- Email: [email protected]
Resident Should Know
Genetics and Child Neurology: What Every Trainee/Resident Should Know Vinodh Narayanan, MD The training of residents in child neurology varies from one center to another, being influenced to a large extent by the nature and volume of the clinical practice at a specific center and the expertise of the faculty. There is no doubt that there is an undercurrent of genetics in everything we do as child neurologists, sometimes explicit and sometimes implicit. In this article, we highlight a fundamental set of concepts, principles, methodologies, and learning tools/resources of which every child neurology trainee should have some knowledge. We may eventually arrive at a child neurology curriculum that might be continuously revised and maintained (perhaps through the Child Neurology Society) and serve as a template for individual training programs. Semin Pediatr Neurol 18:81-84 © 2011 Elsevier Inc. All rights reserved.
R
matosis 1, tuberous sclerosis, Rett syndrome, and Angelman syndrome) allows us to make the right diagnosis, determine if this is a simplex or familial case, and counsel the family regarding the risk of recurrence. In others, we are less certain and may not be able to differentiate between autosomal versus sex-linked recessive inheritance (2 affected boys in a family) or sex-linked recessive versus mitochondrial inheritance patterns (maternal transmission). Although we might be able to estimate the maximum possible risk of recurrence in a given family, it is probably best to use the expertise of trained genetic counselors. One important complicating phenomenon is that of gonadal mosaicism. This idea may be considered in settings in which there are 2 or more affected children and neither parent harbors a mutation, at least by blood test. The hypothesis is that gonadal tissue in one of the parents contains germ cells of both mutant and normal genotypes. Although linkage analysis and positional cloning have largely been replaced by array comparative genomic hybridization (array CGH) and whole-genome or whole-exome sequencing, these are still useful research tools. We are still uncovering the genetic basis of rare, recessively inherited syndromes by the study of consanguineous families and identifying regions of homozygosity as candidate regions.
esidents in child neurology strive to be excellent diagnosticians, to understand the pathologic processes affecting their patient at many levels, to define the etiology of disease, and to formulate a treatment plan. It has become the norm to aim for an understanding at the genetic and molecular level. Could the patient I am dealing with now have a genetic disorder? Is there a risk of recurrence in this family? The importance of obtaining a careful medical history has been emphasized to us from the start of medical school. This is absolutely essential in order for the clinician to formulate an image of the clinical problem; decide on its nature (temporal course, area of the nervous system that is primarily affected); and start thinking about etiology, pathogenesis, and treatment. Every time we evaluate a child with a neurologic disorder, these questions should be considered. A detailed family history and construction of a pedigree should be part of every new patient encounter in child neurology. This is placed in a framework of the following recognized inheritance patterns: autosomal dominant, autosomal recessive, sex-linked inheritance (recessive and dominant), and mitochondrial inheritance. In practice, we are most often faced with singleton cases or families with at most 2 affected siblings and have to recognize the possibility that this may be a genetic disorder, rather than being the result of an environmental insult. In some cases, the phenotype (eg, neurofibro-
Basic Concepts
From St Joseph’s Hospital and Medical Center, Barrow Neurological Institute, Phoenix, AZ. Address reprint requests to Vinodh Narayanan, MD, St Joseph’s Hospital and Medical Center, Barrow Neurological Institute, 500 W Thomas Rd, Suite 400, Phoenix, AZ 85013. E-mail: [email protected]
Some concepts (DNA to RNA to protein) are familiar to all students, but many new ideas may not be second nature to everyone. A typical gene contains multiple exons and introns. Transcription of the gene is regulated by control elements (promoters, enhancers, and suppressors). The primary RNA
1071-9091/11/$-see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.spen.2011.05.012
81
V. Narayanan
82 transcript is processed in the nucleus by splicing machinery, removing intronic sequences, and adding a poly-A segment at the 3=-end and “cap” structure at the 5=-end, resulting in mature messenger RNA (mRNA). Through a combination of these processes (ie, different promoters, different transcription start sites, alternative splicing of exons, and multiple polyadenylation sites), a single gene might give rise to thousands of different mRNAs. The mRNA typically includes segments that code for a protein and noncoding regions at the 5=- and 3=-end (5=-untranslated region and 3=-untranslated region). Mature mRNA is exported to the cytoplasm for translation into proteins. Proteins are further modified (either during or after translation) and targeted to the appropriate cellular compartment. Genes encode proteins, ribosomal RNAs, transfer RNAs (tRNAs), and a class of noncoding RNAs called microRNAs (miRNAs); miRNAs can regulate the stability of specific target mRNAs and can also modulate translation of target mRNAs.1 Mitochondria are a unique cellular organelle in that they contain their own DNA, transcription, and translational machinery. The mitochondrial genome has a slightly different genetic code than nuclear DNA. Mitochondrial DNA, a 16.5kbp circular DNA molecule much like a bacterial chromosome, contains genes that encode 13 polypeptides, a full complement of tRNAs corresponding to the 20 amino acids, and 2 ribosomal RNA genes. The remaining proteins in mitochondria (proteins that are in the outer or inner mitochondrial membrane, proteins in the intermembrane space, and proteins within the mitochondrial matrix) are encoded by nuclear genes, are translated in the cell cytoplasm, and then are targeted to specific mitochondrial compartments. Thus, there are several potential genetic mechanisms that result in mitochondrial dysfunction including (1) mitochondrial DNA mutation, (2) mutation of a nuclear gene encoding a protein involved in mitochondrial biogenesis and maintenance (DNA replication, transcription, protein translation, fission, fusion, import of proteins), and (3) mutation of a nuclear gene encoding a specific structural protein or enzyme (mutation could result in mis-targeting or altered function). There are a number of different classes of mutations that can cause disease. These include cytogenetically demonstrable chromosomal defects (ie, aneuploidy, balanced or unbalanced translocations, and large deletions or duplications); deletions; duplications; inversions; small deletions and insertions that shift the reading frame; missense, and nonsense point mutations. The study of the genetic basis of fragile X syndrome led to the discovery of a new class of mutations, trinucleotide repeat expansions. This has led to an understanding at the molecular level for phenomena, such as anticipation (earlier onset disease or worse phenotype with each generation), nonexpressing carrier in a dominant disorder (premutation), and meiotic instability. Nucleotide repeat expansions are associated with many neurologic disorders. In some cases, the nucleotide repeat expansion occurs in a noncoding gene segment, affecting transcription or RNA metabolism (fragile X). In others, the expansion occurs within a protein coding segment, often resulting in expansion of a
polyglutamine or polyalanine tract within the protein. This causes a dramatic change in protein structure and function. It is important to consider how mutations associated with specific classes of inherited disease cause cellular dysfunction. For diseases that exhibit an autosomal dominant inheritance pattern, the phenotype is expressed in heterozygotes. In some cases, the mutant RNA or protein might acquire a novel pathologic cell function (gain of function mutation). For instance, myotonic dystrophy type 1 is caused by the expansion of a CTG repeat located in the 3=-untranslated region of the DMPK gene. The mutant form of DMPK mRNA appears to have a toxic effect on cells, interfering with the splicing or translation of other RNA species. Spinocerebellar ataxia type 1 is caused by expansion of a CAG trinucleotide repeat within a protein coding segment of the ATXN1 gene. This mutation results in expansion of a polyglutamine (poly-Q) tract in the ataxin protein. As in several poly-glutamine diseases, mutant proteins aggregate in the nucleus forming nuclear inclusions and affecting cellular processes that normally degrade misfolded proteins (ie, chaperones, ubiquitin, and proteasomes). In other cases (eg, retinoblastoma, neurofibromatosis, or tuberous sclerosis), the targeted gene might encode a tumor suppressor. Patients who harbor a germline mutation in such a gene are heterozygous (haploinsufficient) in all cells. Tumors that are associated with these syndromes appear because of a “second-hit” somatic mutation that inactivates the remaining normal allele. In the case of most chromosomal defects (ie, duplications and deletions), we infer that the phenotype is a result of an abnormality of gene dosage. Most genetic syndromes that are associated with copy number variations (CNVs) fall into this category. For diseases that exhibit an autosomal recessive inheritance pattern, the phenotype is expressed only when both alleles are mutated (homozygotes or compound heterozygotes). These disorders are a result of loss of function of the encoded protein. Most metabolic disorders that result in an enzyme deficiency fall into this category. It is implied that heterozygous carriers do not express a disease phenotype because having 50% of normal levels of the encoded protein (haploinsufficiency) is sufficient for normal cellular function. This may not hold true when other factors (epigenetic modifications) affect transcription of the normal allele.
Phenotype to Genotype When approaching a child with a genetic disorder, the initial task facing us is to accurately define the neurologic phenotype and work toward determining the genotype. It is important for the neurologist to recognize neurologic and morphologic features characteristic of certain conditions. These include Rett syndrome and its variants, Prader-Willi syndrome, Angelman syndrome, fragile X syndrome, Down syndrome, Williams syndrome, and so on. Consultation with a clinical geneticist expert in dysmorphology is an important initial step. Based on the clinical phenotype, we can formulate a strategy for genetic testing that quickly allows us to confirm the precise diagnosis. There are tools and
Genetics and child neurology databases (OMIM, http://www.genetests.org and http:// www.Simulconsult.com) available through the Internet that help the clinician develop a differential diagnosis and quickly access what is known about a particular phenotype.
Genetic Testing An understanding of the reasons for genetic testing, the screening tools that are commonly used, their particular advantages and limitations, implications of genetic testing results, and their respective costs are probably important for every child neurologist. In some cases, the existing knowledge about the genetic basis of a particular syndrome (tuberous sclerosis complex for instance) may suggest that genetic testing in a particular patient may not help in further refining the diagnosis or guiding treatment. In most cases, we are dealing with broadly defined phenotypes (ie, developmental delay, microcephaly, dystonia, and ataxia) in which a genetic molecular diagnosis is important for a number of reasons. A precise diagnosis may allow us to counsel the family about the expected course of disease and prognosis. It also allows for genetic counseling and helping the family in future family planning. Once we make a precise genetic/molecular diagnosis, we can draw on the advances in neurosciences to better understand the pathogenesis of the disease. This permits us to avoid factors that may aggravate or exacerbate symptoms or use novel approaches to treatment (mGluR inhibitors for fragile X). What is the next step when the neurologic phenotype is not sufficient to suggest a small number of specific diagnoses? We then must turn to genetic screening tools. There is still a place for the high-resolution karyotype (cytogenetic testing), but this has largely been replaced as a genetic screening tool by array CGH. Array CGH is very sensitive and specific for detecting variations in genomic copy number (ie, duplications and deletions), with a resolution that is determined by the density of probes on the array. Clinical laboratories may differ in the precise method or platform that is used for array CGH; many use an array of synthetic oligonucleotides designed to hybridize against most of the human genome, some use oligonucleotides that are designed to detect single nucleotide polymorphisms located throughout the genome, and some use large cloned DNA segments (bacterial artificial chromosome clones). One important limitation of this technique is that rearrangements that do not change gene dosing, such as balanced translocations, will be missed. With the increased use of this screening tool, we have also learned about copy number polymorphisms in the normal population. Thus, the identification of a CNV in a patient does not necessarily mean this genetic variant is the cause of the neurological phenotype. One follow-up test suggested in most clinical reports in which a CNV has been identified in a patient is to look for the same CNV in both parents. A de novo CNV (ie, one not found in either parent) may have greater pathologic significance than if one of the parents had the same CNV as the patient. However, sometimes even inherited CNVs may affect the child and parent differently. This is because they differ completely with regards to the homologous chromosome.
83 We are already well into the era of whole genome screening using techniques like array CGH for genotyping. With advances in technology, the cost of sequencing an entire genome has plummeted and may soon be comparable to current costs for array CGH. Although the costs of whole genome or whole exome sequencing may make it feasible in every patient, these are still in the realm of research tools, and much work needs to be done to understand how best to use these methods when it comes to our patients. At this time, sifting through the gigabits of data generated by whole genome sequencing requires a team of bioinformatics and genomic scientists.
Epigenetics Epigenetics refers to a level of control of genes and their expression that goes beyond the primary sequence of the gene and its control elements. Epigenetic tags or marks alter the expression of genes. Two genes (for instance, the 2 alleles on different chromosomes) may have identical DNA sequences, but their expression might be differentially regulated by chemical methylation of CpG dinucleotides, a reaction that is performed by site-specific DNA methylases. Methylation may alter the binding of proteins to the DNA (ie, transcriptional factors and proteins that modulate the local topology of the chromosome) and render a particular gene transcriptionally active or inactive. Epigenetic modification of DNA may also lead to changes in the topological structure of DNA/nucleosome altering access by transcriptional protein complexes. The most dramatic example of such epigenetic modulation of transcription is X inactivation. In all females, one of the 2 X chromosomes is completely silenced in every diploid cell in the body. Which of the 2 X chromosomes (paternally inherited or maternally inherited) is transcriptionally active is a stochastic event and varies from cell to cell. Thus, normal females are a mosaic of 2 cell populations, each expressing one of the 2 X chromosomes. In most women, the pattern of X inactivation is random, with approximately 50% of the cells expressing the paternal X chromosome and 50% of the cells expressing the maternal X chromosome. In a minority of women (about 10%), the X inactivation pattern is skewed, so that one of the 2 X chromosomes is preferentially expressed in ⬎80% of the cells. This is of clinical importance when a female is a carrier for a mutant gene on the X chromosome; they usually do not manifest symptoms of the disease. A phenotypically normal female who has multiple affected girls with Rett syndrome may have extremely skewed X inactivation.2 In such cases, the expression of the mutant gene might result in a growth disadvantage to the cell applying a negative selection, with the end result being that most cells express the normal X chromosome (skewing of X inactivation). An example of this phenomenon is incontinentia pigmenti in which cells expressing the mutant X chromosome are eliminated by apoptosis and the fraction of cells expressing the mutant X chromosome decreases with age. When a carrier female shows skewed X inactivation with a majority of cells expressing the mutant chromosome (as might happen in
V. Narayanan
84 muscle disorders, such as Duchenne muscular dystrophy), the carrier female might exhibit the phenotype of the disease.3,4 A second clinically important example of epigenetics is the mechanism of disease in Prader-Willi syndrome (PWS) and Angelman syndrome (AS). In the majority (⬃70%) of cases of PWS, the mutation is a microdeletion within chromosome 15q11-13. Similarly, the majority (70%-75%) of cases of AS is also caused by a microdeletion within chromosome 15q1113. How can such a microdeletion within a single region of chromosome 15 cause such different phenotypes? The study of these 2 disorders led to the discovery that this region is “imprinted.” The maternal and paternal alleles may have the exact same complement of genes and the same DNA sequence but are not equivalent when it comes to gene expression. Some of the genes in this cluster are silenced on the maternal chromosome, whereas a complementary set of genes are silenced on the paternal chromosome. The expression of genes from both the maternal and paternal chromosomes is required for normal brain development. When the microdeletion is located on the paternal chromosome, only a subset of genes is transcribed from the maternal allele, and the patient develops PWS. The same phenotype results when both chromosomes 15 come from the maternal germ cell (maternal uniparental disomy) or when there is a primary defect in the control of the imprinting process. When the microdeletion is located on the maternal chromosome, only genes from the paternal chromosome are transcribed, resulting in the AS phenotype. Paternal uniparental disomy accounts for 3% to 5% of cases of AS, whereas the remaining are caused by point mutations in the maternal copy of the AS gene (UBE3A). Methylation of DNA appears to be a critical mechanism for genetic imprinting.
Treatment of Genetic Disorders Ultimately, we need to pay particular attention to disorders in which a diagnosis may allow us to alter the course of the disease. Diseases that are due to the absence of an enzyme (nonneurologic form of Gaucher disease, caused by glucocerebrosidase deficiency, and Pompe disease, caused by acid maltase deficiency) may be treated by enzyme replacement
therapy. In some cases, early diagnosis (by metabolic or genetic testing) may permit treatment with organ transplantation, stem cell transplantation, or transplantation of cell after ex vivo genetic engineering. Specific genetic strategies can be developed to treat certain classes of mutations. The injection of antisense oligonucleotides that induce skipping of mutant exons in Duchenne muscular dystrophy may convert the disease into a milder (Becker) form. Drugs that promote stop codon readthrough are being tested in genetic disorders caused by nonsense mutations. Viral vectors (adeno-associated virus) carrying minigenes that correct the underlying defect or that express miRNAs that target the mutant mRNA are being studied as potential treatments. Advances in our understanding of the biological function of molecules encoded by the mutant genes (altered signaling pathways) allow us to consider novel treatments for previously untreatable conditions. Examples include the use of metabotropic glutamate receptor (mGluR) inhibitors for fragile X mental retardation, mammalian target of rapamycin (mTOR) inhibitors for tuberous sclerosis, and Ras inhibitors for neurofibromatosis. We are in a very exciting period in which advances in neurosciences and genomics will lead to novel approaches to treating developmental brain disorders. Child neurologists have to understand these new approaches and use all these tools appropriately in treating our patients and their families.
References 1. Feero GW, Guttmacher AE, Collins FS: Genomic medicine—An updated primer. N Engl J Med 362:2001-2011, 2010 2. Sirianni N, Naidu S, Pereira J, et al: Rett syndrome: Confirmation of X-linked dominant inheritance, and localization of the gene to Xq28. Am J Hum Genet 63:1552-1558, 1998 3. Puck JM, Willard HF: X inactivation in females with X-linked disease. N Engl J Med 338:325-328, 1998 4. Plenge RM, Stevenson RA, Lubs HA, et al: Skewed X-chromosome inactivation is a common feature of X-linked mental retardation disorders. Am J Hum Genet 71:168-173, 2002
Internet Resources http://www.genetests.org http://www.simulconsult.com University of Utah Genetic Science Learning Center: http:// learn.genetics.utah.edu OMIM: http://www.ncbi.nlm.nih.gov/omimglossary of terms
R
matosis 1, tuberous sclerosis, Rett syndrome, and Angelman syndrome) allows us to make the right diagnosis, determine if this is a simplex or familial case, and counsel the family regarding the risk of recurrence. In others, we are less certain and may not be able to differentiate between autosomal versus sex-linked recessive inheritance (2 affected boys in a family) or sex-linked recessive versus mitochondrial inheritance patterns (maternal transmission). Although we might be able to estimate the maximum possible risk of recurrence in a given family, it is probably best to use the expertise of trained genetic counselors. One important complicating phenomenon is that of gonadal mosaicism. This idea may be considered in settings in which there are 2 or more affected children and neither parent harbors a mutation, at least by blood test. The hypothesis is that gonadal tissue in one of the parents contains germ cells of both mutant and normal genotypes. Although linkage analysis and positional cloning have largely been replaced by array comparative genomic hybridization (array CGH) and whole-genome or whole-exome sequencing, these are still useful research tools. We are still uncovering the genetic basis of rare, recessively inherited syndromes by the study of consanguineous families and identifying regions of homozygosity as candidate regions.
esidents in child neurology strive to be excellent diagnosticians, to understand the pathologic processes affecting their patient at many levels, to define the etiology of disease, and to formulate a treatment plan. It has become the norm to aim for an understanding at the genetic and molecular level. Could the patient I am dealing with now have a genetic disorder? Is there a risk of recurrence in this family? The importance of obtaining a careful medical history has been emphasized to us from the start of medical school. This is absolutely essential in order for the clinician to formulate an image of the clinical problem; decide on its nature (temporal course, area of the nervous system that is primarily affected); and start thinking about etiology, pathogenesis, and treatment. Every time we evaluate a child with a neurologic disorder, these questions should be considered. A detailed family history and construction of a pedigree should be part of every new patient encounter in child neurology. This is placed in a framework of the following recognized inheritance patterns: autosomal dominant, autosomal recessive, sex-linked inheritance (recessive and dominant), and mitochondrial inheritance. In practice, we are most often faced with singleton cases or families with at most 2 affected siblings and have to recognize the possibility that this may be a genetic disorder, rather than being the result of an environmental insult. In some cases, the phenotype (eg, neurofibro-
Basic Concepts
From St Joseph’s Hospital and Medical Center, Barrow Neurological Institute, Phoenix, AZ. Address reprint requests to Vinodh Narayanan, MD, St Joseph’s Hospital and Medical Center, Barrow Neurological Institute, 500 W Thomas Rd, Suite 400, Phoenix, AZ 85013. E-mail: [email protected]
Some concepts (DNA to RNA to protein) are familiar to all students, but many new ideas may not be second nature to everyone. A typical gene contains multiple exons and introns. Transcription of the gene is regulated by control elements (promoters, enhancers, and suppressors). The primary RNA
1071-9091/11/$-see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.spen.2011.05.012
81
V. Narayanan
82 transcript is processed in the nucleus by splicing machinery, removing intronic sequences, and adding a poly-A segment at the 3=-end and “cap” structure at the 5=-end, resulting in mature messenger RNA (mRNA). Through a combination of these processes (ie, different promoters, different transcription start sites, alternative splicing of exons, and multiple polyadenylation sites), a single gene might give rise to thousands of different mRNAs. The mRNA typically includes segments that code for a protein and noncoding regions at the 5=- and 3=-end (5=-untranslated region and 3=-untranslated region). Mature mRNA is exported to the cytoplasm for translation into proteins. Proteins are further modified (either during or after translation) and targeted to the appropriate cellular compartment. Genes encode proteins, ribosomal RNAs, transfer RNAs (tRNAs), and a class of noncoding RNAs called microRNAs (miRNAs); miRNAs can regulate the stability of specific target mRNAs and can also modulate translation of target mRNAs.1 Mitochondria are a unique cellular organelle in that they contain their own DNA, transcription, and translational machinery. The mitochondrial genome has a slightly different genetic code than nuclear DNA. Mitochondrial DNA, a 16.5kbp circular DNA molecule much like a bacterial chromosome, contains genes that encode 13 polypeptides, a full complement of tRNAs corresponding to the 20 amino acids, and 2 ribosomal RNA genes. The remaining proteins in mitochondria (proteins that are in the outer or inner mitochondrial membrane, proteins in the intermembrane space, and proteins within the mitochondrial matrix) are encoded by nuclear genes, are translated in the cell cytoplasm, and then are targeted to specific mitochondrial compartments. Thus, there are several potential genetic mechanisms that result in mitochondrial dysfunction including (1) mitochondrial DNA mutation, (2) mutation of a nuclear gene encoding a protein involved in mitochondrial biogenesis and maintenance (DNA replication, transcription, protein translation, fission, fusion, import of proteins), and (3) mutation of a nuclear gene encoding a specific structural protein or enzyme (mutation could result in mis-targeting or altered function). There are a number of different classes of mutations that can cause disease. These include cytogenetically demonstrable chromosomal defects (ie, aneuploidy, balanced or unbalanced translocations, and large deletions or duplications); deletions; duplications; inversions; small deletions and insertions that shift the reading frame; missense, and nonsense point mutations. The study of the genetic basis of fragile X syndrome led to the discovery of a new class of mutations, trinucleotide repeat expansions. This has led to an understanding at the molecular level for phenomena, such as anticipation (earlier onset disease or worse phenotype with each generation), nonexpressing carrier in a dominant disorder (premutation), and meiotic instability. Nucleotide repeat expansions are associated with many neurologic disorders. In some cases, the nucleotide repeat expansion occurs in a noncoding gene segment, affecting transcription or RNA metabolism (fragile X). In others, the expansion occurs within a protein coding segment, often resulting in expansion of a
polyglutamine or polyalanine tract within the protein. This causes a dramatic change in protein structure and function. It is important to consider how mutations associated with specific classes of inherited disease cause cellular dysfunction. For diseases that exhibit an autosomal dominant inheritance pattern, the phenotype is expressed in heterozygotes. In some cases, the mutant RNA or protein might acquire a novel pathologic cell function (gain of function mutation). For instance, myotonic dystrophy type 1 is caused by the expansion of a CTG repeat located in the 3=-untranslated region of the DMPK gene. The mutant form of DMPK mRNA appears to have a toxic effect on cells, interfering with the splicing or translation of other RNA species. Spinocerebellar ataxia type 1 is caused by expansion of a CAG trinucleotide repeat within a protein coding segment of the ATXN1 gene. This mutation results in expansion of a polyglutamine (poly-Q) tract in the ataxin protein. As in several poly-glutamine diseases, mutant proteins aggregate in the nucleus forming nuclear inclusions and affecting cellular processes that normally degrade misfolded proteins (ie, chaperones, ubiquitin, and proteasomes). In other cases (eg, retinoblastoma, neurofibromatosis, or tuberous sclerosis), the targeted gene might encode a tumor suppressor. Patients who harbor a germline mutation in such a gene are heterozygous (haploinsufficient) in all cells. Tumors that are associated with these syndromes appear because of a “second-hit” somatic mutation that inactivates the remaining normal allele. In the case of most chromosomal defects (ie, duplications and deletions), we infer that the phenotype is a result of an abnormality of gene dosage. Most genetic syndromes that are associated with copy number variations (CNVs) fall into this category. For diseases that exhibit an autosomal recessive inheritance pattern, the phenotype is expressed only when both alleles are mutated (homozygotes or compound heterozygotes). These disorders are a result of loss of function of the encoded protein. Most metabolic disorders that result in an enzyme deficiency fall into this category. It is implied that heterozygous carriers do not express a disease phenotype because having 50% of normal levels of the encoded protein (haploinsufficiency) is sufficient for normal cellular function. This may not hold true when other factors (epigenetic modifications) affect transcription of the normal allele.
Phenotype to Genotype When approaching a child with a genetic disorder, the initial task facing us is to accurately define the neurologic phenotype and work toward determining the genotype. It is important for the neurologist to recognize neurologic and morphologic features characteristic of certain conditions. These include Rett syndrome and its variants, Prader-Willi syndrome, Angelman syndrome, fragile X syndrome, Down syndrome, Williams syndrome, and so on. Consultation with a clinical geneticist expert in dysmorphology is an important initial step. Based on the clinical phenotype, we can formulate a strategy for genetic testing that quickly allows us to confirm the precise diagnosis. There are tools and
Genetics and child neurology databases (OMIM, http://www.genetests.org and http:// www.Simulconsult.com) available through the Internet that help the clinician develop a differential diagnosis and quickly access what is known about a particular phenotype.
Genetic Testing An understanding of the reasons for genetic testing, the screening tools that are commonly used, their particular advantages and limitations, implications of genetic testing results, and their respective costs are probably important for every child neurologist. In some cases, the existing knowledge about the genetic basis of a particular syndrome (tuberous sclerosis complex for instance) may suggest that genetic testing in a particular patient may not help in further refining the diagnosis or guiding treatment. In most cases, we are dealing with broadly defined phenotypes (ie, developmental delay, microcephaly, dystonia, and ataxia) in which a genetic molecular diagnosis is important for a number of reasons. A precise diagnosis may allow us to counsel the family about the expected course of disease and prognosis. It also allows for genetic counseling and helping the family in future family planning. Once we make a precise genetic/molecular diagnosis, we can draw on the advances in neurosciences to better understand the pathogenesis of the disease. This permits us to avoid factors that may aggravate or exacerbate symptoms or use novel approaches to treatment (mGluR inhibitors for fragile X). What is the next step when the neurologic phenotype is not sufficient to suggest a small number of specific diagnoses? We then must turn to genetic screening tools. There is still a place for the high-resolution karyotype (cytogenetic testing), but this has largely been replaced as a genetic screening tool by array CGH. Array CGH is very sensitive and specific for detecting variations in genomic copy number (ie, duplications and deletions), with a resolution that is determined by the density of probes on the array. Clinical laboratories may differ in the precise method or platform that is used for array CGH; many use an array of synthetic oligonucleotides designed to hybridize against most of the human genome, some use oligonucleotides that are designed to detect single nucleotide polymorphisms located throughout the genome, and some use large cloned DNA segments (bacterial artificial chromosome clones). One important limitation of this technique is that rearrangements that do not change gene dosing, such as balanced translocations, will be missed. With the increased use of this screening tool, we have also learned about copy number polymorphisms in the normal population. Thus, the identification of a CNV in a patient does not necessarily mean this genetic variant is the cause of the neurological phenotype. One follow-up test suggested in most clinical reports in which a CNV has been identified in a patient is to look for the same CNV in both parents. A de novo CNV (ie, one not found in either parent) may have greater pathologic significance than if one of the parents had the same CNV as the patient. However, sometimes even inherited CNVs may affect the child and parent differently. This is because they differ completely with regards to the homologous chromosome.
83 We are already well into the era of whole genome screening using techniques like array CGH for genotyping. With advances in technology, the cost of sequencing an entire genome has plummeted and may soon be comparable to current costs for array CGH. Although the costs of whole genome or whole exome sequencing may make it feasible in every patient, these are still in the realm of research tools, and much work needs to be done to understand how best to use these methods when it comes to our patients. At this time, sifting through the gigabits of data generated by whole genome sequencing requires a team of bioinformatics and genomic scientists.
Epigenetics Epigenetics refers to a level of control of genes and their expression that goes beyond the primary sequence of the gene and its control elements. Epigenetic tags or marks alter the expression of genes. Two genes (for instance, the 2 alleles on different chromosomes) may have identical DNA sequences, but their expression might be differentially regulated by chemical methylation of CpG dinucleotides, a reaction that is performed by site-specific DNA methylases. Methylation may alter the binding of proteins to the DNA (ie, transcriptional factors and proteins that modulate the local topology of the chromosome) and render a particular gene transcriptionally active or inactive. Epigenetic modification of DNA may also lead to changes in the topological structure of DNA/nucleosome altering access by transcriptional protein complexes. The most dramatic example of such epigenetic modulation of transcription is X inactivation. In all females, one of the 2 X chromosomes is completely silenced in every diploid cell in the body. Which of the 2 X chromosomes (paternally inherited or maternally inherited) is transcriptionally active is a stochastic event and varies from cell to cell. Thus, normal females are a mosaic of 2 cell populations, each expressing one of the 2 X chromosomes. In most women, the pattern of X inactivation is random, with approximately 50% of the cells expressing the paternal X chromosome and 50% of the cells expressing the maternal X chromosome. In a minority of women (about 10%), the X inactivation pattern is skewed, so that one of the 2 X chromosomes is preferentially expressed in ⬎80% of the cells. This is of clinical importance when a female is a carrier for a mutant gene on the X chromosome; they usually do not manifest symptoms of the disease. A phenotypically normal female who has multiple affected girls with Rett syndrome may have extremely skewed X inactivation.2 In such cases, the expression of the mutant gene might result in a growth disadvantage to the cell applying a negative selection, with the end result being that most cells express the normal X chromosome (skewing of X inactivation). An example of this phenomenon is incontinentia pigmenti in which cells expressing the mutant X chromosome are eliminated by apoptosis and the fraction of cells expressing the mutant X chromosome decreases with age. When a carrier female shows skewed X inactivation with a majority of cells expressing the mutant chromosome (as might happen in
V. Narayanan
84 muscle disorders, such as Duchenne muscular dystrophy), the carrier female might exhibit the phenotype of the disease.3,4 A second clinically important example of epigenetics is the mechanism of disease in Prader-Willi syndrome (PWS) and Angelman syndrome (AS). In the majority (⬃70%) of cases of PWS, the mutation is a microdeletion within chromosome 15q11-13. Similarly, the majority (70%-75%) of cases of AS is also caused by a microdeletion within chromosome 15q1113. How can such a microdeletion within a single region of chromosome 15 cause such different phenotypes? The study of these 2 disorders led to the discovery that this region is “imprinted.” The maternal and paternal alleles may have the exact same complement of genes and the same DNA sequence but are not equivalent when it comes to gene expression. Some of the genes in this cluster are silenced on the maternal chromosome, whereas a complementary set of genes are silenced on the paternal chromosome. The expression of genes from both the maternal and paternal chromosomes is required for normal brain development. When the microdeletion is located on the paternal chromosome, only a subset of genes is transcribed from the maternal allele, and the patient develops PWS. The same phenotype results when both chromosomes 15 come from the maternal germ cell (maternal uniparental disomy) or when there is a primary defect in the control of the imprinting process. When the microdeletion is located on the maternal chromosome, only genes from the paternal chromosome are transcribed, resulting in the AS phenotype. Paternal uniparental disomy accounts for 3% to 5% of cases of AS, whereas the remaining are caused by point mutations in the maternal copy of the AS gene (UBE3A). Methylation of DNA appears to be a critical mechanism for genetic imprinting.
Treatment of Genetic Disorders Ultimately, we need to pay particular attention to disorders in which a diagnosis may allow us to alter the course of the disease. Diseases that are due to the absence of an enzyme (nonneurologic form of Gaucher disease, caused by glucocerebrosidase deficiency, and Pompe disease, caused by acid maltase deficiency) may be treated by enzyme replacement
therapy. In some cases, early diagnosis (by metabolic or genetic testing) may permit treatment with organ transplantation, stem cell transplantation, or transplantation of cell after ex vivo genetic engineering. Specific genetic strategies can be developed to treat certain classes of mutations. The injection of antisense oligonucleotides that induce skipping of mutant exons in Duchenne muscular dystrophy may convert the disease into a milder (Becker) form. Drugs that promote stop codon readthrough are being tested in genetic disorders caused by nonsense mutations. Viral vectors (adeno-associated virus) carrying minigenes that correct the underlying defect or that express miRNAs that target the mutant mRNA are being studied as potential treatments. Advances in our understanding of the biological function of molecules encoded by the mutant genes (altered signaling pathways) allow us to consider novel treatments for previously untreatable conditions. Examples include the use of metabotropic glutamate receptor (mGluR) inhibitors for fragile X mental retardation, mammalian target of rapamycin (mTOR) inhibitors for tuberous sclerosis, and Ras inhibitors for neurofibromatosis. We are in a very exciting period in which advances in neurosciences and genomics will lead to novel approaches to treating developmental brain disorders. Child neurologists have to understand these new approaches and use all these tools appropriately in treating our patients and their families.
References 1. Feero GW, Guttmacher AE, Collins FS: Genomic medicine—An updated primer. N Engl J Med 362:2001-2011, 2010 2. Sirianni N, Naidu S, Pereira J, et al: Rett syndrome: Confirmation of X-linked dominant inheritance, and localization of the gene to Xq28. Am J Hum Genet 63:1552-1558, 1998 3. Puck JM, Willard HF: X inactivation in females with X-linked disease. N Engl J Med 338:325-328, 1998 4. Plenge RM, Stevenson RA, Lubs HA, et al: Skewed X-chromosome inactivation is a common feature of X-linked mental retardation disorders. Am J Hum Genet 71:168-173, 2002
Internet Resources http://www.genetests.org http://www.simulconsult.com University of Utah Genetic Science Learning Center: http:// learn.genetics.utah.edu OMIM: http://www.ncbi.nlm.nih.gov/omimglossary of terms