L4: Segmental Duplications and Fusion Transcripts

Question 1

Colette’s current research line is regarding the role of human-specific duplicated genes in neurological disorders. Regarding evolution, in what two ways can this be studied?

Answer 1

Humans and prior: What genetic changes have occurred in the human genome during evolution? How might these affect human brain development?

Humans: Is there genomic variation in modern humans that exists as part of ongoing evolution? Could these changes have given our species an increased susceptibility to certain diseases?

Question 2

On what four levels foes genomic evolution roughly take place?

Answer 2

1) Changes in the coding part of genes
2) Changes in how genes are regulated
3) Changes in (epi)genomic environment?
4) Creation of new genes / deletion of existing genes

Question 3

Name four ways in which the creation of new genes or deletion of existing genes could occur

Answer 3

• Retrotransposition of processed pseudogenes
• De novo genesis from non-coding DNA
• Gene duplication
• Gene fusion

Question 4

How may de novo genesis from non-coding DNA occur?

Answer 4

Couldn’t previously encode a protein but undergoes some sequence mutation which allows translation to occur

Question 5

How can gene duplication occur?

Answer 5

Error during replication

Question 6

How may gene fusion occur?

Answer 6

Deletion causes fusion of two genes which create a new transcript

Question 7

To what extent do non-orthologous regions between human and chimp account for in our differences?

Answer 7

Non-orthologous regions between human and chimp account for 3% of genetic differences

Question 8

Do segmental duplications account for any differences in our genome between humans and chimps?

Answer 8

Yes, there is no overlap in the human specific SD with chimps whatsoever

Question 9

What is one major way that duplications have functioned as a mechanism of genome evolution?

Answer 9

There have been several whole-genome duplication (WGD) events in the genome of the vertebrate ancestor

the ancestral chordate had seventeen chromosomes. Subsequent to one WGD resulting in 34 chromosomes, a second WGD plus chromosomal fusions resulted in 54 chromosomes at the base of the vertebrates. Additional chromosome fusions occurred in the lineages leading to euteleostomes (bony vertebrates) and amniotes (including birds and mammals). A third WGD occurred in the teleost lineage.

Question 10

What is the relevance of these WGDs for evolution?

Answer 10

Opens the floodgates for a whole new wave of evolutionary changes

Question 11

What is a much more common duplication event than WGD?

Answer 11

Much more frequent are smaller duplications of subsections of the genome: segmental duplications:
Genomic regions greater than 1 kilobase (but usually 100s of kbp to mega bp), with greater than 90% identical sequence.

Question 12

What is the relevance of these segmental duplications for our genome evolution?

Answer 12

These events have been the origin of many new protein-coding genes during evolution (NOTCH2NL, ARGHAP11B etc arise from this)

Question 13

What event during mitosis can incur duplications or deletions in the genome?

Answer 13

Non-homologous crossover can cause duplications/deletions in the genome: Unequal crossing over leads to a reciprocal duplication and deletion

Question 14

Where in the genome does non-homologous crossover often occur?

Answer 14

Often involved with regions displaying high levels of sequence similarity / repetitive DNA. By mistake lined up and swapped genetic material (often when this happens is around same sequence across people- could lead to disorders). Leads to swap and deletion of large portions of DNA. Therefore they are not randomly distributed; occur in these hotspots- in certain clusters or loci. Some homologous sequence which triggers exchange of genetic material

Question 15

Why is non-homologous crossover relevant to evolution?

Answer 15

An important source of novel genetic material and new genes! (Often leads to negative effects but rarely can lead to something evolutionary beneficial)

There was a ‘burst’ of duplications in the human-great ape lineage; Duplications were highest in the ancestral branch linking humans and great apes

Question 16

How frequent are deletions as opposed to duplications?

Answer 16

3-fold excess of duplications over deletions

Question 17

What are duplications in NHC linked to?

Answer 17

Duplications are linked to segments of DNA known as core duplicons

Question 18

What can these segments of DNA called core duplicons lead to?

Answer 18

Segments of DNA called core duplicons result in the serial accumulation of segmental duplications. Results in increasingly larger duplication blocks (100s of kbp)

Question 19

What pattern is seen in evolutionarily younger duplicated segments?

Answer 19

Evolutionarily younger duplicated segments are located at increasing distances from the core

Question 20

What is the relevance of core duplicons to segmental duplications?

Answer 20

Most human-specific segmental duplications are located around core duplicons

Question 21

Give an example of a human-specific family of genes arising from segmental duplication

Answer 21

The morpheus gene family, also named the nuclear pore inter-acting protein (NPIP) family, is one of the best studied human core duplicon gene families. It originated from one gene present on chromosome 20 in macaques and expanded in the great ape–human (NPIP) gene family on lineage along chromosome 16 through segmental duplications. It can be subdivided into two distinct subfamilies, NPIPA and NPIPB, which mostly differ with respect to exon 5 and the structure of amino acid repeats in the C-terminus

Question 22

How is the NPIP gene relevant to evolution?

Answer 22

It has not been possible to identify paralogs outside of primates, i.e. it appears tobe a newly evolved or rapidly evolving gene. In fact, the Morpheus gene family was shown to be one of the most rapidly evolving gene families during hominoid evolution

Question 23

Describe how segmental duplications can be evolutionarily beneficial

Answer 23

Segmental duplications can include genes or parts of genes. There are many examples in which a complete or partial gene duplication has created an entirely new human-specific gene over the course of evolution.

Genes arising from duplication events are not subject to strong selection, so can accumulate mutations at a greater pace. Segmental duplications also accelerate evolution by providing homologous sequences that encourage further rounds of duplication. Many genes located in highly duplicated regions serve critical neurodevelopmental functions.

Question 24

How may an entirely new gene arise from a duplication? What else could happen regarding the gene following a duplication? (4)

Answer 24

Mutations could accumulate on the duplicated copy so that it takes on a novel function OR degenerates and becomes a pseudogene OR the duplicated and parental copy undergo subfunctionalisation.

If only part of the gene is duplicated, it might already have a different function to its parental gene.

If it is placed in a new genetic context, it may be expressed in different spatial or temporal context.

It could be adjacent to other genes so that fusion transcripts are formed from transcriptional readthrough (Might be placed upstream of an existing gene, you can get something where the first gene starts to be transcribed and you get a fusion transcription)

Question 25

If a segmental duplication is not evolutionarily beneficial, what else can arise from it?

Answer 25

Segmental duplications can lead to serious disorders: Highly repetitive and unstable regions are much more prone to deletions and duplications. Unequal crossover events can lead to the loss of genes that have acquired a critical new function.

Question 26

Give seven examples of disorders arising from segmental duplications from one family of disorders

Answer 26

Copy number variations in multiple human-specific genes have been linked to neurodevelopmental disorders:
* Developmental delay
* Autism
* Schizophrenia
* Epilepsy
* Spinal muscular atrophy
* Microcephaly
* Macrocephaly
* Others…

Question 27

Describe the figure shown in the slides depicting how evolutionary innovation can lead to genomic instability

Answer 27

Increasing genomic complexity can lead to segmental duplications which can further this complexity and later give rise to new genes and functionality. This can also give rise to genetic deletions and duplications which can then give rise to neurological disorders.

Question 28

Why do we stand to benefit from knowing the evolutionary history of these duplicated regions?

Answer 28

The human genome has many regions that have undergone a high amount of segmental duplications or other genomic rearrangements over evolution. These regions are also prone to continued rearrangements and these can cause genetic disease
Knowing the evolutionary history of these regions can help us understand what goes wrong in these diseases

Question 29

How common are fusion transcripts?

Answer 29

While structural rearrangements can place whole genes or parts of genes in new contexts so new “fusion transcripts” arise, “fusions” between two adjacent genes are very common

Basically all pairs of adjacent genes show some level of transcriptional readthrough, but usually at very low levels (~100 times less than the two full-length genes). But in some cases, such fusion genes could become evolutionarily beneficial and take on a novel function

Question 30

Where has there been a lot of work done on fusion transcripts (field)

Answer 30

There has been much focus on the role of fusion transcripts in cancer, but they are also abundant in normal tissues (e.g. brain).

Question 31

In a karyoplot showing fusion transcripts, what do long loops typically depict?

Answer 31

(The long-range loops shown on the karyoplot likely come from DNA sequences that have undergone structural rearrangements, and their new location is not properly marked in the reference genome.)

Question 32

In what two ways can fusion transcripts vary?

Answer 32

Fusion transcripts vary in expression level and frequency in the population

Some are widespread, others are found in very few individuals

Question 33

What clinical relevance do fusion transcripts then have?

Answer 33

Could differences in fusion transcript expression underlie disease susceptibility in some cases?

Question 34

What is the relevance of segmental duplications to fusion transcripts?

Answer 34

There is a tendency for fusion transcripts to cluster in regions enriched with segmental duplications, although they are also found in non-duplicated regions.

Question 35

What region of the genome was selected for the lecture due to its richness in segmental duplications?

Answer 35

17q21.31 is a region enriched with segmental duplications

Question 36

What is significant about the variation of this region in the population? (what was originally thought)?

Answer 36

The region can occur in direct orientation (H1) or inverted orientation (H2) in the human population (25%). Contains multiple genes implicated in neuronal functioning including CRHR1, MAPT and KANSL1.

E.g H1:
CRHR (=>) MAPT(=>) KANSL1
KANSL1 (=>) CRHR (<=) MAPT(<=)

Question 37

Why is this H1 and H2 description of 17q21.31 not completely accurate?

Answer 37

There are 8 structural haplotypes (sets of DNA variations, or polymorphisms, that tend to be inherited together) within this region. This reflects that there is a lot more variation hidden when considering one reference human genome.

Question 38

Describe the 8 haplotypes present in 17q21.31

Answer 38

Region containing CRHR1 and MAPT: [g]
Region containing KANSL1: [y]
Region adjacent to that containing KANSL1 in original models: [b]

H1’: Three haplotypes:
[g >] [< y] [b] [b]
[g >] [< y] [b]
[g >] [< y] [b] [b] [b]

H1D: Two haplotypes:
[g >] [< y] [y] [b] [b]
[g >] [< y] [y] [b] [b] [b]

H2’: Two haplotypes:
[y >] [< g] [b]
[b] [y >] [< g] [b]

H2D: One haplotype:
[b] [y >] [< g] [y] [b]

Question 39

What syndrome is associated with 17q21.31?

Answer 39

17q21.31 microdeletion syndrome (Koolen de Vries syndrome):
* Developmental delay, intellectual disability
* Cheerful, social disposition
* Distinctive facial features
* Epilepsy
* Hypotonia
* Cardiac, kidney and skeletal abnormalities

Question 40

Where lied the difficulty in detecting segmental duplications in people?

Answer 40

When sequencing for the reference genome, fragments of DNA ~300,000 bp were Sanger sequenced and assembled together (“contigs”). The consensus sequence was taken and transposable elements or SNPs not alligned with the consensus were not integrated into the reference genome.

The method of assembly does not allow for correct mapping of loci containing recent duplications. The reference genome is used, however, when examining the genome and sequencing a given persons DNA; take DNA and map to reference genome.

Question 41

Describe how we can detect variation in segmental duplications between people (but not orientation)

Answer 41

Whole genome sequencing can be used and coverage/ density plots can show the number of reads that were mapped in that location in the reference genome. Regions with increased / reduced coverage indicate duplications / deletions

In areas of duplication you should see a rise in the density plot. This is meant to represent how many sequences have mapped to that region of DNA, can see how many copies someone has in a given location. Doesn’t tell us where they are on the genome. Analysing coverage from sequencing data therefore tells us about copy number, but not about orientation and arrangement of segmental duplications

Question 42

How can copy number and orientation be detected?

Answer 42

Copy number and orientation can be detected with fluorescence in situ hybridisation (FISH). FISH uses fluorescent DNA probes to target specific chromosomal locations within the nucleus, resulting in colored signals that can be detected using a fluorescent microscope.

Question 43

How was the evolutionary history of the 17q21.31 locus reconstructed and what was found? (3)

Answer 43

Aligning sequences between different haplotypes allows us to estimate the time to most recent common ancestor and reconstruct the evolutionary trajectory of the region.

Then H2 (“inverted”) orientation is the ancestral haplotype. The inversion occurred independently in humans and chimpanzees. This regions is prone to recurrent inversions.

H2 only: New world & old world monkeys

H1/H2 polymorphic: Orangutang, Gorillas, chimpanzees, humans

Question 44

How are there patterns in the evolutionary history of the 17q21.31 locus within humans?

Answer 44

From our H2 ancestor in western Africa, there was a H1-H2 diverge 2.3 million years ago which migrated south in Africa. H1’ carriers also migrated to east Africa.

There was also an African-European H2 diverge with the out-of-africa migration. Included in this was a H2’-H2D diverge 1.3 million years ago and a H1’-H1D 250K years ago. H1’ carriers migrated to asia and and the other migrated to europe ????

Question 45

Has there been selection of particular 17q21.31 haplotypes during evolution?

Answer 45

When comparing genetic sequences across the locus, we see that H2D individuals are extremely similar to each other (extremely low genetic diversity). This is suggestive of a recent bottleneck followed by a population expansion or selective sweep.

Question 46

Describe specifically what changes in 17q21.31are attributed to Koolen de Vries syndrome and what is observed in neurons of patients with this disease

Answer 46

Microdeletion syndrome is caused by loss of the gene KANSL1. Most of the symptoms of Koolen de Vries syndrome have been attributed to loss of KANSL1. KANSL1 is part of an epigenetic modifying complex, which influences gene expression via H4K16 acetylation. Koolen de Vries syndrome neurons show synaptic defects due to oxidative stress and proliferation defects.

Question 47

Is this region of the genome linked to any other diseases apart from 17q21.31 microdeletion/ Koolen de Vries syndrome? Explain

Answer 47

17q21.31 is also linked to risk of neurodegenerative diseases: Genome-wide association studies (GWAS) look at single nucleotide polymorphisms (SNPs) throughout the genome, and assess which of these are statistically linked to disease. There are certain SNPs found more often on H1 haplotypes, and others found more often on H2.

H2-associated SNPs have been linked to:
o Reduced incidence of Parkinson’s disease, Alzheimer’s disease, and progressive supranuclear palsy
o Increased fecundity (reproductive potential)
o Larger intracranial volume

Thus it could be said that a forward orientation has an increased risk compared to reverse orientation.

Question 48

What is more likely than these being SNPs according to Colette?

Answer 48

If some has a forward and someone has a backward they don’t recombine; suggesting they have been evolving separately.

The more likely underlying factor than individual SNPs are that these SNPs, once inherited, combined with something else in the locus.

Question 49

How could this alternative manner of variation in 17q21.31 have come about?

Answer 49

It could have been due to segmental duplication:

Most people with the protective H2 haplotype are H2D, i.e. have an extra segmental duplication (giving a partial KANSL1 duplication) not found in the reference genome. Perhaps this duplication contains some genetic material that affects neurodegenerative disease risk.

Question 50

Why don’t people with H1 have the same risk then, since H1D also has a partial KANSL1 duplication arising from segmental duplication?

Answer 50

The proportion of H1 people that are
H1D is much lower, which could explain why the H1 group has higher risk overall

30% of H1 individuals have a partial KANSL1 duplication (H1D) while over 95% of H2 individuals have a partial KANSL1 duplication (H2D).

SNPs in H1 are associated with increased risk while SNPs in H2 individuals are associated with increased risk. Maybe something in this duplication carries something to do with neurodegenerative diseases

Question 51

What differences are there in the transcription of H1D and H2D? (2)

Answer 51

1. There is one full length transcription of the KANSL1 gene in all haplotypes. The partial duplication in H1 covers the KANSL1 exon 1, 2 and 3 before being truncated while H2D only covers exon 1 and 2.

No duplication:
[g >] [< y] [b] [b]

H1D:
[g >] [< y] [y] [b] [b]

H2D:
[b] [y >] [< g] [y] [b]

2. KANSL1 genetic duplications also create different fusion transcripts in H1D and H2D

Question 52

Describe how the KANSL1 genetic duplications create fusion transcripts

Answer 52

They looked in RNA sequencing data for all sequencing reads containing 3’ end of KANSL1 exon 3 and found that KANSL1 exon 3 fused to the ARL17A/B exon but out of frame.

They also looked in RNA sequencing data for all sequencing reads containing 3’ end of KANSL1 exon 2 and found that exon 2 fused to the KANSL1 alternative exon but out of frame BUT it also fused to the LRRC37A3 in frame.

They found that a small proportion of H1’-H1’ individuals could possess any of these fusions.

All H1D-H1D individuals possessed the exon 3 ARL17A/B fusion and none of the others.

H2D-H2D individuals all possesed the alternate out of frame fusion and a large proportion possessed the in frame LRRC37A3 fusion and none possessed the ARL17A/B fusion.

Therefore the KANSL1 fusion transcripts result from the fusion of KANSL1 exon 2 or 3 with exons of other nearby genes, or novel exons and they are strongly linked to specific 17q21.31 haplotypes (H1D and H2D).

Question 53

How strongly expressed are these fusion transcripts?

Answer 53

These fusion transcripts are highly expressed, sometimes as highly as the normal full-length KANSL1 mRNA (ARL17A/B (out-of- frame) in H1D-H1D individuals). Buttt the OOF novel exon was 50% of expression and LRRC37A3 was around 10% of expression in H2D-H2D individuals.

The potential consequences of these transcripts for protein expression depend on whether the fusion is in-frame or out-of-frame.

Question 54

How were these findings regarding the KANSL1 fusion transcripts validated?

Answer 54

These KANSL1 fusion transcripts can also be validated by PCR amplification from RNA extracted from people with the segmental duplication. This was carried out via reverse Transcriptase (RT)-PCR (below) followed by Sanger sequencing.

Question 55

How does the genomic assembly of H2D lend insight to these findings?

Answer 55

Long-read sequencing shows us the full exon structure of the fusion transcripts. In the H2D alternate genome assembly, KANSL1 is upstream of LRRC37A3, segmental variation has placed part of KANSL1 upstream of this other gene, and transcriptional readthrough generates the fusion transcript

Question 56

How would we explore whether these KANSL1 fusion transcripts have a function?

Answer 56

The next step is expressing these fusion transcripts in neurons and/or organoids, to establish their potential function. From there we can do Rna-seq, ChIP-seq, electrophysiology etc

Question 57

Relate the earlier figure regarding increased genomic complexity to these new findings

Answer 57

Increasing genomic complexity can lead to segmental duplications such as those which occurred repeatedly in the 17q21.31 locus; originating from a core duplicon.

This can further this complexity and later give rise to new genes and functionality: Partial KANSL1 duplication and fusion transcripts
* Protection against neurodegenerative diseases?
* Another role in neurodevelopment?

This can also give rise to genetic deletions:
Due to the high level of sequence similarity in nearby segmental duplications

This can also give rise to duplications:
17q21.31 microdeletion / Koolen de Vries syndrome