Oxygen-sensing 3D printed skin grafts: printing psECM and checking stiffness under compression

Chronic wounds have a habit of looking stable on the surface while the tissue underneath is doing something else entirely. Oxygen is part of that story, but it is rarely measured where it matters: inside the wound bed and inside whatever graft or dressing gets put on top.

A proof of concept study from researchers at Technion – Israel Institute of Technology builds a workflow around that gap. The team 3D printed an oxygen-sensing 3D printed skin graft from porcine skin extracellular matrix bioink (psECM bioink), embedded oxygen microsensors for electron spin resonance oxygen sensing (ESR oxygen sensing), and used a CellScale MicroTester to run compression testing of hydrogels at 37°C so the printed constructs had a measured stiffness and handling baseline rather than an assumed one.

What they were trying to build

Before the rheology plots and sensor decay curves, it helps to keep the workflow in your head. The aim was a personalized wound graft 3D printing pipeline that starts with porcine skin, converts it into an ECM ink, prints a graft that matches chronic wound geometry, and then measures oxygen inside the printed material rather than inferring it from the surface.

If you’re newer to this space, we put together a quick primer on how 3D bioprinting workflows typically get assembled (bioink, support strategy, and post-print handling)

Turning porcine skin into a printable ECM material

They start where most ECM-based graft workflows start: decellularize, then prove you actually decellularized.

Figure 1 below does three jobs at once. Panel A shows DNA content dropping in psECM compared with native porcine skin tissue. Panels B and C use H&E staining as a quick visual check that this is no longer intact skin structure. Panels D and E switch to SEM, less about aesthetics and more about whether the processing leaves behind a porous collagenous network that still looks like an ECM scaffold.

This part of the paper is easy to gloss over, but it is doing a lot of work for the rest of the story. If the input material is not convincingly decellularized, it becomes hard to interpret anything downstream, from printing behaviour to cell response.

psECM bioink behaviour: gelation and crosslinking choices that show up later

The authors prepared psECM bioinks at different concentrations and tracked viscoelastic behaviour over time and frequency. If you have worked with ECM-derived hydrogels, the time sweep will feel familiar. Storage modulus rises, but the system does not always settle into a tidy plateau on the timescale you would like for printing, handling, and post-processing.

They also used EGCG as a crosslinker and explored how crosslinking changes stability and stiffness. That crosslinking decision ends up mattering later, because once you embed microsensors and print patient-specific shapes, the constructs have more ways to fail mechanically. A cube that sags or fragments is annoying. A graft that loses integrity during implantation is a different problem.

(Their rheology and thermal analysis are summarized in Figure 2 of the publication. It is a dense figure, so I am leaving it out of the blog, but it is the technical basis for why they select specific ink concentrations and crosslinking conditions.)

3D bioprinting and patient-specific geometry

The researchers now shift from “material prep” to “does it print and does it hold together.”

In the figure below, panels C and D are the most accessible demonstration: they printed organ-shaped models and then stepped into patient-specific wound geometry using a scanned diabetic ulcer and a CAD-derived print. It is not a clinical workflow yet, but it shows the concept clearly enough that you can imagine how a clinic might request a graft shape rather than trimming a generic sheet.

Panel B tracks construct weight over time as a simple stability proxy. It is not a perfect measure of mechanical stability, but it answers a real question: do these things stay intact over days in an aqueous environment, especially when crosslinked.

How the MicroTester was used in this study

The mechanical testing in this paper is not a “bonus characterization” section. It sits in the middle of the workflow as a constraint on everything else: printing, handling, implantation, and even sensor placement.

For the mechanical assay, the authors printed 5 mm × 5 mm × 5 mm psECM cubes and crosslinked them with different EGCG concentrations. They then used a CellScale MicroTester to perform displacement-controlled compression at 37°C, using a compression plate configuration suited to small, soft constructs. Young’s modulus was extracted from the loading curve (reported in Figure 3A), providing a direct stiffness comparison between crosslinking conditions.

In practice, this kind of micro-scale compression test is doing two useful things at once:

1. It gives a handling and integrity baseline. If crosslinking increases modulus and reduces construct drift in water, that is relevant before you even start thinking about implantation.
2. It anchors the oxygen-sensing concept to a physical material state. If you are going to interpret oxygen readings inside a graft, you still need to know what kind of mechanical environment the graft is presenting to cells and to surrounding tissue. In soft ECM constructs, small formulation changes can move stiffness enough that “same print, different behaviour” becomes a real outcome.

Testing soft hydrated materials at physiological temperature tends to change what you measure more than people expect, and we’ve run into that same issue in other soft polymer systems as well.

Embedded oxygen microsensors and ESR oxygen sensing in printed grafts

The part that makes this more than an ECM bioink paper is the sensor integration. They embedded LiNc-BuO oxygen microsensors at controlled depths inside printed cubes, then read oxygen through the construct using electron spin resonance oxygen sensing.

The figure below is both the setup and the signal reality check. Panel A shows the ESR measurement approach in a controlled atmosphere. Panels B and C compare microsensors positioned at 1 mm vs 0.5 mm depth from the construct surface. Depth matters here in an unglamorous way. The deeper sensors produced noisier reads, which is a practical constraint if you imagine scaling from cubes to thicker graft geometries.

Cytocompatibility: do the sensors disturb the graft-cell interface?

They evaluated cytocompatibility using keratinocytes and fibroblasts cultured with the psECM system, with and without embedded microsensors.

Figure 5A in the publication tracks relative viability over time, comparing plate controls to bioink, and to bioink containing two microsensor concentrations. Panels B and C add live/dead images at day 1 and day 14.

The results read as “workable, but not neutral.” The presence of microsensors does not collapse viability in vitro, but the curves do not sit exactly on top of each other either, and concentration dependence shows up. That is probably the honest outcome of adding a sensing layer into a biomaterial: it can be compatible, but it is rarely invisible.

In vivo: a useful reality check, and not a perfectly clean story

The in vivo section is where the “proof of concept” framing earns its keep. Feasibility is one thing. Tissue response is another, especially when you introduce a foreign sensor phase into a graft that is meant to integrate.

The figure below compares implantation outcomes for bioink alone, bioink plus microsensors, and an alginate control. Panel A shows the implant appearance after three weeks. Panel B summarizes blood cytokine levels at one and three weeks. Panels C and D show H&E and macrophage staining, which makes the “what did the tissue do?” question less abstract.

One thing that stands out is how quickly “sensor placement” stops being a nice-to-have optimization and turns into an immunology variable. The authors discuss design strategies like keeping sensors away from the wound-facing surface, which feels like an iteration you only arrive at after seeing tissue response rather than predicting it.

Where this leaves oxygen-sensing 3D printed skin grafts

It is tempting to talk about this work as if it solves wound monitoring. It does not. It gives a workable stack: a decellularized psECM ink that prints, a crosslinking strategy that shifts stiffness and stability, a sensor integration approach that can be read by ESR, and early compatibility data that is encouraging but not frictionless.

If you are building ECM-based skin grafts or exploring patient-specific wound graft 3D printing, the useful lesson is that oxygen sensing is not an “extra module.” It changes constraints: geometry, thickness, sensor depth, surface exposure, and the way you interpret biocompatibility. In the middle of all that, having a measured stiffness baseline from compression testing of hydrogels helps keep the rest of the interpretation grounded.

Citation

Zur, Y.; Hayam, R.; Almog, N.; Kovrigina, I.; Baruch, L.; Blank, A.; Machluf, M. 3D Printing of Oxygen-Sensing ECM-Based Skin Graft for Personalized Treatment of Chronic Wounds—A Technological Proof of Concept. J. Funct. Biomater. 2026, 17, 28. https://doi.org/10.3390/jfb17010028

About the CellScale MicroTester

The CellScale MicroTester is a micro-scale mechanical testing system commonly used for compression and tensile testing of small, soft biological samples where conventional mechanical frames are oversized or lack sensitivity. In biomaterials and tissue engineering work, it often shows up when researchers need to measure stiffness changes in hydrogels, ECM constructs, micro-scaffolds, soft tissues, and small printed specimens, especially when testing in hydrated conditions or at physiological temperature.

One place micro-scale compression becomes especially practical is when the sample is simply too small or too fragile for conventional grips and platens, like cell clusters and compact microtissues.

In this study, MicroTester-based compression testing provided a practical stiffness comparison across EGCG crosslinking conditions in a format that matched the printed construct scale. That matters because it keeps the mechanical readout tied to the actual printed graft geometry rather than relying on bulk gel assumptions.

Learn more about the MicroTester micro-scale mechanical tester.