in excel tables that can be downloaded and used by researchers. For example, there is increasing information on frequency of mutations or disease in specific ethnic groups (4). Efficacy of a new treatment may be demonstrated earlier or more robustly if the mutations and their effects on disease progression of the participants are understood. Going forward, functional data for each mutation can begin to be incorporated, as available.

Other databases exist through international cooperation, enabling collection of natural history data for all NCL types and genotype-phenotype data through databases DEM-CHILD (www.dem-child.eu) (77, 78). There are disease rating scales (79–81) to follow disease progression. This is increasing understanding of the genetic spectrum of NCL disease as well as provide necessary control data for use in future clinical trials (77).

Most genes that cause NCL disease in humans are probably identified. This, combined with the broader range of associated phenotypes now described, has shown that the genetic picture is considerably more complex than was first envisioned at the start of the genetic era of the NCL. The functions of all NCL genes and thereby disease mechanisms are not yet known. As understanding increases overlap with other rare and common diseases, such as retinal dystrophies may indicate shared disease mechanisms (82).

The gene dosage or the specific mutations show correlation with clinical phenotype. Some variation in clinical phenotype is therefore explained by differing levels of residual protein function. However, variation between families and even siblings shows that co-inheritance of other genetic variations could influence disease phenotype. It is still unclear whether the underlying pathogenic mechanisms are partly shared between classic NCL forms and the alternative disease forms.

The era of genomic medicine is approaching, where genomic information will be used to design the best clinical care for an individual. For the NCL, personalised treatment approaches will be tailored to the underlying mutation and the genetic background of each patient. An early example is the design and delivery of an oligonucleotide therapy for a child with CLN7 disease (83).

Therapeutic development beyond current palliative treatments is advancing slowly. This relies on continued collection of natural history data for the broadening NCL spectrum to provide a control cohort to aid design of future clinical trials. The first approved treatment is for children with classic late infantile CLN2 disease which delivers recombinant protein directly into the brain at regular intervals. For the best long-term clinical benefit for any NCL disease, treatment must begin as early as possible, before any symptoms, which requires rapid and earlier diagnosis using genotype. This may be facilitated by advances in DNA-based approaches that allow future newborn screening (84, 85).

SM devised, interpreted the data, and wrote the review. EG collated the data on mutations, genes and phenotype, and contributed to the writing. All authors contributed to the article and approved the submitted version.

This work was funded by the Medical Research Council (MRC), award MR/V033956. SM (ORCID:0000-0003-4385-4957) reports support by Biomarin for maintaining the NCL mutation database. The support of UCL and the MRC that provides funding to the MRC Laboratory for Molecular Cell Biology University Unit at UCL (award code MC_U12266B) toward lab and office space is acknowledged. All research at Great Ormond Street Hospital NHS Foundation Trust and UCL Great Ormond Street Institute of Child Health is made possible by the NIHR Great Ormond Street Hospital Biomedical Research Centre. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.