Family affair
Associate Professor Ruth Arkell* is revealing the mysteries of the ZIC family of genes, which have been coordinating the development of the nervous system in vertebrate embryos for hundreds of millions of years.
The ZIC family of genes has been a team through time and tide since Adam was a teleost fish around half a billion years ago. Their role throughout this stretch of time has been to collectively shape the nervous system during very early embryonic development. The question is: have the five genes that code for ZIC proteins been conserved over this enormous time span because they are functionally interdependent? And if so, how do they interact?
Australian National University molecular embryologist Dr Ruth Arkell spoke about her work investigating this mystery at the 13th annual Hunter Meeting in March. According to Arkell, modern fish have seven zic genes, two more than their land-going descendants. That’s unusual, given that the modern vertebrate genome is thought to have arisen when the entire genome of a direct fish ancestor of vertebrates was doubled, then redoubled, before the first tetrapod vertebrate made landfall some 400 million years ago.
In humans, the genes are spread around. We have two tandem pairs of ZICs on chromosomes 3 and 13, while the unpaired gene is on the X-chromosome. But that’s not necessarily odd; in some strains of Drosophila even the ancient Hox body-segmentation genes, which are conserved in four co-linear clusters on different chromosomes of species as divergent as insects and mammals, have been separated and scattered across the fly genome with no apparent effect on embryonic development.
“The highly conserved, co-linear linkage of the Hox genes was thought to maintain tight control over the timing and spatial pattern of expression,” says Arkell. “But apparently, it doesn’t have to be like that.”
However, while mouse Zic genes exhibit overlapping patterns of expression that recapitulate their genomic arrangement, Arkell has wondered for a long time whether there’s more to their durable linkage than conserved expression. Her team’s research indicates they may indeed be linked by function.
They have worked on three of the mammalian ZIC genes - the ones expressed at a critical stage of embryogenesis called gastrulation - the paired ZIC2 and ZIC5 genes, and the unpaired ZIC3. These are all of interest seeing as they are closely linked to a range of developmental disorders.
ZIC5 is of interest because of its association with ventricular defects in the brain and hydrocephalus. ZIC2 defects are associated with a rare and usually lethal congenital disorder called holoprosencephaly, in which the embryonic forebrain remains undivided instead of forming two hemispheres. At an extreme, holoprosencephaly results in cyclopia, an unsettling congenital defect where the face is absent and the skull develops a single orbital cavity on the midline of the skull.
“In humans the frequency of holoprosencephaly at birth is around one in 10,000, and only one in 16,000 babies with the condition survives the immediate post-natal period,” she says. “But while holoprosencephaly is very rare at birth, the frequency during early embryogenesis is very high: about one in 250.”
ZIC3 defects are associated with another congenital disorder called heterotaxy, characterised by disordered development of usually asymmetric body organs such as the heart, the lungs and the kidneys. These disease associations are likely only just the tip of the iceberg when it comes to defining the role of the ZIC proteins in embryonic development and adult homeostasis.
Cumulative effects
According to Arkell, one of the things we tend to forget is that the effects of genetic defects tend to be cumulative over time. “A subtle gene defect may at birth be of little consequence, causing only an unrecognised weakness or susceptibility in some aspect of cell function. Then feedback mechanisms may amplify the effect through life, resulting in the onset of serious disease later in life.
“One of the most exciting aspects of the new genomics capabilities is that we will be able to link mutations in genes that give rise to subtle functional differences to diseases. Already the first inklings of ZIC involvement in adult onset diseases are beginning to emerge.
“The Hunter Meeting is about cell biology, not embryology, so I talked about the way in which ZIC proteins appear to be able to switch from their roles in transcription, to become co-factors in gene suppression.”
Arkell’s talk particularly focused on her team’s investigation of the mechanisms behind ZIC’s Janus faces. “Post-transcriptional modification of the proteins seems to be involved, and we think the modification involves a modification called SUMOylation,” she says.
SUMO refers to small ubiquitin-like modifier. Small molecules of ubiquitin ‘tag’ lysine residues of other proteins, directing them to a variety of cellular fates, including the proteasome, where they are recycled. Small SUMO molecules - those around 10 kilodaltons - modify the activity of proteins by targeting the same lysine residues as ubiquitin. SUMO, also like ubiquitin, forms chains by ‘tagging’ itself.
Ubiquitin, as its name suggests, is ubiquitous in cells, and its multiple functions are now reasonably well understood. But SUMO’s workings are a long way from being understood. “The activity of every transcription factor is likely to involve SUMOylation in some way,” she says. “But SUMOylation also seems to be associated with almost every type of cellular change one could imagine.
“We’ve been working primarily on SUMOylation of ZIC5, because we have identified a mouse strain in which the alteration of one amino acid produces a phenotype that resembles complete loss of function of ZIC5 in human hydrocephalus.”
The mutation, at the third position a wild-type lysine codon, changes it to an arginine. “It’s is about as conservative a change as you can get in a protein, because lysine and arginine have the same size and charge, and you would think it would be a silent mutation. Yet, in the whole mouse genome, this conservative change in just one amino acid, it partially abolishes ZIC5 function.”
The change is catastrophic for the orderly development of the cranial neural tube, which forms the central nervous system, and the neural crest, which forms the peripheral nervous system.
SUMOylation error
Arkell believes that post-translational SUMOylation of the wild-type lysine residue at that position is essential for the protein to perform its normal function as a transcription activator. The substituted arginine residue prevents SUMOylation at that position, and the protein becomes an inhibitor, repressing downstream gene networks that ZIC5 would normally activate. The resulting, partial abolition of ZIC5 function disrupts development of the central and peripheral nervous systems.
Could SUMOylation and the apparent alteration it brings in ZIC protein function be part of the reason the ZICs have stuck together? “We have evidence from lots of mutational studies in mice that the ZIC proteins somehow regulate the expression levels of their linked partners,” says Arkell.
“In this scenario, we usually think of the proteins as being able to cross-regulate each other’s transcription. But we wonder if, in the case of the ZICs, the genetics can be explained by the proteins cross-regulating each other - rather than their level of expression.
“One way that a protein can regulate a related protein’s function is by dominant-negative interference, which involves the two protein forms competing with each other. For example, many genes produce transcripts that are alternately spliced to produce proteins of different size. In some cases a shorter, or truncated, version of the protein can compete with and decrease the function of the full-length protein.”
Arkell suggests that, in the case of the ZIC proteins, the non-SUMOylated form of the protein may act as a dominant negative, preventing the SUMOylated form activating gene expression. Co-expression of the linked proteins, along with differential SUMOylation, may allow the two proteins forms to regulate each other’s activity.
***************************************************
Zic's role in cancer
According to Arkell, the functional redundancy of the ZIC genes has hampered studies of the networks they regulate: knock out one ZIC and the others tend to compensate, in varying degrees, for its absence. However, loss-of-function mutants in mice exhibit a range of phenotypes that hint at their involvement of ZIC genes in multiple signalling pathways - including cancers in which Wnt signalling defects are known to be involved.
Wnt signalling defects have also been implicated in a variety of cancers. In a paper in the International Journal of Biochemistry and Cell Biology last year, Arkell and ANU colleagues Radiya Ali and Helen Bellchambers suggested that in addition to causing a range of congenital disorders in mice and humans, the involvement of ZIC genes as factors or co-factors in maintaining pluripotency in stem cells hints at a possible a role in cancer.
They note that dysregulated Wnt signalling is associated with a growing range of cancers. In particular, mutations that affect Wnt signalling are associated with more than 90% of colorectal carcinomas. The ANU researchers suggested that an understanding of the mechanisms of ZIC-mediated Wnt inhibition may lead to new ways of preventing the uncontrolled proliferation of cancerous cells.
***************************************************
AI-designed DNA switches flip genes on and off
The work creates the opportunity to turn the expression of a gene up or down in just one tissue...
Drug delays tumour growth in models of children's liver cancer
A new drug has been shown to delay the growth of tumours and improve survival in hepatoblastoma,...
Ancient DNA rewrites the stories of those preserved at Pompeii
Researchers have used ancient DNA to challenge long-held assumptions about the inhabitants of...
