Cellular organization of the intestinal crypt: The inner lumen of the humancolon is comprised of a single cell columnar epithelial layer, arranged into fingerlike invaginations, known as the crypts of Lieberkühn. The self-renewing intestinalstem cells (ISC’s) at the crypt base give rise to a proliferating progenitor populationof transit-amplifying cells (TA). As they migrate towards the crypt/surfaceepithelium border, these TA cells terminally differentiate to acquire either asecretory cell or absorptive enterocyte fate. They finally undergo senescenceor apoptosis on reaching the surfaceepithelium, hence balancing the continuous proliferation of the ISC’s. Thishierarchical crypt architecture and well-orchestrated upwards columnar movementof cells enables rapid epithelial self-renewal.
Adult intestinal stem cells: Twoputative populations The stem cell compartment of the small intestine has beenwidely studied and is a model for the adult intestinal crypt, while the coloniccrypt stem cells remain poorly defined. Two populations of stem cellshave been postulated to exist within the intestinal crypt: the crypt base columnar cells(CBC’s) and the +4 cells. Cheng and Leblond identified CBC’s,which were actively cycling cells that populated the crypt base.
Given theirmultipotent and self-renewal capabilities, they were postulated to be intestinalstem cells. (Cheng and Leblond, 1974). Barker et al, 2007 showed that thesecells, which were interspersed between Paneth cells (intestine) or goblet-likecells (colon), uniquely expressed a Wnt target gene, LGR5 (Leucine-rich-repeat-containingG-protein-coupled receptor 5).
This was visualised by creating a knock-inallele, wherein an EGFP-IRES-Cre-ERT2 cassette was inserted downstream of thestart codon of the LGR5 gene and EGFP expression was detected in LGR5+ cells.Furthermore, lineage tracing using an Lgr5-EGFP-ires-CreERT2/Rosa26RlacZmouse model revealed that the crypt base stem cells could generate multiple epithelialcell lineages as well as maintain long-term self-renewing capacity. This wasevidenced by the presence of Lac-Z labelled cells in the crypt-villi axes evenafter 6-10 months (Barker, 2007). In comparison with the small intestine, LGR5+cells of the colon showed slower cycling kinetics, taking longer to generate Lac-Zstained crypts. This, perhaps, reflects the difference in epithelial turnoverrate, which is slower in the colon. (Barker, 2007).
Following the identificationof LGR5, gene signature studies have revealed various markers that label CBCcells. Ascl2, a Wnt target gene, that encodes a transcription factor involvedin stem cell fate determination was expressed at higher levels in crypt basestem cells of the human small intestine and colon, when compared to TA andPaneth cells. (Van der Flier, 2009). Similarly, Olfm4, was also established asa marker for CBC’s in the mouse small intestine and was later found to beexpressed in the human colon as well (Van der flier, 2009) The + 4 stem cells, characterized by the expression ofBMI1, TERT, HOPX and LRIG1, were originally identified by Potten et al asactively dividing, DNA label retaining cells. However, they are now consideredto represent a relatively quiescent stem cell pool. Lineage Tracing of BMI1 – regeneration, slow-cycling Intestinal stem cell niche:Homeostasis is maintained by major signalling pathways Wnt signalling: Pathway Secreted Wnt glycoproteins act aslong and short distance morphogens, establishing Wnt gradients which play anintegral role in directing cell proliferation and fate determination; bothduring development and in regulating adult tissue homeostasis. The canonical Wnt pathway(described in Figure 2), which promotes the nuclear localization of B-cateninand its subsequent interaction with the Tcf family of transcription factors(Tcf1,Tcf3, Tcf4 and Lef1) to influence gene expression, is a key player indriving homeostasis of the intestinal crypt.
In the absence of Wnt, free B-cateninin the cytoplasm is sequestered by a destruction complex consisting of coreproteins including Axin, APC, CK1a and GSK3B. CK1a and GSK3B phosphorylate B-catenin,leading to its degradation. The binding of a Wnt ligand to Frizzled, a seven-passtransmembrane receptor and its co-receptor lipoprotein receptor related protein(LRP) activates the canonical Wnt pathway. Dishevelled proteins are recruited to thecytoplasm, which inhibit the destruction complex.
Cytosolic B-catenin,therefore accumulates and is translocated into the nucleus, where it interactswith the TCF/Lef complex to displace the transcriptional repressor, transducin-likeEnhancer of split (TLE), which is the mammalian homolog of Groucho, therebyactivating Wnt target genes. In the absence of nuclear B-catenin, TCF3 and TCF4are bound to TLE, which recruits histone deacetylases (HDAC’s) (Brantjes et al,2001).This represses the chromatin state of the downstream target genes, preventingtheir transcription. (Chenet al,2001). When B-catenin binds to the TCF/Lef complex, it interactswith chromatin activators such as CBP/p300 and Brg1, which in turn inducetarget gene transcription.
(Barkeret al,2001) ZNF3/R-SPO FUNCTION/LGR5 Wnt signalling drivesintestinal crypt homeostasis The involvement of Wnt signallingin preserving intestinal architecture was proposed in early studies by Korineket al,1998, who showed that disruption of Tcf7l2,the gene encoding Tcf4, resulted in the loss of crypt formation and stem cell maintenancein murine embryonic small intestine. This was further substantiated in experimentsby Van et al, 2012, where floxed Tcf4 mouse models were generated to inactivateTcf4 function. These Tcf4LoxP mice were crossed with a tamoxifeninducible CreErt2 VillinCreert2 line and ki67 expression was used to detectcell proliferation within the crypt. A loss of ki67+ cells was seen 7 daysafter cre-induction had occurred, within the colonic crypts. Similar results wereobserved in the small intestine, where the number of proliferating cellsreduced within 3 days of cre-activation.
To determine if the proliferatingcells that were lost were in fact stem cells, Olfm4 (intestinal stem cellspecific marker) expression was studied. It was seen that within 7 days ofcre-induction, all Olfm4 expressing cells had been eliminated. Furthermore,Lgr5 expression was lost after 7 days, whereas Bmi1, which is not a Wnt targetgene, was not affected.
This, therefore, highlights the importance of downstreamfactors of the canonical Wnt pathway in maintaining intestinal stem cells. Cryptbased stem cells express high levels of nuclear B-catenin (van de Wetering etal. 2002) and deletion of B-catenin led to loss of intestinal crypts and gobletcells (Ireland et al, 2004). Furthermore, inhibition of the Wnt pathway by overexpressionof Dickkopf (Dkk), a Wnt antagonist, caused the intestinal architecture tobreakdown. LGR5/LGR4 DELETION Transcription factors involvedin Wnt signalling Notch signalling is essentialfor cell fate determination Notch signalling plays a key rolein mediating cell proliferation and cell differentiation as blocking of thenotch pathway leads to a loss of proliferating cells and differentiationtowards a secretory lineage. This mechanism is determined by Notch ligands andreceptors that are present on adjacent cells, through which neighbouring cellsinfluence the fate of each other.
Inmammals, four Notch genes have been identified, which encode a singletransmembrane receptor each, known as Notch 1-4. These receptors in turn,interact with one of five Notch ligands – Delta-like (Dll) 1,3 and 5 andJagged-1 and 2. In the intestine, Notch 1 and 2, are expressed, albeit in a redundantfashion, as only inactivation of both receptors together has a phenotypicimpact. High levels of Notch 1 are localized in the proliferating crypt-basestem cells, in both the colon and the small intestine. Dll1 and Dll4, alongwith Jagged-1 are found at the bottom of the intestinal crypts, while Jagged-2is found in cells underneath absorptive colonocytes (Sander et al, 2004). Upon ligand binding, y-secretase(and other proteases) cleaves the Notch receptor, thus freeing the Notchintracellular domain (NICD), which is then translocated into the nucleus. Once,in the nucleus, NICD binds to a transcriptional regulator, CSL (CBP/RBP-Jk).
This relieves transcriptional repression by CSL on Notch target genes, hence activatingthe canonical notch pathway. This allows Notch target genes, Hairy/Enhancer ofSplit (Hes) 1,5 and 7, HEY1,5 and HEYL to be transcribed. The intestinalepithelium of Hes1 null mice showed an increase in goblet cells and Paneth cells,but a decrease in absorptive cells, indicating that Notch signalling isrequired for absorptive cells specification and represses the formation ofsecretory cells (Jensen et al, 2000). The Hes family act as transcriptionalfactors which repress transcription of basic helix-loop-helix (bHLH) transcriptionfactors including ATOH1, Math1 and HATH1, that are involved in inducing cell differentiationand lineage specification. Math1 is highly expressed in secretory cells, andthe intestinal epithelium of mutant Math1 mice consist of only absorptivecells, demonstrating its role in secretory cell specification.
Anothertranscription factor, kruppel-like factor 4 (Klf4) that is repressed by Hes1activity, is essential for terminal differentiation of goblet cells, asdeletion of Klf4 in mice, caused a 90% reduction in goblet cells in the colon(Katz et al,2002). Zheng et al, 2009, showed that blocking Notchsignalling via a ?-secretase inhibitor significantly increased Klf4 mRNA levelsas well as the number of Klf4-positive cells (goblet-cells), in both the smallintestine and colon. However, Pellegrinet et al, 2012 argued that ,while lossof function of Dll1 caused an increase in goblet cell numbers and decreasedproliferative progenitors, inactivation of Klf4 did not decrease goblet cellnumbers. Furthermore, pharmacological notch inhibition in both, a control andKlf4 mutant mouse model, resulted in excessive goblet cells and a breakdown ofthe proliferative crypt, thus questioning the role of Klf4 in notch-mediatedgoblet cell differentiation. Notch signalling maintainsintestinal stem cells In addition to repressing thesecretory cell phenotype, notch signalling has been implicated in maintainingthe proliferative state of intestinal progenitors and stem cells.Co-localisation of Hes1, with the intestinal stem cell marker Musashi-1 (Msi1)in cells at the bottom of the crypt, indicated the presence of notch activityin CBC’s.
A similar observation was made at the +4 stem cells in the smallintestine. Lineage tracing studies using a tamoxifen induced Notch1 Cre-mousemodel (NIP1::CreERT2) revealed that, after eight months, labelledcells occupied the entire crypt (and villus) and expressed markers of all thedifferentiated cell types in the adult intestine (Pellegrinet et al, 2012). This showed that cells expressing Notch1activity were capable of self-renewal and differentiation, therefore,expressing stem-cell properties. A lower number of Notch1 labelled cells werefound in the colon as compared to the small intestine. This phenomenon issimilar to the aforementioned observation in Lgr5 lineage tracing experiments,where slower cycling rates were attributed to colonic CBC’s (Barker etal,2007). The discovery of Lgr5 as a unique marker of CBC’s paved way for the identificationof Notch activity in intestinal stem cells as well. Dll1 and Dll4 double knockoutand RBP-J inactivated mice, along with increased goblet cell numbers, lost theexpression of stem cell markers including Lgr5, Olfm4 and Ascl2.
Furthermore,proliferating Ki67-positive cells in the crypts were ablated (Pellegrinet etal, 2012).