Expression of mucosal chemokines TECK/CCL25 and MEC/CCL28 during fetal development of the ovine mucosal immune system

Abstract

CCL25/TECK and CCL28/MEC are CC chemokines primarily expressed in thymic dendritic cells and mucosal epithelial cells. The cognate receptors of CCL25 and CCL28, CCR9 and CCR10, respectively, are mainly expressed on T and B lymphocytes. In human, mouse and pig, CCL25 and CCL28 play a key role in the segregation and the compartmentalization of the mucosal immune system through recruitment of immune cells to specific locations. However, little is known about their role in the ontogeny of the mucosal immune system during fetal development. In the present paper, we report the cloning and the sequencing of ovine CCL25, CCL28, CCR9 and CCR10 and the subsequent assessment of their mRNA expression by q-polymerase chain reaction in several tissues, including thymus, gut-associated lymphoid tissue and mammary gland, from young and adult sheep and in the fetal lamb during the development of the immune system. CCL25 mRNA was highly expressed in thymus and gut while CCL28 mRNA was more expressed in large intestine, trachea, tonsils and mammary gland, especially at the end of gestation. These results are consistent with observations in other species suggesting similar roles for these chemokines in sheep. In fetuses, mRNA of CCL25, CCL28 and their receptors are expressed early in the thymus and mucosal tissues, including the small intestine and the nasal mucosa. Furthermore, their expression increased towards the end of gestation. Consequently, we hypothesize that CCL25 and CCL28 play an important role in the lymphocyte colonization of fetal tissues, enabling the development of a functional immune system.

Introduction

Chemokines are involved in regulation of inflammation, leucocyte trafficking, embryogenesis, haematopoiesis and immune cell differentiation.^1,2 Chemokines are highly basic heparin-binding proteins of 70–125 amino acids with molecular masses ranging from 6 to 14 000, which exert their functions by binding specific seven-transmembrane G-protein-coupled receptors located on the cell surface.^3,4 The CC chemokine ligand 25 (CCL25) was first identified in the mouse and human thymus and is also known as thymus-expressed chemokine (TECK).⁵ In the mouse and the pig, CCL25 is constitutively expressed in crypt epithelial and endothelial cells of the intestine and in dendritic cells of the thymus. It is involved in recruitment of T precursor cells to fetal thymi and T- and B-cell homing to the small intestine (SI).^5–13 In concert with other receptors such as integrin α4β7, which interacts with mucosal addressin cellular adhesion molecule-1 (MadCAM-1) expressed on high endothelial venules (HEV), gut lamina propria (LP) and in the gut-associated lymphoid tissues (GALT), this chemokine is responsible for lymphocyte recruitment to the gastrointestinal tract.^14,15 Chemokine receptor 9 (CCR9), the main receptor for CCL25, is specifically expressed on a subset of gut-homing T cells expressing integrin α4β7, immunoglobulin A (IgA)-secreting cells from gastrointestinal organs as well as on central nervous system microglia.^{5–7,12,13,16–20} Hosoe and colleagues¹⁷ reported that lamina propria lymphocytes (LPL) and intraepithelial lymphocytes (IEL) migrated toward both the small intestine and colon, and desensitization of CCR9 with TECK/CCL25 or anti-TECK/CCL25 antibody significantly inhibited these adhesions in the SI. In mice, migration of antibody-secreting cells (IgA-ASCs) towards CCL25 appears to contribute to localization of IgA-ASCs in the SI.^{11,19,21–24} Indeed, an inhibition of IgA plasmablast recruitment to the intestinal LP has been recently shown in rotavirus infected mice following combined treatment with anti-CCL25 and anti-CC chemokine ligand 28 (CCL28) antibodies and CCR9-deficient mice, which have a reduced number of IgA-ASCs in the LP of the SI.^21,24,25 Moreover, increased expression of CCL25 by TNF-α was observed in the LP of the SI.¹⁷ Together, these results demonstrate the important role of CCL25 in the migration of lymphocytes under non-inflamed as well as inflamed conditions.

CCL28, also called mucosae-associated epithelial chemokine (MEC), is expressed in most human, mouse and pig mucosal tissues including the salivary gland, mammary gland (MG), small and large intestines and trachea, where it appears to be predominantly produced by epithelial cells.^10,26,27 Cells expressing CCR10, a receptor known previously as orphan G-protein-coupled receptor GPR2, are mainly IgA and IgM plasmablasts, some T lymphocytes and some bone marrow cells.^27–32 These cells are thought to migrate into tissues in response to CCL28 and α4β1^high/vascular cell adhesion molecule-1 (VCAM-1) interaction.^27,29–32 Additionally, MEC also supports migration through the chemokine receptor 3 (CCR3).²⁶ Consistent with this, MEC attracts eosinophils that express CCR3 in addition to lymphocyte subsets.²⁶ In the mouse and pig, the up-regulation of CCL28 expression in the MG at the end of gestation and during lactation is consistent with the hypothesis that CCL28 plays a key role in the recruitment of IgA-ASCs into the MG, allowing the passive transfer of IgA antibodies from mother to infant.^10,33 Hence, both CCL25 and CCL28 play essential roles in intestinal homing of IgA-ASCs and T cells and with their respective receptors serve to segregate and compartmentalize the mucosal immune system.^21,34,35

Small ruminants, which include sheep and goats, are important agricultural species with an estimated population of 1·7 billion. Sheep have been used extensively as animal models for investigating the physiology of reproduction, endocrinology, cardiovascular function, pulmonary function and the immune system. An understanding of host functions, relative to disease and extensive knowledge of the physiology of this small ruminant contributes much to our ability to work effectively with this species for agrarian or scientific purposes. In fact, the sheep model has proven extremely useful for immunological research when addressing questions regarding the physiology of the immune system.³⁶

Acquisition of immunological competence involves the ordered emergence of lymphoid organs during fetal development. The syndesmochorial placentation in sheep provides a protected environment for the fetus, allowing the development of lymphoid tissues in a milieu devoid of external antigens including antibodies from the ewe. The shaping of the early immune system in the sheep is thus determined by factors in the fetus itself. Ovine and human fetuses display similar patterns of lymphoid development in thymus, spleen, lymph nodes and Peyer's patches (PP) and share similarities in the appearance of circulating leucocytes in blood.^37–39 The development of the lymphoid system in fetal lambs occurs over a protracted period of time (gestation period 150 days).³⁹ The thymus, which is the primary organ for T cells in sheep, appears as early as at the 35th day of the gestation and T cells undergo a rapid expansion over the ensuing days of development.⁴⁰ Indeed, it has been shown that sheep T cells begin to recirculate at around mid-gestation (75th day) in an antigen-free environment in the absence of circulating immunoglobulins.⁴¹ At the early stage of gestation (50–55 days) no functional lymphatic system exists and the first lymph nodes to appear in the fetus, the cervical superficial and the prescapular lymph nodes, can be discerned at 55–60 days of gestation.^39,42 The mediastinal lymph nodes (MDLN) also appear early and in the next few days nodes such as the portal, subiliaci and jejunal develop.⁴³ A connection of lymph nodes to peripheral and central lymphatics is established at 65–70 days gestation. This is when a functional blood and lymph network is established. Sheep PP are present in the jejunum and the ileum. The ileal PP, which are the major site for B-cell production in sheep, are the last of the organized lymphoid tissues to develop in the sheep fetus.⁴⁴ Ileal PP can be first identified around the 100–105th day of gestation while jejunal PP can already be identified around the 75th day.⁴⁴ Thus, at the end of gestation all the components of the mucosal immune system are developed and the fetal lamb is already able to mount a protective response against foreign antigens as observed in previous studies.^45–47 While chemokines are essential components of the immune and central nervous systems, little is known about their distribution during fetal development and their role in the recruitment of lymphocytes to the mucosal and the mucosa-associated lymphoid tissues (MALT) in fetus.

In sheep, CCL25, CCL28 and their cognate receptors were hitherto unknown and there was no data available about their expression in the different organs. In this paper we report the cloning of ovine CCL25, CCL28, CCR9 and CCR10 and their mRNA expression levels in several tissues, including thymus, gut, GALT and MG, from young and adult sheep and in the fetus during the development of the immune system. The results show an early expression of CCL25 and CCL28 mRNA during fetal development potentially implicating these chemokines in the recruitment of lymphocytes to specific locations in the developing fetus.

Materials and methods

Animals and sample collections

Suffolk sheep were obtained from the Department of Animal and Poultry Science (University of Saskatchewan, Saskatoon, Saskatchewan, Canada) and were cared for in accordance with the Guidelines of the Canadian Council for Animal Care. Selected tissues were taken from 13 adult ewes (one animal at the 25th day and two animals every 20 days from 45th to 145th day of gestation), five lambs (1 month old) and 14 fetuses (two fetuses every 20 days from 25th to 145th day of gestation). The sheep were clinically healthy. Tissues were cut in five 3 × 3 mm pieces, laid flat, washed with ice cold phosphate-buffered saline (PBS; Gibco-BRL, Burlington, ON, Canada), snap-frozen in liquid nitrogen and stored at −80° in RNA later (Sigma-Aldrich, Oakville, ON). Twenty-five-day-old whole fetuses were collected because of their very small size, while in older fetuses organs were more developed, allowing the collection of 3 × 3 mm pieces from the selected tissues. Small intestinal samples free of Peyer's patches were collected and treated as described above.

Cell isolation

Peripheral blood mononuclear cells (PBMC) were isolated using a 54% Percoll density gradient (Amersham, Piscataway, NJ) from peripheral blood collected from three 1-year-old sheep. Cells were washed once with PBS (Gibco-BRL, Burlington, ON) supplemented with 2 mm ethylenediaminetetra-acetic acid (EDTA, 325 g) and twice in PBS alone (150 g). Cells were then counted using a haemocytometer, assessed for viability using the trypan blue (Gibco-BRL) exclusion method and resuspended in PBS 2 mm EDTA supplemented with 0·5% bovine serum albumin (Sigma-Aldrich). Cells were further separated by magnetic cell sorting (MACS; Miltenyi Biotec, Auburn, CA) using immunoglobulin G1 (IgG1) mouse anti-ovine CD5 (Serotec, Oxford, UK) and IgM mouse anti-ovine CD21 (VMRD, Pullman, WA), and anti-mouse IgM and anti-mouse IgG1 microbeads (Miltenyi Biotec), according to the manufacturer's instructions. The purity of separated cell populations was verified by flow cytometry with monoclonal antibodies (mAb) specific for sheep CD4, CD8, CD72, CD5 and CD21. mAb specific for sheep CD4, CD8 and CD72 were produced from hybridomas.^48,49 mAb specific for sheep CD5 and CD21 were purchased from Serotec and VMRD. Goat F(ab′)₂ anti-mouse IgG1–fluoroscein isothiocyanate (FITC), goat F(ab′)₂ anti-mouse IgG2a–FITC and goat F(ab′)₂ anti-mouse IgG2b–FITC were purchased from Southern Biotechnology Associates Inc. (Birmingham, AL). Goat anti-mouse IgM–FITC was purchased from Cedarlane (Hornby, ON, Canada). Flow cytometric analyses were restricted to viable cells by excluding cells stained with propidium iodide (2·5 µg/ml; FL3). Specific mAb staining was determined by subtracting cells reacting with isotype-matched and concentration-matched (1–10 µg/ml) irrelevant mAbs (Caltag Laboratories, Burlingame, CA). All samples were analysed with a FACScan (BD Biosciences, Mountain View, CA) flow cytometer and the CellQuest program (BD Biosciences) was used for data acquisition and analysis.

Preparation of mRNA and cDNA

Total RNA was isolated from homogenized tissues using Trizol reagent (Invitrogen, Carlsbad, CA) and RNeasy Mini Kit (Qiagen, Mississauga, ON, Canada). Then RNA samples were treated with DNAse I Amp Grade (Invitrogen; 1 U/µg of RNA). The absence of genomic DNA contamination was verified by using prepared RNA as template in q-polymerase chain reaction (PCR). RNA was quantified by determining optical density at 260 nm (OD₂₆₀) and the OD₂₆₀/OD₂₈₀ ratio was calculated to assess purity. RNA quality was further assessed by capillary electrophoresis (Agilent 2100 Bioanalyzer). To generate the cDNA, RNA sample (4 µg) was incubated in a final volume of 40 µl with dinucleotide triphosphate (dNTP, 0·5 mm final each; Omniscript kit, Qiagen), 0·5 µg oligo dT (Invitrogen), RNAse out inhibitor (20 U) (Invitrogen), Omniscript reverse transcriptase (RT, 8 U; Invitrogen) and 1× buffer RT (Invitrogen). The reaction was allowed to proceed for 1 hr at 37° and then heat inactivated at 93° for 5 min. The cDNA generated was stored at −80°.

Cloning and sequencing of CCL25, CCL28, CCR9 and CCR10

BLAST searches of the GenBank ESTs database using known sequences of human and murine TECK/CCL25, MEC/CCL28, CCR9 and CCR10, identified expressed sequence tags (EST) for bovine CCL25, CCL28 and CCR9 (BE808113, CN794090, CK772673) and Bos taurus chromosome 19 genomic contig for CCR10 (NW929520). Primers (Table 1) were designed using Clone Manager (Scientific & Educational Software, Cary, NC) and the PCR products were cloned using Zero Blunt® TOPO® PCR Cloning Kit for Sequencing from Invitrogen. Inserts were sequenced and their homology to corresponding chemokine and chemokine receptor gene sequences determined with Clone Manager (Scientific & Educational Software). Partial nucleotide sequences between primers used for amplification of CCL25, CCL28, CCR9 and CCR10, based on results from five clones, were submitted to GenBank (CCL25: NM-001040290, CCL28: DQ181939, CCR9: NM-001040286 and CCR10: NM-001040287).

Table 1.

Primers used in this study, purpose, sequence, annealing temperature of primer sets (°C), and expected PCR fragment sizes (bp)

Primer name	Purpose	Primer sequence	Annealing temperature (°C) of primer set	cDNA PCR product (bp)
CCL25	cloning	(S) ATCCTCAAGGTAGGTGCCAGAAGGG	54	519
		(AS) GGTAGATCCTGGAGTAGTG
CCL28	cloning	(S) AGATGGGAATGCAGCAGACAGGACT	64	414
		(AS) GATCCCTGGTTGAGGACCTTCTCTA
CCR9	cloning	(S) GCCTTGAGCACCAGGTACTTG	62	1217
		(AS) TTCCAAAAGGACCGCCCACCTC
CCR10	cloning	(S) CCGGTGGCTTTACCTCAGAAAC	64	1142
		(AS) AGCCCGCCCCATCCTCTTGATTC
CCL25-RT	qPCR	(S) AGGCCCAGAGTTACTATCGC	61	132
		(AS) TCTTCATCCCAGCCTGAACC
CCL28-RT	qPCR	(S) AGCGCAGAAGAGTCTGTGGTC	59	151
		(AS) TTCCTGCCTTTCCCGATGTG
CCR9-RT	qPCR	(S) GCGCAGATGTGGAGGCAGAA	56	142
		(AS) CCACGGTGCAGATAGCAGTT
CCR10-RT	qPCR	(S) TGCCATCTCTGGCCTCTAC	60	164
		(AS) ACAACAGCCACACGATGAC
GAPDH-RT	qPCR	(S) ATGCCTCCTGCACCACCA	60	76
		(AS) AGTCCCTCCACGATGCCAA
SDHA-RT	qPCR	(S) CATCCACTACATGACGGAGCA	60	90
		(AS) ATCTTGCCATCTTCAGTTCTGCTA

Messenger RNA expression analysis using real-time PCR

Real-time PCR (qPCR) was performed using cDNA synthesized as described before. cDNA was combined with primer/probe sets and IQ SYBR Green Supermix (Bio-Rad, Hercules, CA) according to the manufacturer's recommendations. All primers were designed using Clone Manager (Scientific & Educational Software; Table 1) and were purchased from Invitrogen. The qPCR conditions were 95° for 3 min, followed by 45 cycles with denaturation at 95° for 15 s, annealing temperature (Table 1 and Fig. 1) for 30 s and elongation at 72° for 30 s. Real time assays were run on a Bio-Rad iCycler iQ (Bio-Rad). The specificity of the qPCR reactions was assessed by analysing the melting curves of the products and size verification and sequencing of the amplicons. To minimize sample variation, we chose tissue samples of similar size and location and used identical quantities of high quality RNA with no signs of degradation. Samples were normalized internally using the average cycle threshold (CT) of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and succinate dehydrogenase (SDHA) as references in each tissue.⁵⁰ GAPDH and SDHA were selected as the reference genes because of the lowest variation among a panel of six tissues including cerebrum, cerebellum, obex, spleen, mesenteric lymph node (MLN) and ileum.⁵⁰ A standard curve was generated using linearized plasmids containing chemokine and chemokine receptor DNA as template so that data could be expressed as transcript copy number per 50 ng total RNA. The correlation coefficients of the CCL25, CCL28, CCR9 and CCR10 standard curves were 0·995, 0·996, 0·993 and 0·998, respectively, and the concentration of the test samples were calculated from the standard curves, according to the formula y =–M × Ct + B, where M is the slope of the curve, Ct the point during the exponential phase of amplification in which the fluorescent signal is first recorded as being statistically significant above background and B the y-axis intercept. Only Ct values <40 were used for calculation of the qPCR efficiency from the given slope according to the equation: qPCR efficiency = (10^[−1/M] − 1) × 100. All qPCRs displayed efficiency between 92 and 100%.

Figure 1.

Amplification of GAPDH, SDHA, CCL25, CCL28, CCR9, and CCR10 mRNA by real-time PCR. After 45 cycles of amplification, the qPCR products were run on a 1·5% agarose gel and stained with ethidium bromide. One DNA molecular weight marker was used (Fermentas). The qPCR product sizes are presented in Table 1.

Statistical analysis

Data for the comparison of differences in mRNA expression between tissues in fetuses, lambs and ewes are expressed as copy numbers. After logarithmic transformation, most of the data were normally distributed as confirmed by Shapiro–Wilk normality test (using Statistix 7.0®, Analytical Software, Tallahassee, FL). Because the data were paired and normally distributed, group medians were compared using Student's paired t-test (using GraphPad Prism® software version 3.00, GraphPad Software Inc., San Diego, CA). Paired, non-normally distributed data were analysed using the Wilcoxon signed rank test (exact). Differences between groups were considered significant when P < 0·05. Only the most relevant comparisons are presented in figures and in the text. In fetuses and pregnant ewes, linear regression and coefficient of determination where calculated using GraphPad Prism® software version 3.00. The slope was considered significantly different from zero when P < 0·05.

Results

Cloning of ovine CCL25, CCL28, CCR9 and CCR10

We report here the cloning of ovine CCL25, CCL28, CCR9 and CCR10. The ovine CCL25 cDNA sequence was found to be 459 nucleotides in length, encoding a predicted precursor protein of 152 amino acids (GenBank accession number NM-001040290). At the protein level, 56, 71 and 65% identity was found to human, pig and mouse sequences, respectively. The ovine CCL28 cDNA sequence was found to be 318 nucleotides in length without the sequence of the peptide signal encoding a protein of 105 amino acids (GenBank accession number DQ181939). At the protein level, 78, 84 and 74% identity was found to human, pig and mouse sequences, respectively, when comparing the protein without peptide signal. The ovine CCR9 and CCR10 cDNA sequence were found to be 1104 and 1095 nucleotides in length, encoding a precursor protein of 367 and 364 amino acids, respectively (GenBank accession number NM-001040286 and NM-001040287). The receptors were highly conserved at the protein levels. The CCR9 protein, showed a 91, 89 and 91% identity to human, pig and mouse sequences, respectively, and CCR10 displayed a 93, 96 and 89% identity to human, pig and mouse sequences.

Expression of CCL25, CCL28, CCR9 and CCR10 mRNA in various tissues from 1-month-old lambs and adult ewes

Because of their crucial role in lymphocyte trafficking, we assessed mRNA expression of CCL25, CCL28, CCR9 and CCR10 in various tissues from young lambs and adult ewes.

Different mucosal and lymphoid tissues such as the SI, the large intestine (LI), the PP, the MLN, the lung, the trachea, the MDLN, the pharyngeal tonsil, the palatine tonsil, the retropharyngeal lymph node (RPhLN) and the thymus were collected from five 1-month-old lambs and the mRNA expression of the two chemokines and their receptors was analysed by qPCR. CCL25 mRNA was detected in all tissues (Fig. 2); however, the highest levels of expression were detected in the SI, the PP and the thymus. For all other tissues, levels of expression were very low.

Figure 2.

Expression of CCL25, CCL28, CCR9, and CCR10 mRNA in various tissues from one-month-old lambs. Expression of CCL25, CCL28, CCR9, and CCR10 mRNA were measured in small intestine (SI), large intestine (LI), Peyer's patch (PP), mesenteric lymph node (MLN), lung, trachea (TR), mediastinal lymph node (MDLN), pharyngeal tonsil (Ph tonsil), palatine tonsil (Pa tonsil), retropharyngeal lymph node (RPhLN), and thymus (THY). cDNA was synthesized from 50 ng of RNA and subjected to RT–qPCR using primer/probe sets for CCL25, CCL28, CCR9, CCR10 and GAPDH. Levels of expression are expressed as copy number per 50 ng total RNA. Values and median from five animals are shown in each tissue. All groups were compared to SI in CCL25 graph, to thymus in CCR9 graph and to LI in CCL28 and CCR10 graphs. *P < 0·05, **P < 0·01 (Student's t-test and Wilcoxon signed rank test (exact)).

The expression of CCL28 mRNA was high in the SI, LI, trachea, and pharyngeal and palatine tonsils, with the highest expressions observed in the LI (Fig. 2). In this tissue, the expression of CCL28 mRNA was significantly higher than in the in the SI (P < 0·05) and all the other tissues except the pharyngeal and the palatine tonsils. CCL28 mRNA was also detected in lymph nodes and thymus. Nevertheless, mRNA levels of expression were extremely low in these tissues.

CCR9 mRNA was highly expressed in the thymus (Fig. 2). The expression of CCR9 mRNA was also detected in the lymph nodes and the SI but contrary to CCL25 expression, no significant differences between the selected tissues were observed.

The highest levels of expression of CCR10 mRNA were observed in the LI and the RPhLN. The levels of expression observed in the LI were significantly higher than those observed in the PP, the MLN and the trachea (P < 0·05). In all other tissues, levels of expression were similar.

In summary, the expression of CCL25 mRNA in lambs was high in the SI and the thymus. The highest expression of CCL28 mRNA was reported in the LI. Levels of expression of CCR9 mRNA were extremely high in the thymus and highest levels of expression of CCR10 mRNA were detected in the LI.

Messenger RNA expression was also assessed by qPCR in various adult tissues consisting of: (i) pieces of a digestive organ, the ileum/jejunum and associated lymphoid tissues, PP and MLN; (ii) pieces of a secondary lymphoid tissue, the spleen; (iii) pieces of the lung; and (iv) pieces of the MG which is implicated in the transfer of the maternal immunity to the young.

Similar to lamb tissues, the highest levels of expression of CCL25 mRNA were detected in the adult SI (ileum/jejunum). In the PP, the expression was also high while in other collected tissues (MLN, spleen lung and MG) they were extremely low. The levels of expression observed in the ileum/jejunum and PP were significantly higher than those observed in the other tissues (P < 0·01). The highest levels of CCR9 mRNA expression were detected in the PP, the ileum/jejunum and the MLN (Fig. 3). The levels of expression of CCR9 mRNA were significantly higher in the PP than in the spleen, the lung and the MG (P < 0·05).

Figure 3.

Expression of CCL25, CCL28, CCR9, and CCR10 mRNA in various tissues from adult sheep. Expression of CCL25, CCL28, CCR9, and CCR10 mRNA were measured in ileum/jejunum (Ile/Jej), Peyer's patch (PP), mesenteric lymph node (MLN), spleen, lung, and mammary gland (MG). cDNA was synthesized from 50 ng of RNA and subjected to RT–qPCR using primer/probe sets for CCL25, CCL28, CCR9, CCR10 and GAPDH. Levels of expression are expressed as copy number per 50 ng total RNA. Values and median from 13 animals are shown in each tissue. All groups were compared to Ile/Jej in CCL25 graph, to PP in CCR9 graph, to Ile/Jej in CCL28 graph and to PP in CCR10 graph. *P < 0·05, **P < 0·01 (Student's t-test and Wilcoxon signed rank test (exact)).

CCL28 mRNA was highly expressed in the adult ileum/jejunum and in the MG (Fig. 3). In the PP, the levels of expression were still high but significantly lower than in the ileum/jejunum (P < 0·05). Expression in other tissues was very low (P < 0·01). The trend was different for CCR10 (Fig. 3), as the highest levels of expression were measured in the PP followed by ileum/jejunum and the spleen. In the MLN, CCR10 mRNA was detected but it was significantly less than in the PP (P < 0·05). In the lung, expression was similar to what we observed with CCL28, the levels of expression were very low, significantly lower than the PP (P < 0·01) and inconsistent.

Having analysed the expression of CCL25, CCL28, CCR9 and CCR10 in various tissues from lambs and adult ewes, we were interested in identifying the cells that express the respective receptors. Consequently, two general lymphocyte markers were selected for sorting PBMC, namely CD5 and CD21. The CD5 receptor,^51,52 a type I glycoprotein expressed by T cells and a subset of B cells, is thought to play a significant role in modulating antigen receptor signalling. CD21⁵³ was initially identified as a B-cell differentiation antigen expressed on mature B cells. In PBMC, mRNA expression was extremely low for CCL25, barely detectable for CCL28 and high for CCR9 and CCR10, with levels of expression comparable to what we observed in MLN (data not shown). In CD5⁺, CD5^–, CD21⁺ and CD21^– enriched cells, the levels of expression of CCR9 and CCR10 mRNA were not significantly different amongst the populations (data not shown) suggesting that these two receptors were similarly expressed both in blood B and T cells.

Expression of CCL25, CCL28, CCR9 and CCR10 mRNA in fetal tissues during fetal development

To investigate the role of these chemokines in the development of the MALT, we analysed the expression of CCL25, CCL28, CCR9 and CCR10 mRNA in fetal tissues during gestation. To this end, we collected pieces of thymus, duodenum, ileum/jejunum, spleen, lung and nasal mucosa from fetal lambs at different times during gestation and measured the mRNA expression of CCL25 and CCL28 chemokines and their cognate receptors CCR9 and CCR10 (see Figs 5 and 6).

Figure 5.

Expression of CCL28 and CCR10 in various tissues collected from fetuses at different times during gestation. Expression of CCL28 and CCR10 was measured in thymus, duodenum, ileum/jejunum, spleen, lung, and nasal mucosa from fetuses at different times of gestation (45, 65, 85, 105, 125, and 145 days). cDNA was synthesized from 50 ng of RNA and subjected to RT–qPCR using primer/probe sets for CCL28, CCR10, and GAPDH. For each graph, linear regression was calculated and P and r² values indicated beside the regression line.

Figure 6.

Expression of CCL28 mRNA in the mammary tissues of pregnant ewes at different times of the pregnancy. Expression of CCL28 was measured in the mammary tissue from two ewes at each time during pregnancy (45, 65, 85, 105, 125, and 145 days). cDNA was synthesized from 50 ng of RNA and subjected to RT–qPCR using primer/probe sets for CCL28 and GAPDH. The linear regression has been calculated and P and r² values indicated beside the regression line.

At 25 days of gestation, expression of CCL25 and CCR9 mRNA was not detected consistently (data not shown). In contrast, higher levels of expression of CCL25 mRNA were detected at 65 days of gestation in the thymus and at 100 days of gestation in the duodenum, the ileum/jejunum and to a lower extent in the spleen (Fig. 4). In these tissues, a significant increase in the production of the chemokine mRNA was observed during fetal development with levels of expression measured at 125 days similar to those observed in 1-month-old lambs (P < 0·05 in duodenum and <0·01 in thymus, ileum/jejunum and spleen. The coefficients of determination were: 0·63, 0·45, 0·92 and 0·82, respectively; Figs 4 and 5). In the thymus, expression of CCL25 mRNA was still low at 45 days of gestation, with corrected CT around 28, and was high at 65 days, with corrected CT around 22, indicating that induction of expression for this chemokine mainly took place after the 6th week of gestation. Unlike in the thymus, the duodenum, the ileum/jejunum and the spleen, levels of expression of CCL25 mRNA in the lung and in the nasal mucosa stayed very low and inconsistent throughout gestation. For CCR9, there was a significant increase in the mRNA levels of expression in the thymus (P < 0·05, r² = 0·47) and the ileum/jejunum (P < 0·01, r² = 0·68) but no significant increase in the duodenum and the spleen even though the levels of expression in these tissues were higher at the end of gestation (Fig. 4). Both CCL25 and CCR9 had higher levels of expression of mRNA and were detected earlier in the thymus (already around the 65th day) than in the gut or in the spleen (around the 85th day in the ileum/jejunum and around the 105th day in the spleen). In the duodenum and the ileum/jejunum, CCR9 mRNA expression showed a strong increase between 100 and 125 days.

Figure 4.

Expression of CCL25 and CCR9 in various tissues collected from fetuses at different times during gestation. Expression of CCL25 and CCR9 was measured in thymus, duodenum, ileum/jejunum, spleen, lung, and nasal mucosa from fetuses at different times of gestation (45, 65, 85, 105, 125, and 145 days). cDNA was synthesized from 50 ng of RNA and subjected to RT–qPCR using primer/probe sets for CCL25, CCR9, and GAPDH. For each graph, linear regression was calculated and P and r² values indicated beside the regression line.

Unlike CCL25, the levels of expression of CCL28 mRNA in the thymus was low throughout gestation, while there was a significant increase with time in the duodenum (P < 0·01, r² = 0·6), the ileum/jejunum (P < 0·01, r² = 0·75), the spleen (P < 0·01, r² = 0·62) and the nasal mucosa (P < 0·01, r² = 0·88; Fig. 5). As observed for CCL25, higher levels of expression were detected after the 85th day of gestation in the SI and the spleen but also in the nasal mucosa. For the cognate receptor of CCL28, CCR10, expression levels were significantly increased in the ileum/jejunum (P < 0·01, r² =0·64) and in the spleen (P < 0·01, r² = 0·63) but no significant increase in the duodenum and the nasal mucosa (Fig. 5) was observed. In these tissues, the levels of expression of the receptor stayed low even at the end of the gestation. Similar to CCR9 (Fig. 4), a pronounced increase in the mRNA expression of CCL28 and CCR10 was observed in the duodenum and the jejunum/ileum between 100 and 125 days (Fig. 5). Interestingly, for CCL28, the levels of expression observed in the MG from the ewes were higher at the end of gestation (Fig. 6). The increase during gestation was significant (P < 0·01, coefficient of determination: 63%; Fig. 6).

In summary, we observed a progressive and significant increase during fetal development in the expression of both CCL25 and CCR9 mRNA in the thymus and the ileum/jejunum and a significant increase of CCL25 mRNA not accompanied with CCR9 mRNA in the duodenum and the spleen. We observed both an increase in the mRNA expression of CCL28 and CCR10 in the ileum/jejunum and in the spleen and an increase not accompanied with CCR10 in the duodenum and the nasal mucosa.

Discussion

The main finding of the present study is that the mRNA expression of ovine CCL25 and CCL28 and their receptors takes place very early during fetal development and progressively increases towards birth suggesting a contribution of these chemokines in the early development of the mucosal immune system. We also show that CCL25 is mainly associated to the intestinal mucosal tissues and thymus, while CCL28 is associated with epithelial mucosal surfaces. Furthermore, CCL28 mRNA expression in the MG progressively increases during the gestation indicating the important role of this chemokine in ensuring passive transfer of maternal antibodies to the offspring.

As observed in humans, mice and pigs, high levels of CCL25 mRNA were detected in the SI and the thymus of lambs.^{5,7,10,11,17,19} Similar to pigs, we detected expression of CCL25 mRNA in all the other selected tissues, albeit at very low levels, suggesting that this chemokine may be produced by cells other than epithelial cells, such as dendritic cells.⁵ However, given the Monte Carlo effect⁵⁴ and because mRNA levels of expression are not always directly correlated with the protein expression, hypothesis regarding the biological significance of low mRNA levels in the protein production have to be considered cautiously. Concerning the cognate receptor of CCL25, CCR9, high levels of expression of the mRNA were detected in the thymus from the lambs and in the ileum/jejunum, the PP and the MLN from the adult ewes. These results are consistent with previous observations in other species.^{9,10,13,24,55} For CCL28, mRNA was detected in every mucosal tissue tested. Nevertheless, the highest expression levels were observed in the LI, the tonsils and the MG. This observation is consistent with previous studies in human, mouse and pig showing high levels of expression of CCL28 in the colon, the palatine and pharyngeal tonsils, the trachea and the MG.^10,26,27,56 Our finding that CCL28 had slightly higher expression levels in the pharyngeal than in the palatine tonsils is consistent with results obtained in swine studies.⁵⁶ However, there was not a significant difference between the expressions in the palatine and the pharyngeal tonsils given the huge variation observed among lambs. This variation may be caused by an inflammation mediated expression of CCL28 in some tonsils. Recently, a study in human liver has shown that epithelial inflammation process can induce high production of CCL28.⁵⁷ Concerning CCR10, the highest levels of expression were detected in the LI and in the RPhLN in lamb tissues and in PP of ewes.

In fetuses, the highest levels of expression of CCL25 mRNA were observed in the thymus, the duodenum, the ileum/jejunum and more interestingly in the spleen. In particular, high expression levels of CCL25 were found in the thymus of 65-day-old fetuses. In fact, these expression levels were similar to expression levels observed in adult sheep suggesting that CCL25/CCR9 contribute to the development of the ovine thymus. Indeed, the ovine thymus starts to develop around the 35th day of gestation but continues to develop during gestation.^40,58 Similarly, it was recently reported in mice that the in vitro attraction of T-lymphoid progenitor cells to the fetal thymus rudiment is mediated by chemotactic signals and is impaired by neutralizing antibodies specific for CCL21 and CCL25.⁸ Moreover, in mice deficient for CCL21 or CCR7, which is the receptor for CCL21, the number of fetal thymocytes was lower than that in normal mice until embryonic day 14·5 (E14·5).⁸ Furthermore, mice deficient for CCR9 exhibited a three-fold decrease in total thymocyte cellularity until E17·5.²⁵ and double deficiency of CCR7 and CCR9 prevented early fetal thymus colonization before vascularization.⁹ Consequently, it seemed that the co-ordination between CCL21/CCR7- and CCL25/CCR9-mediated chemokine signals was essential for guiding fetal thymus colonization before thymus vascularisation.⁹ In human, an unique tissue-specific multistep adhesion cascade mediating the homing of lymphoid-committed progenitors to the thymus has been described. This cascade involves P-selectin/PSGl-1 interactions, signalling of CCL25 through its receptor CCR9, and activation-induced binding of the integrins α4β1 and αLβ2 to their respective ligands, VCAM-1 and intracellular adhesion molecule-1.⁵⁹ Thus, these findings together suggest that CCL25 plays an important role in the development of the fetal thymus. Similarly to what we observed in the thymus, there was also a progressive increase in the expression of CCL25 mRNA in the ileum/jejunum. However, expression was delayed until 85 days gestation with levels of expression similar to those detected in adults. Again, these results are consistent with previous studies showing that fetal small intestine produces CCL25 and PP starts to develop later than the thymus and the other secondary lymphoid tissues.^26,44 Indeed, in the jejunum, primordia of PP can be first detected at 60–75 days gestation and lymphoid follicles after 75 days. Ileal PP can be first identified around the 100–105th day of gestation.⁴⁴ Contrary to the thymus and the ileum/jejunum, the gradual increase in the expression of CCL25 mRNA in the duodenum was not clearly accompanied by an increase of CCR9 mRNA expression. It may be caused by the absence of PP in this section of the gut, by a delay in the migration of cells expressing CCR9 to this organ or by a down-regulation in the expression of CCR9 after cells arrive in the duodenum LP. It has been shown in mice that IgA-ASCs express high levels of CCR9 in the MLN and PP, but down-regulate CCR9 once they are located in the SI.²⁴ However, as observed in the ileum/jejunum, there was also a strong, but statistically not significant increase in the CCR9 mRNA expression in the duodenum between 100 and 125 days. This sharp increase could be a direct consequence of the establishment of the PP in the ileum and the jejunum around 100 days and the consecutive lymphocyte migration to the duodenum LP. The same observation was made for CCR10 and CCL28 in digestive tissues. In addition, we also detected a gradual increase in the expression of CCL25 mRNA in the spleen. This is in contrast to previous observations and suggests a potential role of CCL25 in the trafficking of CCR9 cells to this secondary lymphoid tissue. Unexpectedly, we did not detect any increase in the expression of CCR9 mRNA in the spleen suggesting either a delay in the migration of CCR9 cells to the spleen or an alternative function for CCL25, like the attraction of lymphocytes expressing chemokine receptor 11 (CCR11) to this organ. Indeed, CCR11, which is primarily expressed in heart, lung and SI has been found to be capable of binding several chemokines including CCL25.^60–62 In conclusion, significant levels of expression of CCL25 mRNA were detected in the fetal thymus, SI and spleen suggesting the implication of this chemokine in the colonization of these tissues with a first population of immune cells. This hypothesis is supported by the detection of high levels of lymphocyte CCR9 mRNA expression following CCL25 mRNA expression in the fetal thymus and ileum/jejunum.

In fetuses, levels of CCL28 mRNA expression were generally low except in the nasal mucosa at the end of gestation where they reached similar values to those observed in adult ileum/jejunum. However, we detected a significant increase in the production of this chemokine mRNA in the duodenum, the ileum/jejunum, the spleen and the nasal mucosa. Regarding the expression of its cognate receptor, CCR10, results showed a significant increase in the mRNA production in the spleen and in the ileum/jejunum but not in the duodenum and in the nasal mucosa. In the duodenum, this could be the result of the small number of copies of CCL28 mRNA in this organ at the end of the gestation, potentially not sufficient to attract lymphocytes. It could also be explained by a delay in the migration of cells expressing CCR10, by a down-regulation in the expression of CCR10 once cells are located in the duodenum LP and epithelium, or by the interaction of CCL28 with its other receptor, CCR3, or with a receptor not yet identified. Indeed, CCL28 and CCL25 are both known to interact with at least two receptors. CCL28 interacts with CCR3 and CCR10, while CCL25 interacts with CCR9 and CCR11.^25,26 Both could interact with additional receptors. It is also important to note that conversely a chemokine receptor, like CCR10, can also interact with several chemokines, including CCL27 and CCL28.²⁷ These various interactions between chemokines and receptors may explain some observed discrepancies between chemokines and chemokine receptor mRNA expression. In addition the kinetics of lymphocyte migration in response to chemokines has to be further assessed. In conclusion, significant levels of expression of CCL28 mRNA, potentially associated to the recruitment of immune cells, were detected in the fetal SI, spleen and particularly nasal mucosa. Similar to CCL25, CCL28 is already produced in fetuses where it may induce the migration of CCR10 cells to specific locations.

In summary, we have cloned CCL25, CCL28 and their receptors in sheep and assessed gene expression in tissues from the lamb, the ewe and the fetal lamb at different stages of the gestation. Previous studies have shown that the immune system of the fetal lamb is able to mount an efficient mucosal immune response before birth.^45–47 Our results demonstrate that mRNA of CCL25, CCL28 and their lymphocyte receptors are expressed early in the thymus and mucosal tissues, including the gut LP and the nasal mucosa, before birth. Consequently, we postulate that CCL25 and CCL28 could play an important role in the lymphocyte colonization of fetal tissues contributing to the development of a functional immune system.