| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print October 6, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Institute of Human Virology, University of Maryland, Baltimore, Maryland, USA; and
Division of Infectious Diseases, Department of Medicine, University of Maryland Medical Center, Baltimore, Maryland, USA
1Correspondence: Institute of Human Virology, University of Maryland Medical System, 725 West Lombard Street, Baltimore, MD 21201, USA. E-mail: stamatos{at}umbi.umd.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in activated cells. The down-regulation of IFN- occurred at the transcriptional level. Thus, sialidase activity in activated T lymphocytes contributes to the hyposialylation of specific cell surface glycoconjugates and to the production of IFN-.
Key Words: activation • sialic acid • cytokines
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sialidases are enzymes that influence cellular activity by removing terminal sialic acid residues from glycolipids and glycoproteins. Four genetically distinct forms of mammalian sialidase have been characterized, each with a predominant cellular localization (lysosomal, cytosolic, or plasma membrane-associated) and substrate specificity [25 26 27 28 29 30 31 32 33 ]. Lysosomal sialidase (Neu1) has a catabolic role in desialylating glycoproteins and glycolipids in lysosomes [34 ] but is also present in extralysosomal locations at the periphery of activated lymphocytes [35 ]. Neu3 is associated with the plasma membrane [29 , 30 ] and by preferentially desialylating gangliosides, is believed to have a regulatory role in cellular activation, differentiation, and transformation [15 16 17 18 ]. The cytosolic sialidase (Neu2) can desialylate glycoproteins and gangliosides [28 ] and appears to have a role in myoblast differentiation [36 ]. The function of the recently characterized Neu4 sialidase has not been established [31 32 33 , 37 ].
Endogenous sialidase activity increases in cells of the immune system during cell activation [13
, 20
, 38
39
40
41
42
] or differentiation [43
]. In one study, the expression of Neu1 and Neu3 was shown to be up-regulated in activated, unfractionated human lymphocytes [40
]. In contrast, other investigators reported that activated murine CD8 lymphocytes expressed increased amounts of only Neu3 [44
]. Although specific immune functions of Neu1 and Neu3 in human cells remain to be determined, Neu1 activity in activated murine lymphocytes was associated with increased production of IL-4 but had no effect on the production of Th1 cytokines IL-2 or IFN- [20
]. Desialylation of gangliosides GM1 and/or GM3 by Neu1 was postulated as an early event in the signaling pathway leading to IL-4 production [20
, 45
]. Neu1 and Neu3 are also expressed in monocytes during differentiation into macrophages, and the activity of Neu1 was specifically up-regulated [43
]. As desialylation of monocytes by exogenous neuraminidase activated the ERK1/2 and p38 MAPK signaling pathways and led to increased production of IL-6, IL-1ß, MIP-1
, and MIP-1ß [21
, 46
], it is possible that Neu1 and/or Neu3 activities in differentiating monocytes are involved in these processes.
Multiple approaches have been used to study the role of endogenous sialidase activity in cell function [17 , 20 , 22 , 45 , 47 ]. The association of Neu1 with IL-4 production was established in lymphocytes from the SM/J mouse strain, which is partially deficient in Neu1 [20 , 48 ]. Silencing the expression of Neu3 with small interfering RNAs showed the importance of this enzyme in the development of the generating growth cone of neurons [47 ]. Polyclonal antibodies prepared against Clostridium perfringens neuraminidase inhibited in vivo the transendothelial migration of neutrophils into the bronchoalveolar compartment of mice intranasally challenged with IL-8 [22 ]. Chemical sialidase inhibitors, such as the sialic acid analogue 2,3-dehydro-2-deoxy-N-acetylneuraminic acid (DANA), have also been used to reverse the changes in cell activity that are associated with the action of sialidases [17 , 20 ]. 4-Guanidino-2-deoxy-2,3-dehydro-N-acetylneuraminic acid (zanamivir) is another inhibitor, claimed to be specific for influenza virus neuraminidase [49 ], which is available for the prophylaxis and treatment of infection by influenza virus [50 , 51 ]. Although this compound has recently been shown to inhibit the activity of purified Neu3 sialidase as efficiently in vitro as DANA [52 ], it has not yet been used to study the role of mammalian sialidases in cells grown in culture.
In this report, we characterize the endogenous sialidase activity of freshly isolated and activated human T lymphocytes and evaluate its role in production of cytokines. In particular, we show that Neu1 and Neu3 sialidases are expressed in activated lymphocytes, that the expression of only Neu1 is specifically up-regulated, and that Neu1 is present on the surface of freshly isolated and activated CD4 and CD8 cells. Furthermore, we show that zanamivir and DANA inhibit endogenous sialidase activity of activated lymphocytes grown in culture, as evidenced by an altered sialylation pattern of cell surface proteins, and that this inhibition of sialidase activity results in reduced production of IFN- RNA and protein.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Culture conditions for purified lymphocytes
Freshly isolated T lymphocytes were suspended at 1 x 106 cells/ml in RPMI-1640 medium containing 10% heat-inactivated FCS (both from Gibco, Invitrogen, Carlsbad, CA) and were maintained at 3 x 106 cells/well in six-well tissue culture plates (Costar, Corning Inc., Corning, NY) at 37°C in a humidified 5% CO2 incubator. Wells of the tissue culture plates contained immobilized anti-CD3 and anti-CD28 IgG (both from Calbiochem, San Diego, CA) and were prepared by exposure to a 1 ml solution of each IgG at 1 µg/ml for 15 h at 4°C. After 48 h of culture, cells were transferred to a new plate without immobilized IgG, medium was supplemented with recombinant human (rh)IL-2 (Roche, Nutley, NJ) at 20 units/ml, and cells were harvested at the indicated times for analysis.
In experiments where endogenous sialidase activity was inhibited, freshly isolated T lymphocytes (3x106) were seeded in wells of an untreated, six-well plate in RPMI medium containing 10% FCS for 18 h in the presence or absence of 2 mM zanamivir (GlaxoSmithKline, Research Triangle Park, NC) or 1 mM DANA (Calbiochem). Cells were activated after 18 h by transferring cells and culture medium to wells of a new six-well plate that contained immobilized anti-CD3 and anti-CD28 IgG as indicated above. After 4 days in culture, medium was collected for cytokine measurements by ELISA, and cells were harvested for analysis of mRNA expression and sialic acid content.
Differentiation of purified monocytes into mature dendritic cells (DC)
DC were generated in culture from monocytes that were purified from human PBMC by negative selection using the StemSep monocyte enrichment protocol (Stem Cell Technologies). Purified monocytes were maintained at 2.5 x 106 cells/well in six-well tissue culture plates in RPMI-1640 medium containing 10% FCS and 1 mM pyruvate (Gibco, Invitrogen), 1% MEM nonessential amino acids and 50 µM 2-ME (Sigma-Aldrich, St. Louis, MO), and rhGM-CSF at 50 ng/ml and rhIL-4 at 50 ng/ml (R&D Systems, Inc., Minneapolis, MN). Immature DC were differentiated further into mature DC by addition of LPS (from Escherichia coli, Sigma-Aldrich) at 1 ng/ml to the culture medium after 4 days in culture and incubation for an additional 24 h. Cells were harvested after 5 days in culture by gentle scraping with a polyethylene cell scraper (Nalge Nunc International, Rochester, NY).
Measurement of sialidase activity
Sialidase activity of freshly isolated lymphocytes and of activated cells, which were maintained in culture for 2, 24, 48, and 120 h, was determined using the artificial substrate 2'-(4-methylumbelliferyl)--D-N-acetylneuraminic acid (4-MU-NANA; Sigma-Aldrich) and mixed bovine brain gangliosides (Calbiochem) as described previously [43
]. In the assay using 4-MU-NANA, activity was measured in 5 x 106 solubilized, freshly isolated lymphocytes and in lymphocytes that were activated for 2 h. At subsequent time points, activity in 2 x 106 activated cells was measured. In the ganglioside assay, 1 x 107 cells were used to measure sialidase activity in freshly isolated cells and in lymphocytes activated for 2 h. Sialidase activity against gangliosides was also measured in 5 x 106 cells maintained in culture for 24 h and in 2 x 106 cells in culture for 48 h and 120 h after stimulation. The units of activity represent sialidase activity per milligram of protein. A unit of sialidase activity was defined as the amount of enzyme that released 1 nmole sialic acid per hour at 37°C. Protein concentration was measured by the Bradford method using a Bio-Rad protein assay kit (Bio-Rad, Hercules, CA).
The ability of zanamivir to inhibit sialidase activity in vitro was determined using enzyme from solubilized human monocyte-derived DC. Cells (2x106) were suspended in 0.20 ml solution containing 0.1% Triton X-100, 0.05 M sodium acetate, pH 4.4, and 0.1% BSA (Pentex bovine albumin fraction V, Miles Inc., Kankakee, IL) in the presence of increasing concentrations (0, 0.01, 0.1, 1, and 2 mM) of zanamivir. Exogenous 4-MU-NANA or mixed bovine brain gangliosides were not added to the reaction mixture, and thus, the free sialic acid detected in this assay is that released from cellular sialylconjugates (i.e., where activity reflects the release of sialic acid from endogenous cellular sialylconjugates). After a 60 min incubation at 37°C, the reaction mixture was microfuged 2 min at 12,000 g to remove cellular debris, and 0.02 ml of each supernatant was analyzed for sialic acid content using a Dionex DX600 chromatography system (Dionex Corporation, Sunnyvale, CA) as described previously [43 ]. One unit of sialidase activity was defined as the amount of enzyme that liberated 1 nmole sialic acid per hour at 37°C. The amount of activity measured in each sample was corrected based on protein concentration to represent activity per milligram of protein.
Characterization of cell surface and intracellular proteins
Proteins on the surface of lymphocytes were biotinylated and separated from intracellular proteins using a Pinpoint cell surface protein isolation kit (Pierce, Rockford, IL) as per the manufacturer’s suggested protocol. Biotinylated surface proteins from 1 x 107 freshly isolated or activated (Day 5) lymphocytes were collected directly in 0.20 ml SDS-PAGE gel loading buffer. The nonbiotinylated proteins (intracellular fraction) were collected in 0.50 ml lysis buffer provided in the kit. Proteins from a portion of each sample (0.03 ml from the biotinylated sample and 0.01 ml from the nonbiotinylated sample) were resolved by electrophoresis on a 10% SDS/polyacrylamide gel and electrotransferred by a semi-wet method to a Sequi-Blot polyvinyldifluoride membrane (Bio-Rad) and probed with polyclonal rabbit anti-Neu1 IgG as described previously [43
]. Blots were also probed with 0.5 µg/ml rabbit polyclonal IgG against protective protein/ cathepsin A (PPCA) (kindly provided by Dr. Alessandra d’Azzo, St. Jude Children’s Research Hospital, Memphis, TN) and 0.15 µg/ml rabbit anti-ß actin mAb (Cell Signaling Technology, Danvers, MA).
Determination of cell surface sialic acid
Lymphocytes were washed three times in PBS, resuspended at 1 x 107 cells in PBS, pH 7.4, containing 0.5% BSA, and were incubated for 5 h at 37°C with 500 mU/ml bacterial neuraminidase (crystalline, type X, from C. perfringens, Sigma-Aldrich). The insoluble cell material was collected by centrifugation for 2 min at 10,000 g, and the sialic acid released into the supernatant was quantitated by HPLC using a Dionex as described above.
Treatment of activated lymphocytes with purified Neu1
Lymphocytes were purified from PBMC, activated by exposure to anti-CD3 and anti-CD28 IgG, and grown in culture for 4 days as described above. Cells were collected, resuspended at 1 x 107 cells in 1 ml PBS, pH 7.4, containing 0.5% BSA and in 0.15 M sodium acetate, pH 4.4, containing 0.5% BSA, and mock-treated or treated with 7.5 µg/ml of purified murine Neu1 sialidase alone or in combination with 3.0 µg/ml protective protein/cathepsin A (both enzymes were kindly provided by Dr. Alessandra d’Azzo, St. Jude Children’s Research Hospital). Cells were incubated in a CO2 incubator at 37°C for 1 h, washed three times with PBS, pH 7.4, and stained with FITC-labeled Erythrina cristagalli (ECA) and evaluated by flow cytometry as described below.
Isolation of RNA and real-time RT-PCR
Freshly isolated and activated lymphocytes were harvested at the indicated times, and total RNA was isolated using an RNeasy mini kit (Qiagen, Valencia, CA) following the protocol suggested by the manufacturer. The RNA preparation was treated with DNase I (Gibco, Invitrogen) at 37°C for 30 min to remove contaminating DNA. DNase was then removed by binding to Blue Sorb DNase affinity slurry (Clonogene, St. Petersburg, Russia). Semiquantitative, real-time RT-PCR was performed using a QuantiTect SYBR green RT-PCR kit (Qiagen, Valencia, CA) with an ABI sequence detection system (ABI Prism 5700) as described previously [43
]. To detect gene expression of Neu1 (GenBank Accession X78687), Neu2 (GenBank Accession NM_005383), Neu3 (GenBank Accession Y18563), Neu4 (GenBank Accession NM_080741), ST3Gal III (GenBank Accession NM_174966), ST3Gal IV (GenBank Accession NM_006278), ST6Gal I (GenBank Accession NM_173216), IFN- (GenBank Accession NM_000619), and 18S rRNA (GenBank Accession X03205), the following primers were selected using DNAsis Max software (Hitachi, Japan) and were synthesized by Qiagen (Germantown, MD): Neu1 [forward, nucleotide (nt) 854–873] 5'ATGAATGCCAGCCCTATGAG3' and (reverse, nt 983–964) 5'TCACGGGGCCTTAGTGTATC3' yielding a 130-bp product; Neu2 (forward, nt 194–213) 5'AGGTTCAGTGGCAAGCTCAG3' and (reverse, nt 320–303) 5'CCAGGGATGGCAATGAAG3' yielding a 127-bp product; Neu3 (forward, nt 1010–1028) 5'GGTCCAGTGCAGAGGTCAT3' and (reverse, nt 1154–1137) 5'TCTGCAAAGGCCAGGAAG3' yielding a 145-bp product; Neu4 (forward, nt 451–470) 5'TGCAGGACTGGGCCACATT3' and (reverse, nt 610–591) 5'CATCGCTGTAGAAGGCGAAG3' yielding a 160-bp product; ST3Gal III (forward, nt 677–696) 5'GCGTTCTTGCCAACAAGTCT3' and (reverse, nt 800–782) 5'GTGATGCGCAGTGTCGTTT3' yielding a 124-bp product; ST3Gal IV (forward, nt 301–320) 5'GCAGAGAGCAAGGCCTCTAA3' and (reverse, nt 437–418) 5'AGATCCTCACTCCCCTTGGT3' yielding a 137-bp product; ST6Gal I (forward, nt 917–937) 5'CCTCTGAATGGGAGGGTTAT3' and (reverse, nt 1053–1034) 5'TGCGTCATGATCATCGATTT3' yielding a 137-bp product; IFN- (forward, nt 404–424) 5'TGGAGACCATCAAGGAAGACA3' and (reverse, nt 551–532) 5'TCAGCCATCACTTGGATGAG3' yielding a 148-bp product; and 18S rRNA (forward, nt 1279–1298) 5'CGGACAGGATTGACAGATTG3' and (reverse, nt 1397–1378) 5'ATGCCAGAGTCTCGTTCGTT3' yielding a 119-bp product. Primer and template concentrations were optimized to assure equal efficiency of each PCR reaction.
Semiquantitative analysis was based on the cycle number [comparative threshold cycle (CT)] at which the SYBR green fluorescent signal crossed a threshold in the log-linear range of RT-PCR, indicating the relative amount of starting template in each sample. The fold change in expression of Neu1 and Neu3 RNAs in freshly isolated lymphocytes compared with activated cells was normalized to the expression of 18S rRNA and was calculated by the equation:
 |
Staining reagents and flow cytometry
Immunofluorescent detection of proteins on the surface of lymphocytes was determined by resuspending cells at 2 x 106 cells in 1 ml PBS containing 2% heat-inactivated human AB serum (HS; Gemini Bioproducts, Calabasas, CA) and reacting cells at 4°C for 30 min with PE-conjugated mAb to CD3, CD4, CD8, CD25, and isotypic control IgG (all mAb were from BD PharMingen). Where indicated, cells were costained with polyclonal rabbit IgG to Neu1 at 100 µg/ml or to Neu3 at 50 µg/ml. The polyclonal rabbit anti-Neu1 IgG (kindly provided by Alexei Pshezhetsky, University of Montreal, Canada) were generated by immunizing rabbits with rhNeu1 sialidase and were described elsewhere [53
]. Rabbit polyclonal anti-Neu3 IgG were generated by immunizing rabbits with a synthetic peptide corresponding to amino acids 109–128 of the human Neu3 sialidase, as referred to previously [43
]. Bound rabbit IgG were detected using 8 µg/ml FITC-conjugated goat anti-rabbit F(ab')2 fragment (Alexa Fluor® 488, Molecular Probes, Eugene, OR) at 4°C for 30 min. Where indicated, cells were reacted with 2 µg/ml FITC-conjugated lectins ECA or Arachis hypogaea (PNA; EY Laboratories, San Mateo, CA). ECA binds to the exposed galactose on glycoconjugates bearing a Galß1-4GlcNAc motif; PNA binds to Galß1-3GalNAc. Following incubation with antibodies or lectins, cells were washed with 2 ml PBS containing 2% HS and fixed with 1% paraformaldehyde. Cells were analyzed using a Becton Dickinson FACSCaliber (Mountain View, CA), and data were analyzed using FlowJo data analysis software (Tree Star, Ashland, OR).
Determination of cytokine levels
ELISA was used to determine cytokine levels in the medium from cells grown in the presence or absence of DANA or zanamivir using kits for IFN-, IL-2, IL-4, and IL-10 (R&D Systems).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Neu1 is present on the surface of freshly isolated and activated lymphocytes
Lysosomal Neu1 sialidase was shown recently to localize to the periphery of PBMC, which were activated with Con A [35
]. To determine whether activation of CD4 and CD8 human lymphocytes through the TCR resulted in translocation of lysosomal Neu1 to the outer cell surface, intact, freshly isolated and activated lymphocytes were evaluated by flow cytometry after staining with antibodies to CD25, CD4, CD8, and Neu1. Exposure of freshly isolated lymphocytes to anti-CD3 and anti-CD28 IgG resulted in activation of nearly 70% of cells within 24 h as seen by up-regulation of cell surface CD25, a marker for early lymphocyte activation (Fig. 2A
and 2B
, middle panels). When costained with antibodies against Neu1, most of these activated cells (CD25+) also expressed an increased amount of Neu1 on the cell surface after 24 h and 48 h of cell activation when compared with activated cells stained with preimmune rabbit IgG (Fig. 2
, A vs. B, middle and right panels). A small percentage of freshly isolated cells (12.5%) stained weakly for cell surface Neu1 (Fig. 2B
, left panel), as did 
16% of cells that did not express detectable amounts of CD25 after 24 h and 48 h following exposure to anti-CD3 and anti-CD28 IgG (Fig. 2B
, middle and right panels). CD4 and CD8 lymphocytes were activated by exposure to anti-CD3 and anti-CD28 IgG and growth in the above culture conditions, and between 74% and 91% of these cells expressed increased amounts of Neu1 on their surface when analyzed 24 h and 48 h after cell activation (Fig. 2C
and D
, middle and right panels).
|
|
Sialidase activity during lymphocyte activation is associated with desialylation of specific glycoconjugates on the cell surface
To determine whether the sialic acid content of glycoconjugates on the surface of lymphocytes changed after lymphocyte activation, freshly isolated and activated lymphocytes were stained with FITC-labeled lectins PNA and ECA, which bind to the exposed galactose of specific, desialylated glycoconjugates, and were evaluated by flow cytometry. PNA and ECA have different binding specificities for terminal glycomoieties: PNA binds to Galß1-3GalNAc, and ECA binds to Galß1-4GlcNAc. Desialylated glycoconjugates were present on the surface of freshly isolated lymphocytes as seen by the binding of PNA and ECA to these cells (Fig. 4A
and 4B
, Day 0, shaded areas). The amount of these lectins bound to the cell surface increased after cell activation (Fig. 4A
and 4B
, Day 4, shaded areas). The median fluorescence intensity (MFI) of PNA- and ECA-stained, activated lymphocytes on Day 4 compared with freshly isolated cells (Day 0) increased from 7.31 to 87 and 111 to 854, respectively. The binding specificity of these lectins was confirmed by a reduction in binding of each lectin when cells were stained in the presence of galactose.
|
|
To confirm that the reduction in ECA-binding sites on glycoconjugates on the surface of activated lymphocytes was not the result of a down-regulation in expression of Neu1 or Neu3 sialidases and/or an up-regulation in expression of sialyltransferases (ST3Gal III, ST3Gal IV, and ST6Gal I) of cells grown in the presence of DANA and zanamivir, the relative amount of RNA encoding these enzymes in control cells and cells grown in the presence of DANA and zanamivir was determined. There was an equal or a slightly greater amount of RNA encoding Neu1 and Neu3 in relation to 18S rRNA in cells grown in the presence of DANA or zanamivir (P values ranged from 0.07 to 0.48) in comparison with control cells (Fig. 6 ). In addition, there was no significant increase in the amount of RNA encoding the three sialyltransferases, which are capable of adding sialic acid to ECA motifs on glycoconjugates (P values >0.05; Fig. 6 ). Thus, the sialidase activity of Neu1 and/or Neu3 in activated lymphocytes contributes to the hyposialylation of glycoconjugates that possess potential ECA binding sites.
|
|
, IL-4, IL-10, or IL-12 was detected in the medium of cells after 4 days in culture (data not shown). In contrast, IFN- was robustly produced in these cells with 842 pg/ml present in the medium when measured on Day 4 after cell activation (Fig. 8A
, Control). When activated lymphocytes were grown in the presence of zanamivir or DANA though, the amount of IFN- present in the medium of cells after 4 days was reduced to 514 and 341 pg/ml, respectively (both P values <0.001). Inhibition of sialidase activity affected the production of IFN- at the transcriptional level. In cells grown in the presence of DANA and zanamivir, the amount of RNA encoding IFN- was reduced to 62% of its level in control cells (both P values <0.005; Fig. 8B
). These results demonstrate that removal of sialic acid from glycoconjugates on the surface of activated human lymphocytes promotes the production of the Th1 cytokine IFN-.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our findings differ quantitatively from a previous report that described a two- to threefold increase in Neu1 and Neu3 in human lymphocytes stimulated with anti-CD3 and anti-CD28 IgG [40 ]. The difference in results may be related to the cyclic cell stimulation protocol used in that study. A conflicting pattern of Neu1 expression was detected in murine splenic CD8 lymphocytes in which Neu1 RNA expression declined fourfold when measured 72 h after stimulation with anti-CD3 IgG [44 ]. Although human and murine sialidases are presumed to have similar functions, this discrepancy may be related to differences between species or in conditions for activating and culturing cells (i.e., exogenous IL-4 was added to the medium of murine cells). It is of note that the expression of Neu1 is up-regulated in monocytes as they differentiate into macrophages [43 ] and DC (unpublished results).
Neu1 sialidase is present in lysosomes, where it catabolizes glycoproteins and glycolipids. Yet, the function of Neu1 in influencing production of IL-4 in mice by desialylating cell surface gangliosides suggested that it might also be present on the cell surface [20 ]. A previous study used confocal immunofluorescent microscopy to show that Neu1 localized to the cell periphery in Con A-stimulated PBMC [35 ], but microscopy was performed on cells that had been permeabilized prior to staining with anti-Neu1 Ig. Thus, localization on the outer cell surface was not shown convincingly. A recent study, though, using immunofluorescent microscopy of intact human aortic smooth muscle cells demonstrated the presence of Neu1 on the outer cell surface [59 ]. We show in our report that Neu1 is present on the surface of intact CD4 and CD8 lymphocytes using anti-Neu1 IgG and flow cytometry. Although the intensity of staining with anti-Neu1 IgG increases in activated cells, Neu1 appears also to be present on the surface of unactivated CD25neg cells. The MFI of unactivated cells stained with anti-Neu1 IgG is slightly greater than the MFI of these cells stained with preimmune IgG (Fig. 2A and 2B , left and middle panels), and thus, the minor shift in intensity of staining in our analysis does not make it possible to determine whether all freshly isolated lymphocytes express Neu1 on the cell surface. Using biotin to label proteins on the surface of intact cells and to separate them from intracellular proteins, we show that Neu1 is clearly present on the surface of freshly isolated and activated lymphocytes. There was a clear, qualitative difference between Neu1 on the cell surface and intracellular Neu1; namely, surface Neu1 was tightly associated with the 20- or 32-kDa subunit of PPCA. It is of note that we did not detect PPCA on the surface of intact, freshly isolated, or activated lymphocytes using anti-PPCA polyclonal IgG and flow cytometry (unpublished results), suggesting that the epitope(s) recognized by the IgG are not readily accessible on the surface of intact cells or that the IgG recognize only denatured, linear epitopes. The mechanism for Neu1 translocation to the cell surface has not been established, although phosphorylation of the carboxyl terminus in activated cells appears to be involved [35 ]. It remains to be determined whether a larger proportion of total cellular Neu1 is present on the surface of activated versus resting lymphocytes, implying a specific mechanism for translocation.
An important question raised by our results is whether cell surface Neu1 actively desialylates cell surface glycoconjugates. We show in this report that exogenous, purified Neu1 is capable of removing sialic acid from glycoconjugates on the surface of activated lymphocytes at pH 4.4 but not at pH 7.4. In addition, we have not been able to detect sialidase activity in vitro in activated lymphocytes or even in DC, which express greater amounts of sialidase activity, when tested at pH 7.4 using 4-MU-NANA as substrate (unpublished data). This does not preclude Neu1 activity at the cell surface though. It is possible that lysosomal Neu1 may travel to the cell surface in endocytic vesicles that fuse with the cell membrane and release acidic contents into an extracellular microenvironment. Thus, there may be a transient acidic pocket in which Neu1 could be active. Alternatively, there may be factors yet to be described that change the pH requirements for Neu1 on the cell surface (e.g., association with other proteins). Plasma membrane-associated Neu3 also has an acidic pH optimum when assayed in vitro, but it has been shown to desialylate gangliosides on the surface of neighboring cells, presumably in a neutral pH extracellular milieu [60 ]. It is also possible that the activity of Neu1 may occur at intracellular sites for proteins and gangliosides that are targeted to the cell surface. If Neu1 is functionally active on the cell surface, it is possible that the two sialidases (Neu1 and Neu3), likely with differing substrate specificities, work in concert to modulate the sialic acid content of cell surface and extracellular glycoconjugates.
Changes in the sialic acid content of cell surface proteins during lymphocyte activation have been demonstrated using lectins that recognize terminal galactose moieties of glycoconjugates [39 , 44 ]. Consistent with these reports, we show that the number of binding sites for PNA (Galß1-3GalNAc) and ECA (Galß1-4GlcNAc) increases on the surface of activated lymphocytes. The increase in PNA- and ECA-binding sites in activated cells compared with resting lymphocytes is likely not all attributable to sialidase activity but may be partly related to down-regulation in sialyltransferase activity. Indeed, others have shown that increased binding of PNA to the surface of activated CD8 T lymphocytes, a result of hyposialylation of CD45, was related to down-regulation of sialyltransferase ST3Gal I, rather than to up-regulation of sialidases [44 ]. This conclusion was based partly on the lack of effect of DANA on PNA binding in those cells. We also found in our study that growth of activated lymphocytes in the presence of DANA has little effect on the amount of PNA bound to cells. In contrast, the amount of ECA bound to these cells was reduced significantly. Thus, sialidase activity likely contributes to the hyposialylation of an increased number of glycoconjugates that contain the ECA motif in activated lymphocytes. It is to be noted that the ECA- and PNA-binding sites comprise only a portion of the total potential cell surface sialylation sites, and thus, what occurs at these sites may not represent global changes in sialylation on the cell surface. Indeed, we show in this report that the total amount of cell surface sialic acid increases during activation of lymphocytes.
DANA is an effective inhibitor of a broad range of sialidases and has been shown to alter the activity of lymphocytes and other cells in culture [17 , 20 ]. Conversely, zanamivir was designed to inhibit specifically influenza virus neuraminidases and was shown initially to be a poor inhibitor of mammalian sialidases [49 ]. We demonstrate in this report that zanamivir is indeed an effective inhibitor in vivo of sialidase activity of human lymphocytes. In fact, zanamivir was as effective as DANA in reversing the hyposialylation of cell surface glycoconjugates containing the ECA motif and in increasing the total amount of cell surface sialic acid in activated lymphocytes. This finding in vivo is supported by a recent report that showed that zanamivir was as active as DANA in inhibiting the in vitro activity of Neu3 sialidase, which was purified from human neurons [52 ] and by our demonstration in this report that zanamivir efficiently inhibits the endogenous sialidase activity (Neu1 and Neu3) from DC when assayed against endogenous sialylconjugates. The prior demonstration that zanamivir was a relatively ineffective inhibitor of mammalian sialidases compared with DANA was likely attributable to use of the artificial substrate 4-MU-NANA rather than the more physiologic, endogenous sialyconjugates that we used. Changes in sialic acid content of glycoconjugates containing the ECA-binding motif can be mediated by sialidases or by the ST3Gal III, ST3Gal IV, and ST6Gal I sialyltransferases [61 , 62 ]. The specificity of each sialidase inhibitor used in culture with lymphocytes in our study was supported by the insignificant differences in the change of expression of Neu1 and Neu3 and of the relevant sialyltransferase genes when cells were grown in the absence or presence of inhibitor.
Lymphocytes respond to diverse extracellular stimuli by producing specific cytokines to potentiate immune activities. The proinflammatory cytokine IFN- is produced by activated Th1 lymphocytes and has antimicrobial, antiproliferative, and immunomodulatory effects [63
]. The conditions for cell activation and growth used in our studies led to brisk IFN- production but only minimal expression of other Th1 (IL-2, IL-12) or Th2 (IL-4, IL-10) cytokines. Inhibition of sialidase activity resulted in a significant reduction in synthesis of IFN-. The minor amount of other cytokines produced by activated lymphocytes made it difficult to determine whether sialidase activity also affected their production. In other studies in murine lymphocytes, sialidase activity and desialylation of cell surface glycoconjugates have been implicated in the production of IL-4 [20
, 45
] and of Th1 and Th2 cytokines [64
]. In each of these studies, though, factors in addition to sialidase activity may have contributed to cytokine production. The role of Neu1 in production of IL-4 was demonstrated in lymphocytes that were purified from Neu1-deficient SM/J mice [20
]. However, these cells retain 30% residual Neu1 activity [48
], thus raising the possibility that the reduced production of IL-4 was related to other genetic features of these mice. In the second study, murine lymphocytes produced Th1 and Th2 cytokines when exposed in a mixed cell reaction to the viral neuraminidase in influenza virus-infected DC [64
]. In this system, lymphocyte-DC interactions and the action of an exogenous viral neuraminidase activity likely influenced the ultimate cytokine profile. Thus, in contrast to these reports, our results demonstrate directly in human lymphocytes that endogenous sialidase activity (Neu1 and/or Neu3) promotes the production of IFN-, most likely by modulating the content of cell surface sialic acid.
Sialylated gangliosides on the surface of murine lymphocytes are involved in the regulation of intracellular signaling pathways and thus, help regulate the production of specific cytokines [65
, 66
]. The involvement of Neu1 in desialylation of gangliosides was previously suggested to be related to the production of IL-4 in murine lymphocytes [20
, 45
]. Although the up-regulation of Neu1 in activated lymphocytes suggests a role in desialylation of glycoconjugates and in cytokine production, Neu3, which is present on the cell surface in lipid rafts [67
, 68
], preferentially desialylates gangliosides. The expression of Neu3 does not change significantly during lymphocyte activation as evidenced by the in vitro sialidase assay, but its functional activity in vivo may be modulated by interactions with other regulatory molecules. Thus, Neu1 and Neu3 have the potential to desialylate cell surface gangliosides and consequently, to activate intracellular signaling pathways, such as JAK-STAT, which mediate induction of IFN- [63
]. Neu1 and Neu3 sialidases may also activate intracellular signaling, which leads to cytokine production, by desialylating cell surface glycoproteins, such as TCR, CD4, CD8, CD43, CD45, IFN- receptor, and ligands for siglecs [3
, 4
, 23
, 24
]. Future experiments using lymphocytes from individuals with genetic lysosomal storage disorders of sialidosis (primary deficiency of Neu1) and galactosialidosis (secondary Neu1 deficiency resulting from defective PPCA) may help establish the specific substrates for Neu1 and Neu3 and the contribution of each enzyme to the production of IFN-.
In addition to influencing the production of IFN-, Neu1 and/or Neu3 sialidases may play a role in some of the other immune functions of human lymphocytes. This prospect raises the possibility that inhibition of the activity of cellular sialidases with anti-sialidase antibodies or pharmacologic inhibitors may have therapeutic value in treating inflammation and infection. It should be noted though that sialidases act in a cellular milieu in which sialic acid is continuously being added back to glycoconjugates by sialyltransferases. Thus, sialyltransferases may compete with or complement the action of sialidases to remodel cell surface glycoconjugates. Establishing the role of human sialidases during activation of lymphocytes clearly presents great challenges.
|
ACKNOWLEDGEMENTS |
|---|
Received November 27, 2005; revised August 30, 2006; accepted September 11, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
