Notre Dame Team 3D-Prints Blood Capillary Networks Thinner Than a Hair

Why Capillaries Are the Hardest Part

Biologists have been trying to bioprint functional tissue for decades. The problem has never been the big stuff. Muscle fibers, organ-shaped scaffolds, even simple skin layers all have precedents. The actual blocker is the plumbing. Every cell in your body needs oxygen and nutrients delivered through capillaries: vessels so small that 10 of them side by side would barely span a single sheet of paper. Replicating that network inside a printed structure has been out of reach, until now.

A team at the University of Notre Dame, led by Yanliang Zhang and doctoral student Yuxuan Liao, just published work on the cover of Nature Chemical Engineering showing they can bioprint capillary channels under 10 micrometers in diameter. These channels were then lined with living endothelial cells, the same cells that form the barrier inside real human blood vessels. The cells attached, spread, and formed a functional barrier without leaking.

How They Did It

The method is a hybrid. The tissue-mimicking gel matrix is laid down by extrusion, the same basic approach used in most desktop bioprinters. But the capillaries themselves come from a second process: aerosol jet printing. AJP deposits fine threads of a sacrificial gelatin material inside the matrix. Once the full structure is printed, it goes into warm water. The gelatin dissolves, leaving behind hollow channels of precise dimensions.

The clever part is in the control. Small changes in ink flow rate and sheath gas flow rate change the final channel size. The team built a machine learning framework that figures out the right parameter combination for any given target diameter. That is how they reliably hit sub-10-micrometer channels, which is well below what manual calibration would produce consistently.

The result is hierarchical vascular networks across one, two, and three dimensions. Channels as wide as several hundred micrometers branch down to channels barely wider than a red blood cell. That branching pattern is what makes real vascular systems functional, and it is what has been missing from prior bioprinting work.

What It Means Right Now

The immediate applications are drug testing platforms. Lab-grown tissue models with real capillary barriers can be used as organ-on-a-chip systems to test whether a drug crosses a tissue boundary safely and at what concentration. That has direct implications for personalized medicine, where a patient's own cells could be used to build a testing platform before treatment begins.

The work is NIH-funded, and the team has already secured additional NIH support with collaborators at Harvard Medical School to build a more advanced version of the bioprinter. The long-term goal is a fully autonomous system that can print functional organs: hearts, kidneys, livers. No timeline has been given for that. The near-term goal of organ-on-a-chip drug testing is closer, and it is a genuinely useful application in its own right.

For 3D printing watchers, the significance is broader than bioprinting specifically. The core technique, combining two printing processes with ML-driven parameter optimization, is transferable. Any application requiring fine channel structures in a soft matrix benefits from the same approach. That spans tissue engineering, but also microfluidics, soft robotics, and certain composite manufacturing methods.

The Bottom Line

The breakthrough here is not that researchers printed something that looked like capillaries. It is that they printed something that behaves like capillaries under biological testing. Living cells formed a proper barrier, which is the functional definition of a working blood vessel. That distinction is what separates this work from prior demonstrations that stalled at the visual stage. Capillary-scale bioprinting is now a reproducible engineering process, and drug testing platforms will be the first place it shows up commercially.