Surface Glycan Pattern of Canine, Equine, and
Ovine Bone Marrow-Derived Mesenchymal
Stem Cells
Salvatore Desantis,
1* Gianluca Accogli,
1Antonio Crovace,
1Edda G. Francioso,
1Alberto Maria Crovace
2Abstract
The use of bone marrow-derived mesenchymal stem cells (MSCs) for clinical and experimental studies is increasing, but full characterization of MSCs in veterinary spe-cies is hindered by the variability in spespe-cies-specific cell surface marker expression and antibody cross reactivity. Recent studies demonstrated that the glycans in the glycocalyx of MSCs are promising candidates as cell biomarkers. In the present study, we analyzed the glycocalyx of canine MSCs (cMSCs), ovine MSCs (oMSCs), and equine MSCs (eMSCs) using a cell microarray procedure in which MSCs were spotted on microarray slides and incubated with a panel of 14 biotinylated lectins and Cy3-conjugated strepta-vidin. The signal intensity was then detected using a microarray scanner. The lectin-binding signals indicated that the MSC surface of the investigated species contained both N- and O-linked glycan types, with N-glycosylation predominating over O-glyco-sylation and fucoO-glyco-sylation being more abundant than sialylation. Relative quantification revealed an interspecific difference between these glycans. In addition, cMSCs expressed more a2,3-linked sialic acid (MAL II), terminal lactosamine (RCA120), and a1,6 and
a1,3 fucosylated oligosaccharides (PSA, LTA); oMSCs exhibited more T antigen (Jaca-lin), GalNAca1,3(LFuca1,2)Galb1,3/4GlcNAcb1 (DBA), chitotriose (succinylated WGA), and a1,2-linked fucose (UEA I); and eMSCs showed a higher density of a2,6 sialic acids (SNA) and high mannose N-glycans (Con A). Using cell microarray meth-odology, we have for the first time demonstrated differences in the glycosylation pro-files of cMSC, oMSC, and eMSC surfaces. These results could be valuable as resources and references for MSC differentiation and molecular remodeling in clinical cell-based therapy and tissue engineering studies. VC 2017 International Society for Advancement of Cytometry
Key terms
mesenchymal stromal cells; dog; horse; sheep; glycoconjugates; glycomics; cell microarray
B
ONE marrow (BM)-derived mesenchymal stem cells (MSCs) are non-hematopoietic adult multipotent cells characterized by their plastic adherence, capac-ity to differentiate into a variety of tissues (including fat, bone, and cartilage) and to migrate to diseased organs (1,2), strong immunosuppressive properties, and secre-tion of regenerative factors (3). Therefore, MSCs are promising for applicasecre-tion in cell-based therapy.Human MSCs (hMSCs) express several surface markers, the most prominent of which, CD73, CD90, CD105, CD166, and Stro-1, are not specific (4). Therefore, a deeper knowledge of surface markers for precise labeling and tracking of MSCs is of paramount importance, especially in regenerative medicine.
The MSC surface is coated with a glycocalyx, a dense layer composed of many different complex carbohydrates, which represents the interface between the cell and
1Section of Veterinary Clinics and Animal Productions, Department of Emergency and Organ Transplantation (DETO), University of Bari Aldo Moro, Bari, Italy 2Dottorato di Ricerca in Sanita e Scienze
Sperimentali Veterinarie, University of Perugia, Perugia, Italy
Received 3 May 2017; Revised 24 July 2017; Accepted 23 August 2017
Grant sponsors: ONEV (Omica e Nanotec-nologie applicate agli Esseri Viventi per la diagnosi di malattie) MIUR PONa3 00134–n.254/R&C 18/05/2011 (2013–14), and Fondo di Sviluppo e Coesione2007–13 – APQ Ricerca Regione Puglia ‘Pro-gramma regionale a sostegno della spe-cializzazione intelligente e della sostenibilita sociale ed ambientale – FutureInResearch’
*Correspondence to: Salvatore Desantis, Department of Emergency and Organ Transplantation (DETO), University of Bari Aldo Moro, S.P. Casamassima Km 3, 70010 Valenzano (BA), Italy. E-mail: salvatore.desantis@uniba.it Published online 14 September 2017 in Wiley Online Library (wileyonlinelibrary. com)
DOI: 10.1002/cyto.a.23241
VC2017 International Society for Advancement of Cytometry
asparagine. The O-linked (mucin-type) oligosaccharides clas-sically contain a reducing terminal N-acetylgalactosamine (GalNAc) that is O-glycosidically linked to serine or threo-nine. The composition and structure of glycoconjugates can be decoded with several techniques. Lectins are particularly well suited for discriminating glycoconjugates because of their specificity and ability to distinguish sugar isomers, as well as branching, linkage, and terminal modifications, of complex glycans (11). Lectins have been used for hMSC glycomic anal-ysis in several techniques, such as flow cytometry (12), lectin histochemistry (7), and lectin microarray (13,14). These tech-niques are laborious and time-consuming, however, and the ability to measure large sample sets is limited. Membrane gly-coproteins are known to be difficult to handle and immobili-zation of functional proteins on protein arrays is a real challenge, as the immobilized proteins must be presented in their native functional state. This is a difficult task with trans-membrane proteins because the optimal environment for the stabilization of cell surface proteins is the cell membrane itself. Lectin microarray methodology has been used to investigate the surface of human embryonic stem cells (hESCs) (14) and induced pluripotent stem cells (13,14) and was performed on extracted glycoproteins from cell membrane fractions. Thus, cell microarrays prepared by direct printing of cells on a microarray slide is a promising alternative for in situ charac-terization of cell surface antigens. More recently, our group developed a cell microarray method for the in situ profiling of a mammalian cell surface glycosylation pattern (15,16). Cell microarray analysis enables high-throughput and sensitive analysis of a large set of biological samples and provides a snapshot of cell profiling.
The use of MSCs in veterinary medicine for clinical and experimental studies is increasing, and long-term studies have shown no adverse effects with the administration of MSCs in several mammals. The full characterization of MSCs from vet-erinary species (including dog, horse, and sheep) is hindered, however, by variability in species-specific cell surface marker expression and antibody cross reactivity. Although CD29, CD44, and MHC I were highly expressed in canine, ovine, and equine MSCs (17–19), CD 73 and CD90 expression decreased from ovine to canine to equine MSCs (17–20)
To date no reports of MSC glycan pattern profiling in veterinary species are available. Therefore, in this study, we investigated the glycan pattern of canine, ovine, and equine MSCs to highlight species-specific glycan markers using a cell microarray procedure. This could be of value, considering that the glycomic analysis of stem cells has led to a novel
informed consent was obtained from the owners.
The BM samples were collected from the iliac crest of healthy dogs (n 5 3, Pinscher breed, 10- to 13-month old), sheep (n 5 3, Bergamasca breed, 3-year old), and horses (n 5 3, Italian saddle breed, 6- to 7-year old), in accordance with the veterinary medicine guideline cited above regarding procedures for the use of BM-derived cells in clinical cases. All animals were considered generally healthy on the basis of clinic visits and laboratory analyses. Canine and equine BM samples were collected from patients to be submitted to cell therapy for the treatment of local lesions. In particular, the dogs were affected by Legg-Calve`-Perthes disease and the horses presented localized tendon lesions. The sheep were involved in an experimental study for which the authorization of the Ministry of Health was obtained (N.483/2015-PR).
The collecting procedure was performed with the follow-ing protocols for sedation and local anesthesia in sheep and horses and sedation and general anesthesia in dogs:
Dogs: The dogs were pre-anaesthetized with aceproma-zine (0.02 mg/kg i.v.), and induction and general anesthesia was maintained with propofol (5 mg/kg i.v.). The BM was harvested with an 18-gauge Jamshidi needle. A 20 ml heparin-ized (2,500 I.U./20 ml BM) syringe was attached to the needle to collect the BM.
Sheep: Diazepam (0.05 mg/kg i.v.) and local anesthesia with lidocaine chlorhydrate 2% (20 ml) was used in the sheep around the tuber coxae. After aseptic preparation of the skin, a 14-gauge Jamshidi needle was inserted in the tuber coxae to a depth of 3–4 cm. A 20 ml heparinized (2,500 I.U./20 ml BM) syringe was attached to the needle to collect the BM.
Horses: The horses were sedated with detomidine (20 mg/ kg i.v.) and butorphanol (10 mg/kg i.v.) and restrained in a standing stock. Following aseptic skin preparation and subcu-taneous infiltration of 20–30 ml 2% lidocaine, the BM was harvested using a 14-gauge Jamshidi needle. A 50 ml heparin-ized syringe (2,500 I.U./20 ml BM) was attached to the needle to collect the BM.
Cell Culture
The BM samples were diluted 1:1 in phosphate-buffered saline (PBS) (EuroClone S.P.A., Milano, Italy) and then strati-fied 1:1 on Lympholyte-H (gradient 1.077 g/ml) (EuroClone S.P.A., Milano, Italy) and centrifuged at 380 g for 30 min. The mononuclear fraction of each sample was collected, rinsed in PBS, and centrifuged at 380 g for 10 min. The cells were seeded in flasks at a concentration of 4–5 3 106cells/cm2in
DMEM (Dulbecco’s Modified Eagle’s Medium (EuroClone S.P.A., Milano, Italy) to which was added 1% penicillin/strep-tomycin, a 5% solution of 200 mML-glucosamine, and 10%
fetal bovine serum, and then incubated at 378C in a humid 5% CO2atmosphere. Non-adherent cells were discarded after
3 days and adherent cells were cultured until they reached the number necessary for the microarray analyses (1 3 106cells/
ml). The cultures of bone marrow stem cells (BMSCs) at the third passage (Fig. 1) were trypsin-treated in a routine manner and, after collection in centrifuge tubes, the detached cells were rinsed three times in PBS and centrifuged at 380 g for 10 min.
Cell Microarray Fabrication
A cell microarray was prepared according to the method of Accogli et al. (2017) (16). Briefly, after centrifugation, the pellet was resuspended in approximately 50 ml of printing buffer PBS (0.01 M, pH 7.4) (about 1 3 106 BMSCs) and transferred by spotting (in three replicates) onto three-dimensional thin film-coated glass slides (Nexterion Slide H; Schott, Jena, Germany). Spotting was performed with a non-contact microarray printing robot (sciFLEXARRAYER S1; Sci-enion, Berlin, Germany). Each sample was spotted in 10 repli-cates per line within each array (Fig. 1).
Approximately 500 pl of cell suspension was spotted per one spot. The transfer efficiency and reproducibility of the
printing process was assessed by a test array of 400 replicates (20 3 20 spots) generated in a single run. Printing of the 0.01 M PBS buffer was also performed to check the back-ground signal produced by the lectin incubation. The slides were then held in a humidity chamber (50–70%) at 378C for 1 h to ensure sufficient attachment of the cells to the slide sur-face. The unoccupied surface of the slides was blocked with 1 M ethanolamine dissolved in 0.01 M PBS with 0.05% Tween 20 (PBST) at room temperature for 1 h. Blocked slides were gently washed with PBST (two rinses, 2 min each).
Cell Microarray-Lectin Binding Procedure
A multi-well incubation chamber was applied to the sur-face of the spotted slides, in line with the arrays that were cre-ated during the printing process. The biotinylcre-ated lectins that were used, along with their concentration and sugar specific-ity, are listed in Table 1. Each lectin (Vector Laboratories, Bur-lingame, CA) was diluted in PBST at an optimized concentration to obtain the highest specific signal with the lowest background (Table 1) and loaded directly onto the samples for all 12 subarrays. Samples were allowed to react with 50 ml of lectin solution at room temperature for 1 h. Lec-tins were then removed by gentle immersion of the slides in
Figure 1. Morphological characteristics of canine (A), ovine (B), and equine MSC cultures at the third passage observed under a phase contrast microscope. MSC monolayers are composed of spindle-shaped cells. Scale bars: 60 lm.
Table 1. Lectins used, their sugar specificities, and the inhibitory sugars used in control experiments
LECTIN ABBREVIATION SOURCE OF LECTIN CONCENTRATION (lg/ml) SUGAR SPECIFICITY INHIBITORY SUGAR
MAL II Maackia amurensis 15 NeuNAca2,3Galb1,3(6NeuNAca2,6) GalNAc
NeuNAc SNA Sambucus nigra 15 Neu5Aca2,6Gal/GalNAc NeuNAc Jacalin Artocarpus integrifolia 15 Galb1,3GalNAc b-D-Gal PNA Arachis hypogaea 25 Galb1,3GalNAc b-D-Gal RCA120 Ricinus communis 20 Galb1,4GlcNAc Gal
GSA I-B4 Griffonia simplicifolia 20 aGal Gal
DBA Dolichos biflorus 25 GalNAca1,3(LFuca1,2)Galb1,3/4GlcNAcb1 D- GalNAc
SBA Glycine max 20 a/bGalNAc D-GalNAc Con A Canavalia ensiformis 15 aMan>aGlc Methyl-a-Man succWGA Triticum vulgaris 15 bGlcNAc D-GlcNAc GSA II Griffonia simplicifolia 20 D-GlcNAc D-GlcNAc PSA Pisum sativum 20 L-Fuca1,6GlcNAc a-L-Fuc
UEA I Ulex europaeus 20 L-Fuca1,2Galb1,4GlcNAcb a-L-Fuc LTA Lotus tetragonolobus 25 aL-Fuc a-L-Fuc
Abbreviations: Fuc, fucose; Gal, galactose; GalNAc, N-acetylgalactosamine; Glc, glucose; GlcNAc, N-acetylglucosamine; Man, mannose; NeuNAc, N-acetylneuraminic (sialic) acid; succ, succinylated WGA.
PBST and immediately washed in PBST for 5 min. Subse-quently, each subarray was incubated with Cy3-conjugated streptavidin (Jackson Immuno Research Laboratories, Balti-more, PA) at 0.5 lg/ml in PBST for 15 min. Redundant Cy3-conjugated streptavidin was removed and the slides were washed with PBST and then with distilled water (4 min each). Lastly, residual water was removed using a slide centrifuge at 1,600 g (Arrayit Corporation, Sunnyvale, CA).
Scanning and Data Analysis
Images of the stained cell microarray slides were taken with the InnoScanVR
710 fluorescent scanner (Innopsys, Car-bonne, France) set at the appropriate excitation wavelength for Cy3 (570 nm) (Fig. 1). The resolution was set at 3 lm for quantification purposes. The slide images obtained in the lectin-based cell microarray analysis were evaluated using Mapix5.5.0 software (Innopsys, Carbonne, France) and the fluorescence intensity signals of each spot were measured by
background subtraction. The raw numeric values, correspond-ing to the detected intensity of spots from cell microarrays, were normalized by considering the mean of the pixel-by-pixel ratio of pixel-by-pixel intensity wavelength (570 nm) with the median background intensity subtracted (ratios >100 and <0.01 were excluded when calculating this value). The numeric values obtained were reported as the value of fluores-cence intensity relative to the average number of cells per spot. Relative quantification of N- and O-linked glycosyla-tions, as well as sialylation and fucosylation, was calculated from peak intensities of each glycan among the intensities of the total lectin used (10).
Lectin Histochemistry
Microarray results were validated using lectin histochem-istry, which is the conventional method used for the in situ detection of carbohydrate residues in glycan chains. MSCs seeded on poly-L-lysine-coated glass slides were fixed by air
Figure 2. Raw signals of lectin-based cell microarray analysis of canine, ovine, and equine MSCs. Each sample was spotted in 10 repli-cates within each array. s-WGA, succinylated WGA. [Color figure can be viewed at wileyonlinelibrary.com]
drying and incubated for 1 h in the dark with the same panel of lectins used for the cell microarray. Slides were subse-quently rinsed in the same buffer and mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA). Controls included (1) incubation of spotted MSCs with the buffer lacking lectins, (2) incubation of spotted MSCs with each lectin in the presence of its hapten sugar (0.05 M in Tris-buffered saline), and (3) incubation of spotted buffer (PBS without cells) with each lectin solution (only for cell microar-rays). All control experiments showed negative reactions. Slides were observed and photographed at 203 magnification under a light photomicroscope (Eclipse Ni-U, Nikon, Japan) equipped with a digital camera (DS-U3, Nikon, Japan). The images were analyzed with the image-analyzing program NIS Elements BR (Vers. 4.20) (Nikon, Japan).
Statistical Analysis
The intensity values were subjected to statistical analysis with SPSS software (IBM, Italy). The normality of sample dis-tributions was tested using Shapiro-Wilk tests. For each lectin, the comparison between the intensity values among canine, ovine, and equine MSCs was performed with a one-way analy-sis of variance (ANOVA) for independent parameters followed by Tukey’s post hoc multiple comparison test to assess the sig-nificant intensity differences between samples. Data are given as means 6 standard error of the mean. The statistical signifi-cance was set at P < 0.05.
R
ESULTSCell Microarray Method
The average number of cells per spot was as follows: 12.5 6 2.8 canine MSCs/spot, 10.5 6 1.3 ovine MSCs/spot, and 12.8 6 2.2 equine MSCs/spot. Thus, the differences in
signal intensities between the MSCs of the three species reflect real differences between the cells. Canine, ovine, and equine MSCs spotted on the cell microarray slides reacted with all of the lectins that were used, but with different signal intensities (Fig. 2).
The lectin binding patterns of canine, ovine, and equine MSCs are shown in Figure 3. To take the signal intensity into account, we divided the MSC-lectin binding intensities into three grades: strong binding (signal intensity 2,500), medium binding (signal intensity 1,250), and weak binding (signal intensity < 1,250). According to these grades, for dogs, there were 2 (Con A and PSA), 3 (Jacalin, RCA120, DBA), and
8 (MAL II, SNA, PNA, GSA I-B4, SBA, succinylated WGA,
GSA II, UEA I, LTA) lectins of strong, medium, and weak bindings, respectively; for sheep, there were 1 (Con A), 3 (Jacalin, DBA, PSA), and 10 (MAL II, SNA, PNA, RCA120,
GSA I-B4, SBA, succinylated WGA, GSA II, UEA I, LTA)
lec-tins of strong, medium, and weak bindings, respectively; and for horses, there were 1 (Con A), 4 (SNA, Jacalin, DBA, PSA), and 9 (MAL II, PNA, RCA120, GSA I-B4, SBA, succinylated
WGA, GSA II, UEA I, LTA) lectins of strong, medium, and weak bindings, respectively (Table 2).
Comparison of MSC fluorescence intensity signals dis-played five different lectin binding patterns (Fig. 3): 1) MAL II, RCA120, and PSA were significantly stronger in dogs
(P < 0.05, ANOVA test) than in sheep and horses; 2) Jacalin, DBA, succinylated WGA, and UEA I were stronger in sheep (P < 0.05, ANOVA test) than in the other two investigated species; 3) SNA and Con A were stronger in horses (P < 0.05, ANOVA test) than in dogs and sheep; 4) equine SNA, PNA, and GSA I-B4significantly differed (P < 0.05, ANOVA test)
from those of canine and ovine; and 5) canine MAL II signif-icantly differed (P < 0.05, ANOVA test) from that of ovine
Figure 3. Differential fluorescent signals of lectin-based cell microarray of the glycocalyx in canine, ovine, and equine MSCs. Identical lowercase letters represent statistical non-significance (ANOVA test, P > 0.05). s-WGA, succinylated WGA. [Color figure can be viewed at wileyonlinelibrary.com]
and equine. No interspecific difference was observed with GSA II incubation.
The species-specific relative abundance of N-linked, O-linked, fucose, and sialoglycans is shown in Figure 4. Gener-ally, a gradual decrease in N-linked, O-linked, fucose, and sialo glycans on the canine, ovine, and equine MSC surface was observed. Interestingly, the glycocalyx of equine MSCs exhibited no difference in fucose and sialo glycans.
Lectin Histochemistry
Confirmation of cell microarray results was performed using lectin histochemistry. This methodology provided com-parable results to the lectin-based cell microarray procedure, a representative of which is shown in Figures 5 and 6. Only equine MSCs revealed different intensity staining with Con A, which was lower than that for canine MSCs, and with UEA I and LTA, which were not statistically different from those for canine and ovine MSCs, respectively (Fig. 4). In addition, light microscopy gave information about the culture homoge-neity of the MSCs under investigation.
D
ISCUSSIONThe cell glycocalyx plays an important role in stem cell growth, differentiation, communication with the environment, and trafficking. Considering the species-specific variability of the MSC surface marker, as well as the cell and species specific-ity of the glycocalyx composition, a comprehensive profile of MSC surface glycans could reveal MSC biomarkers that are of value in the facilitation of basic research, clinical studies, and experimental studies, including cell-based therapy and tissue engineering in veterinary animals. In the present study, we characterized the glycan profile of the MSC glycocalyx in dogs, sheep, and horses, which are largely used animal models for clinical and experimental studies. The investigation was per-formed by means of cell microarray technology.
The lectin-binding signals indicated that the MSC surface of the three investigated species contained a complex glycan pattern with high mannose N-linked glycans predominating over O-glycans, which were expressed at low levels. This result agrees with that reported in other glycomic studies on hMSCs (7,8,10,12). O-glycans contained the core 1 disaccharide (Galb1,3GalNAc) (named T antigen) (Jacalin and PNA reac-tivity), the terminal aGalNAc and Tn antigen (the simplest mucin O-glycan made by a single GalNAc linked to serine or threonine) (SBA), the oligosaccharides terminating with GalNAca1,3(L-Fuca1,2)Galb1,3/4GlcNAcb1 (DBA), and the
disialyl T-antigen sequence NeuNAca2,3Galb1,3(6 NeuNAca2,6)GalNAc (MAL II). Statistical analysis of the
Figure 4. Relative quantification of N-linked, O-linked, fucosy-lated, and sialylated glycans in canine, ovine, and equine MSCs. [Color figure can be viewed at wileyonlinelibrary.com]
binding signals revealed that O-glycans containing a2,3-linked sialic acid and T antigen were more expressed on the glycoca-lyx of canine and ovine MSCs than on that of equine MSCs. Strong a2,3-sialic acid expression has been detected on the surface of hMSCs (7). It has been reported that a2,3-linked sialic acid may regulate MSC adhesion to the extracellular matrix (10). Moreover, a2,3-sialylation of the CD44 protein, which is a ubiquitous cell membrane protein, allows MSC trafficking to bone by the binding of carbohydrate receptors, such as E-selectin (9). Tn and T antigens have also been shown on the surface of hESCs (21). In addition, DBA binding sites have been observed on the surface of murine ESCs and their expression has been considered a cell surface marker that is tightly associated with the pluripotent state (22). Recent stud-ies also demonstrated the presence of O-linked glycans termi-nating with GalNAc (SBA staining) on the surface of equine amniotic mesenchymal cells (23). The role of O-glycans on the surface of canine, ovine, and equine MSCs is unknown. However, O-glycans form a smooth surface layer, form a highly hydrated surface, promote the binding of water and salts, provide protection from protease degradation, may reg-ulate cell-surface receptor stability and cell adhesion, may con-stitute a viscous physical barrier between the environment and the cell, and may serve as a trap for bacteria (24). Tumor-rejection antigenic O-glycans (Tra-1–60 and Tra1–81) are commonly used as markers for the identification of specific stem cell types (25,26).
N-glycans of the canine, ovine, and equine MSC glycoca-lyx were high in mannose and biantennary complex types (Con A). The presence of a relatively high amount of high mannose N-glycans in the total glycan pool is a common feature of hMSC glycomes (7,8,10). This finding agrees with the constant presence of intracellular glycoprotein synthesis machinery, such as the rough endoplasmic reticulum and the Golgi apparatus, described in the ultrastructural studies of several mammals,
including ovine MSCs (see Ref. 27). N-glycans are implicated in viability or growth of cultured cells, cytokine binding, cell-cell interaction, and cell-cell migration (28). N-glycans of the glyco-calyx of canine, ovine, and equine MSCs were highly fucosy-lated, mainly in the presence ofL-Fuca1, 6GlcNAc (PSA) and
lower density of Fuca1,2Galb1,4GlcNAcb (UEA I) and a1,3-linked fucose (LTA). Interestingly, hMSCs also have highly fucosylated N-linked glycans (10), showing the abundant pres-ence of a1,6-linked fucose (8). Glycocalyx N-glycans in our investigated MSCs also contained terminal lactosamine and Neu5Aca2,6Gal/GalNAc (RCA120 and SNA) (29). Abundant
sialylated N-glycans have been furthermore detected in cultured hMSCs (7,8,10) that express a2,6-sialylation on the surface of undifferentiated cells (7,10). Multi-fucosylated sialylated N-gly-cans, together with the fucosylated and sialylated O-glycan, are likely candidates as carriers of sialyl-Lewisx (sLex) in hMSCs (7). sLex is best known as a selectin ligand and for its involve-ment in leukocyte homing and cancer metastasis (30). In addi-tion, sLex has been reported to mediate fucosylation-dependent homing of MSCs into BM (9).
Comparison of relative glycosylation quantification revealed a progressive reduction in N-glycans, O-linked glycans, fucosylated glycans, and sialylated glycans in the MSC glycoca-lyx of the investigated species. Only equine MSCs showed the same degree of fucosylation and sialylation. We focused our attention on these glycan types because their differences have been observed on the surface of hMSCs during differentiation into specific cells, such as osteogenic cells (7) and adipogenic cells (8), and cell-matrix adhesion mechanism (10).
In-depth analysis of the differential profiling of the glyco-calyx of canine, ovine, and equine MSCs showed significant interspecific differences in glycopattern. Canine MSCs showed a predominant expression of O-linked sialo glycans with a2,3-linked sialic acid (MAL II) and sialo N-glycan with terminal lactosamine (RCA120) and a1,6 fucosylated oligosaccharides
Figure 5. Lectin fluorescent signals of histochemically stained canine, ovine, and equine MSCs. Identical lowercase letters represent sta-tistical non-significance (ANOVA test, P > 0.05). s-WGA, succinylated WGA. [Color figure can be viewed at wileyonlinelibrary.com]
(PSA); ovine MSCs expressed a significantly higher density of asialo O-glycans terminating with T antigen (Jacalin) and GalNAca1,3(LFuca1,2)Galb1,3/4GlcNAcb1 (DBA), as well as
N-glycans containing chitotriose (succinylated WGA) and
a1,2-linked fucose (UEA I); and equine BM-MSCs expressed a higher density of a2,6 sialic acids (SNA) in N-glycan (Con A). Compared with that of horses, canine, and ovine MSCs had in common the same low density of a2,6 sialic acids (SNA) and
Figure 6. Lectin histochemistry staining of canine, ovine, and equine MSCs. This figure shows similar binding patterns of FITC-labeled Con A, RCA120, SNA, DBA, UEA I, and LTA to cell microarray results (compare with Fig. 3). Nuclei are stained in blue with DAPI. [Color
high density of Galb1,3GalNAc (PNA) and aGal (GSA I-B4).
Ovine and equine MSCs expressed a lower density of a2,3-linked sialic acid (MAL II) than did canine MSCs. Interestingly, only GSA II, specific for glycan terminating with GlcNAc, bound the glycocalyx of canine, ovine, and equine MSCs with a similar intensity signal. This suggests that GSA II binders are a stable marker of the canine, ovine, and equine MSC glycocalyx. The presence of a species-specific glycocalyx glycan pattern has also been observed in ESCs. For instance, DBA and UEA I binding sites are present on the surface of hESCs (14) but not in the murine ESC glycocalyx (22). The findings were con-firmed by lectin histochemistry, which is the conventional method used for in situ detection of carbohydrate residues in the glycan chain. However, a comparison between lectin histo-chemistry and cell microarray results indicates that the cell microarray is an unbiased methodology that gives more infor-mation about the glycocalyx composition of MSCs.
In conclusion, this study revealed that the glycosylation of canine, ovine, and equine MSCs has species-specific charac-teristic features. In particular, canine MSCs expressed more a2,3-linked sialic acid (MAL II), terminal lactosamine (RCA120), and a1,6 and a1,3 fucosylated oligosaccharides
(PSA, LTA); ovine MSCs exhibited more T antigen (Jacalin), GalNAca1,3(LFuca1,2)Galb1,3/4GlcNAcb1 (DBA), chitotriose
(succinylated WGA), and a1,2-linked fucose (UEA I); and equine MSCs showed a higher density of a2,6 sialic acids (SNA) and high mannose N-glycans (Con A). The different glycan patterns were clearly detected using cell microarray-lectin binding technology, confirming that this procedure is a useful, simple, and sensitive tool for obtaining accurate infor-mation on cell surface glycosylation. We did not detect a species-specific glycomarker of canine, ovine, and equine MSCs, perhaps because a larger number of lectins is needed. However, these results could provide a useful starting point for the detection of species-specific MSC surface biomarkers in future experimental studies dealing with the differentiation of veterinary species MSCs into specific cells (adipogenic, chondrogenic, osteogenic), as well as in cell-based therapy and regenerative medicine. The detected glycan pattern may provide a phenotype-specific marker in varying culture condi-tions and passage numbers as these two factors influence sur-face glycan expression patterns.
A
CKNOWLEDGMENTSThe authors thank Dr. Peter Gemeiner and Dr. Jaroslav Katrlık (Department of Glycobiotechnology, Slovak Academy, Bratislava) for their support in the development of this micro-array methodology. In addition, we thank Dr. Nicola Antonio Martino for his assistance in capturing images of cultured cells. We confirm that there are no known conflicts of interest associated with this publication.
L
ITERATUREC
ITED1. Ji JF, He BP, Dheen ST, Tay SS. Interactions of chemokines and chemokine receptors mediate the migration of mesenchymal stem cells to the impaired site in the brain after hypoglossal nerve injury. Stem Cells 2004;22:415–427.
2. Yu J, Li M, Qu Z, Yan D, Li D, Ruan Q. SDF-1/CXCR4-mediated migration of trans-planted bone marrow stromal cells toward areas of heart myocardial infarction through activation of PI3K/Akt. J Cardiovasc Pharmacol 2010;55:496–505. 3. Caplan AI, Correa D. The MSC: An injury drugstore. Cell Stem Cell 2011;9:11–15. 4. Kolf CM, Cho E, Tuan RS. Mesenchymal stromal cells. Biology of adult mesenchymal
stem cells: Regulation of niche, self-renewal and differentiation. Arthritis Res Ther 2007;9:204.
5. Lanctot PM, Gage FH, Varki AP. The glycans of stem cells. Curr Opin Chem Biol 2007;11:373–380.
6. Takahata M, Iwasaki N, Nakagawa H, Abe Y, Watanabe T, Ito M, Majima T, Minami A. Sialylation of cell surface glycoconjugates is essential for osteoclastogenesis. Bone 2007;41:77–86.
7. Heiskanen A, Hirvonen T, Salo H, Impola U, Olonen A, Laitinen A, Tiitinen S, Natunen S, Aitio O, Miller-Podraza H, et al. Glycomics of bone marrow-derived mesenchymal stem cells can be used to evaluate their cellular differentiation stage. Glycoconj J 2009;26:367–384.
8. Hamouda H, Ullah M, Berger M, Sittinger M, Tauber R, Ringe J, Blanchard V. N-gly-cosylation profile of undifferentiated and adipogenically differentiated human bone marrow mesenchymal stem cells: Towards a next generation of stem cell markers. Stem Cells Dev 2013;22:3100–3113.
9. Sackstein R, Merzaban JS, Cain DW, Dagia NM, Spencer JA, Lin CP, Wohlgemuth R. Ex vivo glycan engineering of CD44 programs human multipotent mesenchymal stromal cell trafficking to bone. Nat Med 2008;14:181–187.
10. Kang J, Park HM, Kim YW, Kim YH, Varghese S, Seok HK, Kim YG, Kim SH. Con-trol of mesenchymal stem cell phenotype and differentiation depending on cell adhe-sion mechanism. Eur Cell Mater 2014;28:387–403.
11. Sharon N, Lis H. History of lectins: From hemagglutinins to biological recognition molecules. Glycobiology 2004;14:53R–62R.
12. Wilson KM, Thomas-Oates JE, Genever PG, Ungar D. Glycan profiling shows unvar-ied N-glycomes in MSC clones with distinct differentiation potentials. Front Cell Dev Biol 2016;4:52.
13. Tateno H, Toyota M, Saito S, Onuma Y, Ito Y, Hiemori K, Fukumura M, Matsushima A, Nakanishi M, Ohnuma K, et al. Glycome diagnosis of human induced pluripotent stem cells using lectin microarray. J Biol Chem 2011;286:20345–22053.
14. Toyoda M, Yamazaki-Inoue M, Itakura Y, Kuno A, Ogawa T, Yamada M, Akutsu H, Takahashi Y, Kanzaki S, Narimatsu H, et al. Lectin microarray analysis of pluripotent and multipotent stem cells. Genes Cells 2011;16:1–11.
15. Accogli G, Desantis S, Martino NA, Dell’Aquila ME, Gemeiner P, Katrlık J. A lectin-based cell microarray approach to analyze the mammalian granulosa cell surface gly-cosylation profile. Glycoconj J 2016;33:717–724.
16. Accogli G, Lacalandra GM, Aiudi G, Cox SN, Desantis S. Differential surface glyco-profile of buffalo bull spermatozoa during mating and non-mating periods. Animal 2017;7:1–9.
17. Russell KA, Chow NH, Dukoff D, Gibson TW, LaMarre J, Betts DH, Koch TG. Char-acterization and immunomodulatory effects of canine adipose tissue-and bone marrow-derived mesenchymal stromal cells. PLoS One 2016;11(12):e0167442. 18. Khan MR, Chandrashekran A, Smith RK, Dudhia J. Immunophenotypic
characteri-zation of ovine mesenchymal stem cells. Cytometry Part A 2016;89A:443–450. 19. Clark KC, Kol A, Shahbenderian S, Granick JL, Walker NJ, Borjesson DL. Canine
and equine mesenchymal stem cells grow in serum free media have altered immuno-phenotype. Stem Cell Rev 2016;12:245–256.
20. Paebst F, Piehler D, Brehm W, Heller S, Schroeck C, Tarnok A, Burk J. Comparative immunophenotyping of equine multipotent mesenchymal stromal cells: an approach toward a standardized definition. Cytometry Part A 2014;85A:678–687.
21. Wearne KA, Winter HC, O’Shea K, Goldstein IJ. Use of lectins for probing differenti-ated human embryonic stem cells for carbohydrates. Glycobiology 2006;16:981–990. 22. Nash R, Neves L, Faast R, Pierce M, Dalton S. The lectin Dolichos biflorus agglutinin
recognizes glycan epitopes on the surface of murine embryonic stem cells: A new tool for characterizing pluripotent cells and early differentiation. Stem Cells 2007;25: 974–982.
23. Lange-Consiglio A, Accogli G, Cremonesi F, Desantis S. Cell surface glycan changes in the spontaneous epithelial-mesenchymal transition of equine amniotic multipo-tent progenitor cells. Cells Tissues Organs 2015;200:212–226.
24. Brockhausen I, Schachter H, Stanley P. O-GalNAcglycans. In: Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essen-tials of Glycobiology, 2nd ed. New York: Cold Spring Harbor Laboratory Press; 2009. pp. 105–127.
25. Muramatsu T, Muramatsu H. Carbohydrate antigens expressed on stem cells and early embryonic cells. Glycoconj J 2004;21:41–45.
26. Schopperle WM, DeWolf WC. The TRA-1–60 and TRA-1–81 human pluripotent stem cell markers are expressed on podocalyx in in embryonal carcinoma. Stem Cell 2007;25:723–730.
27. Desantis S, Accogli G, Zizza S, Mastrodonato M, Blasi A, Francioso E, Rossi R, Crovace A, Resta L. Ultrastructural study of cultured ovine bone marrow-derived mesenchymal stromal cells. Ann Anat 2015;201:43–49.
28. Stanley P, Schachter H, Taniguchi N. N-Glycans. In: Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essentials of Glycobiology, 2nd ed. New York: Cold Spring Harbor Laboratory Press; 2009. pp. 101–114.
29. Geisler C, Jarvis DL. Effective glycoanalysis with Maackia amurensis lectins requires a clear understanding of their binding specificities. Glycobiology 2011;21:988–993. 30. Magnani JL. The discovery, biology, and drug development of sialyl Lea and sialyl
