Calico’s Adam Freund tells us about using a network approach to aging – and that PAPP-A does preach.

It’s ARDD week; the annual Aging Research & Drug Discovery Meeting is underway, online due to COVID-19 and completely free to attend. Ahead of his presentation on Friday, Adam Freund, Principal Investigator at Calico, took some time to speak to us about his work at Google’s research and development biotech company.

Longevity.Technology: OK Google, what’s Calico? Well, the company states that it is “not a traditional biotechnology company” and its ambitious message is that it wants “to better understand the biology that controls aging and lifespan … to use the knowledge we gain to discover and develop interventions that enable people to live longer and healthier lives.” We kicked off by asking Dr Freund what area of aging and age-related diseases has Calico’s biggest focus at the moment.

Adam Freund: Our top-level goal is to develop interventions that delay aging, but to test such interventions, we have to be able to measure aging. This is easier said than done – the gold standard, lifespan, takes a long time and is relatively information-poor. There are molecular and cellular changes that occur with age, but it’s not always clear which are the most relevant readouts. We’d like to measure aspects of physiological decline, but current healthspan assays take a lot of time and effort, and even then tend to be pretty noisy.

To address those limitations, we’ve spent a lot of time developing innovative tools and novel analyses for quantifying physiological decline in mouse models. We emphasize automated, longitudinal monitoring and multi-dimensional time-series analysis.

“I prefer to take a multi-dimensional approach, measuring many different aspects of aging. Any single biomarker, or even a small set of biomarkers, is unlikely to capture the complexity of the system …”

On the intervention side, one area of focus for my lab is IGF signaling. This was a pretty straightforward choice – reduced IGF signaling is the most validated anti-aging intervention known (slows aging from worms to mammals, with the largest effect sizes ever reported). There are challenges with targeting this pathway, of course – dose-limiting toxicity, endocrine feedback, lack of biomarkers, just to name a few – but we think we’ve identified a viable therapeutic strategy, which I’ll talk about during the conference.

Longevity.Technology: What emerging discoveries and techniques is Calico utilising in its Longevity research (that you are able to share of course)?
Adam Freund: I’m excited about using outbred mice for intervention testing. We’re clearly not the first people to think of this, but we’ve embraced the concept. Outbred mice are somewhat more resource-intensive than inbred mice because they have more variability (although some of that can be reduced by the longitudinal monitoring I mentioned above), but we think they’re worth it.

As we’re all painfully aware, many published results fail to replicate. I think that a big fraction of what’s being called irreproducibility is actually a lack of generalizability. In other words, the results might repeat under the exact same conditions, but alter those conditions just a little and it’s a different answer. For mouse studies, strain background is an important condition, and we worry about results from a single, homozygous-at-all-loci genotypes not being generalizable. Outbred mice help us avoid this. Voelkl et al., 2020 is a great paper on the value of diversifying study samples and conditions.

“… If even a fraction of the effect we see in worms, flies, or mice translates to humans, it could add years to lifespan and healthspan.”

Longevity.Technology: You said that the best advice you ever received was “learn to fail as quickly as possible” – have you been able to apply this philosophy to your Longevity research at Calico?
Adam Freund: It’s always weird to learn you’ve been thoroughly Googled! In science, this means doing the killer experiment as quickly as possible, rather than dancing around the key questions in the name of exploration. It’s scary to put one’s most cherished ideas to the test, but in the search for truth, it’s better than the alternative.

When we first started working on PAPP-A [pregnancy associated plasma protein A], we had no experience with it, so we started by testing its most essential characteristic for our purpose – the ability to modulate IGF signaling. We biochemically confirmed that activity and so we decided to move forward with the program, but if we had found the opposite, we would have stopped right there and moved on to the next thing.

Longevity.Technology: What do you think is the best way to quantify longitudinal decline – are there key biomarkers that you’re addressing?
Aging manifests at all levels of biological organization (i.e. molecules, cells, tissues, organs, organ-systems, and whole organisms), and measuring aging at each level has pros and cons. Molecular and cellular data provide mechanistic insight and can point to new therapeutic targets, but it can be hard to know if effects are truly relevant to the organism (e.g. does delaying mutation accumulation delay decline in organ function)?

Organ-level and physiological data provide health relevance, but it can be hard to tease out mechanism – good for testing putative targets, less good for target discovery. My lab focuses on developing tools for measuring organism-level decline because we think the state of the art is lacking and robustly testing putative targets is rate-limiting in the field.
Irrespective of the level of phenotyping (e.g. cellular or physiological), I prefer to take a multi-dimensional approach, measuring many different aspects of aging.

Any single biomarker, or even a small set of biomarkers, is unlikely to capture the complexity of the system, so if we focus on a small subset, we run the risk of falling prey to Goodhart’s law: “any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes”. In other words, when a measure becomes a target, it ceases to be a good measure. Looking at many components of the system simultaneously helps us avoid overemphasizing any particular endpoint.

Longevity.Technology: Have you been able to bring your network approach to aging (focusing on organism-level phenotypes that change with age) into your research at Calico? Is it driving results?
Adam Freund: Absolutely – it’s the framework that drives development of the automated phenotyping platform mentioned [above]. We use a conceptually similar approach for cell and molecule data as well, with the eventual goal of merging data from multiple layers of biological organization. We’re currently using these platforms to evaluate and compare and contrast putative aging interventions.

Longevity.Technology: What product in Calico’s pipeline excites you the most and how will it change the face of Longevity?
Adam Freund: For the last five years I’ve been working on PAPP-A, and I think it’s a really interesting target. With newer areas of focus like senescent cells, NAD, and mTOR, IGF signaling has fallen out of favor in the field (search “IGF signaling, aging” in Pubmed and look at the results by year), but inhibition of IGF signaling remains the most efficacious aging intervention we know of in preclinical models. If even a fraction of the effect we see in worms, flies, or mice translates to humans, it could add years to lifespan and healthspan.

The 7th Annual Aging Research and Drug Discovery Meeting is taking place virtually, 1 September 2020 to 4 September. To find out more, visit the ARDD website.