Background

The intestinal mucosa is an active host-microorganism interaction interface, and when innate and humoral immune defenses mechanisms fail, infection and disease result. Understanding virulence mechanisms of pathogenic microorganisms, and their consequences, provides an alternative target for disease prevention and mitigation [1]. In vitro models provide advantages for studying virulence mechanisms, given the controlled environment and ease of sample collection. In vitro organ culture (IVOC) involves excision of a small volume of tissue from a healthy donor, and its subsequent culture in appropriate media in the laboratory for a number of days. IVOCs preserve important tissue characteristics, such as different cell types and tissue architecture, while providing a controlled and accessible environment to study host-pathogen interactions [2, 3]. To prevent acidification of the culture media and maintain tissue viability, intestinal IVOCs require a hyperoxic gas environment [4,5,6]. Sealed chambers are usually placed within incubators for culture, an expensive setup that requires extensive bench space and is potentially an explosion hazard due to the high oxygen content [6, 7]. In addition, this setup becomes a challenge when studying host interactions with anaerobic pathogens, due to the hyperoxic environment. Here, we describe and evaluate a 3D–printed device to allow the use of different gas mixtures in the culture media, compared to the mucosal aspect of the tissue.

Methods

3D–Device design, printing and assembling

A 3D modelling software (123D Design, AutoDesk Inc., CA) was used to design and develop the 3D–printed device in silico. The device is divided into two parts; one for tissue support and liquid media flow (bottom), and one for the gas media flow (top). To reversibly seal the top-bottom junction, a rubber O-ring of 14.5 mm diameter was used (Fig. 1). The design was exported as two STL files and printed at a commercial printing facility in Saskatoon, Saskatchewan (Fig. 1, Additional file 1). Printing was performed on a MakerBot Replicator Mini + printer (MakerBot Industries, NY), at 500 μ resolution, using an acrylonitrile butadiene styrene (ABS, MG Chemicals, BC) filament. Post-printing cleaning consisted of manually removing any non-used filaments and other printing debris. Due to the small size of the device, 3D–printing the four hollow tubing adapters was impractical. Thus, a 0.3969 mm (1/64″) drill bit was used to create an opening in each adapter, from the tip to the inside of the device. The rubber O-ring (4.4 cm diameter) was placed on the top part of the device to seal the top-bottom junction (Fig. 2).

Fig. 1
figure 1

Blueprint of the culture device depicting views from side (a), front (b), top (c)

Full size image
Fig. 2
figure 2

3D–printed culture device. The top is reversibly attached to the bottom using a latex O-ring. The device is shown assembled ready for use (a) and disassembled (b) showing the tissue support grid where explants are placed for culture

Full size image

Colon segment collection and explant preparation

Pigs (n = 2) were humanely euthanized at approximately 8 weeks of age for reasons unrelated to this study. Tissue culture procedures and culture media preparation have been previously described [5]. Once explants (0.5 cm × 0.5 cm, n = 3/device) were ready for culture, they were placed mucosa side up on top of a polystyrene foam (VWR, Mississauga, ON) on the designated grid within the 3D–printed device (Fig. 3. The foam was sterilized by fully submersing it in a 70% ethanol / 30% MilliQ water solution 24 h prior to explant culture, followed by 1 h drying period on a sterile surface within a Biosafety cabinet. Prior to tissue placement, the 3D–device was connected to two electric peristaltic pump (Fig. 4, Cole-Parmer, AO-73160-32, Chicago, IL) and flushed with a commercial disinfecting solution (Virkon, DuPont, Wilmington, DE) for 5 min. Next, sterile NaCl solution (0.85%) was used to wash the 3D device and attached tubing for another 5 min. All tubing used was standard sterile intravenous tubing (Lifeshield – primary set, Hospira, Lake Forest, IL). Finally, feeding pump was set at 10 mL/min, connected to the bottom part of the 3D–device (Fig. 4) and a reservoir containing tissue culture media (100 mL/device) supplemented with antibiotics (KBM-Gold Ca2+ and Phenol red free BulletKit (catalog number 195769), Lonza, Walkersville, MD). The second pump, the recover pump, was set at a higher rate to prevent the device from over flowing (15 mL/min). The liquid media was bubbled with medical grade mixture of 99% O2, 1% CO2 flowing from a compressed gas cylinder connected to a regulator, at a 3 L/min rate. For this experiment, the top gas inlets received atmospheric air. Both culture media reservoir and 3D–device were kept immersed in 2 cm high distilled water bath at 37 °C.

Fig. 3
figure 3

Culture device suggested setup. The top of the cap was intentionally removed to allow visualization of the culture grid. Suggested placement of liquid and gas media tubing is shown by coloured arrows

Full size image
Fig. 4
figure 4

Diagram of culture setup. Tissue culture media (red) is placed in a reservoir bubbled with 99% O2 gas. A feeding peristaltic pump is used to pull media out of the reservoir into the culture device bottom (B), bathing the explant (orange form). A vertical wall in the bottom portion of the chamber maintains the depth of the liquid media at the tissue support grid A recovery pump is used to pull the media out of the device back into the media reservoir. A gas mixture of choice (green) flows with low pressure into the top part of the device (A) from a compressed source

Full size image

Explant fixation and analysis of H&E stained sections

To assess tissue viability, explants were placed in 10% buffered formalin for tissue fixation at 0 h (n = 2), 8 h (n = 6), 18 h (n = 6) and 24 h (n = 6). Within 48 h, the explants were paraffin embedded and stained using Haematoxylin and Eosin (H&E), as per routine protocols. Histological analysis of explant morphology was carried out on H&E stained sections by an observer (MC) blinded to the identity of the samples using optical microscopy at 20× magnification. The presence and quality of superficial epithelium, crypts, goblet and necrotic/apoptotic cells within crypts were recorded.

Results and discussion

This study describes a low-cost, 3D–printed device used for intestinal explant culture without the need of a hyperoxic chamber or incubator. 3D–printing required approximately 2 h including manual removal of leftover filament and printing debris. No leakage was observed after 24 h of continuous liquid media circulation through the circuit, and the media was maintained constantly at 37 °C.

Porcine colon explants sampled at 0 h, immediately before placement on the culture device, did not show any histological abnormalities. A total of 66% (4/6) of samples collected at 8 h displayed normal architecture, with the presence of crypts, goblet cells and normal epithelial turnover. Sloughing off the superficial epithelial layer increased at 18 h (50%, 3/6) of samples) and 24 h (66%, 4/6), possibly as a response to the culture environment. At both latter time points, a catarrhal exudate accumulated on the apical side of the explants. After 24 h in culture, explants also displayed distension of glands, perhaps due to the lack of clearance of accumulated mucus from the apical surface. Representative H&E stained explant sections are shown on Fig. 5.

Fig. 5
figure 5

H&E stained sections of porcine colon explants cultured using the 3D–printed device. Overall, normal tissue architecture is observed until 24 h after initial culture setup, with progressively increasing sloughing off the superficial epithelial (yellow arrow) layer beginning at 18 h. Distension of crypt glands is observed at 24 h (green arrow)

Full size image

Although explants were only kept viable for a maximum of 24 h in this experiment, this period is suitable for the majority of studies focused at pathogen-host interactions [8,9,10]. In addition, the ability to manipulate the gas media that explants are exposed can benefit studies aimed at intestinal physiology [11] and nutrition [12] by simulating more closely the luminal gas environment normally observed in vivo. Although providing continual circulation, this setup uses considerably more culture media compared to a 6-well culture plate setup (3 mL vs 100 mL) to continuously provide explants with nutrients and with metabolite clearance by dilution. If culture takes place outside a biosafety cabinet, the setup is prone to contamination.

Conclusions

The design described here consistently supported porcine colon explants for up to 8 h with minimal disruption of tissue viability. It also provides an alternative culture method that does not require the use of a hyperoxic chamber, and is particularly suited for situations where a gas environment other than atmospheric is needed on the mucosal aspect of the explant.