Accelerating the Development of a Non-Hormonal Male Birth Control Pill 💊
Sim & Diba: Progress Update #1
Hey there! 👋
My name is Simran, and I’m working with my partner Diba to accelerate the development of a non-hormonal male contraceptive pill. This newsletter is being written as a public progress update for the past few months.
Where we started - 2021 Hackathon 👩🏽🔬👩🏻🔬
Diba and I’s journey started during a hackathon, towards the end of 2021. We had 24 hours to come up with a novel idea, and build whatever we could in the time given.
Everyone in the team (Diba + Liesl + Anastasija + I) were interested in reproductive biology, so we started exploring problems in that space.
We were all collectively angered by the unpleasant side effects in female hormonal birth control pills (nausea, headaches, migraines, weight gain, increased risk of depression). These effects occur because the pills contain artificial estrogen and progesterone which affect the body’s hormone levels.
The team started brainstorming solutions to the problem.
Unfortunately, a female pill cannot be non-hormonal. Females are born with already differentiated sex cells called primary oocytes that mature into secondary oocytes which can be fertilized. A non-hormonal pill’s approach will require the morphological change of immature sex cells. In females, this would require deforming their primary oocytes which cannot be replenished- leading to sterilization.
Males still have their stem sex cells after birth, which is why they can produce 1000 sperm per second. A non-hormonal contraceptive pill for sperm can be made. But how would it work?
We can deform sperm by inhibiting the protein that helps the maturation of spermatocytes into sperms during spermatogenesis. An example target protein in sperm maturation pathway is SPEM1 but there’s many others as well. SPEM1 specifically is a protein that is very important when it comes to giving sperms their mature shape. SPEM1 plays a role in detaching cytoplasm from the spermatid nucleus and flagellum neck region, a step that results in the straightening of the sperm. Targeting this protein with the right small molecule compound would cause severe deformation of sperm → causing it to lack motility due to their morphology.
So, if we can make a non-hormonal birth control pill for sperm producers, why isn’t it on the market? 💸
After doing a root cause analysis on why this pill doesn’t exist yet, we identified 2 major root causes:
1) Lack of funding from Big Pharma → Burden of funding is on governments & non profits.
Big pharma has been more reluctant to fund male contraception because:
There are yet to be set guidelines by the FDA for a male contraceptive
Since birth control is a drug with no medical benefit, it must be proven that long term use of the product will not affect young healthy people.
2) High safety expectation for contraceptives in general. Clinical trials take a very long time to conduct safety and efficacy tests.
Preclinical testing and clinical testing take the most amount of time
Contraceptives need to be used for a year or two before they can be proven effective.
The expectations for the safety of a birth control are much higher than any other drug.
We started to tackle the second problem, and came across drug repurposing.
Drug repurposing is a strategy for identifying new uses for approved or investigational drugs that are outside the scope of the original medical indication.
FDA already approved = less testing + less time for development.
Repurposed drugs are generally approved sooner (3-12 years), in comparison to de novo drugs (10 - 16 years) at a reduced cost of 50-60%.
Also, while 10% of new drug applications get approval to go to market, 30% of repurposed drugs are approved.
Sildenafil (Viagra) is a very well known, successful example of repurposing. It was introduced into the market as a drug to treat angina + hypertension, and repositioned as a drug for erectile disfunction. There are numerous other drugs which have been re-profiled to treat other medical conditions, such as minoxidil, aspirin, valproic acid, and methotrexate.
Under the route of repurposing, Diba and I first did a manual literature review.
We tried a compound → target, and target → compound analysis.
A compound → target analysis is essentially looking at why a previously approved drug is causing reversible male infertility & then optimizing that compound to cause sterility not infertility. A target → compound analysis is the inverse. Finding a sperm maturation/development pathway and identifying approved compounds that have shown to target it.
As a result, we found several candidates for compounds, one being nifedipine- a calcium ion channel blocker used for angina and high blood pressure. However, we saw that the compound had been given up on without explanation, and had cardiac implications. It would require more safety tests, defeating the purpose of our solution.
We realized that manual review for thousands of literature pieces was unrealistic, and hopped towards a more comprehensive approach using computational methods. We wanted to do this to find a drug with a more specific pharmacological mechanism, that could be buried under layers of paper work.
How would you do this? We hypothesized to get a dataset of compounds tested against a sperm maturation target with determined bioactivity metrics, and train an ML model to predict the bioactivity of FDA approved molecules
We tried to find datasets on ChEMBL- specifically of compounds tested against certain spermatogenesis targets. Unfortunately, there are no publicly available data sets for contraception.
When we tried to get data from private institutions, there were a lot of legal hurdles- and even if we spend time working for them, it’s uncertain when we will get access. There’s a lack of open sourcing in the field.
Learning about Knowledge Graphs 📈
We pondered on the question, “If there’s no dataset of compounds- can we make a dataset of disease/compound relations?”
We found the idea of knowledge graphs, which was inspired by a company called Benevolent AI.
A knowledge graph is a network with combined entities of various types. These entities are connected by numerous types of relationships. Both entities and relationships can also carry additional attributes.
At a high level- in drug discovery, entities in a knowledge graph can be things like compounds, proteins, and genes. Relationships consist of molecular interactions, gene-functional associations, and drug-target interactions among multiple others. The entities and relationships are supported by literature and put together by natural language processing.
Benevolent AI uses machine learning models to predict and validate the most biologically relevant, progressible target hypotheses. Their ML models show large scale predictions for disease targets, with evidence from multiple sources of data.
Impressive case study → They used knowledge graph to search for FDA approved drugs to repurpose in the treatment of COVID-19 infections, specifically to block the viral process. A drug that was previously used to treat rheumatoid arthritis (baricitinib) was identified within 48 hours with their technology. It’s now approved by the FDA as a treatment for Coronavirus.
While this research has been incredibly interesting, Diba and I have always worked in deep science. After talking to many of our mentors, and those using NLP for drug discovery- we made a decision to take a step back from this solution. By this summer, our goal is to make a tangible impact on the field- we can learn to build a prototype, however the prerequisites will take majority of our time.
Lack of Funding 💰
If you recall, the two original root causes were a) lack of funding and b) high safety expectations.
As of recently, we have started to look at lack of funding. Without the support of big Pharma, getting clinical trials for the new birth control pill candidates off the ground will be a difficult task. We are looking to the DeSci movement on blockchain led by Molecule and VitaDAO and planning to build a decentralized autonomous organization (DAO) dedicated to funding research for male contraceptives. A DAO would be able to get more investors involved than solely non-profits and governments, patents held as IP-NFTs would promote open-sourcing and sharing of data amongst research and increase awareness about contraceptive research among the general public.
The first project would be to gather enough funding for the two upcoming clinical trials for a male birth control pill. The next steps for us is to build the DAO and a community. If you know anyone working in DeSci, biotech investors interested in contraceptives, sperm contraceptive researchers, we’d really appreciate a warm intro.
There has definitely some pivoting in the last few months, but through the wins and losses- there have been constants of learning and growth. We’ve both enjoyed the iteration, feedback, and ability to connect with such interesting + talented people.
We are grateful for the opportunities we’ve gotten; We spoke at Male Contraceptive Initiative’s Youth Webinar on Wednesday, joined their Youth Advisory Board, and have been in contact with several institutions who are working on solving the same problem as us.
Our summer plan is to spend a couple weeks in a wet lab setting with a team working on a compound for a sperm contraceptive. We’ll keep you updated on that ;)
Diba and I are excited for the next chapter of our project.
See you soon! Thanks for the support. ❤️