Jeff Moffitt’s path to biology began with lasers, not microscopes. Trained as a physicist, he was drawn to the beauty of building instruments that reveal the unseen. Today, as a professor at Boston Children’s Hospital, Moffitt leads a lab that merges engineering ingenuity with biological discovery. His team designs advanced imaging technologies that make it possible to see how cells and microbes organize in space—turning the invisible architecture of life into something measurable. By pairing this innovation with questions in mucosal immunology and the microbiome, Moffitt’s work is redefining how we understand the dynamic relationships that sustain health and drive disease.
Where are you from? What’s your professional background, and what set you on your current path?
Jeff Moffitt: Big question! I’m a physicist by training. I started out fully intending to do high-energy physics. As an undergrad I spent a lot of time in that world, then went to grad school planning to be an experimental accelerator physicist. I worked at CERN and at the Cornell accelerator, and when I arrived at UC Berkeley I expected to keep going down that road. But early in my PhD a few things shifted. I fell in love with biophysics and transitioned from high energy to biophysics, doing my PhD with Carlos Bustamante at Berkeley. Carlos pioneered optical methods to manipulate and monitor single molecules.
I began building optical tweezers—microscope systems that focus a laser to a diffraction-limited spot to trap small objects like polystyrene beads. You can attach those beads to single DNA molecules and use them as “handles.” As you move the beads with the laser, you can directly calculate the force on the beads—and therefore on the DNA.
My PhD focused on developing an assay to detect the motion of a molecular motor that translocates along DNA. The system was the double-stranded DNA bacteriophage φ29, which infects Bacillus subtilis. The phage forms a protein capsid first, then uses a powerful motor at one portal to pull its genome inside. The packing reaches crystalline densities, with internal pressures estimated at tens of atmospheres—energy later used for genome injection during infection.
That motor is a pentameric ring ATPase. When I started, there were open questions about how the five identical subunits coordinate. We wanted optical tweezers sensitive enough to resolve each increment of DNA translocated—the step size. Because the rise per base pair in B-form DNA is 3.4 Å, we needed angstrom-scale resolution in real time. We built a tweezers instrument that could do that, followed the packaging in real time, and worked out the mechanochemistry—how the motor uses ATP to drive compaction.
Why I gravitated to this? Biology’s complexity is beautiful. And, unlike high-energy physics, biophysics is a scale where one person can build the instrument, do the wet-lab work, and the theory to interpret the measurements.
When did you know you wanted to be a scientist? Was there a single moment?
Moffitt: No single moment. I’ve always been an engineer at heart—I like understanding how things work and building things. I enjoyed science through school and gravitated toward physics in college because it nicely merges math and mechanistic thinking. It felt natural to continue.
Many leaders we talk to come from MD or biology-heavy backgrounds. How does physics training shape how you frame biological questions and design experiments?
Moffitt: Biology is the science of complexity; you won’t reduce most of it from first principles. You still have to learn the biology. Where physics helped me most is in the experimental culture: define what you want to measure, then build the instrument. Molecular biology has been “kit-ified,” which is convenient, but it can limit imagination—you frame questions around what’s available. Physics trained me to say, “If it doesn’t exist, let’s build it,” and to iterate until it works. That mindset has guided my career, and it’s something I consciously instill in my lab.
There’s also a philosophical piece. Physics often deals with simple systems—an electron has charge, mass, spin. That is it. As such, this simplicity allows theory that can match experiment with astonishing precision. Growing up in that environment gives you a faith that things are understandable. In biology, even if some problems resist simple descriptions, that confidence helps you keep pushing.
Give us a two-to-three-minute elevator pitch: what does your lab do and what questions drive you?
Moffitt: We’re both a host–microbe/mucosal immunology lab and a microscopy technology development lab. Our niche is questions at mucosal surfaces where current technology can’t yet give good answers. Then we build the microscopy to make those answers possible.
We develop genome-scale microscopy: methods that retain the strengths of high-resolution imaging—cell morphology, architecture from nanometers to centimeters—while capturing the molecular complexity you associate with sequencing (transcriptomes, genomes). In short, we aim to do sequencing-scale readouts with a microscope. Most of our biology centers on what new insights are possible if we can measure more molecular complexity in situ at the host–microbe interface.
What does microscopy give you that dissociated single-cell or bulk sequencing doesn’t?
Moffitt: Several things:
- Spatial context. Tissues aren’t well-mixed bags of cells; organization drives emergent behavior. We get cell type/state information and who neighbors whom, plus morphology—critical for explaining tissue function.
- Avoidance of dissociation artifacts. Not all cells dissociate or survive prep equally, and some change their state. Imaging fixed tissue in situ avoids those issues and lets us capture populations that are hard to recover by dissociation (large, fragile, or adherent cells like stromal cells or megakaryocytes).
- High detection efficiency. Sequencing workflows have losses at extraction, reverse transcription, etc. Our fluorescent, single-molecule readouts can achieve very high capture efficiency, which is crucial for low-abundance transcripts (certain receptors, transcription factors) where a few molecules matter.
Net: we reconstruct spatial organization, maintain morphology, avoid dissociation bias, and detect low-copy transcripts—often revealing cell states/types or spatial niches missed by other methods.
You’ve also emphasized intracellular spatial organization and the problem of signal crowding. How do you resolve dense RNA regions?
Moffitt: Intracellular localization is vital—many processes depend on where RNAs reside (development, stress responses, motility, synaptic plasticity). Our genome-scale readouts rely on optical barcodes created through repeated staining and imaging cycles; barcodes map to single RNA molecules, so we must resolve single-molecule spots. Optical resolution can limit that, especially in microbes where RNA densities are ~100× higher than in eukaryotes.
We addressed this using expansion microscopy (pioneered by Ed Boyden’s lab) to physically expand samples—diluting signals and overcoming crowding. We routinely expand tissue by ~1000× in volume, which lets us resolve densely packed RNAs, including in bacteria.
How many genes can you measure, and what are the trade-offs of ultra-high multiplexing?
Moffitt: Our probes are targeted oligos; barcodes can be arbitrarily long, so in principle we can target hundreds to hundreds of thousands of RNAs. Practically, as you target more RNAs, density increases and single-molecule resolution becomes harder—solvable with expansion, at the cost of throughput (fewer cells imaged per run).
Imaging can still be very high throughput—we routinely generate multi-million-cell datasets—so there’s room to balance multiplexing and cell count. In practice we often use thousands of genes, which lets us profile millions of cells, detect rare states, and interrogate spatial niches.
Because we’re targeted, we design panels around questions, not everything-omics by default. If we’re studying pathogenic E. coli in the gut, we load panels with cytokines, chemokines, receptors (~300 genes), matrix remodeling (collagens, proteases, inhibitors), morphogens (Wnts, BMPs, receptors), etc. Whole pathways go in. If we later have new questions, we build a new panel rather than trying to measure everything at once. As the field commercializes, learning to ask well-matched questions is key; these methods shine when you choose the right targets.
You generate a lot of custom data and still balance biology with technology development. How do you manage that?
Moffitt: My projects typically start with a biological question we can’t answer yet, then we build the technology, and finally we use it to answer the question. That arc keeps the technology anchored in biology—ensuring it at least gets used once—and it trains scientists across both tech-dev and application. Choices you make in tech-dev shape the biology you can do, and vice versa. It’s a rewarding feedback loop and, I think, a great training experience.
How do you get these technologies into other labs’ hands? Collaboration, open science, commercialization?
Moffitt: All of the above.
- Collaboration. We work broadly beyond our core host–microbe focus, pairing with experts in other tissues. That spreads know-how—experimental design, analysis, and mindset. Some early collaborators now run spatial programs of their own.
- Open sharing. We build our microscopes, fluidics, software, compute, storage—everything—and publish protocols, open-source software, and updates. Many labs have successfully replicated the methods.
- Commercialization. Academia optimizes for “works and is novel,” not “robust and push-button.” A company can “sand off the edges,” engineer reliability, and build training pathways for users who can’t invest years building from scratch. I co-founded Vizgen with Xiaowei Zhuang to commercialize MERFISH. It’s gratifying to see MERFISH used across species and tissues—Arabidopsis, planaria, fish, and extensively in mouse brain and beyond. Collaboration and open tools scale only so far; commercialization brings true access at scale.
Pivoting to clinical impact: why focus on the host–microbe interface and mucosal immunology? Did this grow out of your earlier bacteriophage work?
Moffitt: In a way. As a physicist moving up in complexity, I went from molecules to cells; E. coli felt like biology’s “hydrogen atom.” My postdoc explored bacterial cytoplasmic structure and RNA organization, and some ideas that were used in MERFISH started there. I also worked with microfluidics and grew interested in microbial communities—most bacteria don’t live in test tubes.
Surrounded by early microbiome pioneers, I saw the opportunity. When I started my lab, the gut was a natural place: rich microbial communities, clear relevance to health and disease, and incredibly diverse host biology—development, nervous system, immunity, stroma. It let us push spatial transcriptomics on both microbes and host.
At Boston Children’s, we’ve been moving toward the clinic—mouse models and human samples in inflammatory bowel disease (IBD), for example—leveraging the clinical expertise here to make that translation. The amazing clinicians that are our colleagues have made this pivot to translational work possible.
What is the biggest challenge in moving from mouse to human gut tissue?
Moffitt: Two big ones:
- Logistics. Human tissue collection is nontrivial. You need a deeply committed, well-coordinated clinical team. We’ve been fortunate to partner with Scott Snapper and colleagues—they’re phenomenal.
- RNA preservation. MERFISH images RNA, which is fragile. Much of histology wasn’t designed for RNA integrity. Protocols that work in mouse under controlled conditions don’t necessarily translate to clinical workflows. We spent substantial effort optimizing fixation, processing, and sectioning for human samples, iterating closely with our clinical partners. That work was essential—and now we have robust solutions.
Microbiome identification is hard when references are incomplete. With targeted probes, how do you handle unknown diversity?
Moffitt: It depends on the question. We’ve leaned on two major community advances:
- Defined, whole-genome–sequenced microbial consortia in germ-free mice. Many labs have built communities—from 8 to ~100 species—where every member is cultured and sequenced. These recover key host phenotypes and give exquisite experimental control. For targeted imaging these controlled communities are exceedingly powerful as you know exactly what’s there.
- Metagenomic reconstruction. Powerful pipelines can assemble genomes from samples. For clinical translation, one can match biopsies to fecal samples, reconstruct the patient-specific microbiome, and then design targeted probe sets accordingly. Not a solved problem, but the foundation is strong—and that makes it an exciting technical frontier with clear biological payoff.
Looking ahead a few years, what big biological questions are you most excited to tackle?
Moffitt: A central one:
What spatial structures exist within commensal gut communities, how conserved are they, and how do they influence stability, dynamics, and interactions?
Even in the gut—with peristalsis and flow—there are sharp gradients of antimicrobials, oxygen, pH, and nutrients. We’ve had few tools to study how those gradients shape community structure and function. We’re building those tools now.
Inflammation physically remodels the environment, and there’s growing appreciation that pathogen “success” often involves exploiting host-created niches. Understanding the triad—commensals, pathogens, and host—in space is a major focus.
Fecal microbiota transplants (FMTs) can be lifesaving and have even shown cancer impacts. How might this reshape treatment, and where does spatial biology fit?
Moffitt: FMT highlights the potential but isn’t likely the endgame. Most envision designed, quality-controlled microbial consortia rather than donor stool. For that to work, we need to understand engraftment: which organisms can occupy which spatial niches and why. Spatial constraints will influence stability, competition, and co-occurrence. Our goal is to build the tools to quantify how space governs interactions and to inform the design of engraftable, therapeutic communities.
If you weren’t a scientist?
Moffitt: I’d be an engineer. I love building things.
Greatest scientific advance of the last decade?
Moffitt: The explosion of precise, efficient, rapid genome engineering—CRISPR and its many derivatives. Pairing defined perturbations with genome-scale spatial readouts is incredibly powerful. You can make exactly the change you want, in the cell type you want, and then ask how all surrounding cells respond in situ.
Any advice for young scientists interested in spatial optics/spatial biology?
Moffitt: Dive in. It’s still the wild west—plenty of room to make lasting contributions. Be bold, be brave, and roll up your sleeves. It’s not too late; it’s exactly the right time.