The study asks how two-photon polymerization can produce open 3D fibrous cell-culture scaffolds without unintended directional geometry from elongated voxels and proximity effects. Using OrmoComp, the authors experimentally characterized line writing, developed uniformly spaced elliptical shell designs and calibrated laser power by scaffold level. Electron microscopy measured near-unity beam aspect ratios for nominal 3, 5, 7 and 10 micrometer designs. After two hours in fibronectin-coated cages, NIH 3T3 fibroblasts showed increasing nuclear volume with beam width and a plateau between 7 and 10

Micro and Nano Engineering · 2026

A laser’s smallest writing unit is elongated. Compensate for its shape, and you can build a more controlled world for a cell.

01 The geometry problem

Build a cell’s 3D world.

Two-photon lithography writes a scaffold inside a photosensitive material. But the laser voxel is longer along the optical axis: a horizontal beam can print taller than it is wide.

Choose a target beam width

Axial elongation can introduce unintended directionality. Apply the paper’s writing strategy to reveal the measured print.

Interactive schematic: scaffold dimensions, voxel elongation and cell shapes are illustrative. Animation rotates the view. Measurements shown after correction are from Figure 4 (mean ± SD).

Draw against the distortion.

For 5, 7 and 10 µm targets, the researchers wrote elliptical cylindrical shells with uniformly spaced lines. These shells were then flood UV-cured to complete the beams.

Give each line room.

Uniform 0.3 µm line spacing reduces local excess exposure. The 3 µm design uses multiple lines, rather than a shell. The order of writing also changes how nearby beams develop.

Make geometry a choice.

Measured horizontal sections have aspect ratios close to one, and vertical poles have comparable widths. Beam width becomes a controllable variable for studying cells in an open 3D scaffold.

02 Light meets depth

Write smarter, level by level.

The same laser setting does not produce the same beam at every height. In this setup, the researchers increased power toward the top to compensate for optical distortion.

At the bottom level, the reported writing power is 11.0 mW.

Refractive-index mismatch and spherical aberration change the focus with depth. This calibration is specific to OrmoComp and the paper’s inverted air-objective setup.

03 From scaffold to cell

Look at the cells.

NIH3T3 mouse fibroblasts were placed in fibronectin-coated cages and fixed two hours after seeding. Choose a beam width to compare the real microscopy.

Compare the adhesive landscape

Cells follow the struts.

On the thinner beams, elongated cells align with the scaffold’s struts. The measured nuclear-volume trend rises as beam width increases from 3 to 7 µm.

More adhesive contact area may allow more spreading and cytoskeletal tension—a possible explanation for the nuclear-volume trend.

Geometry and cell response are linked here; contact area is the proposed explanation.

Nuclear volume levels off.

Nuclear volume increases across the 3, 5 and 7 µm groups, then plateaus between 7 and 10 µm. The 2D reference has a mean nuclear volume of 1650 ± 843 µm³ .

Cells touching the thick outer frame or glass were excluded from the scaffold analysis. The image and distributions below are measured results from Figure 5.

↗ The evidence

Follow the experiment.

From the laser’s writing footprint to the cells’ nuclei: inspect the original figures and the supplementary observations.

A controlled stage for new questions.

These isometric scaffolds create a route to studying how 3D geometry relates to cell migration, YAP/TAZ mechanosensing and transitions toward myofibroblast behavior. These are future research directions.

Two-photon lithography-based fabrication of 3D fibrous scaffolds: Addressing anisotropic voxel constraints for the design of bio-inspired scaffolds

Aída Gabriela Fernández Contreras, Inès Paysan, Chloé Viat, Ayla Biallas, Olivier Stephan, Boudewijn van der Sanden, Jean-Hervé Tortai and Alice Nicolas. Micro and Nano Engineering 32 (2026), 100388. DOI: 10.1016/j.mne.2026.100388.

Source: HAL manuscript and supplementary material. Figures © Fernández Contreras et al., shared under CC BY-NC 4.0. Figure panels are extracted; microscopy comparison panels are cropped from Supplementary Figure S5, otherwise unchanged. Graphical abstract: AI-generated conceptual illustration informed by Figures 3, S4 and S5. Interactive geometry and cell shapes are explanatory schematics.

Fernández Contreras et al. · HAL manuscript and supplements · CC BY-NC 4.0.