Based on the article, what are the known causes of SRS and BWS? Which of these causes are genetic, and which are epigenetic?
- BWS: Caused by a 160kb inverted duplication (genetic); including ICR2 and the most 5’ 20 kb of KCNQ1OT1, gain of methylation at ICR1 site (epigenetic), loss-of-function defect of the trans-acting factor (genetic).
- SRS: Caused by a 1.2Mbase inverted duplication of the whole 11p15.5 imprinted gene cluster (genetic). This can result in greater methylation.
Consult Figure 8 to remind yourself of what/where ICR2 is, then consider the data shown in Figure 5B. What do they show, and what do they tell us about the methylation state of the ICR2 region in individuals I-4, II-4 and III-6?
Figure shows methylation of ICR2 regions on both paternal and maternal chromosomes in different individuals. Patients with a duplicated paternal region in their maternal chromosome have one copy unmethylated and the other copy methylated with occasional unmethylated dinucleotides (which resembles the control – for the methylated copy). I-4 (normal phenotype) – duplicated paternal region but inside the paternal chromosome – same methylation pattern as above (difference is the copy on the paternal chromosome). Impaired imprinting results from having the unmethylated copy incorporated into the maternal chromosome.
Notice how I-4, II-4 and III-6 all have the same number and methylation pattern of ICR2 ‘loci’. How can their difference in terms of having vs. not having BWS be explained?
On the maternal chromosome, ICR2 is methylated, KCNQ1OT1 is not transcribed, and the flanking imprinted genes are expressed. In the BWS patients, the 160kb duplication results in a copy of ICR2 that is unmethylated, that makes the maternal gene hypomethylated. I-4 does not have BWS because the paternal ICR2 is always unmethylated, so KCNQ1OT1 is always transcribed and flanking genes are silenced.
Explain what Figure 7B shows and how you interpret the data.
ChIP and qPCR were performed to detect the KCNQ1OT1 transcript and its interaction with chromatin at both chromosomes. Primers were designed to specifically determine how much of the transcript was present from each of the maternal and paternal chromosomes. The figure shows that all tested individuals (BWS patients and a control) had similar levels of KCNQ1OT1 transcript interacting with the paternal chromosome (error bars overlap). However, BWS patients had much higher levels of the KCNQ1OT1 transcript interacting with the maternal chromosome.
One of the authors’ hypothesis is that many of the physical phenotypes associates with the BSW patients are due to reduced expression of CDKN1C. Propose two possible mechanisms that would explain how the duplication of ICR2 in these patients causes a reduction in the expression of CDKN1C . Based on what you know about Airn, Igf2r, and slc22a3, which of the two hypotheses is most likely and why?
ICR2 is a long non-coding gene that overlaps with the promoter for KCNQ1OT1. This gene is found in a protein coding gene, KCNQ1, at an antisense orientation. KCNQ1OT1 acts to silence centromeric domain genes such as CDKN1C, a cell growth inhibitor. ICR2 is normally methylated on the maternal chromosome, KCNQ1OT1 is not transcribed and the flanking genes are expressed.
The researchers observed hypomethylation in the maternal, duplicated ICR2 region in BWS patients. The duplicated ICR2 may not have been sufficient for establishing proper imprinting. The KCNQ1OT1 is not silenced and it acts to suppress CDKN1C expression.
In BWS patients, the 160 kb duplication included most of the 5’ region (20 kb) of KCNQ1OT1, as well as the ICR2. It is possible that this 20 kb region could display a gain of function mutation, by allowing KCNQ1OT1 to bind an enhancer with higher than normal affinity. It is also possible that a the inverted region could cause KCNQ1OT1 to overcome the actions of of an inhibitor and continue to silence CDKN1C in its presence.
Like KCNQ1 and KCNQ1OT1, Airn and Igf2 are also overlapping genes with antisense orientation. Airn is a long non-coding RNA like KCNQ1OT1 and they both act to silence flanking genes. When Airn is deleted and Igf2 is intact, Igf2 is expressed because it is no longer being silenced. Slc22a3 would also be expressed. The first hypothesis is more likely because it is well documented that ICR2 is hypomethylated in BWS patients so it would make sense that the correct imprinting is not established, leading to errant expression of KCNQ1OT1.
After reading this paper, how do you think clinical papers describing just a few patients can contribute to our understanding of the regulation of developmentally relevant genes?
This paper highlighted that many types of mutations can have an effect on a disease phenotype. Researchers found in cis mutations in 20% of the BWS patients with gain of methylation at ICR1, but also a loss-of-function defect of a trans-acting factor in a familial case with multiple ICR hypomethylation (12 – 16). This finding demonstrates that both genetic and epigenetic factors can contribute to disease and that the cause of disease may vary between cases. In addition, they saw that mutations of CDKN1C accounted for 5% of the BWS cases, suggesting that multiple different genes can be involved in the development of the disease as well. With this in mind, I think that a few patients can still contribute greatly to our understanding of the regulation of developmentally relevant genes since each individual may have a different molecule or factor that is defective and causing their disease phenotype. This way, every involved gene or component that we discover brings us closer to understanding the pathways that regulate development.
This paper required a lot of data interpretation in order to gain a complete understanding of the researchers’ findings. Therefore, I would say that the most significant gain I obtained from this paper would be in my data interpretation skills. What researchers discovered was a cluster of imprinted genes on one chromosome that were associated with Silver-Russell syndrome (SRS) and Beckwith-Wiedemann syndrome (BWS). The cluster of genes was proved to be divided into two domains with independent imprinting control regions (ICRs). In this paper, Chiesa et al. highlight two maternal microduplications. The first is an inverted and in cis duplication of a gene cluster resulting normal genomic imprinting. The second is a 160 kb duplication also inverted and in cis, but resulting in imprinting of a different chromosomal region. The first inversion results in the SRS phenotype, while the seconds leads to BWS. Chiesa’s team support the hypothesis that the ICR2 sequences they researched in these diseases are not sufficient for establishing imprinted methylation and that perhaps some other property which may be orientation-dependent, is needed instead.