Protocols for Analyzing the Role of Paneth Cells in Regenerating the Murine Intestine using Conditional Cre-lox Mouse Models

The epithelial surface of the mammalian intestine is a dynamic tissue that renews every 3 - 7 days. Understanding this renewal process identified a population of rapidly cycling intestinal stem cells (ISCs) characterized by their expression of the Lgr5 gene. These are supported by a quiescent stem cell population, marked by Bmi-1 expression, capable of replacing them in the event of injury. Investigating the interactions between these populations is crucial to understanding their roles in disease and cancer. The ISCs exist within crypts on the intestinal surface, these niches support the ISC in replenishing the epithelia. The interaction between active and quiescent ISCs likely involves other differentiated cells within the niche, as it has previously been demonstrated that the ‘‘stemness’’ of the Lgr5 ISC is closely tied to the presence of their neighboring Paneth cells. Using conditional cre-lox mouse models we tested the effect of deleting the majority of active ISCs in the presence or absence of the Paneth cells. Here we describe the techniques and analysis undertaken to characterize the intestine and demonstrate that the Paneth cells play a crucial role within the ISC niche in aiding recovery following substantial insult.


Introduction
The luminal surface of the mammalian intestine features repeating units of crypts and finger like projections, termed villi, which protrude into the lumen. This surface is a continuous sheet of epithelia which undergoes complete self-renewal approximately every 3 -4 days 1 . This dynamic tissue is supported by a population of rapidly cycling stem cells (ISCs; also known as crypt base columnar cells), which were initially identified by their expression of the Lgr5 gene 2,3 . These cells exist in a specialized niche at the bottom of the crypts of Lieberkuhn. Initially, the discovery that ISCs were rapidly cycling was discordant with the prevailing idea that a stem cell was quiescent in nature. Previous to the identification of the Lgr5 + ISC it was postulated that a population of quiescent label retaining cells at the +4 position, relative to the base of the crypt, were the ISCs 1 . Recent research has now reconciled these observations by demonstrating that primarily there is a pool of equipotent cycling ISCs in each crypt whose fate are regulated by its neighbors 4,5 . In the event they are lost these can be replaced by quiescent cells that ordinarily are committed to the secretory lineage but can revert to ISCs if the ISC population is damaged 6 . ISC neighbors can either be ISCs or their daughter cells. The ISCs produce naïve daughter cells which multiply and differentiate into the specialized cell types that comprise the epithelial sheet which lines the intestinal lumen 1 . The goblet, enteroendocrine, enterocytes, tuft and M cells migrate upwards to the luminal surface where they provide various absorptive and regulatory functions, however, the Paneth cells remain at the bottom of the crypt where they exist intermingled with the ISCs. In recent years it has been demonstrated that a proportion of the naïve daughter cells destined for a secretory lineage are quiescent label retaining Lgr5 lo cells capable of reverting to an ISC upon injury 6,7 .
Due to its importance in crypt regeneration a priority was placed on understanding the interactions between the ISCs and its neighbors, particularly the Paneth cells. The Paneth cells play a crucial role in the niche which supports the ISCs 8 . In addition to bactericidal products the Paneth cells produce signaling molecules that activate the pathways which govern ISC renewal or differentiation. Previous studies showed that the Lgr5 + ISCs could only exist when they could compete for essential niche signals provided by their daughter Paneth cells 8 . These studies examined the role of Paneth cells on normal Lgr5 + ISCs and not in a situation where they are damaged and require replenishment from an Lgr5 lo population.
To understand intestinal biology and model disease we examine the functional role of cells and/or genes using transgenic mouse models 9,10 . Frequently these models utilize cre-lox technology to conditionally modify gene(s) 9,10 . The study by Fevr et al. 12 , demonstrated loss of stem cells and intestinal homeostasis. Whereas the Ireland et al. 14 study reported that following a reduction in cell viability the crypt-villus axis was repopulated from wild type cells expressing CatnB. The major difference in these studies was the promoter used to express Cre in the intestinal epithelia. The Fevr et al., study used the villin gene promoter linked to the estrogen receptor which can be activated by administering tamoxifen (vil-Cre-ER T2 ) 15,16 . In contrast Ireland et al., utilized the promoter element of the rat cytochrome P450A1 (CYP1A1) gene to drive Cre expression in response to the xenobiotic β-naphthoflavone (Ah-cre). The characteristics of these different systems generated two hypotheses to account for these different observations. The first that CatnB is more efficiently deleted in the ISC using the vil-Cre-ER T2 system compared to the Ah-cre, thereby reducing the number of ISCs to sub-repopulation levels. Alternatively it was due to differential CatnB deletion in the differentiated cell population. The vil-Cre-ER T2 system targets all epithelial cells of the crypt and villus whereas the Ah-cre system only targets the non-Paneth cells of the ISC niche and crypt. These systems provided ideal tools for examining the behavior of the ISCs and their interaction with the Paneth cells. Here we present several detailed protocols based on how we used these systems to determine that Paneth cells play a crucial role in mediating the intestinal response to injury 17 .

Protocol
Information on all material used is given in . For visual analysis of recombination, via a β-gal stain, cohorts should contain the Rosa26R-lacZ reporter 17 . Cohorts should control for the presence of modified genes and use of induction agents, the size of cohorts required should be estimated using a power analysis. 2. To induce the Ah-cre transgene prepare β-Naphthoflavone (BNF), or for the vil-Cre-ER T2 transgene tamoxifen (TAM) in corn oil to give a working solution of 10 mg/ml. NOTE: The agents should be weighed out in a fume hood using appropriate personal protection. 3. Heat solutions in an amber bottle (BNF is light sensitive) to either 99.9 °C for BNF or 80 °C for TAM in a water bath. 4. Transfer to a heated stirrer set at 100 °C for BNF, or 80 °C for TAM, and stir for 10 min. 5. For BNF repeat 1.3-1.4 until dissolved (can take >1 hr). 6. Aliquot into small amber bottles (~5 ml) and then freeze at -20 °C. Bottles should be discarded after 3 freeze/thaw cycles. Prior to use thaw agents and reheat to appropriate temperature if they have fallen out of solution. Allow to cool to <37 °C prior to injection. 7. Inject the mice intraperitoneally (I.P.) with a dose of 80 mg/kg, for example a 25 g mouse receives 0.2 ml of the appropriate induction agent.
For Ah-cre deliver three injections in 24 hr period, for vil-Cre-ER T2 give one injection per day for 4 days.

Dissection of Intestine for Reporter Visualization and Immunohistochemistry (IHC)
1. Prepare wax plates by combining molten ralwax with mineral oil at 10:1. Pour into 15 cm petri dishes and leave to cool. Prepare X-gal fixative as per Table 1 and store on ice. 2. Remove whole intestine and flush through with ice-cold 1x PBS, as per section 2. Fix intestine by flushing with 25 ml ice-cold X-gal fixative. 3. Using a scissor cut the intestine into 3 -5 equal sections (maximum 5 per plate). Place each section onto wax plate and pin down each end so the section is slightly stretched with the mesenteric line uppermost; trim any excess mesentery. Using a springbow scissors cut the gut longitudinally and pin out along the way. 4. Flood the plate with X-gal fixative to cover the sections and leave for at least 1 hr at 4 °C. Remove X-gal fixative using a 25 ml pipette and wash once with 30 ml of 1x PBS. Cover sections with 30 ml of DTT demucifying solution for 30 -60 min at RT, ideally on a rocking platform. 5. Remove demucifiying solution using a 25 ml pipette and flood the plate with 30 ml of 1x PBS. Using a pasteur pipette wash the intestine sections with the 1x PBS in the plate to remove mucus. 6. Remove 1x PBS with a 25 ml pipette and flood with 30 ml of X-gal stain. Incubate overnight at RT in the dark with gentle agitation on a rocking platform. 7. Following the overnight incubation check that the sections have developed a blue/green stain, if the background color is still white, fresh staining solution can be added and monitored till staining develops. NOTE: Once sections have stained then no further staining can be attempted. 8. Remove the X-gal stain using a 25 ml pipette and flood plate with 30 ml of 1x PBS and leave for 3 min with gentle agitation. Remove pins and pick up, with forceps, the end of an intestine section. Wind the intestine around the forceps to form a "Swiss roll", secure the roll by slightly opening the forceps and putting a 25 G needle through it. 9. Place tissue in a wide mouthed flat bottomed container containing a large excess of neutral buffered formalin fixative, at least 10x the volume of fixative to the volume of tissue. Place samples at 4 °C for at least 24 hr prior to embedding and sectioning.

Extraction of Crypts from Intestine
1. Isolate the first 20 cm of the small intestine, as per section 2. Place the intestine on a clean dissecting surface and using a forceps and scissors remove any attached fat/mesentery. Using a springbow scissors open the gut longitudinally. 2. Using a microscope cover slide, firmly scrape the gut lumen to remove villi and mucus. Using a scissors cut the intestine into ~5 mm pieces and transfer into a 50 ml tube with 25 ml 1x HBSS supplemented with penicillin (100 U/ml) and streptomycin (100 U/ml). 3. Incubate for 10 min at RT. Remove the antibiotic containing media by passing the samples through a 70 µm cell strainer. 4. Place intestinal sections into a fresh 50 ml tube containing 10 ml 1x HBSS and replace the cap. 5. Gently invert twice and remove the 1x HBSS by passing through a 70 µm cell strainer. Further wash the intestinal pieces by repeating 6.4 & 6.5 three times and ensure that the final fraction is relatively clear. 6. Transfer tissue to a fresh 50 ml tube containing 10 ml of EDTA (8 mM)/1x HBSS and leave at RT for 5 min. Shake vigorously (20 -30x) or vortex, pass through a 70 µm cell strainer and transfer tissue pieces to a fresh 50 ml tube containing EDTA (8 mM)/1x HBSS. NOTE: The flow through can either be discarded or retained if analysis of villi epithelia is required. 7. Incubate the tissue pieces on ice for 30 min, shake the sample vigorously (20 -30x) or vortex. Pass the samples through a 70µm cell strainer and retain the flow through as this contains the crypts. 8. Transfer the tissue pieces to a fresh 50 ml tube containing 10 ml of 1x HBSS. Shake vigorously (20 -30x) or vortex, pass through a 70 µm cell strainer and retain the flow through. Repeat one more time to ensure maximum recovery of crypts from intestinal pieces. Combine the flow through fractions and centrifuge at 300 x g for 5 min. Pour off the supernatant and retain the crypt pellet. NOTE: The pellets can be used immediately for culturing (if applicable) or stored at -80 °C prior to standard DNA/ RNA/protein extraction procedures.

Standard Immunohistochemical Visualization
1. Cut 5 µm sections of paraffin embedded tissue onto poly-L-lysine (PLL) slides. Note: A standard protocol for staining with a B-catenin antibody is given below, parameters for other antibodies are given in Table 2. 2. De-wax with 2x 3 min washes in slide baths containing fresh xylene. Rehydrate by passing slides for 3 min through slide baths containing fresh: 100% EtOH (2x), 95% EtOH and 70% EtOH and finally into 1x PBS. 3. Place slides into a slide bath containing citrate buffer (pH 6) and heat 99.9°C for 20 min to retrieve antigens. Allow slides to cool and then wash 3x 5 min in a slide bath containing 1xTBS/T for 5 min. Remove slides from last wash, immediately draw around tissue with a PAP pen, and cover section with a commercial peroxidase block or 1.5% H 2 O 2 (in distilled H 2 O). 4. Incubate for 20 min at room temperature (RT) then wash 3x in slide baths containing fresh 1x TBS/T for 5 min. After washing cover sections, using a pipette, in 5% Normal Rabbit Serum (NRS)/1xTBS/T for 30 min at RT to block non-specific hydrophobic binding of your primary antibody. 1. Remove NRS block with a Pasteur pipette and cover section in B-catenin primary antibody diluted 1:200 with 5% NRS. Wash slides for 3x 5 min in slide baths containing fresh 1xTBS/T. Note -other antibodies will require optimization for dilution and specificity. To ensure specificity of an antibody, appropriate no antibody and isotype control stains should be performed. An isotype control is matched to the host species and isotype of your primary antibody.

Histological Identification of Specific Intestinal Epithelial Cells
1. Enterocytes 1. Prepare sections using section 7 with the villin antibody and conditions described in Table 2.
2. Entero-endocrine cells (Grimelius stain 18,19 ). 1. Prepare sections by following steps 7.1-7.2. Wash slides in a slide bath containing ultrapure water for 3 min. Transfer the slides to a slide bath containing preheated silver solution ( Table 1) and incubate at 60 °C for 3 hr. 2. Remove the slides from the silver solution and place in a slide bath containing freshly prepared preheated reducer solution ( Table 2) at 45 °C for 3 min. Remove the slides and place into a slide bath containing fresh ultrapure water for 3 min. Follow steps 7.6-7.8 from section 7. twice in a slide bath with fresh 1x PBS for 3 min. Discard 1x PBS and, using a pipette, cover section with 4% paraformaldehyde for 20 min on ice. 3. Wash twice in a slide bath with fresh 1x PBS for 3 min. Using a pipette cover sections with Proteinase K solution for 5 min Wash in a slide bath with fresh 1x PBS for 3 min. Using a pipette post-fix sections by covering in 4% paraformaldehyde for 5 min at RT. 4. Wash in a slide bath with DEPC treated H 2 0 for 2 min. Using a pipette cover sections in acetic anhydride solution for 10 min with agitation.

Goblet cells (Alcian Blue stain
Wash in a slide bath with fresh 1x PBS/3 min, followed by 1x saline/3 min. Pass slides through baths containing increasing concentrations of alcohols; 1x 30 sec in 70% EtOh, 1x 30 sec in 95% EtOH, 2x 30 sec washes in fresh 100% EtOH, 2x 2 min in fresh xylene and allow to air dry. 5. Dilute Olfm4 probe 1:100 in hybridization buffer and denature probe by heating to 80˚C for 3 min. Apply 100 μl of probe to each section and cover with parafilm to prevent dehydration of the slide. Incubate overnight in a dark, moist chamber at 65 ˚C. 6. Wash in a slide bath with 5× SSC at 65 ˚C for 15 min. Wash sections in a slide bath twice with fresh 50% formamide/5× SSC/1% SDS for 30 min at 65 ˚C. Wash sections twice in a slide bath in fresh PBT for 10 min, the first at 65 ˚C and the second at RT. Using a pipette cover sections with PBT containing 25 µg RNAse for 45 min at 37 ˚C. Wash sections in a slide bath in PBT for 5 min at RT. 7. Wash sections in a slide bath twice with fresh 50% formamide/5× SSC for 30 min at 65 ˚C. To block, cover sections using a pipette with 10% sheep serum in PBT and store in a dark, moist chamber at RT for 2 -3 hr. 8. Prepare antibody by diluting anti-digoxigenin alkaline phosphatase conjugated antibody at 1:500 with 10% sheep serum in PBT containing 5 mg/ml mouse intestinal powder. Incubate for 3 hr at 4 °C in dark on a rocking platform. Spin down to remove excess intestinal powder and add 3x volumes of 1% sheep serum in PBT to the supernatant. 9. Remove block from slides with a pipette and add 100 μl of the antibody solution to each section, cover with parafilm and incubate in a dark, moist chamber at 4°C overnight. Wash sections in a slide bath 3× with fresh PBT for 5 min. To block sections wash in a slide bath 3× with fresh NTMT buffer for 5 min. 10. To visualize, using a pipette cover each section with BM purple and incubate in the dark at RT for 24 -72 hr until a sufficiently strong colour develops. Wash sections in a slide bath once in PBT and counterstain by immersing in eosin for 1 min. Remove excess eosin by washing sections in a slide bath under a running water for 3 -5 min. Immerse slides in xylene and allow to air dry. Mount under a coverslip using commercial media.  (Figure 3b).

Histological Characterization of Intestinal Epithelia
3. Apoptosis 22,23 . 1. Method 1: Use a high power magnification (e.g. 20X or 40X) to manually count the numbers of apoptotic cells in each crypt.
Apoptotic cells can be identified by cell shrinkage, chromatin condensation, formation of cytoplasmic blebs and apoptotic bodies (Figure 3a). 2. Method 2: Perform an IHC stain (section 7) for Caspase-3 using conditions described in Table 2. Using a high power magnification (e.g. 20X or 40X) manually count the number of positive cells in each crypt to quantitate the cells in the execution phase of apoptosis.
4. Mitosis. 1. Use a high power magnification (e.g. 20X or 40X) to manually count the numbers of mitotic cells in each crypt. Mitotic cells contain condensed DNA material and are typically symmetric and well formed (Figure 3b).

Proliferation.
1. Perform an IHC (section 7) stain for Ki-67 using conditions described in Table 2. Manually count positive cells to quantitate the proportion of crypt cells proliferating.

Comparing ISC Recombination Efficiency in the Ah-cre and Vil-Cre-ER T2 Systems
Use of these cre-lox systems for evaluating the role of Paneth cells, in repopulating the intestine following damage, required characterization of the efficiency of recombination within the ISCs. Using the Rosa26R-lacZ conditional reporter we demonstrated that in both systems 3 days post induction (d.p.i.) there is ~100% recombination in the small intestine (Figure 1a). Quantitating the presence of the recombined allele by qPCR was confounded by the differences in Cre expression patterns between the systems. The vil-Cre-ER T2 system showed a 3.53 fold increase in the presence of the recombined allele compared to the Ah-cre system, due to its expression in a greater proportion of the epithelia 16 . To overcome this we adopted a different strategy that allowed us to directly compare the systems. We induced the mice with different induction regimes and analyzed at 30 d.p.i., at which point LacZ positive crypts and villus represent an ISC recombination event. Using this approach we demonstrated that in both systems, 3 injections of inducing agent (delivered I.P. at 80 mg/kg in 24 hr), recombined in an equivalent number of ISCs despite initial recombination levels being far greater in the vil-Cre-ER T2 system 16 (Figure 1b-d). Further, using DNA extracted from the recombined crypts, qPCR for the recombined alleles demonstrated a non-significant increase in recombination using the vil-Cre-ER T2 system, potentially due to the recombination in the Paneth cells not observed using the Ah-cre system (Figure 1d). Further, staining for epithelia cell types did not indicate any alteration to differentiation pattern, representative images of each cell type investigated is shown in Figure 2e-2h.

Quantification of Crypt Loss
Characterizing the kinetics of recombination in these Cre systems enabled us to analyse the mouse intestine when equivalent numbers of ISCs are recombined. Using the LacZ reporter both systems showed complete loss of recombined (blue) cells at 3 d.p.i. (Figure 2a). As previously reported three days after deletion of CatnB the Ah-cre mice showed partial crypt loss, whereas the vil-Cre-ER T2 mice demonstrated complete destruction of the crypt/villus axis 13,16,24 (Figure 2b-d).

Dynamics of Epithelial Repopulation
Using the techniques above we characterized multiple parameters to enable us to understand this observation. Representative images of the parameters and cell types analyzed using protocols 7 -10 are given in (Figure 3a & 3b). Briefly, the loss of crypts was consistent with the elevated levels of apoptosis displayed in both systems (Figure 3e). However the mitosis, proliferation, crypt cellular height, crypt and expression (not shown) data indicated the Ah-cre system could recover, presumably due to repopulation by un-recombined ISCs (Figure 3c). In stark comparison, the vil-Cre-ER T2 failed to recover despite retaining epithelial crypt cells (Figure 3d).

Characterization of Cellular Phenotypes within Crypt
To understand why the crypts from Ah-cre mice could repopulate whereas the vil-Cre-ER T2 couldn't we characterized the epithelial cells three days after deletion of CatnB. Using in situ hybridization (section 9) and IHC analysis (section 7, 8 & 10) we demonstrated that the crypt cells in the vil-Cre-ER T2 CatnB flox/flox mice were non-proliferative and lacked expression of the ISC marker Olfm4, unlike the crypts in the Ah-cre mice (Figure 4c & f). As the initial characterization had demonstrated that recombination in crypts was equivalent we proceeded to examine the role of the Paneth cells. We performed a dual fluorescent IHC against CatnB and Lyz1 to identify which cells had lost β-catenin and whether they were Paneth cells (Figure 5a-5c). As previously described we demonstrated that all crypt cells are targeted using the vil-Cre-ER T2 system. In       Intestinal tissue powder * * The small intestines of 5 adult mice were combined and homogenized in the minimum volume of ice cold PBS. 4 volumes of ice cold acetone were added to the homogenized intestine, which was mixed thoroughly and incubated on ice for 30 min. This was centrifuged and the pellet was washed using ice cold acetone. This was further centrifuged and the resulting pellet spread onto filter paper and allowed to dry. Once thoroughly dry the material was ground to a fine powder using a pestle and mortar. Fisher Scientific X/0200/21

Discussion
Using conditional cre-lox transgenic mice to dissect the function of genes and cells is a commonly used approach. These models have been used with great success in the intestine to identify and characterize the stem cells 2,4-6 and understand their role in disease 25 . To fully exploit these models requires a comprehensive characterization of the system to enable data to be interpreted correctly. A complete understanding of these systems is difficult to achieve due to genes rarely being specific to a solitary cell type or location, a lack of biological knowledge and inefficiency of the systems used to induce Cre expression. The methods described here demonstrate how we overcome these issues through experimental design and application of existing knowledge. Although we used these methods to answer a specific research question the techniques presented here are generic and can be exploited for any research investigating the murine intestine.

Preparation of Intestinal Tissue
The crucial step for ensuring robust results is the harvesting and processing of the tissue, which needs to be processed in a timely manner and fixation protocols strictly adhered to. As almost all significant issues downstream can be attributed to artefacts associated with the tissue drying out and/or incomplete fixation. Timing is crucial to prevent degradation of tissue architecture and/or nucleic acids and proteins. Incomplete or overzealous fixation can result in loss of histochemical resolution. Incomplete fixation due to insufficient time or sections too thick to allow fixative penetration can result in loss of resolution within the intestinal crypts that can be observed as a "tide mark" upon IHC analysis. Further it is crucial that fixation does not extend for too long, as nuclear β-catenin can diffuse out of the nucleus unless immediately processed and wax embedded following formalin fixation.

Role of Paneth Cells in the ISC Niche
The data presented here effectively show the importance of Paneth cells in crypt regeneration in the adult intestine following ISC loss. However there remained the possibility that Ah-cre spares a population of ISCs that vil-Cre-ER T2 targets. Tian et al. 26 elegantly demonstrated that the Lgr5 hi ISCs are replaced by a population of Lgr5 lo reserve ISCs. It now seems likely that these ISCs are spared in the Ah-cre system due to the reserve population having been identified as secretory cell precursors 6,7 . The importance of the mature Paneth cell in supporting these secretory cell precursors when required to revert to an ISC state remains to be answered. As Paneth cells constitute the ISC niche 8 and play roles in regulating the ISC responses to calorie intake 27 and inflammation 28 it remains likely that their nursing functions will extend to their own precursors.

New Approaches and Technologies to Effectively Model Human Colorectal Cancer
The discovery of the ISC led to the identification of genes which are now being used to generate new mouse models for investigating the roles of genes and cells in intestinal biology and disease, reviewed by Clarke et al 9 .
The only limitations to this technique is the identification of genes to express the Cre protein. Currently ISCs are routinely investigated using conditional transgenic mice based on the Lgr5 gene expression pattern. Mice which express Cre from the Lgr5 promoter have been used to delete Apc, the gene most commonly mutated in colorectal cancer (CRC), demonstrating the ISC as the cell of origin 25 . Selectively deleting other CRC genes in these cells is providing insight into disease progression and spread e.g. PTEN 29 . Further insight into ISC function is being retrieved by specifically ablating Lgr5-expressing cells in mice using a human diphtheria toxin receptor (DTR) gene knocked into the Lgr5 locus 26 . Other strategies use the Tet-O system which enables ongoing reversible expression of mutant proteins 30 . Using these tools to modify gene(s) in different cells 31 and locations 32,33 is used to understand how cancer initiate, progress and metastasize 34 . Alternatively mutagenesis using the sleeping beauty transposon system is identifying new drivers of CRC. The continuing development of mice, techniques and genetic alteration strategies is continuing to develop more patient relevant models.
New methods have been developed for characterization of the intestinal epithelia and ISCs. Characterization of the ratio of epithelial cell types can be achieved using flow cytometry based on the differential expression of lectin and CD24 35 . Potentially the biggest progress in understanding ISC biology and their role in disease will be made using the ex vivo organoid culture system 36 . This system allows normal and malignant ISCs to culture in 3D, where they replicate and differentiate in a more physiologically relevant way. It is hoped that these will enable direct testing of drugs on patient samples in vitro, paving the way for personalized medicine 37 .

Disclosures
The authors have nothing to disclose.