Engineered Matrices Enable the Culture of Human Patient-Derived Intestinal Organoids

Abstract

Human intestinal organoids from primary human tissues have the potential to revolutionize personalized medicine and preclinical gastrointestinal disease models. A tunable, fully defined, designer matrix, termed hyaluronan elastin-like protein (HELP) is reported, which enables the formation, differentiation, and passaging of adult primary tissue-derived, epithelial-only intestinal organoids. HELP enables the encapsulation of dissociated patient-derived cells, which then undergo proliferation and formation of enteroids, spherical structures with polarized internal lumens. After 12 rounds of passaging, enteroid growth in HELP materials is found to be statistically similar to that in animal-derived matrices. HELP materials also support the differentiation of human enteroids into mature intestinal cell subtypes. HELP matrices allow stiffness, stress relaxation rate, and integrin-ligand concentration to be independently and quantitatively specified, enabling fundamental studies of organoid-matrix interactions and potential patient-specific optimization. Organoid formation in HELP materials is most robust in gels with stiffer moduli (G' ≈ 1 kPa), slower stress relaxation rate (t_1/2 ≈ 18 h), and higher integrin ligand concentration (0.5 × 10^-3-1 × 10^-3 m RGD peptide). This material provides a promising in vitro model for further understanding intestinal development and disease in humans and a reproducible, biodegradable, minimal matrix with no animal-derived products or synthetic polyethylene glycol for potential clinical translation.

Keywords: 3D cell culture; adult stem cells; engineered biomaterial; extracellular matrix; intestinal organoid.

Figure 1

HELP matrix for patient‐derived intestinal organoid formation, growth, and passaging. a) Enteroids were generated from intestinal tissue biopsies from human patients. b) Schematic of HELP matrix, which is composed of benzaldehyde‐modified hyaluronan (HA) and hydrazine‐modified elastin‐like protein (ELP). Hyaluronan can engage the CD44 receptors on cells, while recombinant ELP contains an RGD peptide ligand that engages cell integrin receptors. c) Schematic of enteroid passaging techniques. Enteroids can either be dissociated into single cells or directly re‐embedded as fully formed enteroids into a new material at the time of passaging. d) Brightfield and confocal fluorescence micrographs of enteroids in different materials, when dissociated (left) and re‐embedded (right) in these materials. e) Representative brightfield images of enteroids grown on Passage 1 and Passage 5 in HELP, with dissociation into single cells at the time of each passage. f) Growth curves of dissociated enteroids grown from Passage 12 in EHS matrix or Passage 12 in HELP; at each timepoint, a two‐tailed Student's t‐test was conducted, n = 3, n.s. = not significant. g) Enteroid formation efficiency for enteroids grown from single cells in EHS matrix, HELP, or ELP only. One‐way ANOVA with Tukey's multiple comparisons testing; ** = p < 0.01, n = 3, n.d. = none detected. All data shown are mean ± SD.

Figure 2

Differentiation of organoids grown in HELP and EHS matrices. a) Top: Schematic of differentiation experiment timeline. Organoids are cultured from single cells in growth medium for 10 days, followed by 5 days in differentiation medium (see the Experimental Section). Bottom: Confocal micrographs illustrating the progression of organoids from early enteroid to polarized enteroid to differentiated organoid. b) Confocal micrographs demonstrating the observance (left) and absence (right) of intestinal stem cell marker Bmi1 during culture in growth and differentiation media, respectively. c) Confocal micrographs demonstrating the observance of mature intestinal cell subtypes: Paneth cells (Lyz⁺, left), goblet cells (Muc2⁺, middle), and enteroendocrine cells (ChgA⁺, right). d) RT‐qPCR quantification of changes in RNA expression of differentiated cell type markers, compared to cells that were maintained for 15 days in maintenance medium, and relative to control gene BACT. Under the assumption that C _T values are normally distributed, two‐tailed Student's t‐tests were performed on C _T values between differentiated versus maintenance cultures for each gene in each material, respectively. ** = p < 0.01, * = p < 0.05, N = 3 independent experiments, n = 4 technical replicates. Data shown are geometric mean ± geometric SD.

Figure 3

Role of hyaluronan in HELP matrices. a) Flow cytometric analysis of enteroids grown in EHS and HELP matrices 11 days post‐encapsulation for CD44‐positive cells, compared to negative controls, dashed line indicates flow gating. b) Quantification of percentages of CD44‐positive cells in EHS matrix and HELP across three independent experiments. c) Gene expression of CD44 for passage 20 enteroids cultured in HELP for different lengths of time and EHS matrix (all three conditions had 20 total passages, normalized to P20 in EHS). One‐way ANOVA with Tukey post‐hoc testing, n = 4, * = p < 0.05, *** = p < 0.001. d) Storage moduli for HELP and ELP‐PEG formulations as measured by oscillatory rheology. At least three measurements were performed for each material. Two‐tailed Student's t‐test, n.s. = not significant. e) Brightfield and confocal micrographs at 6 days post‐seeding of enteroids grown in HELP and ELP‐PEG matrices, where HA is absent in the ELP‐PEG gels. f) Formation efficiency between enteroids grown in HELP and ELP‐PEG. Two‐tailed Student's t‐test, ** = p < 0.01, n = 3, n.d. = none detected. Data shown are mean ± SD.

Figure 4

Custom tailoring of HELP matrix properties. a) HELP schematic illustrating specification of the RGD ligand concentration by blending different ELP variants in the material. b) HELP schematic illustrating that matrix stiffness is precisely modulated by changing the number of crosslinks (top), while matrix viscoelasticity is tuned by replacing benzaldehyde with aldehyde moieties. c) Shear storage moduli of Stiff Elastic (EL) and Soft EL HELP formulations. Student's t‐test, ** = p < 0.01, N = 3–5. d) Step‐strain–stress relaxation curves comparing EL formulations. e) Cross‐sectional area measurements of enteroids in EL materials at 12 days post‐encapsulation as single cells. Two‐way ANOVA with Tukey multiple comparisons testing, * = p < 0.05, ** = p < 0.01, n = 3. f) Shear storage moduli of Stiff EL and Stiff Viscoelastic (VE) formulations. Student's t‐test, N = 3–5, n.s. = not significant. g) Step‐strain–stress relaxation curves comparing Stiff EL and Stiff VE formulations. h) Cross‐sectional area measurements of enteroids in stiff materials at 12 days post‐encapsulation as single cells. Two‐way ANOVA with Tukey multiple comparisons testing, * = p < 0.05, ** = p < 0.01, n = 3. All data shown are mean ± SD.