NewLimit was founded to significantly extend healthy human healthspan.
We are developing epigenetic reprogramming medicines to treat age-related diseases on our way to a more general use medicine to control aging itself. The core of our approach is rooted in epigenetic control of gene expression. All of our cells experience functional decline as we age, yet we appear to be able to reverse this functional decline through induced pluripotent cell reprogramming and then re-differentiation to the original cell type. Pluripotent reprogramming reverses cellular “age” while also reversing cell type. NewLimit’s aim is to reprogram cellular “age” without altering cell type.
In some ways this is a narrower task that full pluripotent reprogramming, but the result is more nuanced. The cell type remains the same, while the cell’s epigenetic state and functional abilities are changed. We think these kinds of reprogramming factors could make powerful medicines in a subset of diseases where the root cause is due to functional decline in a specific cell type. We are starting our work in T cells and hepatocytes with specific indications in mind as we iterate towards more general purpose medicines to treat broader aspects of aging.
More than a decade ago, scientists discovered that we can rewind the clock of animal development in adult cells by activating just a handful of regulatory genes, “reprogramming” adult cells to their embryonic state. Reprogramming works by rewriting the epigenome, changing the rules that dictate which genes can be turned on and off. Remarkably, reprogramming to an embryonic state reverses many features of aging – reprogrammed cells from young and old animals become nearly indistinguishable.
Early experiments have shown that even partial activation of these embryonic programs can restore function in old cells without erasing their unique identities, improving outcomes in disease and injury models [1]. These results are early and more work is required to validate the magnitude of effect, but they do serve as a hint that robust partial reprogramming is possible. Taken in totality with pluripotent reprogramming, the work suggest that many aspects of aging may be the result of plastic, reversible, changes in the epigenome. If we can construct the right epigenetic program, we could conceivably reverse these changes and restore youthful function to old cells. This rejuvenation could have beneficial impacts across a number of diseases with large, unmet clinical needs.
What’s holding us back? There are both unanswered biological questions and unsolved engineering challenges between us and the first reprogramming therapies. We don’t yet know exactly where partial reprogramming will be most impactful, what programs to execute, or what safety risks reprogramming might present. After those questions are answered, we require a delivery system that can execute our programs in the right cells, at the right time, with the right dose. We outline a few of these open problems below.
None of these challenges are trivial, but all are surmountable.
New Limit’s product discovery and development starts with modeling how specific cell types age, proceeds with discovery of novel reprogramming factors that can restore necessary function, and ends with fitting these factors into a drug product for patients. NewLimit’s scientific team will employ single cell multi-omic tools and machine learning methods to address these outstanding challenges and enable therapeutic development. We plan to initially start with two to three cell types, that address specific indication, with distinct risk/reward profiles. For each indication, we will identify target cell populations where we believe epigenetic reprogramming will be disease-modifying and initiate a reprogramming factor discovery campaign to design a therapeutic program.
Payload discovery: Reprogramming factor campaigns will employ a tiered screening strategy, beginning with pooled single cell genomics screens to find factors that restore functional molecular states [2]. Partial reprogramming strategies as described above may serve as a positive control and offer a useful initial search heuristic, but alternative payloads are desirable to reduce the risk of adverse events. Most reprogramming strategies that achieve a stable cell state change require a combination of factors. Even a small hypothesis space contains an intractable number of combinations, so an exhaustive search is infeasible (e.g. ~$10^5$ combinations of 4 in 40 factors). This has traditionally been a roadblock for the field, but guided search methods from machine learning may allow us to efficiently search the combinatorial hypothesis space and find effective combinations using a tractable number of experiments.
Functional validation: Given hits from these initial screens, we will perform validation using medium-throughput ex vivo assays and finally pre-clinical in vivo disease models. These assays will inherently be unique to each indication program. As we acquire more data spanning each of these assay tiers, we will develop machine learning models to infer later stage assay results from molecular profiles, improving both interpretability and utility of our first-tier screens.
Biological model systems: We believe that performing payload discovery in human cells will improve our chances of reducing human disease burden later on in development. Wherever possible, we will use primary human cells for our payload discovery screens and functional assays, transitioning to animal models for physiological endpoints. We have multiple collaborators interested in providing us with human primary cells in exchange for data.
Focus on disease-modifying endpoints: Each indication program will be evaluated based on pre-defined efficacy milestones and halting criteria. We believe that it’s important to focus our resources on the most promising applications, requiring us to continually re-evaluate our priorities and avoid the sunk costs fallacy. To that end, we will treat our functional assays as the ultimate Key Performance Indicators for each program.
Team Structure: We envision the Scientific Team clustering around four key functions.
Read: Genomics experts developing methods to read epigenetic states in single cells
Write: Molecular biologists developing tools to rewrite the epigenome and engineer cell state
Predict: Computational scientists developing models to enable reprogramming factor discovery
Product: Cell biologists and physiologists developing functional assays for indication programs