Laminin Receptor 37/67LR Regulates Adhesion and Proliferation of Normal Human Intestinal Epithelial Cells

Interactions between the cell basal membrane domain and the basement membrane are involved in several cell functions including proliferation, migration and differentiation. Intestinal epithelial cells can interact with laminin, a major intestinal basement membrane glycoprotein, via several cell-surface laminin-binding proteins including integrin and non-integrin receptors. The 37/67kDa laminin receptor (37/67LR) is one of these but its role in normal epithelial cells is still unknown. The aim of this study was to characterise the expression pattern and determine the main function of 37/67LR in the normal human small intestinal epithelium. Immunolocalization studies revealed that 37/67LR was predominantly present in the undifferentiated/proliferative region of the human intestinal crypt in both the immature and adult intestine. Using a human intestinal epithelial crypt (HIEC) cell line as experimental model, we determined that 37/67LR was expressed in proliferative cells in both the cytoplasmic and membrane compartments. Small-interfering RNA-mediated reduction of 37/67LR expression led to HIEC cell-cycle reduction and loss of the ability to adhere to laminin-related peptides under conditions not altering ribosomal function. Taken together, these findings indicate that 37/67LR regulates proliferation and adhesion in normal intestinal epithelial cells independently of its known association with ribosomal function.

While 37/67LR was the first laminin receptor to be identified [13,14], its characterization is still incomplete and has been complicated by the fact that it is also involved in a variety of other unrelated roles. Indeed, beside its ability to interact directly with laminin through most likely the CDPGYIGSR sequence on the laminin β1 chain short arm, 37/67LR can play additional roles in the cell. Indeed, phylogenetic analysis carried out on 37/67LR found homologues in all kingdoms from archaebacteria to mammals and suggests that it was originally a ribosomal protein that acquired additional novel functions though evolution [11,15]. As reviewed in detail by Nelson et al. [11], the human 37/67LR gene product (UniGene ID Hs. 181357; ribosomal protein name RPSA) has been found in the ribosome of all tissues investigated [16] where it appears to serve as a critical component of the translational machinery [17]. The 37/67 kDa laminin receptor has also been identified as a component of the nuclear machinery where it can bind to both chromatin and the nuclear envelope [18][19][20]. It is noteworthy that 37/67LR can also act as a cell surface receptor for bacteria, viruses and prions [11,12,21]. Consistent with these multiple functions, 37/67LR is not only localized on the cell surface but can also be found in the cytoplasm, perinuclear compartment and nucleus. The alternative "37/67 kDa" nomenclature still used to identify 37/67LR arises from the observation that the gene corresponding to the originally identified 67 kDa laminin-binding protein encodes a 32.8 kDa protein, which migrates at 37 kDa on SDS-PAGE suggesting that the 67 kDa form could result from homo or heterodimerization reactions involving the 37 kDa precursor and fatty acid acylation [22][23][24]. Albeit the proposed precursorproduct relationship, the exact relationship between the 37LR precursor and 67LR remains unclear [11,12,21]. For instance, some antibodies raised against amino-peptides of the 37 kDa sequence failed to recognize the 67 kDa polypeptide in Western blots [25] while the 37LR precursor can be detected on the plasma membrane [25][26][27].
Functionally, 37/67LR has attracted considerable interest since its discovery 30 years ago [28,29]. Indeed, overexpression of 37/67LR has been shown in a variety of cancer cell types where its expression levels have been found to strongly correlate with the risk of tumour invasion and metastasis [30][31][32][33]. 37/67LR may also be of importance in other pathologies including neurodegenerative and angiogenic diseases such as Alzheimer's disease [21] and retinal neovascularisation [34]. The mechanism has not yet been elucidated but recent studies indicate that 37/67LR can prevent apoptosis [35,36] and acts as the cell receptor that mediates the anti-inflammatory and anti-thrombotic activities of epigallocatechin-3-gallate [37][38][39]. Further studies are nevertheless required to fully understand the involvement of 37/67LR in these pathologies [11].
Another intriguing question pertaining to 37/67LR is its role in the normal state. Indeed, very few studies have addressed its extraribosomal function in normal cells [11]. The intestinal epithelium represents a useful system to investigate such a question. Indeed, under physiological conditions, the architecture of the small intestinal mucosa is maintained through a sensitive equilibrium between epithelial cell production and maturation in the crypt compartment and migration along the length of the crypt-villus axis and extrusion at the villus tips [40][41][42]. As epithelial cells migrate upwards they are exposed to compositional changes of the underlying BM components such as collagens, fibronectin and laminins [41,[43][44][45], which have been demonstrated to influence cell behaviours including proliferation, migration, differentiation and cell survival [46][47][48][49][50][51][52]. Interestingly, 37/67LR expression has previously been investigated along the intestinal epithelial crypt-villus axis. While the analysis performed on the human normal intestinal epithelium revealed relatively contradictory staining patterns from weak/absent [53] to predominant expression in the brush border/supranuclear region of the entire epithelium [54] or in the basolateral domain restricted to crypt epithelial cells [55], transcript analysis in the rat intestine revealed higher mRNA levels of 37/67LR in the fetal intestine and adult epithelial crypt cells suggesting that high 37/67LR expression may correlate with proliferation in intestinal cells [33].
In the present study, we sought to determine the role of 37/67LR on normal cell functions using normal human small intestinal specimens as well as a human intestinal epithelial crypt (HIEC) non-transformed non-immortalized cell line to perform experimental analysis. By immunofluorescence, we first established the predominant expression of 37/67LR in the immature intestinal epithelium and in the crypt cells of the adult small intestine and confirmed that these observations can be extended to proliferating HIEC cells. Then, using an siRNA strategy under conditions that down-regulate 37/67LR expression without altering translation, we found a significant reduction of both HIEC cell proliferation and specific adhesion to the laminin-related peptide YIGSR. Taken together, these data establish for the first time that 37/67LR expression in normal intestinal epithelial cells regulates cell proliferation and adhesion, two crucial functions for intestinal epithelial homeostasis.

Tissues
Tissues from the normal adult proximal ileum were obtained from Quebec Transplant. Specimens of small intestine (ileum) were obtained from fetuses ranging in age from 10 to 20 weeks of gestation following legal or therapeutic pregnancy termination with written informed patient consent. No tissues were collected from cases associated with known fetal abnormalities or intrauterine fetal demise. Studies were approved by the Institutional Review Committee for the Use of Human Material of the "Centre Hospitalier Universitaire de Sherbrooke/Faculté de Médecine et des Sciences de la Santé".
In some experiments, pure epithelial and stromal fractions were prepared and analysed as previously described [56].

RNA interference and transfection
Predesigned siRNA sequences targeting 37/67LR (siLR1, 2, 3 and 4) and a non-silencing negative control siRNA (siCtrl) were purchased from Sigma Aldrich (Oakville, ON) and used as previously described [46]. One day prior to transfection, 2x10 5 cells were plated in 30mm dishes (Falcon Plastics) and transfected with siRNAs using the X-tremeGENE siRNA transfection reagent (Roche Diagnostics, Laval, QC). Cells were used in the various experiments 48 hours posttransfection unless otherwise specified.

Real-time RT-PCR quantification analyses
For quantitative PCR experiments, primers for 37/67LR and RPLPO were the same as above. Amplification efficiencies and assessment of differences in gene expression between control and experimental conditions were established according to the Pfaffl mathematical model [69]. Relative mRNA expression levels were established by comparing the levels under experimental conditions to those of control samples and RPLPO was used for normalization. For quantitative evaluation of transcript levels, real-time experiments were performed using an Mx3000P (Stratagene, La Jolla, CA) as previously described [70].

Western blotting analyses
Western blots were performed on SDS-PAGE gels under reducing conditions as previously described [46,49]. Total proteins (50µg/ml) were separated on 12% SDS-PAGE gels, electrotransferred onto a nitrocellulose membrane (BioRad) and probed as described previously [46]. Immunoreactive bands were visualized using the Immobilon Western kit (Millipore) according to the manufacturer's instructions or with AlexaFluor secondary antibodies and visualized with a Molecular Imager® FX equipped with an external laser (BioRad).

Epithelial-stromal dissociation
Pure fractions of epithelial and mesenchymal cells from midgestation ileum were obtained using Matrisperse (BD Biosciences) and tested as previously described [56,60].

BrdU incorporation assays
BrdU incorporation experiments in cells were performed in accordance with the In Situ Cell Proliferation Kit FLUOS® protocol (Roche) as described previously [71].

Cell-cycle progression analysis by laser scanning cytometry
HIEC were seeded onto 12-well plates (Falcon) at 5× 10 4 cells per well 48 h under standard culture conditions with complete medium before methanol fixation and DAPI staining. DAPI-stained cells were scanned with an iCys imaging cytometer (Compucyte, Cambridge, MA) to measure DNA 37/67LR in the Human Small Intestinal Epithelium PLOS ONE | www.plosone.org content using violet diode laser excitation (405 nm) and emission (460 nm) filters for fluorescence detection as previously described [68]. DNA content was measured for at least 3000 isolated nuclei per sample in three separate experiments to assess cell-cycle distribution.
Preparation of CMD, GRGDSPC and CDPGYIGSR surfaces and cells prior to seeding was as described [73]. Briefly, HIEC cells were harvested using PBS/0.5 mM EDTA 48 hours after transfection. Cells were seeded (1×10 5 cells per well) and incubated for 1 h. Non-adherent cells were gently washed away with PBS and the remaining cells were fixed with 2.0% paraformaldehyde (pH 7.4) for 25 min at 4°C and permeabilized with 0.1% Triton X-100. Cell nuclei were stained with DAPI and adherent cells were counted manually in 10 squares of 0.42 mm 2 .

Blocking experiments
HIEC were harvested using PBS/0.5 mM EDTA for 10 min and processed as described previously [71] for blocking experiments. Briefly, HIEC cells were resuspended in serumfree culture medium containing 20 µg/ml of neutralizing antibody or non-immune serum protein for 1h, then seeded in complete culture medium onto 12 well plates and incubated at 37 o C for 24h before being processed for BrdU staining.

Statistical analysis
Results are expressed as mean ± S.E.M. Each experiment was repeated at least three times and representative results are shown. Student's paired t-test and ANOVA using Bonferroni's Multiple Comparison Test were used to analyze the data. Data were considered to be statistically significant at p < 0.05. Statistical calculations were performed using Prism 3.0 Software (GraphPad Software, San Diego, CA).

In situ distribution of 37/67LR in the human small intestinal mucosa
The expression and distribution patterns of 37/67LR along the crypt-villus axis in the normal human fetal and adult small intestinal mucosa were determined by indirect immunofluorescence on cryosections. The specificity of the antibody was determined by Western blot. As shown in Figure  1A, the anti-37/67LR antibody predominantly detected a major band at the expected molecular range of the monomer, i.e. ~37 kDa, in normal HIEC cells ( Figure 1A, lane 1) as well as in Caco-2 cells (lane 2), a colorectal cancer cell lines used here as positive control. No specific band was detected in the 67 kDa range with this antiserum ( Figure 1A) or with three other tested anti-37/67kDA antisera (not shown). At 10 weeks of gestation, crypts are not yet formed [41] and 37/67LR staining was found to be widespread in both the intervillous region and the villus (although at a weaker intensity) ( Figure 1B). As expected at this stage, cell proliferation as evaluated by Ki67 staining was found in the intervillous area and in the lower part of the villus (Figure 1B', arrows and arrowheads, respectively). At 14 and 20 weeks ( Figure 1C and D), 37/67LR staining was gradually restricted to the crypts where the proliferative epithelial cells are confined ( Figure 1C' and D', arrows). In the adult, expression of 37/67LR was also predominantly located to the middle and upper parts of the crypts ( Figure 1E, brackets) that also contains the most active proliferative cell population (transit amplifying cells) as identified with Ki67 ( Figure 1E'). The villus differentiated epithelium (v) as well as the lower part of the crypts (*) containing the differentiated PLA2-expressing Paneth cells ( Figure 1E'') were found to be negative ( Figure  1E). As summarized in Figure 1F based on a series of observation, the 37/67LR was consistently found to be absent from the Paneth/stem cell zone (1). Its expression was maximal in the middle crypt containing the Ki67-positive transit amplifying cells (2) while its expression decreased toward the terminal differentiation compartment (3) and the base of the villus (4).
To confirm this distribution, the expression of 37/67LR was analyzed on purified epithelial and stromal fractions isolated from mid-gestation ileum based on a method established in our laboratory [56]. Semi-quantitative RT-PCR showed the presence of the 37/67LR transcript in both fractions ( Figure  2A). The purity of the fractions was confirmed by the exclusive expression of E-cadherin (epithelial) and tenascin-C (stromal). Analysis of 37/67LR at the protein level confirmed the presence of 37/67LR in its 37 kDa form in both fractions ( Figure 2B) although quantitative assessment showed ~6 times more antigen in the epithelium than in the stroma ( Figure 2C). Taken together, these observations indicate that 37/67LR is predominantly expressed in mucosal epithelial cells and proliferative cells in both the fetal and adult intestine.

Proliferative Human Intestinal Epithelial Crypt Cells Express 37/67LR In Vitro
Considering the distribution pattern of immunoreactive 37/67LR in intestinal crypts in situ, we used the normal human intestinal crypt (HIEC) cell line to further investigate 37/67LR expression in relation to cell proliferation. As shown in Figure 3, 37/67LR expression was observed in actively proliferating HIEC cells (Figure 3A, SC) and a significant reduction was observed in HIEC cells maintained for 5 days at confluence ( Figure 3A, PC). Indeed, while subconfluent HIEC cell growth is linear before confluence, proliferation is totally inhibited after confluence ( Figure 3B).

Knockdown of 37/67LR expression under conditions that do not affect protein translation
As a first step to investigate the function of 37/67LR in HIEC cells, we transiently knocked down its expression by transfection with siRNA targeting the 37/67LR sequence. Analysis of 37/67LR expression after 48 hours in the presence of 40 µM of one of 4 predesigned siRNA sequences (siLR1,2,3 and 4) and siCtrl was performed by Western blot (Fig. 4A). While all 4 sequences tested showed some inhibition of expression relative to control, siLR4 showed more efficiency and reproducibility. Using siLR4, 37/67LR expression was found to be reduced by more than 5 times at both the transcript and protein levels ( Figure 4B) while the appearance and viability of the cells appeared to be altered under these conditions ( Figure 4C).
Because 37/67LR is thought to be required for protein translation [35], we performed a 0-50 µM dose-response analysis of siLR4 efficiency in order to evaluate 37/67LR expression relative to other control gene products such as fibronectin, IGFBP7 and TIMP3 which are constitutively produced by HIEC cells [47,71,74] and used to monitor translation. Control siRNA (siCtrl) dose-response showed that expression of 37/67LR, fibronectin, IGFBP7 and TIMP3 was not altered at any tested concentration ( Figure 5A, left panel; Figure 5B, upper panel; TIMP3 data: 1.0 at 0 µM vs 1.05 ± 0.15 at 20 µM). In contrast, siLR4 showed maximal apparent efficiency at 30 µM and higher for both 37/67LR and fibronectin and 40 and 50 µM for IGFBP7 ( Figure 5A, right panel). However at 20 µM, a ~50% knockdown of the expression of 37/67LR was observed with siLR4 while neither fibronectin, IGFBP-7 nor TIMP3 expression was significantly affected ( Figure 5B, lower panel; TIMP3: 1.0 at 0 µM vs 1.22 ± 0.14 at 20 µM). These results indicated that siLR4 at a concentration of 20µM (siLR4-20) can significantly reduce 37/67LR expression without altering general translation. We then used these conditions to investigate the role of 37/67LR in normal HIEC cells.

Reduction of 37/67LR expression inhibits cell proliferation in G1
To determine whether 37/67LR was directly implicated in cell proliferation, we first evaluated proliferation by BrdU incorporation 24 and 48h following control and siLR transfection. As expected from previous studies, BrdU incorporation under these experimental conditions resulted in ã 30% proportion of positive cells under control conditions [46,71] as also observed herein ( Figure 6A). A significant reduction of cell proliferation was observed at both 24 and 48h after transfection with the siLR4 used at 20 µM ( Figure 6A). To confirm the specificity of the effect, we tested a second siRNA. As shown in Figure 6B, siLR3 used at both 20 and 50 µM had comparable effects on inhibition of BrdU incorporation as siLR4.
To further characterize the involvement of 37/67LR on HIEC proliferation, laser scanning cytometry analysis revealed that the apparent decrease in DNA synthesis observed in Figure 6A was consistent with a significant increase in the percentage of cells in G1 and a ~ 50% reduction in S phase in siLR4-20 treated cells as compared to control ( Figure 6C and D) after 48 h. It is noteworthy that the lack of staining in the Sub-G1 region for both siCtrl and siLR4 ( Figure 6C and D) indicates that apoptosis was not involved.

Reduction of 37/67LR expression inhibits LM-111 (YIGSR)-dependent cell adhesion
Previous studies have identified that 37/67LR binds to the short arm of the laminin β chain via the YIGSR peptide sequence [11]. This characteristic was used to evaluate    expression were found to be stable in the presence of increasing concentrations of siCtrl. With siLR4, 50% knockdown of 37/67LR expression was achieved at 20 µM, a concentration that had no effect on HFN or IGFBP7 expression in contrast to higher concentrations such as 40 and 50 µM which resulted in more than 80% knockdown of 37/67LR expression and inhibition of protein synthesis as observed for HFN and IGFBP7. Relative amounts of 37/67LR, HFN and IGFBP7 were determined by optical densitometry and expressed relative to β-actin (mean ± SEM, n=3 separate experiments, * p < 0.05 relative to 0 µM)).  Cell adhesion assays were performed on low-fouling surfaces on which peptide immobilization supports specific cell-biomaterial interaction while preventing non-specific protein adsorption allowing the discrimination of a specific from a non-specific cell response [72]. As performed previously [73], non-specific adhesion was determined on CMD surfaces which prevent cell adhesion, 37/67LR-mediated adhesion was evaluated on CDPGYIGSR (YIGSR) while maximal adhesion was determined on RGD peptide. Previous studies have shown that HIEC cells attach strongly to RGD peptides [71]. The results showed that reduced 37/67LR expression had no significant effect on binding to the RGD peptide but induced a statistically significant ~50% reduction of adhesion to YIGSR (Figure 7) confirming that 37/67LR is functional for LM-111 adhesion in intestinal crypt cells.

Functional blocking of the 37/67LR inhibits cell proliferation
Taking into consideration the presence of functional membrane 37/67LR receptor for laminin adhesion in HIEC cells, we further investigated the reduction of cell proliferation in 37/67LR depleted cells observed above ( Figure 6) by conducting antibody blocking experiments on wild-type HIEC cells for a 24h period followed by BrdU incorporation. As shown in Figure 8, a significant reduction of BrdU incorporation was noted with the anti-37/67LR MLuC5 azide-free antibody. The effect appears to be specific since a non-immune serum (NI) had no effect. The blocking anti-β1 integrin mAb13 antibody [71] was used as a positive control.

Discussion
In the present study, we first analyzed the expression and function of 37/67LR in normal human intestinal epithelial cells in situ and in vitro. Previous expression studies in the human intestinal mucosa had generated conflicting results [53][54][55].
Our results show that the receptor is predominantly found in the epithelium in the intact intestine, being expressed in the undifferentiated and proliferative epithelial cells of both the developing and adult small intestine. This pattern of expression, which is in agreement with one of the aforementioned studies [55], confirms previous data reported in the developing and adult mouse at the transcript level [33]. Interestingly, the distribution of 37/67LR observed herein appears to be consistent with the expression of its characterized ligand LM-111, which is found in the intervillous area and the crypts before 17 weeks of gestation [48]. At later developmental stages and in the adult, LM-111 is replaced by a complementary gradient of LM-211 and LM-511 in the crypts [45,48,75]. The fact that these laminins contain the β1 chain suggests that they may serve as a ligand for 37/67LR. In this case, it would be interesting to investigate 37/67LR expression in samples from patients with Crohn's disease since a major redistribution of these laminins has been reported in the crypts [76].
The fact that 37/67LR appears to be associated with the undifferentiated and proliferative cell population in the normal intestine is interesting in regard to the up-regulation of its expression in colorectal cancer [77][78][79]. To further investigate this relation, we tested HIEC cells, which are normal human intestinal crypt cells [58,60], for the expression of 37/67LR. To our knowledge, this is the first time that the expression of 37/67LR has been reported in normal human epithelial cells. Interestingly, a significant reduction of 37/67LR was noted when switching HIEC cells from a proliferative to quiescent state. This phenomenon, which has also been observed in endothelial cells [80], is consistent with the low level of 37/67LR expression observed in situ in the non-proliferative cells of the villus and lower crypt. Further analysis of its distribution revealed that a significant proportion of 37/67LR was present in the cytosol and membrane fractions. Such a distribution is consistent with previous immunolocalization studies performed on cells in culture [25] and the multiple cell functions attributed to 37/67LR in the ribosome and plasma membrane [11,12,20,21,35].
As a general strategy to investigate the role of a cell component in HIEC cells [46,47,71], we undertook the knockdown of expression of 37/67LR. A standard protocol resulted in very efficient reduction of the target gene at both the transcript and protein levels and altered the apparent viability of the cells and expression of fibronectin and IGFBP7, which are genes constitutively expressed at high levels in HIEC cells [46,47,74], confirming that the 37/67LR gene product was required for translation and general cell function [17]. Interestingly, dropping the working concentration of the siRNA targeting 37/67LR sequences by half allowed a 50% reduction of 37/67LR at the protein level without significantly altering translation efficiency as tested by the maintenance of fibronectin, IGFBP7 as well as TIMP3 expression at control levels.
A partial 37/67LR depletion protocol was then used to examine the involvement of the receptor in the regulation of cell proliferation and adhesion. A significant reduction in cell cycle progression through the G1 phase was observed as a result of partial depletion of 37/67LR in these normal intestinal cells. As proposed elsewhere [35], these data suggest that threshold levels of 37/67LR expression are required for distinct cellular Figure 8. Blocking 37/67LR reduces cell proliferation. HIEC cells were treated for 1h with neutralizing-blocking anti-37/67LR antibody (MLuC5) or anti-integrin β1 antibody (mAb13) as well as non-immune serum (NI) as negative control before plating. BrdUpositive cells were counted and expressed as percentage of total cells determined by DAPI staining (mean ± SEM, n=3, * p < 0.001; ** p < 0.0001). doi: 10.1371/journal.pone.0074337.g008 functions, lower levels of 37/67LR being sufficient for normal ribosomal function but not proliferation. Taken together, these results indicate that 37/67LR regulates normal cell proliferation.
Finally, the ability of 37/67LR to interact with laminin was investigated. Laminin promotes epithelial cell adhesion in part through the CDPGYIGSR sequence on its β1 chain although its minimum ligand sequence is YIGSR [81]. Interestingly, phagedisplayed mapping revealed that at least two distinct regions of 37/67LR can bind to YIGSR [82]. Partial depletion of 37/67LR in HIEC cells resulted in a significant reduction of cell adhesion to YIGSR confirming that 37/67LR is expressed in the plasma membrane under its functional form. The relatively weak adhesion to YIGSR as compared to RGD was expected considering the expression of a number of specific RGDbinding integrins in HIEC [71] and the fact that 37/67LR may act as a co-receptor with laminin-binding integrins [11], namely the α6β4 integrin [83]. By using another strategy based on the use of blocking antibodies directed toward 37/67LR, we provided evidence for the requirement of a functional form of 37/67LR at the membrane for modulating cell proliferation.
In summary, these results confirmed the predominant expression of 37/67LR in the immature intestinal epithelium and in the crypt cells of both the fetal and adult small intestine.
Functionally, 37/67LR was found to be expressed in significant amounts in proliferating HIEC cells, a line of human intestinal crypt cells. Using an siRNA strategy that allowed partial downregulation of 37/67LR expression leaving translation normal, or by using blocking antibodies, we found a significant reduction of cell proliferation and specific adhesion to the laminin-related peptide YIGSR. Taken together, these data demonstrate that 37/67LR is expressed in normal intestinal epithelial cells where it regulates cell proliferation and adhesion.