1 Molecular and Cellular Biochemistry 300: 29–37, 2007. DOI 10.1007/s11010-006-9343-z
Springer 2007
Pattern expression of glycan residues in AZT-treated K562 cells analyzed by lectin cytochemistry Anna Rita Lizzi,1 Anna Maria D’Alessandro,1 Argante Bozzi,1 Benedetta Cinque,2 Arduino Oratore,1 and Gabriele D’Andrea,1 1
Department of Biomedical Sciences and Technologies, University of L’Aquila, 67100, L’Aquila, Italy; 2Department of Experimental Medicine, University of L’Aquila, 67100, L’Aquila, Italy Received 21 July 2006; accepted 28 September 2006; Published online: 12 April 2007
Abstract The present paper shows that human chronic myeloid (K562) cells exposed 3 h to 20 lM 3¢-azido-3¢-deoxythymidine (AZT) exhibit marked variations of the oligosaccharide moiety of glycoconjugates. These changes were analyzed by confocal fluorescence microscopy, upon incubation of control and AZT-treated cells with biotin–lectin conjugates to visualize cell surface glycans or total glycans after cells permeabilization. In addition, cell fluorescence distribution of the biotinylated lectins, localized with streptavidin conjugates labeled with Alexa Fluor 488, was analyzed by flow cytometry. The results obtained show significant variations on the expression/distribution of membrane surface glycans as detected by both WGA and SNA, two lectins that recognize primarily cellular internal membrane glycolipids. A further interesting result was the significant increase of N-acetylglucosamine linked glycans localized either at the cell surface or intracellularly but only in K562 cells exposed to AZT. On the whole, our data demonstrate that AZT alters both lipid and Nlinked glycosylations thus confirming previous observations, from our laboratory and from other Authors, that the drug impair the nucleotide-sugar import in the Golgi’s lumen. AZT does also alter the O-linked glycosylations that occur in the Golgi complex since these reactions require the incorporation of sialic acid, GlcNAc and GalNAc all of which are sensitive to the drug. Key words: K562 cells, AZT, glycans, confocal fluorescence microscopy, flow cytometry Abbreviations: AZT, 3¢-azido-3¢-deoxythymidine; AZTMP, AZT monophosphate; BSA, Bovine serum albumin; FCS, Fetal calf serum; Gal, Galactose; GalNAc, N-acetylgalactosamine; GlcNAc, N-acetylglucosamine; HRP, Horseradish peroxidase; Man, Mannose; NeuAc, Neuraminic acid (sialic acid); PBS, 20 mM K-phosphate buffer, pH 7.2, containing 150 mM NaCl; PBS-T, 10 mM K-phosphate buffer, pH 7.2, containing 140 mM NaCl and 0.5% (w/v) Tween-20; PNA, Arachis ipogea (peanut) agglutinin; RCA120, Ricinus communis agglutinin; SNA, Sambucus nigra agglutinin; WGA, Triticum vulgaris agglutinin; ConA, Concanavalin A; EcorA, Erythrina.
Address for offprints: G. D’Andrea, Department of Biomedical Sciences and Technologies, University of L’Aquila, 67100, L’Aquila, Italy (E-mail:
[email protected])
30
Introduction Sugar chains play important roles in defining the characteristics of glycoproteins, such as biological activity immunogenicity, pharmacokinetics, solubility, and protease resistance [1, 2]. Previous studies from other laboratories have shown that inhibition of nucleotide-sugar transport by monophosphorylated derivative of AZT (AZTMP), significantly modifies the glycosylation of proteins and lipids. These alterations are likely involved in some of the side effects associated with AZT therapy of HIV-infected patients [3–5]. Among these noxious effects, the ability of the drug to block the early development of blood progenitor stem cells (thus inducing anemia and neutropenia) appears particularly harmful [6–8]. Interestingly, blood progenitor cell differentiation is characterized by significant alterations in the structure of the asparagine-linked oligosaccharides of many cell surface proteins [9]. Several other developmentally regulated events including increases in O-linked glycosylation, modifications in proteoglycan expression and altered glycosphingolipid metabolism also occur together with the maturation of many hematopoietic stem cells [10–12]. Although the precise role of these changes has not yet been fully clarified, the remodeling of surface oligosaccharides by inhibitors of glycosylation may impair the ability of these cells to proliferate and differentiate by altering specific structural determinants required for cell–cell interactions and signal transduction events. Variations in N-linked protein glycosylation cause several effects such as, a reduced synthesis of tri- and tetraantennary glycans, accumulation of biantennary glycans, shortening of polylactosamine chains, and inhibition of sialylation reactions [4, 13]. A decrease in the length as well as in the synthesis of the glycosaminoglycans chondroitin and heparan sulfate was also detected, together with the inhibition of N-acetylgalactosamine and sialic acid incorporation in glycolipids [4]. The marked decrease in sialylation indicates that AZTMP dramatically changes CMP-sialic acid transport into the Golgi complex [14]. By investigating a set of nucleotide monophosphate (NMP) analogs, the same authors evidenced those portions of a NMP, which are essential for interaction with CMP-sialic acid transporter. In particular, it was demonstrated that the carrier is tolerant of changes in many positions, with remarkable exceptions for the 2¢-ara hydrogen of the ribose ring and the lateral groups at C2, N3, and C4 of the base [14]. In the light of these data and of our previous studies on the effects induced by AZT in in vitro systems and in different cell lines [15–20], we focussed our interest on the evaluation of changes in cellular glycans expression on the surface and/or internal membrane of AZT-treated human erythroleukemic cells
(K562) by lectins cytochemistry. It is known that lectins are non-immune proteins or glycoproteins derived from plants, animals, or microorganisms and are specific to terminal or subterminal carbohydrate residues to which they bind non-covalently [21, 22]. Furthermore, lectin specificity for sugar residues has been extensively used as a cytochemical probe to characterize different cells types in various stages of differentiation and maturation [23–26]. A cytochemical lectin-binding study was therefore performed to investigate the distribution and changes of the oligosaccharidic component of the glycoconjugates either at the cell surface and in the intracellular space in AZT-treated erythroleukemia cell line K562. These alterations were visualized by confocal fluorescence microscopy and analyzed by flow cytometry.
Materials and methods Materials Biotinylated labeled lectins (ConA, EcorA, PNA, RCA120, SNA, and WGA) were from Sigma Chemical Co (St. Louis, MO, USA); streptavidine conjugate labeled with Alexa Fluor 488 was from Molecular Probes (Eugene, Oregon, USA); polylysinated slides were from Bio-Optica (MI, Italy). All other chemicals used were reagent grade.
Cells cultures and treatment Human chronic myeloid (K562) leukemia cells were obtained from the American Type Culture Collection (ATCC), maintained in exponential growth in RPMI 1640 medium (pH 7.2), at 37 C in an humidified atmosphere of 5% CO2 in air, supplemented with 10% (v/v) heat-inactivated fetal calf serum (FCS), 2 mM glutamine and 0.1 mg/ ml of penicillin and streptomycin. Cells were seeded at the density of 3 · 105 cells/ml and incubated in the absence or in the presence of 20 lM AZT for 3 h in routine experiments. Other experimental conditions were also checked (AZT concentrations ranging from 2 to 40 lM and cells treatment from 5 min to 48 h), but the results reported in this paper were obtained with the above reported conditions, 3 h cells exposure to 20 lM drug, since in these experimental conditions, the major differences between untreated (control) and AZT-treated cells were evidenced without any effect on cells morphology and cell viability. All the experiments were carried out with cells in the exponential growth phase. Cell viability was determined at different times by trypan blue exclusion assay [27]. Figure 1 shows cell growth in absence (control) and in the presence of 40 lM AZT.
31 obtained; on the contrary, with lower AZT concentrations (i.e., 2, 5, 10 lM) or shorter incubation times (i.e., 5¢, 30¢, 1 h), no significant differences were detected between control and AZT-treated cells (data not shown).
12
cells/mL (x10-5)
10 8 6
Flow cytometry analysis
4
Control and AZT-treated cells, at a density of 1 · 106/ml, were prepared as described above. Biotinylated lectins were localized with streptavidine conjugate labeled with Alexa Fluor 488. Cell fluorescence distribution was analyzed using flow cytometry (FACScan flow cytometer, Becton Dickinson Immunocytometry System, San Jose`, USA, equipped with a CELLQuest Software Program). One thousand cells from each sample were computed, and the mean fluorescence intensity was calculated. All experiments were repeated at least four times and each checked point was in duplicate.
2 0 0
12
24
36
48
60
time, h Fig. 1. K562 cell growth in absence () and in the presence (m) of 40 lM AZT. Cell viability was determined at different times by trypan blue exclusion assay [27].
Confocal fluorescence microscopy K562 cells grown in RPMI medium were incubated at a density of 5·105/ml in the absence or in the presence of 20 lM AZT for 3 h. The cells were washed twice with icecold PBS and fixed in 4% paraformaldehyde at room temperature for 10 min. To visualize the cell surface glycans, the cells were suspended in a blocking solution (3 mg/ml BSA, 7.5 mg/ml glycine in PBS) for 30 min and then incubated 1 h with biotinylated lectin (10 lg/ml). For total glycans visualization, untreated or AZT-treated cells were permeabilized with a blocking solution composed of 80 mM PIPES, pH 6.8, 5 mM EGTA, 1 mM MgCl2, containing 0.05% saponin, 50 mM NH4Cl, 0.02% (w/v) NaN3 for 30 min and then treated 1 h with lectin conjugated to biotin (10 lg/ml). In both conditions, the same set of lectins were used (i.e., PNA, EcorA, RCA120, WGA, Con A, SNA). After three washes with PBS, cells were incubated 1 h with streptavidine conjugate labeled with Alexa Fluor 488, dissolved in the blocking solution. Upon three more washes, the samples were resuspended in a MOWIOL solution, placed on a slide and examined according to [28], with minor modifications. Fluorescence images of cells were obtained with an Olympus Fluoview FV500 confocal laser scanning microscope (objective PLAN-APO 60/1.4 oil), equipped with Blue Argon (488 nm) laser and Green Helium–Neon laser (543 nm). Digital images (0.25 lm thick) were collected at 512 · 512 pixel in size with a pixel size of 0.08 lm. The images were processed by deconvolution, and false color look-up tables were applied to them for the final presentation. With a higher AZT concentration (i.e., 40 lM) or longer incubation times (i.e., 6, 12, 24, 48 h) similar results were
Results Binding of fluorochrome-labeled lectin evaluated by confocal fluorescence microscopy Glycans were localized in K562 cells by fluorescence analysis as described in ÔMaterials and Methods’ Section. The effects of AZT exposure on pattern expression of glycans, either at the cell surface as well as intracellularly, were monitored and are shown in Fig. 2, where the green fluorescence indicates the glycoprotein/lipoglycans localization. For each lectin, the fluorescence pattern showed a membrane surface and intracellular glycans distribution quite similar between control and AZT-treated cells with some remarkable differences. In particular, the more evident variations were found on the expression/distribution of membrane surface glycoproteins’ glycans as detected by both PNA and EcorA – known to bind mainly glycoproteins’ glycans – the first of which recognizes galactose linked as b(1 fi 3) and the second one galactose linked as b(1 fi 4). In fact, binding intensity for PNA significantly increased when cells were treated with 20 lM AZT showing a more decorated membrane surface vesicles (Fig. 2, PNA, top). This indicated a greater exhibition of b-D-galactose residues on the surface membrane and a lower amount of sialic acid residues. On the contrary, control cells resulted more decorated than the treated ones when EcorA was used (Fig. 2, EcorA, top). Moreover, regarding this latter lectin, control permeabilized cells showed a more intense surface area glycoprotein labeling compared to AZT-treated cells. These
32
Fig. 2. Distribution of the glycans in the K562 cells as evidenced by biotinylated PNA, biotinylated EcorA, biotinylated RCA120, biotinylated WGA, biotinylated ConA, biotinylated SNA; (left side: control; right side: AZT-treated; top: glycans distribution on the cellular surface; bottom: glycans distribution after cell permeabilization.
latter displayed vesicles more fragmented than those of control cells (Fig. 2, EcorA, bottom). RCA as well as ConA allowed the labeling of both outside and inside membrane glycoproteins and glycolipids [29, 30]. However, ConA labeling offers a better indication about the glycoprotein/lipid composition of perinuclear RE as well as Golgi membrane; in addition, ConA is known to colocalize with fibronectin a peculiar cytoskeleton glycoprotein [30]. Nevertheless, Con A and RCA120 displayed a
ubiquitous distribution of binding sites. Since no appreciable differences were found between RCA/ConA control cells and RCA/ConA AZT-treated cells (Fig. 2, RCA, Con A, top and bottom), the high expression of both mannose (Man) and galactose (Gal) residues seemed not depend on treatment conditions. As far as WGA and SNA concerns, it is recognized that labeling regards mainly the plasmatic and periplasmatic membrane instead the cytoplasmatic sub-cellular membranes
33 Flow cytometry analysis For detection of total glycoconjugates in K562 cells, a specific set of biotinylated lectins (see Table 1 for acronyms and specificity) as described above, and Steptavidine-Alexa 488 were used. Control and AZT-treated cells (20 lM for 3 h) showed different amount of the glycan portion at level of glycoconjugates. The measures were performed either in intact cells to monitor the distribution at level of the cellular surface, and in permeabilized cells to detect the intracellular localization. Figure 3 shows a typical flow cytometry fluorescence profile and is representative of several similar patterns. As far as the NeuAc linked as a(2 fi 6) (SNA) concerns, it seemed to exhibit an increase of the surface glycoconjugates, as compared to those located intracellularly. The same pattern was observed for the galactose linked b(1 fi 3) (PNA) that appeared increased in cells treated with AZT. On the contrary, the galactose linked b(1 fi 4) (EcorA) decreased upon drug exposure. A further interesting result, was the marked increase of N-acetylglucosamine (WGA) at level of the cellular surface, with respect to that distributed intracellularly after K562 cells exposure to AZT (Tables 2 and 3).
Fig. 3. Typical fluorescence profile of control (black line) and AZT-treated (gray line) cells. K562 cells were treated with or without 20 lM AZT for 3 h, and the appearance of surface or total glycans fluorescence (after permeabilization), was monitored by flow cytometry using SteptavidineAlexa Fluor 488. The pattern shown is representative of four similar ones concerning the glycans distribution after permeabilization detected by biotinylated WGA. Black dashed line: autofluorescence.
Discussion
[31]. In this context, our results showed a more intense labeling of the treated cells than the control ones, especially for the periplasmatic membrane being the difference more evident when SNA was used (Fig. 2, WGA, SNA, top and bottom).
Glycoproteins are widely distributed among different species, cell types, and tissues. Intriguingly, the sugar moieties of mammalian glycoproteins show significant changes in their structures and relative occurrences during growth, development and differentiation and have been involved in many biological functions [32]. Furthermore, glycosylation abnormalities are found in many diseases (e.g., cancer, metastasis, leukemia, inflammatory, and other diseases) and are the result of complex alterations of oligosaccharide assembly by glycosyltransferases, mainly in the Golgi apparatus. In this context, a pivotal role is ascribed to an (aberrant) unusual expression of plasma membrane glycans [for review see 33, 34]. Of course, many other causes, e.g., drug therapy, could bring to an irregular glycans expression. In fact, it is known that AZT treatment dramatically alters
Table 1. Details of lectins used in this study and their sugar specificity Lectin origin
Acronym
Major sugar specificity
Concanavalin A Sambucus nigra
ConA SNA
a-Man NeuAca(2 fi 6)Gal/GalNAc
Erythrina
EcorA
Galb(1 fi 4) GlcNAc
Ricinus comunis
RCA120
b-Gal
Arachis hipogea (peanut)
PNA
Galb(1 fi 3)GalNAc
Triticum vulgaris
WGA
(GlcNAc)2
Table 2. Cell surface glycans distribution analyzed by flow cytometry* Con A
SNA
Erytrhina (EcorA)
Ricinus (RCA120)
Arachis (PNA)
Triticum (WGA)
Sample
Mean
D (%)
Mean
D (%)
Mean
D (%)
Mean
D (%)
Mean
D (%)
Mean
D (%)
C
27
–
169
–
29
–
151
–
8
–
46
–
T
27
–
192
+12
29
–
157
–
11
+15
83
+44
C, control cells; T, AZT-treated cells (20 lM, 3 h). *Values are means of four different experiments in duplicate. SD was
