Executive Insights: The New Frontier in Column Chemistry

Erin Chambers, global vice president and general manager of consumables and lab automation at Waters Corporation, spoke with LCGC International about why column chemistry has become chromatography’s real point of differentiation, particularly for complex biopharma modalities like GLP-1 peptides and lipid nanoparticles. As these molecules become more structurally intricate, analytical methods must evolve just as quickly to keep pace with the demand for speed, sensitivity, and regulatory-grade reliability. By innovating chemistry and particle physics simultaneously rather than separately, Waters was able to cut a GLP-1 impurity profiling workflow from three hours to under an hour. Chambers also discusses how the company builds regulatory readiness into its products from the outset and where she sees artificial intelligence heading in the chromatography community.

You’ve described hardware and software as increasingly table stakes. Can you walk me through what that shift looks like from where you sit, and why the column has become the new frontier of differentiation?

When it comes to instrumentation, there are what I call evergreen needs—things that must constantly evolve: sensitivity, resolution, speed, size. Those needs don’t go away, so there has to be continual innovation on the hardware side. The same holds true for software, which is increasingly being used not just to process data, but to interpret it.

But I’d say the real magic happens in the chemistry, because we can innovate and modify our chemistries to create a precise product—a column or a sample-prep device—that matches the needs of a molecule or a class of molecules. That’s what converts a system into a solution. Instead of just generating data, you can answer a question around a specific molecule or a specific problem.

With the BioResolve Peptide and GTxResolve lipid columns, you’ve talked about making simultaneous advances in both chemistry and particle physics. How rare is that kind of dual breakthrough, and what does it tell us about where column science is headed?

It’s unusual at best—rare, I’d say. It’s hard enough to innovate in one dimension, whether that’s particle physics or surface chemistry, but bringing those two together requires decades of material science expertise: expertise in pore sizes, in base particle chemistry, in ligands, and a deep understanding of how small changes in chemistry or physical properties affect our ability to resolve, retain, separate, and select different modalities.

Gone are the days where you might innovate around particle size or pore size alone. The newer modalities are so complex that engineering a chemistry as a complete system, rather than a collection of individual attributes, is going to be the new standard for separations.

Customer needs sit at the root of innovation, and there’s one example that demonstrates this well. Waters recently released 2 columns: a BioResolve column for peptides, specifically GLP-1s and other insulins, and a GTxResolve column for lipid nanoparticles. Both were born out of very specific needs—the size of the molecules, and the fact that subtle changes impart different requirements on the chemistries. That’s what demanded that we innovate not just around particle physics, but simultaneously around the chemistry; it wasn’t adequate to do one or the other.

One specific example: Customers were spending 3 hours doing a GLP-1 impurity profile, which is critical for safety. We were able to bring that down to under an hour. In an industry where it costs on average 2 billion dollars to bring a drug to market, any time you can save is critical. And beyond that, savings in development can ultimately translate into savings in drug prices and broader availability of these critical therapies.

Waters is investing across GLP-1 peptides, LNPs, mRNA, ADCs, and more, simultaneously. How do you make R&D allocation decisions when the science is evolving so rapidly? What signals are you watching?

What we’re seeing today is a rapid emergence of so many new types of therapies, and it’s so nascent that you could say we don’t know what we don’t know. The things we watch include emerging modalities, scientific breakthroughs, investment trends, and shifts in development pipelines. But probably the strongest signals come when a promising drug class begins to advance into clinical development, and the analytical requirements therefore become more demanding and exacting.

We have very strong relationships with our customers. They come to us asking, “What do I do in this situation?” When we start to see more customers asking the same type of question, that’s a very clear signal that there should be innovation in that space.

We’re watching a lot of different modalities. We’re deeply embedded in GLP-1s and insulin-related compounds, and in lipid nanoparticles. Some of the newer areas we’re watching are bispecifics, trispecifics, antibody-oligo conjugates, DNA-based therapies, and cell-based therapies. There are a lot of risks with many of these, but also a lot of really difficult questions, and it’s those most difficult questions that we’re best positioned to answer.

One of the distinct advantages we have when we see these analytical challenges is our willingness to rethink multiple aspects of the technology we use to solve the problem, rather than optimizing around existing constraints. That only comes from decades of fully integrated manufacturing and understanding the subtlety of every parameter that affects a separation. We’re not just optimizing, we’re fundamentally rethinking the way we answer some of these problems, because we’re sitting front and center with customers. We have a front-row seat to these challenges as they emerge, and that’s really pushing us to think differently.

You’ve talked about reliable data as a precondition for scientific courage—the idea that researchers can only ask bolder questions when they trust their results. Where do you see that dynamic playing out most acutely right now in biopharma development?

I think the fundamental reality here is that if you have questions about your data, a question like, “Is there a coeluting peak?” you can’t move on to ask a question like, “Is there a relationship between this structure, this modification, and the biology or the function?” It’s in biopharmaceutical development that this plays out most acutely.

How much of your product development process is now shaped by anticipating what regulators will require, versus what scientists are asking for today?

The simplest way to answer that is that everything we do is rooted in customer need, derived from deep conversations with customers—questions like, “What do I do here? How do I do this?” That comes from building up those relationships. It’s also critical to anticipate that this ultimately needs to meet certain regulatory demands.

When we talk about being able to trust the data, regulatory readiness is built into our products. It’s not bolted on, and it’s not an afterthought—it comes through a history of quality, reliability, robustness, and batch testing for specific applications. That’s what gives us built-in readiness for regulatory requirements. The two aren’t mutually exclusive, but it’s rooted in the customer need.

How do you build regulatory requirements into the columns when the guidance potentially isn’t fully developed or written yet?

It’s all about the fundamental aspects of quality, reliability, and reproducibility, and being able to trust the data. You want to get the same answer every single time, over years, and that’s something we’ve built.

It’s in the development of the product where we do extremely rigorous testing around reliability and reproducibility, ensuring that batch to batch, time and time again, tested for the specific application it’s intended for, we get the same answer. I’ll give an example with GLP-1s: There are impurities that can differ by a single atom or a spatial arrangement, and the ability to reliably separate such subtle differences, time and time again, is critical. That’s something you need to be able to count on when you’re doing QC and release testing.

What’s your take on artificial intelligence (AI) and how it will affect the chromatography community?

In the near term, I’d say the most immediate uses of AI are things like mining manufacturing and QC data: looking for trends, looking for flags in the data that signal failures, and building predictive diagnostics. Those things are happening today. Another near-term use is generating new hypotheses. Our brains tend to be wired to ask the same types of questions and interrogate data in the same ways, but looking at an aggregate allows a system to ask different questions and create different hypotheses. Right now it’s also a productivity tool; many of us use it to get quick answers or build presentations.

I do see it evolving, especially for our customers, into a growth accelerator. They need to find ways to make R&D more productive.

Longer term, I see it becoming more of a closed-loop learning system that brings together scientific data, AI models, lab experiments, and manufacturing to continuously inform and improve each other—making connections across an aggregate of data that the human brain can’t quite get its head around. One prerequisite for this is having a common, reliable data system; that’s something that could make or break the advancement.

What’s the best piece of leadership advice you’ve received in your career?

I’ve been fortunate to have many great mentors and leaders, and I’d say there are 2 key philosophies I live by, plus a 3-part mantra.

The first philosophy is: Hire people who are smarter and more expert than I am, trust them, and then help them shine—guide them, but get out of their way. I see these incredible people come into Waters, and in an industry that advances as quickly as ours, it’s important to do that, because there’s no way any of us can keep up with every advancement alone. It becomes a virtuous cycle—you bring in good, smart, expert people, they’re trusted by customers, customers trust you to answer their questions, and they come to you first.

The other is building a leadership team that thinks differently than I do. I learned this one day sitting in a project review board. I’m very decisive. We were discussing a new product, and I had a colleague who was thinking very carefully, taking a long time to answer, and I found myself getting a bit frustrated. Then I thought back and realized: That’s exactly what I need. I need somebody methodical, somebody analytical. I need people who are creative, who think blue-sky. Don’t be afraid to surround yourself with people who think differently. Human nature is to surround yourself with people who are the same as you are, but on a leadership team, you have to do the opposite.

And then there are three things I live by: Be optimistic, be pragmatic, and be accountable. My current leader has a great saying: “If not me, then who?” You owe that to yourself, to the people who trust you, and to the enterprise you work within.