Shape-defined solid micro-objects from poly(d,l-lactic acid) as cell-supportive counterparts in bottom-up tissue engineering

Abstract

In bottom-up tissue engineering, small modular units of cells and biomaterials are assembled toward ​larger and more complex ones. In conjunction with a new implementation of this approach, a novel method to fabricate microscale objects from biopolymers by thermal imprinting on water-soluble sacrificial layers is presented. By this means, geometrically well-defined objects could be obtained without involving toxic agents in the form of photoinitiators. The micro-objects were used as cell-adhesive substrates and cell spacers in engineered tissues created by cell-guided assembly of the objects. Such constructs can be applied both for in vitro studies and clinical treatments. Clinically relevantly sized aggregates comprised of cells and micro-objects retained their viability up to 2 weeks of culture. The aggregation behavior of cells and objects showed to depend on the type and number of cells applied. To demonstrate the micro-objects' potential for engineering vascularized tissues, small aggregates of human bone marrow stromal cells (hMSCs) and micro-objects were coated with a layer of human umbilical vein endothelial cells (HUVECs) and fused into larger tissue constructs, resulting in HUVEC-rich regions at the aggregates' interfaces. This three-dimensional network-type spatial cellular organization could foster the establishment of (premature) vascular structures as a vital prerequisite of, for example, bottom-up-engineered bone-like tissue.

Keywords: Bone tissue engineering; Hot embossing/thermal imprinting; Human bone marrow stromal cells; Human umbilical vein endothelial cells; Poly(lactic acid); Self-assembly.

Graphical abstract

Fig. 1

Visual comparison of the (A) traditional gel-based bottom-up TE approach and the (B) novel ‘solid’ particles–based bottom-up TE approach presented in this article. (A) Well-known hydrogel-based bottom-up approaches provide cells with structural support and allow for high control of constructs' architecture by assembling cell-laden building blocks . (B) Here, cells are combined with micro-objects of different shapes and sizes to create millimeter-sized modular tissues. In these modules, the objects are occupying space, which considerably lowers the number of required cells to obtain clinically relevantly sized tissue constructs compared with a cell-only approach (the latter is not shown in the figure). Furthermore, these objects provide surface area for cell attachment. The formed tissue modules can be obtained in various shapes and sizes depending on the mold in which they are allowed to self-assemble. By combining multiple tissue modules, larger, millimeter-sized, and more complex tissue constructs can be created. TE, tissue engineering.

Fig. 2

A schematic representation of the five sequential processing steps to fabricate free-floating PDLLA-based micro-objects. (A) PDLLA was dissolved in DCM (9% w v⁻¹). PDLLA films with a thickness of approximately 5 ​μm were created from this solution on a stainless-steel wafer by dip-coating. (B) Before imprinting, the dip-coated PDLLA film was peeled of the stainless-steel wafer and transferred onto a silicon wafer, which before was spin-coated with a layer of PVA. (C) As a mold, a silicon wafer containing the inverse structures of the micro-objects to be molded and coated with an FOTS antisticking layer was applied. Micro-objects were thermally imprinted under vacuum at a temperature of 130 ​°C. (D) The imprinted micro-objects were then still connected through a residual layer of PDLLA, which was removed by directional oxygen plasma etching. (E) After etching, free-floating micro-objects were obtained by dissolving the sacrificial PVA layer underneath the objects when immersing the wafer in demi water. PDLLA, poly(d,l-lactic acid); DCM, dichloromethane; PVA, poly(vinyl alcohol); FOTS, perfluorooctyltrichlorosilane.

Fig. 3

Micro-objects of different defined shapes such as (A) cubes, (B) donut-, and (C) LEGO™-shaped (porous) objects can be obtained and are shown here in suspension in culture medium to incubate them before cell seeding. A live-dead assay on the full aggregate of 5 ​× ​10⁴ hMSCs with (D) approximately 2000 cubes and (E) approximately 700 donuts shows minimal cell death after 14 days of culture. SEM images show a compact assembly of (F, G) cubes or (H) donuts with hMSCs after 1 week of culture. (F) One aggregate was disrupted on purpose to be able to study the core of the aggregate. It can be seen that the core of the aggregate shows a certain degree of porosity with cells homogeneously distributed between the objects while the outside of the aggregate is fully covered with a dense cell sheet. (H) In some aggregates, micro-objects were found on the outer layer of the aggregate. Yet, in most aggregates, all objects are fully integrated under a dense layer of cells and ECM. (I–J) These results are confirmed by histological analysis where a homogeneous distribution of cells and objects is found in the core of the aggregate, while the exterior shows multiple layers of cellular and extracellular matter. The box in (I) indicates the area displayed at higher magnification in (J). The aggregates in (D) and (E) show a diameter of approximately 600 ​μm. The scale bars represent (A–C, F, J) 100 ​μm, (G) 20 ​μm, (H) 50 ​μm, and (I) 500 ​μm. hMSCs, human bone marrow stromal cells; SEM, scanning electron microscopy; ECM, extracellular matrix.

Fig. 4

(A) Circularity quantified by image analysis of aggregates formed at various object-to-cell ratios of 1:2.5, 1:5, and 1:10 obtained by introducing 10,000, 5000, and 2500 cube-shaped objects per well, respectively (n ​= ​3). (‘hMSCs’) For aggregates of only 2.5 ​× ​10⁴ hMSCs and micro-objects, a significant increase in circularity is found over time and with lower object-to-cell ratios. (‘HUVECs’) For aggregates of 2.5 ​× ​10⁴ HUVECs and objects, a lower object-to-cell ratio increases the circularity as well. Yet, the circularity of the formed aggregate is significantly lower than for hMSCs and does not increase over time. (‘Mix’) When the two cell types are mixed in a 1:1 ratio before addition to the micro-objects, after 9 days, the circularity reaches similar levels as for hMSCs only, indicating that the hMSCs are able to compensate for the limited aggregation behavior of the HUVECs. (‘Added’) However, when 1.25 ​× ​10⁴ HUVECs are added to a preaggregate of 1.25 ​× ​10⁴ hMSCs and micro-objects that was cultured for 1 day, the subsequent aggregation behavior seems to be compromised, even when one would normalize to the absolute number of seeded hMSCs. (B) 3 ​× ​10⁴ hMSCs were stained red and allowed to preaggregate with approximately 3000 cube-shaped micro-objects for 1 day before adding 3 ​× ​10⁴ green-stained HUVECs (final object-to-cell ratio: 1:20). It can be observed that the HUVECs form a layer around the hMSCs within 4 days of culture. Yet, this layer disintegrates after day 10, and the HUVECs show a decrease in viability and number. (C) Four green-stained 4-day-old aggregates of 1 ​× ​10⁴ hMSCs and approximately 500 donut-shaped micro-objects were complemented with 2.5 ​× ​10³ red-stained HUVECs per aggregate and transferred to a hemispherically shaped cell-repellent agarose layer one day later ​and cultured for another 7 days. It can be observed that the HUVECs stay aligned in between the preaggregates. (D) Confocal fluorescence image of the surface of the aggregate shown in (C). The scale bars represent 1 ​mm. The height of the scanned area in (D) is 345 ​μm. * indicates a p-value < 0.05, ** indicates a p-value < 0.01, and *** indicates a p-value < 0.001. hMSCs, human bone marrow stromal cells; HUVECs, human umbilical vein endothelial cells.