#### Demonstration of full spectrum animal model from single mutations is optimistic for gene therapy
Three different Midwestern NF centers with mini swine facilities ([Iowa](https://www.ncbi.nlm.nih.gov/pubmed/29925695), [Minnesota](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6168575/) and [Wisconsin](https://news.wisc.edu/nf1/)) have claimed full-spectrum reproduction of NF1 complications in miniswine, with a single gene mutation. To the extent that a single mutation can reliably produce an NF1 complication, gene therapy becomes attractive.
Using mini swine can rapidly accelerate the testing cycle for either drug or gene therapy. We might also be able to speed up the testing cycle by another order of magnitude by, for example, using NF1 mini swine [organoids ](https://www.drugtargetreview.com/article/48244/what-are-the-pros-and-cons-of-using-organoids/)(or just human tumor organoids).
Recently conditions for gene therapy have become more promising due to [CRISPR](https://www.sciencemag.org/news/2019/08/forget-single-genes-crispr-now-cuts-and-splices-whole-chromosomes).
#### The only known attempted NF1 gene expression therapy failed due to incorrect hypothesis about NF1 +/- availability in tumor microenvironment
I only know of one attempted gene expression therapy for NF1, by David Segal at UCSD. He [attempted an approach, with a negative outcome, 5 years ago](https://apps.dtic.mil/dtic/tr/fulltext/u2/a609751.pdf). In email communication to me, Prof. Segal described the reasons for the negative outcome as follows:
> Our approach tried to take advantage of the natural biology of TALE-type artificial transcription factors. In nature, TALE proteins are already gene activator proteins, and they are produced in bacteria which transmit the TALE proteins to their host. The plan was to engineer TALEs that could turn up the expression of the good copy of NF1 (most people with NF1 inherit one good copy and one mutant copy). Bacteria producing the engineered TALEs would then be injected into the fibromas, where they would remain to pump out the TALE until the tumor went away. They they would be exposed to, and cleared by, the host immune system.
> The first problem was that the natural bacteria of the TALEs does not grown in humans, so we had to use a different strain of bacteria. That already raised concerns by some in our lab safety.
> The second problem was the realization that what actually happens in the fibromas is that the good copy of NF1 is lost, leaving only the mutant copy. That means there was no good copy to increase the expression of. So our plan then switched to targeting the therapy to the cells in the microenvironment of the fibroma. There was what appeared to be pretty convincing evidence in the literature (mostly Science and Cell papers) that the NF1 expression in the microenvironment was absolutely critical for the fibroma to form and persist. That was what we wrote into the grant, and the grant was reviewed and approved. However, as we talked with more and more scientists, they started to cast doubt about the necessity of NF1+/- status in the microenvironment, now that there have been a lot more studies.
> So yes we had a partial success with the methodology, and yes it would in principle work with any type of NF1 mutation.
> However, I think the approach we were taking had no future.
#### A recent startup focuses on enhancing normal tissue expression of neurofibromin 1
One startup, http://www.infixionbio.com/our-approach/ focussess on enhancing the ability of normal tissue to produce neurofibromin 1. To summarize their website:
* NF1 mutations occur on one of 2 alleles in the NF1 gene (autosomal dominant), so there is always a health copy present. That is, there are no observed individuals who are NF1 -/- in non-tumor tissue.
* It has been shown that the lack of a healthy amount of wild-type neurofibromin 1, not the presence of mutated neurofibromin 1, causes the complications seen in NF1
* The approach is to amplify the healthy allele: "Utilizing cutting-edge tools and techniques in gene regulation, gene transcription, protein life-cycle management, pluripotent stem cells, and high-throughput screening, the Infixion Bioscience research focus is to develop a safe and mild drug that enhances the NF1-affected person?s natural ability to both generate and conserve normal Neurofibromin protein produced by the unmutated, or wild-type, allele. This increase in the amount of Neurofibromin has the potential to mitigate many, if not ALL, downstream NF1 symptoms that result from [haploinsuffiency](http://jem.rupress.org/content/191/1/181) of Neurofibromin. Similar to insulin treatments for diabetes, a drug that increases NF1 protein on an ongoing basis has the potential to deliver a functional cure for NF1."
So, their approach is to stimulate additional production of neufibromin 1 by the wild-type allele which, they say, is always present. However, note that Dr. Segal's experience was that the tumor microevironment is NF1-/-, which may limit the effectiveness of this approach, in case the normal tissue expressed neurofibromin 1 cannot find it's way into tumor cells. (Which may be by way of definition of tumor cells.)
#### The case for gene mimetic therapy
The wild-type [neurofibromin 1 protein](https://en.wikipedia.org/wiki/Neurofibromin_1) is a GTPase-activating protein that negatively regulates RAS/MAPK pathway activity by accelerating the hydrolysis of Ras-bound GTP.
As a Hackathon deliverable, can we produce a summary of the receptor sites and signalling pathways acted on by neurofibromin 1?
Then we have two mimetic approaches:
* Identify chemical structures of known drugs which can bind to those receptor sites and act similarly to accelerate the hydrolysis of Ras-bound GTP
* Synthesize neurofibromin 1 directly by way of some organic chemical reaction, or by cloning neurofibromin 1 by some standard protein replication process
Then we need to assess if either a chemical analog or synthesized neurofibromin 1 is deliverable into cells by some method (injection, ingestion, topical cream), and once delivered will it act the same way as wild-type neurofibromin 1 produced by the body. Direct delivery to tumor cells may be necessary to overcome the naturally NF-/- status of the tumor microenvironment, as the presence of a tumor may be defined by the tumor microenvironment's ability to maintain NF-/- status.
#### How does the tumor maintain NF-/- status?
What are the mechanisms by which a tumor excludes NF+/- cells in favor of NF-/- cells? Can this mechanism be defeated? If so, would defeating NF+/- rejection be a therapy, or is it a necessary antecedent to delivery of neurofibromin 1 or a mimetic?
Created by Lars Ericson lars.ericson That's a good idea, we can search for genes which are Ras-GTP hydrolysis activators which are not mutated in the patient samples, and increase their expression. We can search for this in the [whole exome dataset](https://www.synapse.org/#!Synapse:syn20554969) provided for the Hackathon.
For the Hippo idea I have to do some more reading to understand what they mean by the Hippo pathway, which sites get bound by NF1+/+ and not bound by NF1-/-, what short chemical compounds can also bind to those sites (or in your example, what other genes generate proteins which can bind to those sites). For short chemical compounds, what is their toxicity. For proteins, are the corresponding genes unmutated in the whole exome data.
I am bringing a friend, @ikedim, who is also interested in working on the images. For image processing he favors [fast.ai](https://github.com/fastai/fastai). I'm also interested in applying [NVidia RetinaNet](https://github.com/NVIDIA/retinanet-examples). It is not clear how well the data is labelled, which is essential for training, unless you can figure out an [unsupervised approach](https://www.cc.gatech.edu/~jyang375/Jianwei_Yang_files/Unsupervised_Learning_JianweiYang.pdf). I went to look at the [data](https://www.synapse.org/#!Synapse:syn18666641/wiki/594208) just now. It needs a **permissioning request**. If you haven't done so already, now is the time to do that. I just put one in.
So in case the data turns out not be well labelled for training, finding a productive unsupervised approach would be a good Hackathon objective. Sorry I guess I wasn't very clear before! My (naive, un-researched) idea was that there are other proteins (GAPs) that activate hydrolysis of Ras-GTP, which could possibly substitute for NF1. Some of these are discussed in the link I sent. Instead of using the TALE to directly increase NF1 (since maybe there's not even a good copy to express), maybe the TALE could be used to increase expression of other Ras-GAPs instead. Along the same lines, drugs that increase activity of other Ras-GAPs could be identified too. @allawayr and @sgosline probably have a quick answer as to the feasibility of this approach.
Hippo as a modifier of tumors makes a lot of sense and sounds interesting. How do you envision using the datasets to explore it further?
I am interested in analyzing the imaging datasets for this weekend. I think finding predictive or deterministic features could be interesting or further stratifying the data by mutations or other features. Thanks @adnama547. Tales to increase production is along the lines of the David Segal paper above. He reported a negative outcome, due to NF-/- in tumor microenvironment meaning that there was no NF+ to increase the expression of. What do you think of his reasoning for why he failed? What would you do to change his process, or what different process would you employ, to get a better outcome with Tales?
For this weekend, I am focusing on cNF and the direction set by this paper:
* [Spatiotemporal Loss of NF1 in Schwann Cell Lineage Leads to Different Types of Cutaneous Neurofibroma Susceptible to Modification by the Hippo Pathway.](https://www.ncbi.nlm.nih.gov/pubmed/30348677)
Two of the co-authors are guiding us on Synapse, @allawayr and @sgosline . The paper gives a strong theory of CNF and PNF development and also one for bone dysplasia. They suggest that "dampening the Hippo pathway" will cure CNF and look to this paper for Hippo pathway targets:
* [Drug development against the hippo pathway in mesothelioma](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5504108/)
There are more Hippo pathway papers that may also give ideas:
* [The Hippo Signaling Pathway: A Candidate New Drug Target for Malignant Tumors](https://www.ncbi.nlm.nih.gov/books/NBK500335/)
* [Repurposing of Drugs Targeting YAP-TEAD Functions](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6162436/)
* [Safety Considerations in the Development of Hippo Pathway Inhibitors in Cancers](https://www.frontiersin.org/articles/10.3389/fcell.2019.00156/full)
* [Role of Hippo Pathway-YAP/TAZ Signaling in Angiogenesis](https://www.frontiersin.org/articles/10.3389/fcell.2019.00049/full)
* [The Hippo Signaling Pathway: an Emerging Anti-cancer Drug Target](http://www.discoverymedicine.com/Yuquan-Tao/2017/08/the-hippo-signaling-pathway-an-emerging-anti-cancer-drug-target/)
* [Analysis of the role of the Hippo pathway in cancer](https://translational-medicine.biomedcentral.com/articles/10.1186/s12967-019-1869-4)
* [The two faces of Hippo: targeting the Hippo pathway for regenerative medicine and cancer treatment](https://www.nature.com/articles/nrd4161)
* [Hippo pathway mediates resistance to cytotoxic drugs](https://www.pnas.org/content/114/18/E3729)
What do you think? What do you like to focus on in NF1 and what are you most interested in for the Hackathon?
You could look for other Ras GAPs and maybe find drugs that activate them, or design Tales to increase production, to compensate for loss of NF1 as well. https://www.sciencedirect.com/science/article/pii/S0014579397003219 One thought - NF1 tumors are not generally thought to be exclusively NF1-/-, but rather a heterogeneous mix of NF1 -/- tumor cells, and NF1 +/- other cells such as fibroblasts, macrophages, mast cells, etc. Either way, I suspect (and some have observed that) both NF1-/- or NF1+/- cells in the tumor microenvironment have a phenotype different than wild-type cells.
@ferrerm, one of the hackathon mentors, might be a good person to engage on some of the questions regarding your pharmacodynamic questions, i.e.:
>Then we need to assess if either a chemical analog or synthesized neurofibromin 1 is deliverable into cells by some method (injection, ingestion, topical cream), and once delivered will it act the same way as wild-type neurofibromin 1 produced by the body. Direct delivery to tumor cells may be necessary to overcome the naturally NF-/- status of the tumor microenvironment, as the presence of a tumor may be defined by the tumor microenvironment's ability to maintain NF-/- status. Thanks, I'm revising start of thread to incorporate this. A genetic engineering approach or some type of NF1 mimetic is an interesting idea. I think it is likely a very challenging bioengineering problem to deliver or re-express the NF1 protein. I know there's at least one startup looking into this approach though, not sure if you have heard of them or not: http://www.infixionbio.com/
Obviously, as a biotech, they might be tight-lipped on their approach, but perhaps they'd be willing to engage if you reached out? Hard to say.
Drop files to upload
NF1 gene expression therapy and NF1 gene mimetic therapy page is loading…