This article and the accompanying video present our protocol for generating tissue-engineered intestine in the mouse, using an organoid units-on-scaffold approach.
Tissue-engineered small intestine (TESI) has successfully been used to rescue Lewis rats after massive small bowel resection, resulting in return to preoperative weights within 40 days.1 In humans, massive small bowel resection can result in short bowel syndrome, a functional malabsorptive state that confers significant morbidity, mortality, and healthcare costs including parenteral nutrition dependence, liver failure and cirrhosis, and the need for multivisceral organ transplantation.2 In this paper, we describe and document our protocol for creating tissue-engineered intestine in a mouse model with a multicellular organoid units-on-scaffold approach. Organoid units are multicellular aggregates derived from the intestine that contain both mucosal and mesenchymal elements,3 the relationship between which preserves the intestinal stem cell niche.4 In ongoing and future research, the transition of our technique into the mouse will allow for investigation of the processes involved during TESI formation by utilizing the transgenic tools available in this species.5The availability of immunocompromised mouse strains will also permit us to apply the technique to human intestinal tissue and optimize the formation of human TESI as a mouse xenograft before its transition into humans. Our method employs good manufacturing practice (GMP) reagents and materials that have already been approved for use in human patients, and therefore offers a significant advantage over approaches that rely upon decellularized animal tissues. The ultimate goal of this method is its translation to humans as a regenerative medicine therapeutic strategy for short bowel syndrome.
1. Organoid Units Preparation
2. Loading of Polyglycolic Acid Scaffold
3. Implantation into Host Mouse
4. Harvest
Figure 1 shows an overall schema for the protocol documented here. The end result of this protocol is a globe or spherical structure of tissue-engineered murine intestine with a lumen, mucosa, submucosa, and surrounding muscularis. Figure 2A shows a typical globe in comparison to a starting polymer scaffold. Figure 2B displays the same construct sharply bivalved to reveal its lumen. Figure 3 demonstrates a hematoxylin/eosin-stained paraffin-mounted cross section of a typical successful construct after 4 weeks of incubation. In Ref. 4, we were able to produce tissue-engineered intestine successfully 89% of the time (39 out of 44 implants).
An unsuccessful construct is one in which no mucosa forms. It is impossible to judge based upon gross appearance at harvest whether the globe will have mucosa on final histologic analysis. Therefore if biochemical assays are performed on the fresh construct tissue, half of each globe should be fixed and paraffin mounted for histologic analysis to confirm that mucosa is present in each sample. An unsuccessful construct will demonstrate only stroma and fibrosis on hematoxylin/eosin staining.
Figure 1. Schema for the production of tissue-engineered intestine in the mouse. In brief, donor tissue is harvested and processed into organoid units. The organoid units are loaded onto a porous polyglycolic acid scaffold, which is then implanted into a host and allowed to incubate for 4 weeks. The engineered construct is then retrieved and can be characterized via histology or biochemical assays.
Figure 2. Example of tissue-engineered intestine construct harvested at 4 weeks. A: Construct in comparison to starting polymer scaffold. B: The same construct, bivalved to reveal its lumen.
Figure 3. Low-power hematoxylin/eosin micrograph of a typical successful tissue-engineered intestine construct. The labels and arrows indicate the lumen with intestinal mucosa, and adherent host pancreas.
We present a protocol for producing tissue-engineered intestine in the mouse using an organoid units-on-scaffold approach. The most critical steps are those of the organoid units preparation. Care must be taken to adequately clean and mechanically process the tissue, but equal care must be taken not to overdigest or overtriturate the organoid units after the digestion is performed (step 1.11). If this is done, the organoid units can be reduced to single cells, which can be lost in the supernatant of step 1.12, and are unlikely to survive to produce tissue-engineered intestine, as the intestinal stem cell niche would not be preserved in this case.
The small size of the mouse makes it challenging to directly anastomose the tissue-engineered construct to the host animal’s intestine to test its function in vivo. Alternatively, the rat represents a larger model for TESI growth that facilitates in vivo anastomosis and has already been performed for this reason1. Nevertheless, the size of the mouse is not an insuperable obstacle, as others have been able to perform small bowel resection and anastomosis,6 or anorectal surgery,7 in these animals. To address these issues, experiments to characterize the function of our tissue-engineered intestinal constructs in an in vitro fashion are ongoing. Additional limitations to this technique include an 89% success rate in the generation of TESI4. That is, 89% of OU loaded scaffolds will successfully generate TESI. The application of this technique by others may help refine and improve this technique increasing the overall yield to 100%. In addition, this technique could be improved if the total tissue mass generated could be increased. Currently, flow cytometry has demonstrated that the total number of cells present in the harvested TESI construct is 3-fold greater than the number present at implantation (3.07 x 106 ± 0.5 x 106)4. Improving the volume of TESI production is an important step toward translation to therapy.
We note also that the mouse intestine, being very thin-walled and of small caliber, is particularly easy to process in comparison to that of other species such as swine or the rat. In humans, we expect that some of the details of the organoid units preparation steps would have to be modified to both adequately digest the tissue and avoid overdigestion to single cells, in particular the concentrations of dispase and collagenase in the digestion solution and the time of digestion at 37 °C.
The ultimate goal of this technique is a transition to human therapy. Tissue engineering of human intestine from the patient’s own tissue could potentially offer a durable, long-term cure for short bowel syndrome (SBS) with none of the drawbacks of existing therapies. Short bowel syndrome is a morbid condition caused by resection of a significant fraction of the total length of the small bowel, usually greater than 50-75 percent, such that its absorptive capacity is severely reduced and the patient cannot obtain sufficient nourishment from enteral nutrition.2 In children, the most common causes of SBS are massive small bowel resection secondary to necrotizing enterocolitis or malrotation with midgut volvulus.8,9 Also, though less common, SBS can occur in adults because of multiple resections in the setting of Crohn’s disease, or with mesenteric ischemia secondary to vascular disease.10,11 Because SBS patients cannot maintain sufficient nutrition with enteral intake, they may require long-term total parenteral nutrition, which itself can be complicated in children by liver failure and cirrhosis.12 SBS patients therefore endure significant healthcare costs, recently estimated to be on the order of $1.6 million per patient over 5 years.13 The current standard of care for intestinal failure secondary to SBS is intestinal, liver/intestinal, or other multivisceral transplantation, but this confers only a 60% 5-year survival and consigns the patient to a lifelong course of immunosuppressive therapy.14 Further, limited donor organ availability results in an inevitable mismatch in demand and supply and long wait times.15 Therefore, tissue engineering from the patient’s autologous tissue would be an attractive alternative.
The authors have nothing to disclose.
Tracy C. Grikscheit, Erik R. Barthel, and Frédéric G. Sala are supported by the California Institute for Regenerative Medicine (CIRM), grant numbers RN2-00946-1 (TCG) and TG2-01168 (ERB, FGS). Allison L. Speer is a Society of University Surgeons Ethicon scholar. Yashuhiro Torashima is funded by a Children’s Hospital Los Angeles Saban Institute Research Career Development Fellowship.
Name of the reagent | Company | Catalogue number | Comments (optional) |
HBSS | Gibco | 114170-112 | |
Antibiotic-Antimycotic 100X | Invitrogen | 15240-062 | |
Dispase | Gibco | 17105-041 | |
Collagenase Type 1 | Worthington | LS004194 | |
DMEM High Glucose 1X | Gibco | 11995-065 | |
Heat inactivated FBS | Invitrogen | 16140-071 | |
Biofelt 100% PGA | Concordia Medical | FELT01-1005 | For polymer preparation as in Ref. 4 |
Poly-L-lactic acid | Durect | B6002-1 | For polymer preparation as in Ref. 4 |
Type I Collagen, rat tail | Sigma-Aldrich | C3867-1VL | For polymer preparation as in Ref. 4 |
Ketoprofen 100 mg/ml | Fort Dodge Animal Health | 71-KETOI-100-50 | |
LabDiet 5001 rodent chow | LabDiet | 5001 | |
Septra 200 mg / 40 mg per 5 ml, USP | Hi-Tech Pharmacal | 50383-824-16 | |
Isoflurane, USP | Phoenix Pharmaceuticals | 57319-507-06 |