Vitamin D3-binding protein (Ge protein), a serum glycoprotein, Is theprecursor for the macrophage activating factor. Cancer patient seracontain a-N-acetylgalactosamlnldase that deglycosylates Ge protein. Deglycosylated Ge protein cannot be converted to macrophage activatingfactor, leading to Immunosuppresslon. Of 46 oral cancer patients withsquamous cell cardnoma, approximately 22% had greatly reduced precursor activities. The precursor aCtivIty of approximately 61% of these
@nt.swas moderately reduced. The remaining patients (17%) hadprecursor activities equivalent to those of healthy humans. Patients withlow precursor activity of serum Gc protein had high serum a-N-acetylgalactosaminidase activity. In contrast, patients with high precursor activity had low serum a-N-acetylgalactosaminldase activity. Thus, levels ofserum a-N-acetylgalactosaminldase ofindividual patients have an inversecorrelation with precursor activities of their serum Gc protein. Surgicalremoval oftumors resulted In a subtle decrease In serum a-N-acetylgalactosaminidase activity with concomitant Increase In the precursor activityof serum Ge protein. Serum enzyme analysis of nude mice transplantedwith a human oral squamous carcinoma cell line revealed that senima-N-acetylgalactosaminidase activity Is directly proportional to tumor
ew-N-acetylgalactosamlnjdase activity in patient blood
(9—12)and superoxide-generating (14, 15) capacities of macrophages,suggesting a rapid transmission of macrophage activating signal fromlyso-Pc-treated lymphocytes (B and T cells) to macrophages duringthe 30-mn lyso-Pc treatment period. Macrophage activation requiresserum Gc protein and participation of lymphocytes (5, 7, 16—18).Asshown in Fig. la, a trisaccharide composed of N-acetylgalactosamine,
galactose, and sialic acid on Gc protein is hydrolyzed by the lyso-Pcinducible fI-galactosidase of B cells to yield a macrophage proactivating factor, which is in turn converted by the action of a T-cellsialidase to MAP, a protein with N-acetylgalactosamine as the remaining sugar moiety (5—7,18). Thus, Gc protein is a precursor for MAP(5—7,18). In vitro hydrolysis of Gc protein with immobilized @3-galactosidase and sialidase yields a remarkably high-titered MAF,termed GcMAF (6, 7, 18). A minute amount (4—10pg/mouse) ofGcMAF given to mice results in greatly enhanced phagocytic andsuperoxide-generating capacities of macrophages (6, 7, 18).
When we characterized peripheral blood lymphocytes and Gc protein in patients bearing various types of cancer, lymphocytes of allcancer patients were capable of generating MAP. However, the MAFprecursor activity of plasma Gc protein was lost or reduced, leading toimmunosuppression (4). Loss of the precursor activity was due todeglycosylation of Gc protein by a-N-acetylgalactosaminidase activity detected in a variety of cancer patients (4). Because of readilyperceptible primary oral tumors and metastasis (19, 20), we chose oralsquamous cell carcinoma patients to study correlation between theprecursor activity and a-N-acetylgalactosaminidase activity in patient
sera and surgical effect on these parameters. To determine the diagnostic/prognostic value of a-N-acetylgalactosaminidase, correlationbetween serum a-N-acetylgalactosaminidase activity and tumor burden was investigated using a nude mouse model transplanted with ahuman oral squamous cell carcinoma cell line.
MATERIALS AND METHODS
Reagents and Media. Lyso-Pc and p-nitrophenyl N-acetyl-a-D-galactosaminide were purchased from Sigma Chemical Co. (St. Louis, MO). Lymphoprep was obtained from Polysciences, Inc. (Warrington, PA). For in vitromanipulationandcultivationof lymphocytesandphagocytes,0.1%eggalbumm-supplementedRPMI 1640wasused.
Nude Mouse and Cell Line. FemaleBALB/c nu/numice, 4-5 weeks oldweighing —20g, were obtained from the Japan SLC, Inc. (Shizuoka, Japan).Mice were fed Purina Mouse Chow and water ad Iibitwn. The nude mice weretransplanted with a human oral squamous carcinoma cell line, KB.
Collection of Sera from Oral Cancer Patients. All patient blood sampleswere collected in yellow top tubes for serum separation. After centrifugation at1500 X g for 15 ruin, serum samples were assayed for both the precursoractivity of Gc protein and a-N-acetylgalactosaminidase activity.
Histological Analysis and Clinical Staging of Oral Cancer. Biopsy usmor tissue specimens were fixed in 10% formalin, embedded in paraffin, andcut into 4-pin sections. After deparaffinization, they were stained with H&Eand examined histologically. Histological review included all slides of eachcase and evaluation for the following parameters (21): (a) tumor types (squamous cell carcinoma or others); (b) tumor differentiation grade (well, moderately,andpoorly differentiated);and(c) tumor invasioninto lymphovascularspaces.Using roentgenography and scintigram, the patients were grouped intofour stagings based on the tumor-node-metastasis classification of the International Union Against Cancer system (22).
burden. Thus,stream can serve as a dlagnosife/prognoslic Index.
INTRODUCTION
Advanced cancer patients suffer from a deficiency of immunity andincreased susceptibility to infection (1, 2). Susceptibility of cancerpatients to infection and concomitant depression of antibody production become more evident as the disease progresses (3). Recently, adefect in the macrophage activation cascade has been demonstrated inall types of cancer patients (4). Particularly advanced cancer patientshave a severe defect in macrophage activation process. Becausemacrophage activation for phagocytosis and antigen presentation isthe first step in the immune developmental cascade, a defect in themacrophage activation cascade leads to immunosuppression.
Inflammation-primed macrophage activation cascade has been defined as the major process for the production of MAF@ (5—8).Infections and tissue damage induce inflammation, resulting in macrophageactivation (9—12).Membranous lipid metabolites of inflamed tissues,lysophospholipids, are potent macrophage stimulating agents (9—12).Administration of lyso-Pc to mice activates macrophages to phagocytize target antigens or cells via the Fc receptor and to generatesuperoxide (9—13).In vitro treatment of mouse peritoneal cells (mixtare of macrophages and lymphocytes) with lyso-Pc in a serumsupplemented medium for 30 rain, followed by a 3-h cultivation of themacrophages without lymphocytes, markedly enhanced phagocytic
charges. This article must therefore be hereby marked advertisement in accordance with18U.S.C.Section1734solelyto indicatethisfact.
I This study was supported by United States Public Health Service Grant AI-32140 (to
N.Y.).2 To whom requests for reprints should be addressed, at the Laboratory of Cancer
Immunology, Albert Einstein Cancer Center, Korman Research Pavilion B-31, 5501 OldYork Road, Philadelphia, PA 19141. Phone: (215) 456-6335; Fax: (215) 456-3824.
3 The abbreviations used are: Oc protein, human vitamin D3 binding protein; MAF,
Prognostic Utility of Serum a-N-Acetylgalactosaminidase and ImmunosuppressionResulted from Deglycosylation of Serum Gc Protein in Oral Cancer Patients'
Nobuto Yamamoto,2 Venkateswara R. Narapara.ju, and Masahiro Urade
Laboratory of Cancer Immunology and Molecular Biology, Albert Einstein Cancer Center, Philadelphia, Pennsylvania 19141, (N. Y., V. R. N.], and Department of Oral Surgery.Hyogo College of Medicine. Nishinomiya, Hyogo 663, Japan (M. U.]
Macrophago activatingfactor (MAF)Fig. I. Schematic illustration offormation ofMAF(a) and
deglycosylation of Gc protein (b). s, lyso-Pc-treated B cells.
Gc protein (Gc 1)
—cx,
Deglycosylated Gc protein
Analysis of Squamous Cell Carcinoma Antigen In Oral Cancer PatientSara. Oral cancer patient sera were submitted to RIA for SCCAg by themethod of Kato and Torigoe (23). Values of patient SCCAg assayed by RIAare dependent on the quality of monoclonal antibody. Therefore, before
SCCAg RIA of cancer patient sera, healthy human sera were assayed forstandardization and found to be <2 ng/ml of serum.
Assay for MAF Precursor Activity of Serum Ge Protein. Peripheralblood mononuclear cells (mixture of phagocytes and lymphocytes) of healthyhumans can be used for determination of the MAF precursor activity of patientserum Gc protein. Healthy human blood samples were collected in purple-toptubescontainingEDTA to preventcoagulation.A 5-mIbloodsampleand5 ml ofsaline (0.9% NaCI) were mixed and gently laid on a 15-mI centrifuge tubecontaining 3 ml of Lymphoprep and centrifuged at 1000 X g for 15 ruin. Thedense white mononuclearcell layer containing monocytes/macrophages(designated phagocytes) and lymphocytes (B and T cells) was collected using Pasteur
pipettes,washed twice with PBS, suspendedin 0.1% egg albumin-supplementedRPMI 1640,and placed in 16-mmculture wells at a concentrationof 6—9X l0@cells/well.Incubationfor 45 ruin in a 5%CO2incubatorat 37°Callowedadherence of phagocytes to the plastic substrata.Lyso-Pc (I pg/mi) was added to thewellsand incubatedat 37°Cfor 30 rain.The adherentphagocytesand nonadherentlymphocyteswerewashedseparatelywith PBS,admixed,andculturedfor 3 h at37°Cin a medium supplemented with 0.1% patient senim as a source of Ocprotein before assay for superoxidegenerationof the phagocytes.
For superoxidegenerationassay (14, 15), the phagocyteswere washedwithPBS andincubatedin 1 ml of PBScontaining20 @gofcytochromec for 10mm.About 30 mm after addition of phorbol-l2-myristate acetate (5 @g/ml),thesuperoxide-generatingcapacityof the phagocyteswas determinedby spectrophotometricalcytochromec reduction at 550 am. The data were expressedas nanomoles ofsuperoxide produced/min/l0@phagocytesand were taken to representtheprecursoractivity of serum Oc protein of individualpatients.
Detection of a-N-Acetylgalactosaininidase in Patient Bloodstream.Cancerpatientsera(300pJ) andnudemousesera(100s.d)wereprecipitatedwith 70% saturated ammonium sulfate. The precipitates were dissolved in 50mMsodium citrate buffer (pH 6.0) and dialyzed against the same buffer at 4°Cfor 2 h. The dialysates were made up to 1 ml in volume and assayed for theenzyme (4, 24). Substrate solution (250 p1) contained 5 @molof p-nitrophenylN-acetyl-a-D-galactosaminidein 50 mMcitratebuffer(pH 6.0).Thereactionwas initiated by addition of 250 @lof the dialyzed samples, kept at 37°Cfor60 mm, and terminated by adding 200 p1 of 10% trichloroacetic acid. Aftercentrifugation of the reaction mixture, 300 @lof 0.5 MNa2CO3solution wereadded to the supernatant. The amount of released p-nitrophenol was determined spectrophotometrically at 420 nm and expressed as nmol/mg/min.Protein concentrations were estimated by the Bradford method.
Serum a-N-acetylgalactosaminidase activity is extremely stable and highlyreproducible after storage at 4°Cfor more than a month (4). To study the effectof surgical removal of malignant lesions on the serum enzyme level, seracollected at various postoperative periods in several cases allowed us toanalyze changes in the enzyme level.
Histological and Tumor Antigen Analyses of Oral Cancer. Acohort of 46 oral cancer patients with squamous cell carcinoma wasstudied. We began histological examination of 36 patients at the time ofdiagnosis. The remaining 10 patients have already been under various
therapies. As shown in Table 1, carcinoma tissues of 36 newly diagnosedpatients were characterized by degree of differentiation. The majority
(86%) of patients' tumors was well differentiated, whereas only fivepatients' tumors were moderately differentiated. Tumor-node-metastasisstagings of individual patients were also determined.
Scm ofthe individual patients were analyzed for SCCAg by RIA usingmonoclonal antibody (23). These patient sera contained various levels ofthe antigen, ranging from 0.5 to 14.1 ng/ml of serum. No significantcorrelation between SCCAg levels and histology or staging can be seenin Table 1. The majority (23 patients) carried SCCAg <2.0 ng/ml. Onlyeight patients (approximately 26% of this patient population) carriedSCCAg >2.0 ng/ml. Because sara of healthy humans contain up to 2ng/ml of SCCAg, the majority of this patient population falls within theantigen levels of healthy humans. Thus, establishment of sensitive diagnostic and prognostic indices was imperative.
Precursor Activity of Serum Gc Protein in Oral Cancer Patients. When lyso-Pc-treated peripheral blood mononuclear cells
(mixture of lymphocytes and phagocytes) of healthy humans werecultured in a medium supplemented with 0.1% serum of individualpatients for 3 h, the precursor activity of serum Gc protein was greatlyreduced in approximately 22% of this patient population and producedsuperoxide ranging from 0.31 to 1.5 nmol/min/l06 cells as shown inTable 2. Thus, phagocyte activation cannot develop in certain cancerpatients. In this analysis, approximately 61% of the patients had
moderately reduced MAF precursor activities of serum Gc protein, asdemonstrated by phagocytes to produce superoxide ranging from 1.5to 3.7 nmoL/min/106 cells. The remaining cancer patients (approximately 17%) had MAP precursor activities (>4.0 nmol/min/l06phagocytes) similar to those of healthy humans.
Deglycosylation of Gc Protein Is the Cause for Loss of Precursor Activity. Loss of the MAF precursor activity of patient serum Gcprotein appears to be the result of deglycosylation of Gc protein (4,24). Deglycosylated Gc protein cannot be converted to MAP (Fig. lb).Because Gc protein is O-glycosylated (25, 26), deglycosylation ofserum Gc protein appears to be due to the presence of a-N-acetylgalactosaminidase in the patient bloodstream (4). Thus, serum samples of these oral cancer patients were assayed for a-N-acetylgalactosaminidase. As shown in Table 2, patients having a greatly reduced
Table I Site and histological classification of carcinoma. SCCAg, and clinical staging of oral cancer patientsHealthy human sera contain SCCAg ranging from 0.3 to 2.0 ng/ml of plasma.
PatientSCCAg
(ng/ml plasma)C
I .24.94.1
0.80.8I .00.42.2I .20.51.1ND7.0I .00.80.7
0.6I.22.1
NDND2.9
I.00.41.4ND2.22.3I .60.61.2ND0.5I .20.9
Original tumor staging@'
T2N0MØ IIT1N@M0 IllT2N1M0 IIT2NØM() IIT1N@M0 IT2N@M0 IIT2N1M0 IIIT3N@M,J 111T2N(@M0IIT1N@M@) III
T2N@M0 IIT,N@M0 IIT4NQM0 IVT3N@M0 IIIT1N@M() IT1N(@M@J IT2N@M() IIT1N(@M0IT4N1M0 IVT2N@M() II
T2N@M0 IIT3N@M@IllT4N@M() IVT3N1M0 IIIT1N@M() IT,N@M@) II
Tongue, well differentiated 5CCTongue, well differentiated 5CCOral floor, well differentiated SCCLower gingiva, well differentiated 5CCTongue, well differentiated SCCUpper gingiva, well differentiated 5CCBuccal mucosa, well differentiated 5CCTongue,welldifferentiated5CCLower gingiva, well differentiated 5CCMaxillary sinus, well differentiated SCCLower gingiva, well differentiated 5CCTongue, well differentiated SCCOral floor, well differentiated 5CCMaxillary sinus, well differentiated 5CCTongue, well differentiated 5CCOral floor, well differentiated 5CCBuccal mucosa, moderately differentiated 5CCTongue, well differentiated 5CCLower gingiva. well differentiated SCCBuccal mucosa, well differentiated 5CCBuccalmucosa,welldifferentiated5CCOral floor, well differentiated 5CCOral floor, well differentiated5CCTongue, well differentiated 5CCTongue, well differentiated 5CCLower gingiva, well differentiated 5CCUpper gingiva, well differentiated 5CCUpper gingiva, moderately differentiated 5CCTongue, moderately differentiated 5CCTongue, well differentiated 5CCTongue, moderately differentiated 5CCOral floor. moderately differentiated 5CCOral floor, well differentiated 5CCLower gingiva, well differentiated 5CCLower gingiva, well differentiated 5CCBuccal mucosa, well differentiated 5CC
precursor activity of serum Gc protein carried large amounts (3.51—
7.88 nmol/mg/min) of cs-N-acetylgalactosaminidase in their bloodstream. Patients having moderately reduced precursor activity carrieda-N-acetylgalactosaminidase activities ranging from 1.15 to 3.12nmol/mg/min. Patients who had high levels of precursor activity
carried small but significant amounts (0.75—1.15nmol/mg/min) ofa-N-acetylgalactosaminidase in their bloodstream. Sera of healthyhumans contained an extremely low level (approximately 0.29 nmol/mg/mm) of a-N-acetylgalactosaminidase activity. Thus, the enzymeactivity in patient sera clearly showed an inverse correlation with theprecursor activity of the patient serum Gc protein, as can be seen inTable 2. To illustrate this inverse correlation between the precursoractivity and serum a-N-acetylgalactosaminidase activity of all 36patients, the enzyme activity was plotted against the precursor activityof plasma Gc protein. As shown in Fig. 2, serum a-N-acetylgalactosaminidase activity increased as the precursor activity of plasma Gcprotein decreased, and a linear inverse correlation between the precursor activity and plasma enzyme activity is clearly evident atenzyme levels above 1.3 nmol/mg/min. When we followed the precursor activity and serum a-N-acetylgalactosaminidase activity ofthese oral cancer patients for 2 years, regardless of therapeutic background (i.e., various chemotherapies and radiation), the data fall in thesame linear inverse correlation as those values at the time initiallydiagnosed, as shown in Fig. 2. Thus, increase in patient serum a-N-acetylgalactosaminidase activity is exclusively responsible for decrease in the precursor activity of serum Gc protein. However, because a large amount (250 @g/ml)of Gc protein circulates in thebloodstream, small amounts (< 1.3 nmol/mg/min) of serum a-N-
acetylgalactosaminidase hydrolyze only a small fraction of Gc protein, and thus do not affect their precursor activity significantly (seeFig. 2). Nevertheless, these results led us to conclude that deglycosylation of serum Gc protein by a-N-acetylgalactosaminidase is acause for inactivation of the precursor activity of plasma Gc protein,thus leading to immunosuppression.
The Origin of Serum a-N-Acetylgalactosaminidase. We proposed previously that the origin of serum a-N-acetylgalactosaminidase is cancerous cells (4). High a-N-acetylgalactosaminidase activities were detected in tumor tissue homogenates of various organs (4).As shown in Table 3, surgical removal of malignant lesions results ina subtle decrease of plasma a-N-acetylgalactosaminidase activity,with concomitant increase in the precursor activity. However, evenafter surgical removal of primary tumor lesions, the postoperativepatients carried significant amounts of a-N-acetylgalactosaminidaseactivity in their bloodstream, suggesting that the enzyme is derivedmainly from the metastasized tumor lesions. When lymph nodesharboring metastasized cancerous cells became palpable, surgicalremoval of the lymph nodes further decreased the serum enzymeactivity toward that of the healthy control level with concomitantincrease of the precursor activity to the healthy control (Table 3).
Quantitative Correlation of a-N-Acetylgalactosaminidase Activity with Tumor Burden. When a human oral squamous cellcarcinoma (KB) cell line was cultured, a-N-acetylgalactosaminidaseactivity in cultured medium increased as cell counts increased (datanot shown). From 1 week after nude mice were transplanted s.c. with2 X 106@ cells/mouse, we began analyzing serum a-N-acetylgalactosaminidase activity. Serum a-N-acetylgalactosaminidase activity in
Fig. 2. Inverse correlation between the precursor activity of serum Gc protein anda-N-acetylgalactosaminidase detected in patient bloodstream. All follow-up serum specimens were obtained from patients who are not under treatment (at least 2 weeks after avariety of treatments).
creased as the tumor size (measured by weight) of the solid tumorincreased, as shown in Fig. 3. Thus, the enzyme activity was directlyproportional to tumor burden in the host. In support of this finding,serum a-N-acetylgalactosaminidase activity levels in BALB/c mousetransplanted with Ehrlich ascites tumor are proportional to the tumorcell counts in peritoneal cavity (27).
In the present paper, we began investigating oral squamous cell carcinoma to determine the diagnostic/prognostic value ofthe existing tumormarker and histological analysis. Squamous cell carcinoma antigen was
detected at significant levels in only approximately 26% of this patientpopulation. We found no obvious correlation between the carcinoma
antigen and histological analysis or disease staging. Thus, an alternatediagnostic/prognostic assay for oral cancer became desirable. The malignant specific serum enzyme a-N-acetylgalactosaminidase was detected inall stages of the oral cancer patients. This enzyme is universally detectedin a variety of cancer patients, but not in healthy humans (4). Patientserum a-N-acetylgalactosaminidase deglycosylates Gc protein (Fig. 1 b).Deglycosylated Gc protein cannot be converted to MAF. Thus, thisserum enzyme inactivates the MAF precursor activity of Oc protein.Advanced cancer patient serum contains high a-N-acetylgalactosaminidase activity, leading to no macrophage activation (4). Because macrophage activation is the first step in the immune development cascade,such cancer patients become severely immunosuppressed. This mayexplain, at least in part, why cancer patients die from overwhelminginfection, e.g., pneumonia.
HIV-infected patients also carry a-N-acetylgalactosaminidase intheir bloodstream (24). This enzyme seems to play a role in immunosuppression in HIV-infected/AIDS patients. The enzyme in cancerpatients appears to be coded by an oncogene, whereas the enzyme in
Fig. 3. Correlation between serum a-N-acetylgalactosaminidase activity and tumorburden (measured by total weight) in nude mouse transplanted with human oral squamouscell carcinoma (KB) cell line.
HIV-infected/AIDS patients is coded by the viral genome.4 An impairment of immune function of the hosts will enhance opportunitiesfor survival and proliferation of neoplastic cells (28). Thus, AIDSpatients are at extremely high risk for development of various formsof malignant tumors (29).
Our accumulated evidence suggests that secretion of a-N-acetylgalactosaminidase from cancerous cells to the bloodstream appeared to be responsible for deglycosylation of patient serum Gcprotein (4). The present study revealed that the amount of a-N-acetylgalactosaminidase in the bloodstream is directly proportionalto the number of cancerous cells or the size of tumor in the hosts(Fig. 3). Surgical removal of malignant lesions resulted in a subtledecrease in serum a-N-acetylgalactosaminidase activity and concomitant increase in the precursor activity. However, even aftersurgical removal of tumor lesions, the majority of postoperativepatients carried significant amounts of a-N-acetylgalactosaminidase activity in their bloodstream (Table 3), suggesting that theenzyme is derived from the remnant cancerous lesions and/ormetastasized tumor lesions. These remnant cancerous lesions andmetastasis in these postoperative patients cannot be detectable byany other procedures (e.g. , X-ray, Ga67 scintigram, etc). Oralcancer often metastasizes to neck lymph nodes, but less frequentlyinto distant organs (19, 20, 30, 31). When the lymph nodes withmetastasized cancerous cells became palpable, surgical removal ofthe metastasized lymph nodes resulted in subtle decreased enzymelevels toward the healthy control level (see Table 3). Thus, levelsof a-N-acetylgalactosaminidase in cancer patient bloodstreamshould serve as a prognostic index. Unlike various existing tumormarkers, this enzymatic tumor marker is universal among a variety
of cancers (4). We have been using this most sensitive enzymeassay as a prognostic index during the course of radiation therapy
(4) and various other therapeutic procedures for various types ofcancer. Under a variety of therapeutic conditions, serum a-N-acetylgalactosaminidase and the MAF precursor activity of Gcprotein always show an inverse correlation. Furthermore, directproportionality of serum a-N-acetylgalactosaminidase activity totumor burden in the hosts firmly supports the reliability of theserum a-N-acetylgalactosaminidase activity as a prognostic index.
4 N. Yamamoto and V. R. Naraparaju, unpublished observations.
299
REFERENCES
I . Bumet, M. Immunologic factors in the process of carcinogenesis. Br. Med. Bull., 20:154—158,1964.
2. Gross, L. Immunological defect in aged populations and its relationship to cancer.Cancer Res., 18: 201—204,1965.
3. Freedman, H., and Ceglowski, W. Immunosuppression in the etiology of cancer. In:H. Freedman and W. Ceglowski (eds.), Immunological Aspects of Neoplasia, pp.253—291. Baltimore: Williams & Wilkins, 1975.
4. Yamamoto, N., Naraparaju, V. R., and Asbell, S. 0. Deglycosylation of serumvitamin D3-binding protein leads to immunosuppression in cancer patients. CancerRes.,56:2827—2831,1996.
5. Yamamoto, N., and Homma, S. Vitamin D3 binding protein (group-specific component,Gc) is a precursorfor the macrophageactivatingsignal factor from lysophosphatidylcholine-treatedlymphocytes.Proc. NatI. Acad. Sci. USA, 88: 8539—8543,1991.
6. Yamamoto, N., and Kumashiro, R. Conversion of vitamin D3 binding protein (group
specific component) to a macrophage activating factor by the stepwise action of (3-galactosidaseof B cells and sialidaseof T cells. J. Immunol., 151: 2794—2902,1993.
7. Naraparaju, V. R., and Yamamoto, N. Roles of (3-galactosidase of B lymphocytes andsialidase of I lymphocytes in inflammation-primed activation of macrophages. Immunol.Lett.,43: 143—148,1994.
8. Yamamoto, N., and Naraparaju, V. R. Role of mouse vitamin D3-binding protein inactivationof macrophages.J. Immunol.,157: 1744—1751,1996.
9. Ngwenya, B. Z., and Yamamoto, N. Activation of peritoneal macrophages by lyso
phosphatidylcholine. Biochim. Biophys. Acts 839: 9—15,1985.10. Ngwenya, B. Z., and Yamamoto N. Effects of inflammation products on immune
systems: lysophosphatidylcholine stimulates macrophages. Cancer Immunol. Immunother., 21: 1074—1082, 1986.
I 1. Yamamoto, N., and Ngwenya, B. Z. Activation of macrophages by lysophospholipids andether derivatives of neutral lipids and phospholipids. Cancer Res., 47: 2008—2013,1987.
12. Ngwenya, B. Z., and Yamamoto, N. Contribution of lysophosphatidylcholine-treatednonadherent cells to mechanism of macrophage activation. Proc. Soc. Exp. Biol.Med.,193:118—124,1990.
13. Yamamoto, N., Ngwenya, B. Z., Sery, 1. W., and Pieringer R. A. Activation of
macrophages by ether analogues of lysophospholipids. Cancer Immunol. Immunother.,25: 185—192,1987.
14. Yamamoto, N., Kumashiro, R., Yamamoto, M., Willett, N. P., and Lindsay, D. D.
Regulation of inflammation-primed activation of macrophages by two serum factors,vitamin D3-binding protein and albumin. Inf. 1mm., 61: 5388—5391, 1993.
15. Yamamoto. N., Willen, N. P., and Lindsay, D. D. Participation of serum proteins in theinflammation-primedactivationof macrophages.Inflammation,/8: 311—322,1994.
16. Homma, S., Yamamoto, M., and Yamamoto, N. Vitamin D binding protein (groupspecificcomponent,Gc)isthesoleserumproteinrequiredformacrophageactivationafter treatment of peritoneal cells with lysophosphatidylcholine. Immunol. Cell Biol.,71: 249-257, 1993.
17. Yamamoto, N., Homma, S., and Millman, I. Identification of the serum factorrequired for in vitro activation of macrophages: role of vitamin D, binding protein(group specific component, Gc) in lysophospholipid activation of mouse peritonealmacrophages. J. Immunol.. /47: 273—280,1991.
18. Yamamoto, N. Structural definition of a potent macrophage activating factor derivedfrom vitamin D3 binding protein with adjuvant activity for antibody production. Mol.Immunol., in press, 1996.
19. Lindberg, R. Distribution of cervical lymph node metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer (Phila.), 29: 1446—1449,1972.
20. Spiro, R. H., Alfonso, A., Fan, H. W., and Strong, E. M. Cervical node metastasisfrom epidermoid carcinoma of oral cavity and oropharynx. Am. J. Surg., 128:562—567,1974.
21. Shafer, W. G., Hine, M. K., and Levy, B. M. A texthook of oral pathology, 4th Ed.Philadelphia: B. W. Saunders, 1983.
22. Hermanek, P., and Sabin, L. H. (eds.) Union Intemationale contre Ic Cancer: TNMclassification of malignant tumors. Head and neck tumors. New York: SpringerVerlag, 1987.
23. Kato, H., and Torigoe, T. Radioimmunoassay for tumor antigen of human cervicalsquamous cell carcinoma. Cancer (Phila.), 40: 1621—1628,1977.
24. Yamamoto, N., Naraparaju, V. R., and Srinivasula, S. M. Structural modification ofserum vitamin D1-binding protein and immunosuppression in AIDS patients. AIDSRet. Hum. Retroviruses, 11: 1373—1378,1995.
25. Coppenhaver. D. H., Sllenne, N. P., and Bowman, B. H. Post-translational heterogeneity of the human vitamin D-binding protein (group-specific component). Arch.Biochem. Biophys., 226: 218—223,1983.
26. Viau, M., Constans, J., Debray, H., and Motreuil, J. Isolation and characterization ofthe O-glycan chain of the human vitamin-D binding protein. Biochem. Biophys. Res.Commun., 117: 324—331, 1983.
27. Koga, Y., Naraparaju, V. R., and Yamamoto, N. Antitumor effects of vitaminD3-binding protein-derived macrophage activating factor on Ehrlich tumor bearingmice. Cancer Research Proc., 37: 481, 1996.
28. Bumet, F. M. The concept of immunological surveillance. Prog. Exp. Tumor Res., 13:1—27,1970.
29. Dalgleish, A. 0., Beverly, P. C. L.. Clapham, P. R., Crawford, D. H., Greaves, M. F.,and Weiss, R. A. The CD4 (T4) antigen is an essential component of the receptor forthe AIDS retrovirus. Nature (Lond.), 312: 763—767,1984.
30. Probert, J. C., Thompson, R. W., and Bagshow, M. A. Pattems of spread of distantmetastases in head and neck cancer. Cancer (Phila.), 33: 127—133,1974.
31 . Memo, 0. R., Lindberg. R. D., and Fletcher, G. H. An analysis of distant metastasisfrom squamous cell carcinoma of upper respiratory and digestive tracts. Cancer(Phila.),40:145—151,1977.
1997;57:295-299. Cancer Res Nobuto Yamamoto, Venkateswara R. Naraparaju and Masahiro Urade Gc Protein in Oral Cancer PatientsImmunosuppression Resulted from Deglycosylation of Serum
-Acetylgalactosaminidase andN-αPrognostic Utility of Serum
