Over the last two years, a small team in Sheffield have been working hard, trying to generate an EHE zebrafish model. A zebrafish model would be a wonderful addition to EHE research as these fish are cheaper and far quicker to produce, and therefore far lower cost, than mice. However, the project has proved difficult with TAZ-CAMTA1 expression not presenting in endothelial cells. This was surprising as a parallel construct using GFP (green florescent protein) had been successfully introduced.
Dr Fredericus van Eeden, the PI for the project, believes that a number of possible causes for this failure to achieve TAZ-CAMTA1 expression can be hypothesised. There could, for example, be sequence elements in the construct that lead to repression of expression. Codon optimisation was used to try and address this issue but does not change the sequence completely and although the team used two different approaches, it may not have “hit” the relevant bases. If the mechanism of this repression can be identified and understood, the team also hope that it might perhaps be therapeutically exploited in EHE tumours in humans. A detailed understanding of this mechanism is therefore warranted. In addition, understanding of the precise reasons may allow us to relieve this repression and gain epithelial expression.
The first approach will be to examine if there are sequence elements in the construct that prevent expression under control of the fli1a promoter that has been used. The team will systematically add in parts of the TAZ-CAMTA1 coding sequence, and use these to create a set of destination vectors. These will be used to evaluate whether they allow expression in endothelial regions. As a potentially faster assay/alternative the team plan to use these constructs to transfect coronary artery endothelial cells or HUVECs. In these endothelial cells, expression of GFP can again be used as a readout of expression levels. The researchers have collaborators in Sheffield who have a lot of experience with these cells and so have cell culture facilities available. For example, the team will clone either the TAZ or CAMTA1 part of the oncogene behind GFP and see which of the two may block expression. Once they have found which sequence is responsible for a block, they will further subdivide the sequence to home in on potential sequence elements that prevent expression. If a small inhibitory element can be identified, they will collaborate with their bioinformatics core to identify potential binding of proteins or miRNAs that might be responsible for the suppression of the expression.
The second approach has arisen following dialogue with Dr Valerie Kouskoff in Manchester who is also working on EHE. Importantly, Dr Kouskoff found that in embryonic stem cells the TAZ-CAMTA1 expression can lead to downregulation of endothelial markers including fli1 expression (the mammalian equivalent of fli1a). This could therefore explain the negative results as well. In this case, switching to an alternative endothelial promoter (e.g. flt1 or VE-cadherin) is unlikely to solve the problem. To best model the disease, the same promoter should be used as is actively driving TAZ-CAMTA1 in the human condition. It is in all likelihood the TAZ promoter that drives expression. The TAZ promoter has been somewhat characterised in humans and the team have requested this promoter from Dr Kapus. In addition, the team will clone a similar element from the zebrafish and test both of these for driving expression in zebrafish and endothelial cells using GFP initially as a reporter.
Another issue with the use of a TAZ driver is that this will still lead to expression in regions outside the endothelium, as TAZ is rather widely expressed in zebrafish. A parallel and perhaps most important approach will therefore be to combine an appropriate promoter and STOP cassette to the TAZ-CAMTA1 coding sequence, to prevent unwanted expression. The STOP cassette can then be removed, specifically in endothelial cells, and drive expression from this generic promoter at a chosen time, by activation using tamoxifen.
The third element of this year’s work will address the issue faced with previous assays where the team only established that lines were not behaving as expected after creating a line, which takes a significant amount of time (3-4 months). Normally this is fine, because gene expression is not an issue when using well-established promoters like fli1a. However, it is clear from the team’s previous attempts that TAZ-CAMTA1 requires more checks. Therefore, they will also do in vitro experiments before progressing with transgenesis in fish. They will establish a cell-culture system to evaluate the constructs they produce in vitro, in endothelial cells. The plan is to transfect human endothelial cells (arterial and venous) with the relevant constructs, and assess mRNA and protein expression levels. In addition, they will also evaluate some of the previously generated controls to check for expression of the relevant proteins. The team will also evaluate endogenous human fli expression with collaborators in Sheffield who have a wealth of experience with human endothelial cells in vitro, and who have cell culture facilities available within the research department. In addition, they will collaborate further with Dr Kouskoff, and will test some crucial constructs in mouse embryonic stem cells, as Dr Kouskoff has extensive experience with expressing this protein and studying its effects on endothelial differentiation.
The final and fourth element of the year’s work will be to take the learnings of the above assays, and introduce these to the zebrafish, assuming of course that the results are positive. Only when the team are convinced from their in vivo and in vitro experiments that the construct is working as expected, will they progress and raise transgenic lines. If successful, the team hope that with these lines they will be able to: (i) identify the correct expression pattern; (ii) perform a preliminary phenotypic analysis; (iii) verify that the lines activate the TAZ signalling pathway; and (iv) determine the expression levels of TAZ target genes. If successful they will then establish lines from the best performing transmitters, for further analysis.
We thank Dr Van Eeden and his colleagues for their dedicated and detailed analysis to date, and of course wish them every success for the coming year’s work. We are also delighted to see that this project is benefitting from wide collaboration, including with Dr Kouskoff in Manchester and Dr Rubin in Ohio, two other researchers whose EHE research is being funded by the EHE group.