L07 Genetics Whole genome analysis 2

Overview

This lecture contrasts conventional optical karyotyping with modern genome-wide methods for detecting chromosomal imbalance, centred on chromosomal microarray (CMA). A family case series (mosaic chromosome 2 duplication, then Alagille syndrome caused by a chromosome 20 microdeletion) is used to show how CMA is applied, how a CNV’s pathogenicity is established, and how incidental and uncertain-significance findings are handled. The lecture closes with genotype-phenotype mapping in Cri du chat syndrome and general principles for interpreting copy number variants (CNVs).

Chromosomal imbalance: background and detection

• A karyotype is produced by pairing and ordering the chromosomes seen in a raw, unsorted metaphase spread into a numbered karyogram (pairs 1-22 plus X and Y).
• In the interphase (non-dividing) nucleus, each chromosome occupies its own discrete spatial “territory” rather than being randomly intermingled with the others, as shown by FISH-painting experiments (Nature Reviews Genetics, 2001).
• The chromosomal basis of disease rests on genomic balance: chromosome anomalies are common, but not all cause disease. Abnormality rates are 50% of spontaneously aborted fetuses, 10% of stillborn babies, and 0.7% of newborn babies (neonates).
• Microarray analysis of 3400 healthy children shows that apparently healthy people carry many small background deletions and duplications throughout the genome. The central challenge of clinical CNV interpretation is distinguishing pathogenic CNVs from this normal background variation (BMC Medical Genomics, 2021).

Chromosomal microarray (CMA): principle and method

• CMA is a generalised, genome-wide survey for genomic imbalance, also called array Comparative Genomic Hybridisation (CGH). It substantially bridges the gap between chromosome-level (karyotype) and sequencing-level analysis of genetic disorders; optical microscopic karyotyping is now used only rarely.
• Method: uses genomic DNA directly, with no need for metaphase chromosome preparations.
  1. Patient (test) DNA and a normal control genome are differentially fluorescently labelled.
  2. The two are mixed 1:1 and co-hybridised to a spotted DNA array.
  3. Relative DNA content of test versus normal at each locus is read from the red:green fluorescence ratio.
• Plot interpretation: points cluster around a “balance” baseline where test and control DNA content are equal (yellow at the spot level); a deletion produces a cluster of points shifted below baseline (green at the spot level); a duplication produces a cluster shifted above baseline (red at the spot level).
• Resolution: current arrays assay between 1 and 2.5 million individual data points across the genome, giving high-resolution, accurate sizing of deletions and duplications and identifying exactly which genes are affected and which are not.
• A deletion can be mapped onto the genome sequence to ask what it removes beyond coding genes: overlap with a mark of active regulatory activity (H3K27ac) within a deleted interval supports the deletion acting by removing regulatory elements that control a nearby gene’s expression, rather than by removing the gene itself.

Case: mosaic duplication of chromosome 2

• A 12-month-old child (born mid-1970s) presented to a GP with nonspecific global neurodevelopmental delay: not sitting, not rolling, little hand regard, vocalising, low tone, and no family history of neurodevelopmental problems. Conventional optical chromosome analysis was performed.
• Result: a post-zygotic (mosaic) duplication of chromosome 2, karyotype mos 46,XY,dup(2)(q23q35)[5]/46,XY[95] — 5 of 100 cells scored carried the duplicated chromosome 2 (region q23 to q35), the remainder being a normal 46,XY clone.
• The lecture poses two questions arising from a post-zygotic finding: what it implies for sibling recurrence risk in future pregnancies, and what prognostication is possible for the child. [slide does not elaborate on the answers]
• On the family pedigree, this finding is annotated as “dup chr2q31-37” in the affected child.

Case: Alagille syndrome and a chromosome 20 microdeletion

• In 2008, a related child (in the next generation of the same extended family) presented in infancy with conjugated hyperbilirubinaemia, a heart defect, and minor spinal anomalies.
• Neonatal jaundice aside: most healthy babies get some jaundice (“physiological” or breast milk jaundice), lasting a few weeks, caused by unconjugated bilirubin. This must be distinguished from conjugated hyperbilirubinaemia, which implies underlying liver or biliary disease; the two are easy to distinguish biochemically.
• Alagille syndrome is a rare autosomal dominant condition with multiple malformations of the biliary tract, heart, spine and eyes, and variable expressivity. It has a characteristic face (angular, long face, pointed chin, deep-set eyes). Jaundice results from hypoplasia of the biliary tree, a developmental defect. It is caused by mutations in one of two genes, JAG1 or NOTCH2.
• Diagnostic options for Alagille syndrome:
  1. Sequence the JAG1 or NOTCH2 genes.
  2. Look for small-scale deletions or duplications within these genes.
  3. Look for larger deletions or duplications encompassing JAG1 or NOTCH2 — the method of choice for this option is chromosomal microarray.
• CMA in this family identified a 3 Mb deletion on chromosome 20 (chr20:10,828,806-13,606,829) in the affected child.
• Establishing pathogenicity: the father, who himself had a history of craniofacial surgery and renal artery stenosis, was found to carry the same deletion de novo, and the deletion segregated with the phenotype through the family — together supporting it as the likely cause of Alagille syndrome in this family.
• Mapping the deletion onto the genome sequence showed it likely removes key regulatory elements controlling expression of JAG1 (rather than removing JAG1 itself), based on the deleted interval overlapping a peak of the H3K27ac regulatory mark near the gene MKKS.

Interpreting CNVs: incidental findings, susceptibility, and uncertain significance

Testing this family for the Alagille-associated deletion also uncovered additional, incidental CNV findings elsewhere in the pedigree, unrelated to the presenting condition.

The pedigree slides showing these incidental findings use several different shading/hatch patterns (dots, diagonal hatch, cross-hatch) to mark different family members, but no legend defining what each pattern denotes is given in the transcript, so which incidental finding corresponds to which fill pattern cannot be confirmed from the slide content alone.

• One incidental finding was a duplication at chr22q11, illustrating CNVs as susceptibility factors rather than deterministic disease causes: it has a strong association with developmental neuropsychiatric traits (e.g. ADHD, intellectual disability, specific learning disorders), but penetrance is low (around 20%, i.e. only about 1 in 5 carriers develop a phenotype). It is NOT determinative of phenotype and has little use for reproductive decision-making. This illustrates the general difficulty of communicating that a genetic factor can be a susceptibility factor even where a strong statistical association exists.
• Another incidental finding was a “private” deletion on chromosome 11 of uncertain significance. All genomes are replete with rare sequence and copy-number variants, so real caution must surround interpretation of such private variants; family history can be very valuable here. It is unlikely that genome sequence alone will be an accurate prognostic test for many traits, because most traits depend on combinations of other genetic factors together with environment.

Genotype-phenotype correlation: Cri du chat syndrome

• Cri du chat syndrome results from terminal and interstitial deletions of the short arm of chromosome 5 (5p).
• Different sub-regions along 5p correlate with different specific clinical features (speech, the characteristic “cat cry”, intellectual disability, and facial phenotype each map to particular bands along 5p), showing that fine-mapping a deletion against the genome sequence can sharpen genotype-phenotype correlation and prognostication.

Summary principles

• Both high- and low-resolution techniques exist for assessing chromosomal structure.
• High-resolution results must be interpreted in light of background variation, i.e. the normal CNV burden present in healthy people.
• CMA results should be interpreted alongside genome-level knowledge of the content of the regions subject to imbalance (which genes and regulatory elements lie within them).
• A powerful way to interpret a private or unique variant is to ask whether it (a) segregates with the phenotype in the family, or (b) has arisen de novo.
• As with all genetic testing, results can have familial implications.

Self-test

1. Describe how a karyogram is produced from a raw metaphase spread, and explain how chromosome organisation differs in the interphase nucleus.
2. What proportion of spontaneously aborted fetuses, stillborn babies, and newborns show chromosome abnormalities?
3. What central challenge does microarray analysis of large numbers of healthy children reveal about interpreting CNVs?
4. Describe the steps of the chromosomal microarray (CGH) method, from DNA labelling through to the read-out of imbalance.
5. On a CMA plot, how do a deletion, a duplication, and a balanced region appear?
6. Distinguish physiological (breast milk) jaundice from conjugated hyperbilirubinaemia.
7. List the clinical features and genetic causes of Alagille syndrome, and describe the three diagnostic options used to investigate it.
8. Describe the two lines of evidence used to establish that the chromosome 20 microdeletion found in this family was pathogenic.
9. Explain how mapping a deletion onto the genome sequence revealed a likely regulatory mechanism of pathogenicity in the JAG1 case.
10. What does the chr22q11 duplication illustrate about the relationship between a strong genetic association and clinical determinism, and what is its approximate penetrance?
11. What two considerations does the lecture recommend when interpreting a “private” CNV of uncertain significance?
12. Describe how genotype-phenotype correlation was demonstrated in Cri du chat syndrome.
13. Summarise the two general strategies described in this lecture for establishing whether a private or unique CNV is pathogenic.

Answers

Reveal answers

1. A raw metaphase spread shows unsorted, overlapping banded chromosomes as seen under the microscope; pairing and ordering these into numbered pairs (1-22 plus X and Y) produces a karyogram. In the interphase nucleus, by contrast, each chromosome occupies its own discrete spatial territory rather than being randomly intermingled with the others.
2. 50% of spontaneously aborted fetuses, 10% of stillborn babies, and 0.7% of newborn babies show chromosome abnormalities.
3. It reveals that apparently healthy individuals carry many small background CNVs throughout the genome, making it challenging to distinguish pathogenic CNVs from this normal background variation.
4. Patient (test) DNA and a normal control genome are differentially fluorescently labelled, mixed 1:1, and co-hybridised to a spotted DNA array; the relative DNA content of test versus normal at each locus is then read from the red:green fluorescence ratio.
5. A balanced region clusters around the baseline; a deletion shows as a cluster of points shifted below baseline; a duplication shows as a cluster of points shifted above baseline.
6. Physiological/breast milk jaundice is common in healthy babies, lasts a few weeks, and is due to unconjugated bilirubin. Conjugated hyperbilirubinaemia implies underlying liver or biliary disease; the two are easy to distinguish biochemically.
7. Alagille syndrome causes malformations of the biliary tract, heart, spine and eyes with variable expressivity, a characteristic face (angular, long face, pointed chin, deep-set eyes), and jaundice from biliary tree hypoplasia; it is caused by mutations in JAG1 or NOTCH2. Diagnostic options: (1) sequence JAG1/NOTCH2, (2) look for small-scale deletions/duplications within these genes, (3) look for larger deletions/duplications encompassing JAG1/NOTCH2, using chromosomal microarray.
8. The deletion was shown to have arisen de novo in the father, and it segregated with the phenotype through the family.
9. The deleted interval overlapped a peak of the H3K27ac mark of active regulatory activity near the gene MKKS, suggesting the deletion removes regulatory elements controlling JAG1 expression rather than removing JAG1 itself.
10. It shows that a CNV can be a susceptibility factor with a strong statistical association to a phenotype without being deterministic of it; penetrance for the chr22q11 duplication is low, around 20% (about 1 in 5 carriers develop a phenotype).
11. Real caution around interpretation, since all genomes carry rare private variants; and that family history can be very valuable in interpreting them.
12. Different sub-regions along the deleted 5p segment correlate with different specific clinical features (speech, the cat cry, intellectual disability, facial phenotype), each mapping to particular bands along 5p.
13. Ask whether the variant (a) segregates with the phenotype in the family, or (b) has arisen de novo.