Intestinal crypt-derived enteroids from various LF/SC animals grow and expand to high passage

We began working relationships with local animal processing and shelter facilities to obtain tissue samples from the terminal ileum of the small intestine from a variety of LF/SC animals, including: Felis catus (cat), Canis familiaris (dog), Bos taurus (cow), Equus caballus (horse), Sus domesticus (pig), Ovis aries (sheep) and Gallus gallus (chicken – cecum) (Table 1). Multiple independent animals were sampled for each species. After isolation, crypts were cultured in either 50% L-WRN CM or media lacking the WRN factors (D10). The majority of crypt samples cultured in 50% L-WRN CM expanded and grew for several passages and lines could be maintained in 50% L-WRN CM to high passage number over long periods of time (Table 1), whereas none of the isolated crypts cultured in D10 grew to the first passage (Fig. 2A).

Table 1.

Summary of isolation and maintenance of LF/SC organoids

Open in a separate window

Open in a separate window

Fig. 2.

50% L-WRN CM is required for LF/SC enteroid growth and expansion. (A) Enteroids from LF/SC animals cultured with 50% L-WRN CM expand soon after crypt isolation (Day 1), growing to large size after passage (Day 6), and continue to expand and obtain crypt-like morphology at higher passage (white arrows) (P5 and P>10). Lower magnification shows enteroids were maintained at high density throughout passage history (P>10,500 µm). Enteroids fail to grow when cultured in D10 after isolation (Day 6) or when passed to 50% L CM after high passage (P>10) growth in 50% L-WRN CM, no data (N/D). (B) Fold expansion of wells containing LF/SC enteroids. LF/SC enteroids grown in 24-well plates were maintained at high density. The ratio of per well expansion was recorded at each passage throughout the passage history. (C) Number of enteroids expands with each passage. A total of 200 high passage (P>10) enteroids were seeded to two wells and the number of enteroids was counted after 3-4 days of growth, and again after subsequent passage splitting 1:2, n=3. Cat samples are not included in B or C due to the lower passage and density of enteroids. Error bars indicate ±s.d.

Small enteroids could be observed in 50% L-WRN CM within the first day after isolation and increased in number and size by the end of the first week (Fig. 2A). Enteroids from all LF/SC animals expanded and many began to obtain crypt-like folds with subsequent passages (Fig. 2A, P5 and P>10). High passage enteroid lines were maintained at high density (Fig. 2A, P>10,500 µm). Like the mouse enteroids, LF/SC enteroids failed to grow after isolation in D10 media or after passage to 50% L CM when initially grown in 50% L-WRN CM to high passage (Fig. 2A), demonstrating the requirement for the WRN factors to stimulate enteroid growth as D10 and 50% L CM lack the WRN factors. Additionally, LF/SC enteroids can be cryopreserved, thawed, and expanded, providing a banked resource for continued experimentation (Fig. S2). The LF/SC enteroid passage history demonstrates the growth potential of these lines, where the cumulative fold expansion increased over time as enteroids were passed from well to well at high density (Fig. 2B). Likewise, the number of enteroids increased for high passage LF/SC lines when they were passed at a ratio of 1:2 over two passages (Fig. 2C).

Although we have been able to grow one cat intestinal enteroid line to relatively high passage (P18) (Table 1), and have demonstrated the requirement of 50% L-WRN CM for cat enteroid growth (Fig. 2A), cat enteroids cease to expand around P10 and begin to growth arrest around P13-P18 (Fig. 3). We are not certain why this limited expansion is restricted to the cat samples. Enteroids from all other LF/SC species tested here continue to grow and can be maintained at high density well past P10. We did observe a mesenchymal-like cell type growing with the cat enteroids (Fig. 2A, P5) which wane in number around P7-P9 before the enteroids begin to growth arrest, suggesting these cell types are needed for cat enteroid growth. Indeed, culture systems for intestinal cell growth have been developed based on a requirement of co-culturing mesenchymal/stromal components with intestinal crypts (Lahar et al., 2011; Li et al., 2016; Ootani et al., 2009; Spence et al., 2011). In an attempt to rescue the cat enteroids, we supplemented 50% L-WRN CM with various factors that have been shown to induce enteroid growth in other platforms or to be secreted by mesenchymal/stromal cells within the crypt niche (Fig. 3) (Date and Sato, 2015; Roulis and Flavell, 2016). These factors included: ALK5 inhibitor A83-01 and p38 MAP kinase inhibitor SB202190 (Khalil et al., 2016; Sato et al., 2011a), human FGF-4 (Clevers, 2016; McCracken et al., 2014; Spence et al., 2011), human FGF-2 (FGF-basic) (Clevers, 2016), human FGF-10 and nicotinamide (Barker et al., 2010; Boj et al., 2015; Sato and Clevers, 2015), human IGF-I (Reynolds et al., 2014; Simmons et al., 2007), Prostaglandin E2 (PGE2) (Goessling et al., 2009; Jung et al., 2011; Pierzchalska et al., 2012), or mouse Wnt-2b and human Gremlin (Aoki et al., 2016; Stzepourginski et al., 2017). None of these conditions rescued the cat enteroids to significantly improve growth beyond that in 50% L-WRN CM alone (Fig. 3). We continue to test other conditions in an attempt to grow the cat enteroids to higher passage number.

Open in a separate window

Fig. 3.

Supplementation of 50% L-WRN CM with crypt niche factors. Cat enteroids were passed to multiple wells at P12 and cultured with 50% L-WRN CM or with additional crypt niche factor(s) shown to modulate organoid growth in various platforms. The individual cultures were passed every 3-4 days and monitored for growth. Cultures were terminated when the number and growth of the enteroids dissipated (N/A). None of the crypt niche factors rescued the cat enteroids much beyond those grown with 50% L-WRN CM.

Animals

All Louisiana State University (LSU) Institutional Animal Care and Use Committee (IACUC) protocols were followed for mouse tissue collection. All LSU IACUC guidelines were followed for intestinal tissue collection from LF/SC animal carcasses at local processing and shelter facilities, such as not handling live animals or in any way being involved in the euthanasia schedule.

Protein alignments

The human and mouse protein sequence of Wnt3a, R-spondin, and Noggin were used find the LF/SC orthologs using NCBI HomoloGene (https://www.ncbi.nlm.nih.gov/homologene) and/or NCBI BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) against the appropriate genome. The following protein sequences were obtained from NCBI and used to make alignments in UGENE (Okonechnikov et al., 2012) using the Kalign algorithm (Edgar, 2004); Wnt3a: human (NP_149122.1), mouse (NP_033548.1), cat (XP_003980679.1), dog (XP_539327.3), cow (XP_015327344.1), horse (XP_014595070.1), pig (XP_003123669.1), goat (XP_013820403.1), sheep (XP_012026443.1), chicken (NP_001165072.1); R-spondin-3: human (NP_116173.2), mouse (NP_082627.3), cat (XP_003986583.1), dog (XP_005615677.2), cow (NP_001069502.1), horse (ABV31708.1), pig (XP_001926731.4), goat (XP_005684536.1), sheep (XP_004011185.1), chicken (XP_004940314.1); Noggin: human (NP_005441.1), mouse (NP_032737.1), cat (XP_011278221.1 and XP_011278233.1), dog (BAJ24018.1), cow (XP_002695600.1), horse (AAM34779.1), pig (NP_001137163.1), goat (XP_013827246.1), sheep (NP_001167581.1), chicken (AAC83570.1).

Crypt isolation and culturing of enteroids

Murine intestinal crypts were harvested from a 6 week old female BALB/c mouse to test the viability of the 50% L-WRN CM made from the L-WRN cell line. To obtain LF/SC intestine samples, we established working relationships with local animal processing and shelter facilities that routinely euthanize animals. Samples were taken from the terminal ileum of mammal and the cecum of chicken carcasses. Sampling was conducted within 30 min after euthanasia and samples were transported in phosphate buffered saline (PBS) on ice. Several 1 cm² pieces of the intestine were transferred to cryovials and flash frozen in LN₂ to bank control tissue for use in qPCR. Harvesting and isolation of crypts followed previously described protocols (Miyoshi and Stappenbeck, 2013) with minor modification. Briefly, tissue was cut longitudinally, trimmed to 1 cm² pieces, rinsed with PBS, placed in D10, villi were removed by scraping, tissue was minced, digested with collagenase (Sigma-Aldrich, St. Louis, MO, USA, Cat # C0130) at 37°C for 1-1.5 h while vortexing every 10-15 min, crypts were filtered with a 70 µm cell strainer, rinsed with D10 three times, and after final centrifugation suspended in 20 µl Matrigel matrix (Corning, Corning, NY, USA, Cat # 354234) per well and seeded to a 24-well plate. Crypts were then cultured and enteroids maintained in 50% L-WRN CM with 10 µM Y-27632 ROCK inhibitor (BioVision, Milpitas, CA, USA, Cat # 1596) and 10 µM SB-431542 TGF-β inhibitor (BioVision, Cat # 1674). High passage enteroid lines were confirmed by sequencing of the GAPDH gene and were monitored for contamination.

Maintenance of enteroids

Enteroids were passed every 3-4 days, as previously described (Miyoshi and Stappenbeck, 2013) with minor modification. Briefly, Matrigel was washed with PBS, digested with 2.5× trypsin at 37°C for 5 min, enteroids were rinsed in D10, incubated in D10 with 10 µg/ml muramyl dipeptide (Sigma-Aldrich, Cat # A9519) for 10 min at room temperature which has been shown to increase LGR5⁺ stem cell growth (Nigro et al., 2014), then centrifuged (200 rcf), suspended in 20 µl Matrigel per well, seeded to a 24-well plate and grown in 50% L-WRN CM with 10 µM Y-27632 and 10 µM SB-431542.

Additional factors added to 50% L-WRN CM with 10 µM Y-27632 and 10 µM SB-431542 in an attempt to rescue the cat enteroids include: A83-01 (10 µM; BioVision, Cat # 1725-1), SB202190 (10 µM; AdipoGen, San Diego, CA, USA, Cat # AGCR10028M001), human FGF-4 (500 ng/ml; Peprotech, Rocky Hill, NJ, USA, Cat # 100-31), human FGF-2 (FGF-basic) (100 ng/ml; Peprotech, Cat # 100-18B), human FGF-10 (100 ng/ml; Peprotech, Cat # 100-26), nicotinamide (100 µM; Alfa Aesar, Tewksbury, MA, USA, Cat # A15970-22), human IGF-I (100 ng/ml; Peprotech, Cat # 100-11), PGE2 (5 µg/ml; TCI America, Portland, OR, USA, Cat # P1884-1MG), mouse Wnt-2b (100 ng/ml; R&D Systems, Minneapolis, MN, USA, Cat # 3900-WN/CF), human Gremlin (200 ng/ml; Peprotech, Cat # 120-42).

To cryo-preserve the cell lines, Matrigel with enteroids was scrapped with a pipette tip, centrifuged (200 rcf), suspended in D10 with 50% FBS, then mixed with an equal amount of D10 with 20% DMSO, and transferred to a cryovial. Vials were placed in a CoolCell container (Corning) at −80°C for 24 h then transferred to LN₂ storage.

Protein alignments

Protein sequences were aligned in UGENE (Okonechnikov et al., 2012). Details and accession numbers are provided in the Supplementary Materials.

Microscopy

Brightfield images of enteroid in Matrigel were taken with an Olympus IX71 microscope, Olympus U-CMAD3 camera (10×) or a Leica DMIRB microscope, Olympus DP80 camera (2.5×) using the cellSens software (Olympus).

Enteroid counts

Fold expansion of enteroids over time was determined by recording the ratio at which wells were split at each passage throughout the passage history of that cell line, for example two wells of enteroids passed to four new wells (1:2).

To determine enteroid expansion during passage, 200 high passage (P>10) enteroids were seeded in Matrigel matrix to two wells (P1). After 3-4 days of growth the enteroids were counted and passed (P2), split 1:2. Again, after 3-4 days of growth the enteroids were counted.

RNA isolation and qPCR

RNA from enteroids, control tissue, and kidney epithelial cells (CRFK and MDBK) was isolated using the RNeasy Plus mini kit (Qiagen) according to the manufacturer's protocol, with DNase treatment (Qiagen) on the tissue controls samples. The High Capacity RNA-to-cDNA kit (Thermo Fisher 4387406) was used to obtain cDNA from 500 ng of RNA. qPCR negative controls omitted the RT enzyme. qPCR was performed with the PowerUp™ SYBR™ Green Master Mix (Thermo Fisher A25742) with a final primer concentration of 0.5 µM (Table S1) and 10 ng of cDNA, and run on an Applied Biosystems (ABI) 7900HT Sequence Detection System. Data are the average of three independent cDNA samples and technical replicates conducted for each sample per qPCR run. Data were normalized to sample GAPDH values, control tissue values were set to 1 and enteroid values expressed as a ratio to control tissue using the 2-ΔΔCt method (Livak and Schmittgen, 2001). Raw data for three independent cDNA samples and their technical replicates were analyzed in ExpressionSuite Software v 1.0.3 (Thermo Fisher). Biological groups were designated by sample type, with control tissue selected as the biological reference group, and data were normalized using GAPDH as an endogenous control. One standard deviation was used to determine the RQ minimum and maximum (error bars).

We thank the following animal processing and shelter facilities for allowing us to obtain tissue samples from animal carcasses: Roucher's Meat Market, Plaquemine, LA, USA; Iberville Parish Shelter and Animal Control, Plaquemine, LA, USA; Sanderson Farms, Hammond, LA, USA; LSU School of Veterinary Medicine Equine Health Studies, Baton Rouge, LA, USA. We would also like to thank Thaya Stoufflet of the LSU CEIDR Gene Lab Core and Pete Mottram of the LSU PBS Microscopy Core for their expertise and training.