鑑定Tamm-Horsfall醣蛋白的活性作用片段對細胞激素所媒介發炎反應的調控及免疫細胞生物功能的影響

(1)

行政院國家科學委員會專題研究計畫成果報告

鑑定 Tamm-Horsfall 醣蛋白的活性作用片段對細胞激素所

媒介發炎反應的調控及免疫細胞生物功能的影響

研究成果報告(精簡版)

計畫類別：個別型計畫編號： NSC 95-2314-B-002-249- 執行期間： 95 年 08 月 01 日至 96 年 07 月 31 日執行單位：國立臺灣大學醫學院內科計畫主持人：謝松洲共同主持人：余家利計畫參與人員：碩士級-專任助理：陳文萱處理方式：本計畫可公開查詢

中華民國 96 年 11 月 01 日

(2)

行政院國家科學委員會補助專題研究計畫

■ 成果報告

□期中進度報告

鑑定 Tamm-Horsfall 醣蛋白的活性作用片段對細胞激素所媒介發炎反應

的調控及免疫細胞生物功能的影響

計畫類別：■ 個別型計畫 □ 整合型計畫

計畫編號：NSC 95－2314－B－002－249－

執行期間： 95 年 8 月 1 日至 96 年 7 月 31 日

計畫主持人：謝松洲

共同主持人：余家利

計畫參與人員：陳文萱

成果報告類型(依經費核定清單規定繳交)：■精簡報告 □完整報告

本成果報告包括以下應繳交之附件：

□赴國外出差或研習心得報告一份

□赴大陸地區出差或研習心得報告一份

□出席國際學術會議心得報告及發表之論文各一份

□國際合作研究計畫國外研究報告書一份

處理方式：除產學合作研究計畫、提升產業技術及人才培育研究計畫、

列管計畫及下列情形者外，得立即公開查詢

□涉及專利或其他智慧財產權，□一年□二年後可公開查詢

執行單位：台灣大學醫學院內科

中華民國 96 年 10 月 31 日

(3)

摘要

Tamm-Horsfall 醣蛋白是腎小管分泌的重要免疫蛋白分子，並在泌尿系统

的感染防衛扮演重要的角色。在本研究中，我們發現

Tamm-Horsfall 醣蛋白含

有大量的

Sia(2,3)Gal/GalNAc、中量的 β(1,4)GlcNAc oligomers 及 GlcNAc/甘露糖支鏈、少量的甘露糖支鏈，但不含Sia(2,6)Gal/GalNAc 的支鏈。

Tamm-Horsfall 醣蛋白對於人類

腫瘤壞死因子α、免疫球蛋白 G、補体 C1q 及牛血清白蛋白有高的結合親和性，

對於第八介白質僅有中等的結合親合性，對於第六介白質及干擾素γ則僅有弱

的結合親合性。我們進一步探討碳水化合物支鏈及蛋白質核心在蛋白結合及細

胞活化上的角色。Tamm-Horsfall 醣蛋白分別以分解碳水化合物[

neuraminidase

(Nase), β-galactosidase (Gase)]

或蛋白質[

V8 protease (V8),

proteinase K (PaseK)]或醣支

鏈蛋白質

[carboxypeptidase Y (Case),

O-sialoglycoprotein endo- peptidase (Oase)]為

主的酵素分別處理，發現

Tamm-Horsfall 醣蛋白被 V8、 Oase 及 PaseK 裂解

後顯著的減少其蛋白質結合的親和性、減少對於單核球細胞增生的活化、以及

對於多形核嗜中性白血球吞噬功能的活化作用。這些結果顯示蛋白質的核心結

構而非碳水化合物支鏈是

Tamm-Horsfall 醣蛋白多種免疫功能的基礎。

關鍵詞：Tamm-Horsfall 醣蛋白、Lectin 結合活性、細胞激素結合活性、蛋白

質核心、酪支鏈

ABSTACT

Tamm-Horsfall glycoprotein (THP) is synthesized in the particular sites of renal tubules acting as a defense molecule in the urinary system. In the present study, we found that THP contained high amount of Sia(2,3)Gal/GalNAc, moderate amount of β(1,4)GlcNAc oligomers and GlcNAc/branched mannose, and low amount of mannose residues, but no Sia(2,6)Gal/GalNAc, in the side chains of the molecule. THP exhibited high binding affinity with human TNF-α, IgG, C1q and BSA, moderate binding affinity with IL-8, and low binding affinity with IL-6 and IFN-γ. For exploring the role of carbohydrate side-chains and protein core in the protein-binding and cell-stimulating activities, THP was enzyme-digested with carbohydrate-specific [neuraminidase (Nase), β-galactosidase (Gase)], protein-specific [V8 protease (V8), proteinase K (PaseK)] and glycoconjugate-specific [carboxypeptidase Y (Case), O-sialoglycoprotein endopeptidase (Oase)] degrading-enzymes. We found that THP digested with V8, Oase, and PaseK, significantly reduced its protein-binding, mononuclear cell proliferating, and neutrophil phagocytosis-enhancing activities. These results suggest that the intact protein core structure, but not carbohydrate side-chains, is essential for pleotropic functions of THP molecule.

Keywords: Tamm-Horsfall glycoprotein, Lectin-binding activity, Cytokine-binding activity, Protein core, Side-chain glycomoiety

. Introduction

Urinary Tamm-Horsfall glycoprotein (THP), a renal excreted 80-90 kDa GPI-anchored macromolecule, is an important defense protein molecule against microbial invasion in the urinary system [1, 2]. This molecule contains 25-35% different carbohydrate moieties in weight

(4)

and abundant sialic acid [3, 4]. The glycomoiety structure of THP consists mainly of N-linked glycans of di-, tri-, and tetra-antennary types [5, 6] and one N-glycosylation site with high mannose sequences [7]. The high mannose glycans of human THP are carried by Asn251 (8, 10),

Man6GlcNAc2 and Man5GluNAc2 [7, 9] that mediate the interaction with type 1 fimbriated

Escherichia coli [10]. Recently, O-linked chains were found in THP molecule showing the potential interactions with different proteins [11]. Muchmore et al. [12, 13] demonstrated that THP glycomoiety was responsible for binding to recombinant IL-1 and TNF-α. In addition, a very high affinity association between THP and immunoglobulin light chains, complement component 1 and 1q was also observed [14-15]. Functionally, THP can stimulate resting polymorphonuclear neutrophils (PMN), lymphocytes and monocytes [17-19], but inversely suppress mitogen-activated lymphocyte proliferation (20, 21). In certain pathological conditions, urinary THP involves in the pathogenesis of distal nephrons and the urinary tract disorders such as cast nephropathy [22], urolithiasis [23], and tubulointerstitial nephritis [24]. Despite intensive studies in this field, the inter-relationships among glycomoiety, protein core, and biological functions of THP remain to be elucidated. In the present study, normal human THP was digested with different carbohydrate-, glycoconjugate-, and protein-specific degrading enzymes for understanding the role of carbohydrate side-chains and protein core in protein-binding, mononuclear cell proliferating, and PMN phagocytosis-enhancing activities of THP. We found that the intact protein core structure is essential for multiple physiological functions of THP. Materials and Methods

Reagents

Biotin-conjugated lectins specific for particular carbohydrate moieties including MAA [specific for Siaα(2,3)Gal/GalNAc], SNA-I [specific for Siaα(2,6)Gal/GalNAc], GNA (specific for mannose residues), DSA [specific for β(1,4)GlcNAc oligomers] and ConA (specific for GlcNAc/branched mannose) were purchased from E-Y Labs (San Mateo, USA). Bovine serum albumin (BSA), human purified IgG, neuraminidase (Nase), β-glactosidase (Gase), proteinase K (PaseK) and carboxypeptidase Y (Case) were purchased from Sigma-Aldrich Immunochemical Corp. (St. Louis, MO, USA). O-sialoglycoprotein endopeptidase (Oase) was obtained from Cedarlane Laboratories Ltd. (Burlington, NC, USA). Human C1q and V8 protease (V8) were obtained from Merck-Calbiochem (San Diego CA, USA). Human recombinant IFN-γ, TNF-α, IL-6 and IL-8 were obtained from R & D Systems (Minneapolis, MN, USA).

Purification of THG from normal human urine

Urine was collected from normal individuals in clean glass bottles. The purification of THP was carried out as described by Hunt and McGiven [25]. After 3 cycles of 0.58M NaCl precipitation, alkaline distilled water pH 9.0 (adjusted by adding 1N NaOH) dissolution, and centrifugal removal of insoluble substance, the purity and relative molecular weight of the obtained mucoproteins were detected by 10% SDS-PAGE. The THP molecule was identified in Western blot probed by anti-uromucoid antibody (The Binding Site Ltd, University of Birmingham Research Institute, Birmingham, UK). All of the urine donors signed the informed consents to participate the study approved by Institutional Review Board and Medical Ethics Committee, Taipei Veterans General Hospital (VGH IRB No: 94-07-27A), Taiwan.

(5)

enzymes

We followed the method reported by Sherblom et al. [13] to digest THP molecule by different specific enzymes. All incubations were performed at 37ºC for 16h. The concentration of the respective enzyme was: neuraminidase (10 units/ml) in 50mM sodium acetate, pH5.0; β-galactosidase (0.05 units/ml) in 50mM sodium acetate, pH5.0; proteinase K (0.5mg/ml) in 10mM Tris, pH 7.5 with 1mM MgCl2; V8 protease, (1mg/ml) in 50mM phosphate buffer, pH7.8

twice-treatment; carboxypeptidase Y, (enzyme:substrate=1:10) in 0.2M pyridine-acetate buffer, pH5.6; O-sialoglycoprotein endopeptidase (50 g/ml　 ) in PBS, pH 7.2. The enzyme-digested products of THP were intensively dialyzed against alkaline distilled water, pH 9.0 for 24h with frequent changes of dialysate for removing the digested products with molecular weight less than 10kDa. These digested THP products were then heated at 65ºC for 60 min to inactivate the residual enzymes in the mixture. The products were finally lyophilyzed and stored at -20ºC until used. In the mononuclear cell proliferation and PMN phagocytosis assays, the enzymes at the same concentration as used in the THP digestion was heat-inactivated at 65ºC for 60min as a negative control for avoiding the interference from the enzymes per se.

Biotinylation of intact THG and its enzyme-digested products

Intact THP and its enzyme-digested products at a concentration of 2mg/ml were incubated with biotin-labeling buffer containing 0.3mg biotin-7NHS/ml in PBS, pH7.2. (Boehringer-Mannheim Biochemicals, GmbH, Mannheim, Germany) at 4ºC for 1.5h with continuous gently shaking as reported in our previous study [26]. The mixtures were then intensively dialyzed against PBS, pH7.2 to remove the free biotin. The concentration of the biotinylated intact THP and its enzyme-digested products was adjusted to 1mg/ml by BioRad Protein Detection kit and stored at -20ºC until used.

Lectin-binding activity of THP by ELISA

We followed the method reported by Garcia et al. [27]. Briefly, 100μl of intact THP (5μg/ml) or its enzyme-digested products (5μg/ml) dissolved in alkaline distilled water pH9.0 were coated in microwells at 4ºC for 24h. After two washes and the non-specific binding of the microwells were blocked by 1% BSA, 100 μl of the respective commercially available biotin-conjugated lectins (5μg/ml): MAA, SNA-I, GNA, DSA or ConA, were added to the microwells and incubated at room temperature for 1h. HRP-conjugated streptavidin was then added and incubated for another 60 min at room temperature. After color development with NBT/X-phosphate solution, 0.1ml of 0.5M H2SO4 was finally addedtoterminate the reaction. The

binding affinity of THP and its enzyme-digested products with different lectins was measured at OD450nm absorbance by ELISA reader (Dynex Technologist, Chantilly, VA, USA).

Detection of glycoconjugate contents in THP and its enzyme-digested products by DIG (digoxigenin) glycan detection kit

We followed the method reported in our previous study [28]. A commercially available DIG glycan detection assay kit (Boehringer Mannheim GmbH Biochemica, Mannheim, Germany) was used to detect the presence of glycoconjugates in the intact and enzyme-digested products by Western blot. The detailed procedures are described in the manufacturer’s booklet.

The binding activity of intact biotin-THP and its enzyme-digested products with different protein molecules

(6)

Detected by ELISA

MTT test for MNC proliferation

Detection of neutrophil-binding activity of intact THP and its enzyme- digested products by flow cytometry

Detection of PMN phagocytosis-enhancing activity of intact THP and its enzyme-digested products by flow cytometry

Results

Lectin-binding activity of normal human THP

Since THP contains complex carbohydrate side chains in the molecule, the molecule can bind with lectins specific for different carbohydrates in variable degree. As shown in Fig.1-(A), THP binds robustly with MAA [specific for Siaα(2,3)Gal/ GalNAc], DSA [specific for β(1,4)GlcNAc oligomers] and ConA (specific for GlcNAc/branched mannose), modestly with GNA (specific for mannose residues), but no binding with SNA-I [specific for Sia　(2,6)Gal/GalNAc]. These results are quite consistent with the glycosylation of THP molecule that lacks Siaα(2,6)Gal/GalNAc in the side chain structure [8]. It is conceivable that the GalNAc in β(1,4) linkage to galactose in α(2,3) sialylated glycans of THP molecule, formulates the sequence of GalNAcβ(1,4)NeuAcα(2,3)Galβ(1,4)GlcNAc known as Sda antigen in THP [31].

The binding activity of THP with different protein molecules

The binding of THP with different protein molecules including serum proteins (BSA, human IgG, C1q cleavage products), cytokines (IFN-　, IL-6, IL-8 and TNF-　) and cell lysates (PMN and MNC) detected by Western blot is demonstrated in Fig.1-(B). Because BSA (lane 1: 66 kDa) can bind moderately with THP, the bulky bands appearing in IFN-γ (lane 5) and IL-6 (lane 8) are due to plenty BSA contents in the solution as a protein stabilizer (Fig.1-B). The specific banding of IFN-γ (lane 5: 17 kDa) was rather faint. In contrast, the specific banding of IL-6 (lane 8: 20.3 kDa) was quite distinct. We noted that THP could specifically bind with the cleavage products of C1q (lane 4: 30+20 kDa), TNF-α (lane 6: 17.5 kDa), IL-8 (lane 7: 8 kDa), IL-6 (lane 8: 20.3 kDa), IgG (lane 9: 50+23 kDa), and total cell lysates of PMN (lane 2) and MNC (lane 3). The relative binding activity of THP with these protein molecules detected by ELISA is demonstrated in Fig.1-(C). The dose-response binding of TNF-α from 0.1-1 g/ml 　 with THP is demonstrated in Fig.1-(D).

Comparison of protein and glycoconjugate contents among intact THP and its enzyme-digested products

For clarifying the role of carbohydrate side-chains and protein core of THP in the binding activity of THP with different proteins, THP was digested with carbohydrate- degrading (Nase and Gase) (Fig.2-A and B), glycoprotein-degrading (Case and Oase) (Fig.3-A and B), or protein-degrading enzymes (V8 and PaseK) (Fig.3-C and -D). After enzymatic digestion, the intact THP and its degrading products were electrophoresed in 10% SDS-PAGE and stained with Coomasie blue for protein content [(1) of each panel] and stained with DIG glycan detection kit for glycoconjugate contents [(2) of each panel] in Fig.2 and Fig.3. After Nase and Gase digestion, the protein contents of THP were not changed [lanes 3-9 in Fig.2-A-(1) and (2)] whereas the glycoconjugates were reduced or even abolished [lanes 3-9 in Fig.2-B-(2)] compared to intact

(7)

THP (lanes 1 and 2 in all panels). There is no distinct change in protein and glycoconjugate contents in Case-digested THP (Fig.3-A). In contrast, both protein and glycoconjugate contents were moderately changed (Fig.3-B) in Oase-, and remarkably changed in V8, and PaseK-digestion (Fig.3-C, and D).

Comparison of protein-binding activity among intact THP and its enzyme- digested products As demonstrated in Fig.4, the serum protein-binding activity of enzyme-digested

products of THP with C1q cleavage products (panel A), BSA (panel B) and human IgG (panel C) were significantly reduced by Oase, V8 or PaseK digestion. But no change was found by carbohydrate-degrading enzyme (Nase and Gas) or Case digestion compared to intact THP. These results suggest that intact protein core structure, but not carbohydrate side-chains, in THP molecule is required for the entire protein- binding activity of THP.

Comparison of cytokine-binding activity among intact THP and its enzyme digested-products Although the binding of THP with IFN-γ is modest, we could find reduced IFN-　-binding activity in PaseK-digested THP products (Fig.5-A). In addition, the TNF-α, IL-6 and IL-8 binding activities of THP were significantly suppressed after V8-, PaseK-, or Oase-digestion as demonstrated in Fig.5-(B), (C) and (D). The carbohydrate-degrading enzymes (Nase and Gase) did not affect the binding of THP with different inflammatory cytokines although a tendency to reduction. These results further confirm that intact protein core structure is essential for the binding of THP with different inflammatory cytokines. The carbohydrate side-chains of THP seem play a minor role in protein-binding activity of THP molecule.

Comparison of mononuclear cell proliferating activity among intact THP and its enzyme-digested products

As shown in Fig.6, MNC proliferation-enhancing activity of intact THP was decreased by Oase-, V8- and PaseK-digestion, but not by Nase-, Gase- or Case- digestion consistent with serum protein-binding (Fig.4) and cytokine-binding (Fig.5) activities.

Comparison of neutrophil binding and phagocytosis-enhancing activities of intact and enzyme-digested products

For clarifying the affections of carbohydrate side-chains and protein core of THP on PMN functions, two parameters, PMN binding and PMN phagocytosis-enhancing activities, were conducted. We found the PMN binding activity of THP was enhanced after removing sialic acid and 　-galactose [Fig.7-A-(3) and (4)]. In contrast, the proteolytic digestion of THP core structure reduced PMN binding activity of THP [Fig.7-A-(7) and (8)]. The PMN phagocytosis-enhancing activity of THP was in parallel with PMN binding activity that removal of 　-galactose enhanced whereas destruction of protein core reduced the activity [Fig.7-B]. The same experiment using different sets of THP was repeated twice with a similar tendency. These results further support that protein core is essential and glycomoiety is inhibitory toward PMN binding and phagocytosis- enhancing activities mediated by THP.

Discussion

Tamm-Horsfall glycoprotein is the most abundant protein in normal human urine. The daily excretion of THG ranges from 50-200mg in humans [32]. The glycoprotein can interact with viral proteins [1], bacterial structure components [2], bacterial exotoxin [33], immunoglobulin light

(8)

chains [14,15], complement component 1 and 1q [16], proinflammatory cytokine IL-1 [12], IL-2 [34] and TNF-　 [12, 13], and surface membrane proteins on human PMN, lymphocytes and monocytes [17-21]. Immunologically, THG not only activates PMN through binding to a single class of sialic acid-specific cell surface receptors [17] but enhances the release of IL-1, IL-6, and TNF-　 from monocytes responsible for B and T lymphocytes proliferation [20, 21]. Interestingly, THP exhibits large glycosylated heterogeneities over 150 glycan structures, but is mainly sialylated to varying extents in seven glycosylated sites. The carbohydrate part can account for 28% of the total weight of THP [35]. Our purpose is to clarify the role of carbohydrate side-chains and protein core of normal THP in the pleotropic activities of this complex glycoprotein.

At least five original findings were observed in the present study; (a) Normal human urinary THP contains Siaα(2,3)Gal/GalNAc, but no Siaα(2,6)Gal/GalNAc, sugar moiety in the complex glycosylated side-chains judged by lectin-binding activity. (b) The protein-binding strength of THP follows the sequence of TNF-α, IgG, C1q, and BSA (high affinity) > IL-8 (moderate affinity) >IL-6 and IFN-γ (low affinity). (c) The protein-core structure integrity in THP molecule is essential for protein-binding, mononuclear cell proliferating, and PMN phagocytosis-enhancing activities of THP molecule. (d) The carbohydrate-moieties in the side chains of THP also play a role in THP-protein binding and mononuclear cell proliferating activities of THP molecule. (e) The Sia(2,3)Gal/GalNAc sugar moiety in THP probably plays a negative role in protein core-mediated PMN-binding and phagocytosis-enhancing activities of THP. These original observations have not reported in the literatures.

Muchmore et al. [12, 13, 34] demonstrated that the glycomoiety in THP was entirely responsible for binding to recombinant IL-1, IL-2 and TNF-α because these proinflammatory cytokines exhibited lectin-like specificity capable of reacting with THP. Dall’Olio et al. [36] further demonstrated that the immunosuppressive activity of THP on activated lymphocyte proliferation depends on oligosaccharides since removal of the outer sugars abolished the effect. However, Moonen et al. [37] reported that THP could not bind with native cytokines including TNF-　. In addition, the binding of THP to complement 1 and complement 1q depends on the electrostatic evens as demonstrated by the influence of hydrogen concentration and ionic-strength on THP-C1q binding affinity [16, 38]. In this study, we found that enzymatic digestion of THP with neuraminidase or　-glactosidase did not significantly affect the binding activity with different protein molecules, although a tendency of decrease was shown in Fig.4 and Fig.5. On the other hand, the binding activity of THP with different protein molecules was decreased by treatment with two proteases (V8 and PaseK) and O-sialoglycoprotein endopeptodase, but not carboxypeptidase Y. V8 protease is a serine protease that specifically hydrolyzes peptide bonds at the carboxylic end of glutamic acid and aspartic acid residues. Proteinase K is a stable serine protease with much broader substrate specificity than V8 protease. It cleaves protein molecules in the peptide bond adjacent to the carboxyl group of aliphatic and aromatic amino acids that blocks alpha amino groups. These two proteases can effectively digest protein core structure of THP. O-sialoglycoprotein endopeptidase cleaves proteins that are O-glycosylated on serine and threonine resides. Recent studies revealed that O-linked carbohydrate was present in THP molecules from non-pregnant females and males that are potential ligands of L-selectin [11, 35]. The substrates of this glycoprotease include CD34, CD43, CD44, CD45 and P-selectin [39, 40].

(9)

On the other hand, carboxypeptidase Y cleaves N-carbobenzoxyl- phenylalanine- alanine peptide bone but no affection on protein core structure of THP molecule as shown in Fig.3-A. Our results indicate that the intact protein core structure, but not sialic acid or 　-galactose glycomoiety in THP, is essential for THP–protein binding and immune cell-stimulating activities of THP. These observations seem contradictory to those of other authors that carbohydrate moieties in THP are crucial for the binding with lectin-like molecules, such as IL-1, IL-2 and TNF- [12, 13, 3　 4]. Sherblom et al. [13] demonstrated that rIL-2, a lectin-like molecule, shares 27% homology with a 33-residue sequence of the carbohydrate-binding domain of human mannose-binding protein. It is conceivable that IL-2 protein might function preferentially in acidic (pH 4-5) microenvironments to bind with THP molecule. However, Muchmore et al. [12] showed that diacetyl-chitobiose and Man(　1-6)Man(　1-3)- Man-O-ethyl are effective inhibitors for the binding of rTNF with THP. The removal of terminal sialic acid, galactose, and N-acetylhexosamine residues did not affect the binding of THP with TNF-　. Our results are quite consistent with Ying et al. [15] that a 10 amino acid peptide, MQGTHWPPLT, sequence corresponding to the CDR3 region of both and 　　 light chains of IgG molecule is the binding epitope of THP. Bovine serum albumin that contains minimal carbohydrate or lectin-like moiety can bind with THP in moderate degree [41]. Our studies clearly demonstrate that Nase- or Gase-digested products of THP that loss of the terminal sialic acid or galactose/N-acetylhexosamine sugar moieties did not affect its protein-binding activity (Fig.2, 3, and 4) while protein-degrading enzymes remarkably reduced its binding activity with different proteins (Fig.3, 4 and 5). It is conceivable that ionic-strength, electrostatic charge, specific peptide sequence, and other undefined mechanism involve in the binding of THG-protein core with serum proteins (C1q, BSA, IgG) and non-lectin-like pro-inflammatory cytokines (IL-6, IL-8 and IFN- 　Obviously, more studies are required to disclose the real molecular mechanism of THP protein core–protein binding activity.

In addition to bind with soluble serum proteins and proinflammatory cytokines, THP is a potent binder of surface membrane proteins on different blood cells and glomerular mesangial cells [28]. The binding of THP with cell surface molecules subsequently link innate and adaptive immunities via a Toll-like receptor-4-dependent mechanism [42]. Our preliminary data showed that 72, 60, and 30 kDa molecules on human MNC, and 72, 50, 40, and 23 kDa molecules on human PMN are the cognate surface membrane binding proteins for THP (data not shown). Functionally, we note that intact protein core structure is important for mononuclear cell proliferating and PMN phagocytosis-enhancing activities of THP (Fig.6). Different from THP-mediated mononuclear cell proliferation that protein core plays an essential and glycomoiety plays a minor role (Fig.6), the　-galactose moiety in intact-THP molecule may hinder PMN functions since treatment with 　-galactosidase remarkably enhances them (Fig.7). It is believed that the diverse protein-binding activities render THP an important immuno-modulatory factor in the urinary system [43]. In conclusion, THP binds with lectin-like cytokines via its N-linked oligosaccharides. The integrity of protein core in THP is essential for binding with non-lectin protein molecules including BSA, IgG, C1q, IL-6, IL-8 and TNF- 　that ionic strength, electrostatic charge, specific peptide sequence, and certain un-identified mechanism are involved.

(10)

Fig.1 Fig.2

Fig.3 Fig.4

Fig.5 Fig.6