[CANCER RESEARCH57, 2187—2192,June I, 1997)

ABSTRACT

Vitamin D3-binding protein (DBP; human DBP is known as Gc protein)is the precursor of macrophage activating factor (MAF). Treatment ofmouse DBP with hnmobffized @3-galactosidaseor treatment of human Gcprotein with immobifized @-galactosidaseand sialidase generated a remarkably potent MAF, termed DBPMAF or GCMAF, respectively. Thedomain of Ge protein responsible for macrophage activation was clonedand enzymatically converted to the cloned MAF, designated CdMAF. InEhrlich asates tumor-bearing mice, tumor-specific serum a-N-acetylgalactosaminidase (NaGalase) activity increased linearly with time as thetransplanted tumor cells grew in the peritoneal cavity. Therapeutic effectsofDBPMAF, GCMAF, and CdMAF on mice bearing Ehrlich ascites tumorwere assessed by survival time, the total tumor cell count in the peritonealcavity, and serum NaGalase activity. Mice that received a single administration of DBPMAF or GcMAF (100 pg/mouse) on the same day aftertransplantation of tumor (1 x 1O@cells) showed a mean survival time of35 ±4 days,whereastumor-bearing controls had a meansurvival time of16 ±2 days. When mice received the second DBPMAF or GcMAFadministration at day 4, they survived more than 50 days. Mice thatreceived two DBPMAF administrations, at days 4 and 8 after transplanteflon of 1 x@ tumor cells, survived up to 32 ±4 days. At day 4posttransplantation, the total tumor cell count in the peritoneal cavity wasapproximately 5 x 10@cells. Mice that received two DJ3PMAF administrations, at days 0 and 4 after transplantation of 5 x 1O@twnor cells, alsosurvived up to 32 ±4 days, while control mice that received the 5 x 10@ascites tumor cells only survived for 14 ±2 days. Four DBPMAF,GeMAF, or CdMAF administrations to mice transplanted with 5 X 10@Ehrlich ascites tumor cells with 4-day intervals showed an extendedsurvival of at least 90 days and an insignificantly low serum NaGalaselevel between days 30 and 90

INTRODUCTION

Inflammation results in macrophage activation, leading to immunedevelopment. Inflamed tissues release membranous lipid metabolites,lysophosphatidylcholine and other lysophospholipids, that efficientlyactivate macrophages (1—3).Inflammation of cancerous tissues (e.g.,melanoma and bladder cancer) induced by intratumor administrationof BCG (Bacille Calmette Guerin) or other bacterial cells can result inthe regression of local as well as metastasized tumors, suggesting thedevelopment of specific immunity against the cancerous cells (4, 5).Inflamed cancerous tissues release alkylglycerols as well as lysophospholipids, because cancerous tissues contain alkylphospholipids (6—8). Alkylglycerols are approximately 400 times more active macrophage-activating agents than lysophospholipids in terms of the minimaldosages required for optimal macrophage activation (8). This mayexplain why inflammation in cancerous tissues is curative. In fact,macrophages have a potential to eliminate cancerous cells whenactivated (9). Because macrophage activation for phagocytosis and

Received 11/26/96; accepted 4/2/97.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by USPHS Grant RO1 AI-32140 from the National Institute for

Allergy and Infectious Diseases.2 To whom requests for reprints should be addressed, at the Laboratory of Cancer

Immunology and Molecular Biology, Albert Einstein Cancer Center, Korman ResearchPavilion B-31, 5501 Old York Road, Philadelphia, PA 19141.

subsequent antigen presentation is the first step of the immune development cascade, the capacity of the hosts to activate their ownmacrophages with their serum provides an index for their immunepotential (10).

Our accumulated evidence indicates that inflammation-primedmacrophage activation requires the participation of B and T lymphocytes (3, 8, 11—14)and serum DBP3 (human DBP is known as Gcprotein) (12—15).Gd protein (the major subgroup of human Gcprotein) carries a trisaccharide composed of N-acetylgalactosaminewith dibranched galactose and sialic acid termini ( 15—17).As shownin Fig. la, this oligosaccharide is hydrolyzed by membranous f3-galactosidase of inflammation-primed B cells to yield a macrophageproactivating factor, which is in turn hydrolyzed by sialidase ofT-cells to yield a MAF (15—17).Mouse DBP carries a disaccharidecomposed of N-acetylgalactosamine with a galactose terminal. Hydrolysis of this disaccharide by /3-galactosidase of B cells alonegenerates a potent MAF (Refs. 18—20;Fig. lb). Thus, DBP and Gcprotein are precursors for MAFs (15, 18). Incubation of mouse DBPwith immobilized (3-galactosidase or incubation of Gc protein withimmobilized f3-galactosidase and sialidase generates a remarkablyhigh-titered MAF, termed DBPMAF or GcMAF, respectively (16—18,21), whose potency and time course of effect suggest that theseproteins act directly on macrophages and are the most active factorsknown for this process.

A recent important observation was that cancer patients had decreased capability of macrophage activation with their own serumbecause of decreased precursor activity of the serum Gc protein (10).Loss of the precursor activity of the Gc protein was found to be dueto the deglycosylation of Gc protein by serum NaGalase derived fromcancerous cells (Fig. lc; Refs. 10 and 22). Once Gc protein isdeglycosylated, lymphocyte glycosidases (i.e., j3-galactosidase andsialidase of lymphocytes) can no longer convert the deglycosylatedGc protein to the MAF (10). Thus, macrophages cannot be activatedunder such a circumstance. However, GcMAF and DBPMAF canbypass the decapitated macrophage activation cascade and efficientlyactivate macrophages (10, 2 1). Because macrophages have a potentialto kill and eliminate cancerous cells, we studied the efficacy ofDBPMAF, GcMAF, and its cloned derivative (CdMAF) on micebearing Ehrlich ascites tumor. The fate of the transplanted tumor cellsand tumor-specific serum NaGalase as a prognostic index was monitored as the therapy progressed.

MATERIALS AND METHODS

Animals and Tumor

Female BALB/c mice, 7—12 weeks of age, were obtained from The JacksonLaboratory (Bar Harbor, ME). Mice were fed Purina Mouse Chow and waterad libitum. Ehrlich ascites tumor was obtained from the Division of Cancer

Treatment, National Cancer Institute (Bethesda, MD) and maintained by serialpassage through the BALB/c mouse peritoneal cavity.

3 The abbreviations used are: DBP, vitamin D3-binding protein; MAF, macrophage

activating factor; DBPMAF, enzymatically generated DBP-derived MAF: Ge, humanDBP; GcMAF, enzymaticallygeneratedGe protein-derivedMAF; CdMAF, cloned domain of GcMAF responsible for macrophage activation; NaGalase, a-N-acetylgalactosaminidase.

2187

Immunotherapy of BALB/c Mice Bearing Ehrlich Ascites Tumor with Vitamin

D-binding Protein-derived Macrophage Activating Factor'

Nobuto Yamamoto,2 and Venkateswara R. Naraparaju

Laboratory of Cancer Immunology and Molecular Biology, Albert Einstein Cancer Center, Philadelphia, Pennsylvania 19141

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF MAP

aLaJ@Ach

Gd protoin

C.

4@tosase@ofBc&@@IIs* f@ase

@!ch@

@..

—* GaIN @ofTceIIs@IIs

V VAc hMacrophage

activatingfactor(MAF)I

Ib.Mouse DBP@ @@c@Ser

Gal- - @ofBc&@*@IIs*

.. VMacrophage

activatingI

factor (MAF)

Fig. I . Schematic illustration of the formation ofthe MAF derived from Ge protein (a) and mouse DBP(b) and the deglycosylation of Ge protein and mouseDBP (c). a, lysophosphatidylcholine-treatedB cells.

4,/@cetyl

“@a9@..

DeglycosylatedMouse DB

DeglycosylatedGc protein (Gc 1)

Proteins, Enzymes, and Reagents

HumanGcproteinconsistsof threegeneticgroups(proteinpolymorphism):Gclf, Gels, and Gc2. Both Gclf and Gels subtypes of Gel carry sialic acid,whereas 0c2 does not (15). Gel protein (a mixture of Gclf and Gels) and

mouse DBP were purified by vitamin D-affinity chromatography (18, 21, 23).Glycosidases were purchased from Boehringer Mannheim (Indianapolis, IN)and immobilized on Sepharose (21). p-Nitrophenyl N-acetyl-a-D-galactosaminide was purchased from Sigma Chemical Co. (St. Louis, MO). Using

the Limulus amebocyte lysate assay (2), we routinely tested for freedom oflipopolysaccharide contamination in the concentrated stock solutions of enzymes, human Ge protein, mouse DBP, and culture media.

In Vitro Treatment of DBP and Its Cloned Derivative with Immobilizedfi-Galactosidase and Sialidase

Treatment of Mouse DBP by Immobilized Glycosidase. Mouse DBPcarries a disaccharide composed of N-acetylgalactosamine and galactose (18—20). DBP (2 I.Lg) was incubated with immobilized @3-galactosidase (0.1 unit/ml)

in 1 mM PBS by tumbling motion at pH 7.4 for 1 h to yield the MAF,DBPMAF. A small amount (100 pg/mI) of this product activates mouseperitoneal macrophages for 7-fold-increased phagocytic capacities and 15-fold-increased superoxide-generating capacities (18).

Treatment of Gd Protein by Immobilized Glycosidases. Human Gelprotein carries a trisaccharide composed of N-acetylgalactosamine with di

branched galactose and sialic acid termini (15—17,21). Stepwise digestion ofGel glycoprotein by immobilized f3-galactosidase and sialidase yields theMAF, GcMAF. A small amount (10 pg/ml) of this product was required toactivate mouse peritoneal macrophages for 7-fold-increased phagocytic capac

ities and 15-fold-increased superoxide-generating capacities.

Treatment of the Cloned Ge Domain by Immobilized Glycosidases toYield the MAF, CdMAF. Chemical (i.e., cyanogen bromide) and proteolyticfragmentations of Ge peptide (Mr 52,000; 458 amino acid residues) revealed

that a domain containing 85 amino acid residues at the COOH-terminal carriesthe glycosylation site and is responsible for macrophage activation. Thus, thisdomain was cloned via baculovirus vector4 and treated with immobilized

/3-galactosidaseand sialidase to yield the cloned MAF, CdMAF. As shown inFig. 2, the size of CdMAF is 18.5% of the length of the GcMAF peptide.Incubation of mouse peritoneal macrophages with 10 pg Cd.MAF/ml wasrequired for 7-fold-increased phagocytic capacities and 15-fold-increasedsuperoxide-generating capacities.

Treatment of Ehrlich Ascites Tumor-bearing Mice with EnzymaticallyGenerated MAF

Because Ehrlich ascites tumor had been maintained by serial passagethroughthe BALB/cmouseperitonealcavity,the ascitestumorcellsgrewassoon as they were transplanted. To assess the fate of the known cell counts ofthe transplanted tumor, the first administration of MAF should be given 6 hposttransplantation but before the increase in cell counts.

To avoid deglycosylation of MAF by highly concentrated NaGalase in theperitoneal cavity, DBPMAF, GcMAF, and CdMAF were administered i.m. forthe activation of macrophages. Approximately 3.5 h after the administration ofMAF, systemic macrophages (including monocytes) were activated and recruited to inflamed lesions.

The efficacy of DBPMAF, GcMAF, and CdMAF was assessed by timecourse analyses that determinedthe total ascites tumor cell counts in theperitoneal cavity and the serum NaGalase activity after administrations ofGcMAF, CdMAF, or DBPMAF (100 pg/mouse) four times (days 0, 4, 8, and12) at 4-day intervals. At each data point, six mice were anesthetized, and serawere collected from the vena cava inferior for NaGalase assays. For ascites

tumor cell counts in the peritoneal cavity, 5—7ml of PBS were injected into theperitoneal cavity and aspirated with a syringe, and the total number of ascitestumor cells in the peritoneal lavage was determined.

Assay for NaGalase in the Sera of Ehrlich Ascites Tumor-bearing Mice

Proteins in mouse sera (100 s.d)were precipitated by 70% saturated ammonium sulfate. This high-salt precipitation procedure also dissociates the product

inhibitor from the enzyme. The precipitates were dissolved in 50 mxi citratephosphate buffer (pH 6.0) and dialyzed against the same buffer at 4°Covernight. The dialysates were made up to 1 ml in volume and assayed forenzyme activity (10, 22, 24). Substrate solution (250 pi) contained 5 @tmolofp-nitrophenyl N-acetyl-a-D-galactosaminide in 50 mM citrate buffer (pH 6.0).4 Unpublished observations.

2188

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

6 mice Untreatedcontrol 6 micell6 ±2 days6mice DayO 6mice/35 ±4days6 mice Days 0 and 4 6 mice/>5O days6 mice Days 4 and 8 6 mice/32 ±4daysB.

Experiment2 (1 X lO@tumorcells/mouse; 100 pg ofGcMAF/mouse)6

mice Untreated control 6 mice/16 ±2 days6mice DayO 6mice/35 ±Sdays6 mice Days 0 and4 6 mice/>50daysC.

Experiment3 (5 x 1O@tumorcells/mouse; 100 pg ofDBPMAF/mouse)6

mice Untreatedcontrol 6 mice/14 ±2 days6mice DaysOand4 6mice/32 ±4days6 mice Days 0, 4, and 8 2 mice/35 and 38 days

4 mice/>53daysD.

Experiment4 (5 X lO@tumorcells/mouse; 100 pg ofGcMAF/mouse)6

mice Untreated control 6 mice/14 ±2 days6 mice Days 0 and 4 6 mice/32 ±5 days

ANTITUMOR AcrIvrI-Y OF MAP

+1 -->DomainI-16 14K R V L V L L L A @AF G H A L E R G R D Y E K N K V C K E-12 LVLLLALAFGHALERGRDYEKDKVCNE

H14

H I6FSHLGKEDFISLSLVLYSRKFPSGTFEQVSQLVKEVVSLTEACCA14 I6LAMLGKEDFRSLSLI LYSRKFSSSTFEQVNQLVKEVVSLTEECCA

H 61EGADPDCYDTRTSALSAKSCESNSPFPVHPGTAECCTKEGLERKLM 61 E G A D P 1 C Y D I R I S E I S V K S C E S D A P F P V H P G I P E C C I K E G L E R K I

H 106 C N A L I@ H Q P Q E F P 1 Y V E P 1 N D E I C E A F R@ D P K E Y A P4Q F 14W E V S I N14106 C N A L I S H Q P Q E F P 1 Y V E P 1 N D E I C E A F R R D P K G P A D Q F L V E Y S S N

Domain II -->H151YGQAPLSILVSYTKSYLSMVGSCCTSASPTVCFLKERLQLKH LSL14 151 Y G Q A P L P L I VA Y I K N Y I S 14 V G S C C I S A N P 1 V C F V K E R I Q 14 K H L SI

H 196 1 1 1 L S N R V C S Q Y A A Y G E K K S R I S N I I K L A Q K V P 1 A D I E D V L P L A E14 196 1 1 1 14 S N R V C S Q V A A V G K E K S R I S H I I K I A Q K V P 1 A N I E N V I P I A E

H 241 D ! I t(I I S@ C C E S A S E D C 14A K E I P E H I@ K @,,C@ N I S I K N S K F E D C C Q14 241 D F I E I I S R C C E S I S E D C M A S E L P E H I I K I C Q N L S K K N S K F E E C C Q

Fig. 2. Homology ofthe sequences ofhuman Geprotein (11)and mouse (M) DBP. Underlined residues of human Ge protein are not identical to theresidues of mouse DBP. Numbers to the left, aminoacid residue positions. Domain divisions aremarked on the top of the sequences. Bold letters,amino acid residues ofthe cloned CdMAF that spanresidues 374—458of the Ge protein.

H 286 E K I A N D V F V C T V F 14P A A Q I P E I P 0 V E I P 1 N K D V C D P G N T K V H D K VM 286 E N I P 14N I F M C I V F 14P A A E P L 0 1 P A I K I P 1 G K D L C G 0 S T I Q A N D 0 V

Cck4AF - ->H 331 @1F E I S R R I H I P E V F I S K V I E P 1 L K S I G E C C D V E D S I T C F N A K G P 1

14 331 1 F E I S R R I Q V P E V F I S K V I E P 1 I K T L R E C C 0 1 0 D S V A C F S I Q S P I--> Domain III

H 376 1 K K E I S S F I D K G Q E I C A D V S E N I F T E V K K K I A E R I K A K I P D A I P KM 376 1 K R 0 1 1 5 F I E K G Q E 14C A D V S E N T F I E V K K K I A E R I R I K I P N I S P A

H 421 E I A K I V N K R S D F A S N C C S I N S P P 1 V C D S E I D A E I K N I I14 421 E I K D 14 V E K H S D F A S K C C S I N S P P I V C S S Q I D A E 14 I D T I 0 5

The reaction was initiated by the addition of 250 pi of the dialysates, kept at37°Cfor 60 mm, and terminated by adding 200 j.dof 10%trichloroacetic acid.After centrifugation, 300 @&lof 0.5 M Na2CO3 solution was added to thesupernatant. The amount of p-nitrophenol released was determined spectrophotometrically at 420 nm with a Beckman DU600 spectrophotometer, andNaGalase activity was expressed in nanomoles/min/mg protein.

RESULTS

Antitumor Effect of the Enzymatically Generated MAFs(DBPMAF and GCMAF) on Mice Transplanted with EhrlichAscites Tumor. When 6 BALB/c mice were transplanted with 1 X 10@Ehrlich ascites tumor cells and treated 6 h later with 100 pg/mouse of the

Table 1 Therapy of mice bearing Ehrlich ascites tumor with the enzymaticallygenerated MAF (DBPMAF and GcMAF)

No. of mice Posttransplantation treatment No. of mice survived/period―

A. Experiment1 (1 X i0@tumorcells/mouse; 100 pg of DBPMAF/mouse)

enzymatically generated MAF (DBPMAF or GCMAF), they survived forapproximately 35 ±4 days, as shown in Table 1. The control mice thatreceived ascites tumor only died in 16 ±2 days. When these micereceived an additional administration of MAF at 4 days posttransplantation, all 6 mice survived at least 50 days.

Mice that received 2 DBPMAF administrations, at days 4 and 8after transplantation of 1 X l0@ ascites tumor cells, survived forapproximately 32 ±4 days. When the total tumor cell count in theperitoneal cavity was examined at day 4 after transplantation of1 X l0@ascites tumor cells, approximately S X l0@tumor cells weredetected in the peritoneal lavage. When mice received two administrations of DBPMAF or GcMAF at days 0 and 4 after transplantationof 5 X iO@tumor cells, they survived up to 32 ±4 days. The controlmice that received 5 X l0@ascites tumor cells only died in 14 ±2days.

Therefore, 2 administrations of DBPMAF or GcMAF to mice, atdays 0 and 4 after transplantation with 5 X l0@ascites tumor cells,allowed the mice to survive no more than 32 ±4 days. When thenumber of ascites tumor cells in the peritoneal cavity was counted onthe 8th day, approximately 1 X l0@ or 1 X l0@ tumor cells weredetected in the peritoneal lavage from the mice treated with DBPMAFor GcMAF, respectively. As shown in Table 2, at day 12, cell countsin all treated mice increased to approximately 5 X 10@and 5 X l0@cells, respectively. At day 24, the cell count in the peritoneal cavityincreased to more than 2 X l0@ cells in all treated mice. The totaltumor burden of more than 108 ascites cells in the peritoneal cavitykilled mice at 32 ±4 days (Table 1). Thus, more than three MAFadministrations were required for eradication of the ascites tumor. Toassess MAF efficacy and the curative rate of Ehrlich ascites tumorbearing mice, measurement of the fate of the transplanted ascitestumor cells in the peritoneal cavity was required as a prognostic index.

Correlation of Serum NaGalase Activity with Tumor Burden inMice Transplanted with Ehrlich Ascites Tumor CeHs. NaGalaseactivity can be detected in the sera of all cancer patients, but not inhealthy humans (10, 22). When mice were transplanted with S X I0@Ehrlich ascites tumor cells and assayed for serum NaGalase activity,

aSurvivalperiodvaluesrepresentmean±SEofsixmice.Theresultswerereproducedthree times with a BALB/e or Swiss mouse.

2189

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

0 4 8 12 16

Table 2 Thefate of Ehrlich ascites tumor cells in the peritoneal cavity of mice aftertwo administrations ofDBPMAF, GcMAF, or CdMAF at days 0 and 4―

No. of mice Days posttransplantation assay Survived tumor cell counts

A. ExperimentI (5 x l0@tumorcells/mouse: 100 pg of DBPMAF/mouse)

Untreated control6 mice Day 8 1 X l0@cells/mouse6 mice Day 12 1 X l0@cells/mouse

Treated6 mice Day 8 t X lO@cells/mouse6 mice Day 14 5 X l0@cells/mouse6 mice Day 24 2 X l0@cells/mouse6 mice Day 30 >l0@cells/mouseB.

Experiment2 (5 x l0@tumorcells/mouse; 100 pg ofGeMAF/mouse)Untreated

control6 mice Day 8 I X l0@cells/mouse6 mice Day 12 1 X 108cells/mouse

Treated6 mice Day 8 1 X lO@cells/mouse6 mice Day 14 5 X l0@cells/mouse6 mice Day 24 6 X iO@cells/mouse6 mice Day 30 > l0@cells/mouseC.

Experiment3 (5 x l0@tumorcells/mouse; 100 pg ofCdMAF/mouse)Untreated

control6 mice Day 8 1 X l0@cells/mouse6 mice Day 12 1 X l0@cells/mouse

Treated6 mice Day 8 3 x l0@cells/mouse6 mice Day 14 2 X l0@cells/mouse6 mice Day 24 3 X 1ff cells/mouse6 mice Day 30 > l0@cells/mouseLi

Results were reproduced three times with BALB/e or Swiss mouse.

ANTITUMOR ACFIVITY OF MAP

was more effective in eliminating tumor cells than the equivalentGcMAF (100 pg/mouse) therapy. As can be seen in Fig. 4, 4 DBPMAF administrations with 4-day intervals resulted in an extended

survival of over 90 days and an insignificantly low serum NaGalaselevel (1.81 ±0.15 nmol/min/mg) equivalent to that of control mice(that received no transplantation) at 30—90days after tumor transplantation, indicating complete eradication of the tumor. After day 30,no ascites tumor cells were detectable in the peritoneal lavage.

Therapeutic Efficacy of CdMAF for Mice Bearing Ascites Tumor Cells Was Assessed by Serum NaGalase Activity. As shownin Fig. 2, the cloned MAF, CdMAF, contained 18.5% of GcMAF (85amino acid residues at the COOH-terminal). Thus, the amount ofthe immunogenic epitopes nonhomologous to DBPMAF (mouseDBPMAF) in CdMAF would likely be reduced more than 5-fold ascompared with GcMAF. When BALB/c mice received 5 X 10@Ehrlich ascites tumor cells and were treated with 100 pg CdMAF/mouse on days 0 and 4, they survived for 32 ±4 days. The controlmice that received the ascites tumor only died within 14 ±2 days.When these mice were injected with 2 additional administrations ondays 8 and 12 (total of 4 administrations) of CdMAF, the micesurvived for at least 90 days. As in DBPMAF and GcMAF therapy,the most beneficial CdMAF (100 pg/mouse) therapy regimen consisted of 4 treatments given 4 days apart, starting 6 h after transplantation (i.e., on days 0, 4, 8, and 12; Fig. 4). When the total cell countsin the peritoneal cavity and serum NaGalase activity were examinedon days 30, 60, and 90, a low level of serum NaGalase activity(2.31 ±0.19 nmollmin/mg) and approximately 20 Ehrlich ascitestumor cells in the peritoneal lavage were persistently detected.

DISCUSSION

The results of the present study indicate that Ehrlich ascites tumorin BALB/c mice can be efficiently eradicated by macrophage-directedimmunotherapy using DBPMAF. MAFs (i.e., DBPMAF, GcMAF,and CdMAF) can activate macrophages, monocytes, and other phagocytes such as osteoclasts (19, 20) but cannot stimulate B and T cells(17). Moreover, MAFs stimulate the progenitor cells, resulting in adramatic increase in macrophage counts (nearly 30-fold) in the 4 daysafter MAF treatment (19, 20, 26). Thus, the most beneficial therapeutic regimen consisted of four treatments given 4 days apart, startingimmediately after transplantation (i.e., on days 0, 4, 8, and 12).Although 1 or 2 administrations of MAFs reduced the ascites tumorcell counts by more than 1000-fold in the mouse peritoneal cavity,tumor cells began to grow back several days after MAF administra

0Ca‘U

CEE?1UCao'al‘U‘U01@.@@

Z'U

Time(Days)Fig. 3. Time course study of serum NaGalase activity in a BALB/c mouse transplanted

with 5 X lO@Ehrlich ascites tumor cells. The results were reproduced three times, withslightly different time intervals.

they all showed serum NaGalase that deglycosylates mouse DBP (Fig.lc). This serum enzyme activity increased as the ascites tumor cellsgrew in the peritoneal cavity. As shown in Fig. 3, the serum NaGalaseenzyme activity increased linearly with time as the Ehrlich ascitestumor cells grew. Healthy control mice had 1.98 ±0.15 nmoL/min/mgof serum enzyme activity, which is a-galactosidase capable of hydrolysis of the same substrate as that for NaGalase (10). The enzymeactivities beyond that of control mice are considered to be NaGalaseactivity derived from cancerous cells. Thus, serum NaGalase activityis proportional to tumor burden (Fig. 3). In support of this observation,the serum NaGalase activity in nude mice transplanted with humanoral squamous cell carcinoma is proportional to the tumor size (byweight) (22). Thus, the proportionality of the serum NaGalase activityto tumor burden allowed us to use this serum enzyme for prognosticand diagnostic purposes.

Therapeutic Effect of GcMAF on Mice Bearing Ascites TumorCells Was Monitored by Serum NaGalase Activity. When BALB/cmice were transplanted with Ehrlich ascites tumor (5 X l0@ cells/mouse) and treated with 4 administrations of GcMAF (100 pg/mouse)given 4 days apart, starting 6 h after transplantation (i.e., on days 0, 4,8 and 12), they survived for at least 90 days. After 30 days, a very lowlevel of serum NaGalase (2.81 ±0.1 1 nmol/min/mg) was detected(Fig. 4). Although all mice were healthy up to 90 days, approximately90 ascites tumor cells were persistently detected in the peritoneallavage at days 30, 60, and 90.

Antitumor Effect of DBPMAF on Mice Bearing Ascites TumorWas Assessed by Tumor-specific Serum NaGalase Activity.Mouse DBP and human Ge protein are highly conserved and share a78% identical amino acid sequence (Ref. 25; Fig. 2). It is possible thatmice can develop antibodies against nonhomologous epitopes in human protein (GcMAF), which would reduce its effectiveness withregimens involving multiple administration of this agent during protracted time intervals. Indeed, DBPMAF (100 pg/mouse) treatment

0

2190

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF MAF

16

8

0

0

0

1 0 ControlI . GcMAF

I U CdMAF0 0 DBPMAF

0Ca‘U•Ea@CE

Ca

00

ZCU

Fig. 4. Time course analysis of the therapeuticefficacy of DBPMAF, GeMAF, and CdMAF asmonitored by serum NaGalase activity in a BALB/cmouse transplanted with 5 X lO@Ehrlich ascitestumor cells. The results were reproduced threetimes, with slightly different time intervals.

11k 6@l 90

tion. If the ascites tumor growth is repeatedly suppressed by 4 administrations of MAF over 12 days, the ascites tumor did not growback for at least 90 days. In a separate study, we observed that theGcMAF-treated Ehrlich tumor-bearing mice survived for at least 150days. However, very small numbers (< 100) of ascites tumor cellsbecame dormant cells (27—29)in the peritoneal cavity of the GcMAFor CdMAF-treated mice. The effectiveness of multiple administrationof MAFs during protracted time intervals suggests that immunitydeveloped against the tumor. The protracted MAF treatment undersuppression of tumor growth for more than 12 days seemed to berequired for immune development against the cancerous cells. Insupport of this concept, the transplanted Ehrlich ascites tumor cells

cannot grow in mice previously immunized with killed Ehrlich ascitestumor cells.4 In fact, the tumoricidal capacity of macrophages isobserved preferentially via the IgG (Fc-receptor)-mediated pathway(9, 30, 31). Moreover, the MAFs are potent adjuvants in immunizationfor antibody production (21).

All cancer patients carry serum NaGalase that deglycosylates Gcprotein (Fig. lc), resulting in inactivation of the precursor activity ofGe protein (10). GcMAF can bypass the inactivated Ge protein andact directly on macrophages for efficient activation. AdministeredGcMAF in cancer patients circulates rapidly and interacts with systemic macrophages within approximately 30 mm.4 Although GcMAFadministered in cancer patients could be theoretically deglycosylatedby NaGalase, the activity of serum NaGalase complexed with aproduct inhibitor in the blood stream deglycosylates only a very smallfraction of GcMAF during the 30-mm circulation period. TheMAF-primed macrophages require 3 h for the development of fullphagocytic/cytocidal capacity. The half-life of activated humanmacrophages is 5—6days (8). Preliminary studies revealed that

weekly administration of GcMAF (100 ng) to cancer patients showspromising results with a variety of human cancers (32, 33). Becausemammalian DBPs are highly conserved, GcMAF can be used formacrophage activation and cancer therapy in various mammals.

Like human Ge protein in cancer patients (10), the precursoractivity of mouse DBP in cancer-bearing mice was lost or reduced,because mouse DBP was deglycosylated by tumor-derived serumNaGalase (see Fig. lc). GcMAF can bypass inactivated DBP and actdirectly on macrophages for efficient activation. Macrophages ofcancer-bearing mice are capable of being activated, as demonstratedby the activity of macrophages 3.5 h after GcMAF treatment. In thecurrent paper, we showed that a minute amount (100 pg/mouse) ofGcMAF has a potent curative activity for mice bearing Ehrlich ascitestumor. When DBPMAF (100 pg/mouse) was administered to Ehrlichascites tumor-bearing mice, DBPMAF was found to have more curative activity than GcMAF. Four administrations of DBPMAF toEhrlich ascites tumor-bearing mice rendered excellent curative responses.

DBPMAF and GcMAF can be generated only from the glycosylated DBP and Ge protein. The amino acid sequence of mouse DBPis 78% identical to that of the human Ge protein (25). The DBP of allmammalian species carries only one oligosaccharide near the COOHterminal (18). Gel (one of the major Ge isoforms) is 100% glycosylated, whereas only 10% of mouse DBP molecules are glycosylated(15, 18). Because nonglycosylated DBP cannot be converted to

DBPMAF, f3-galactosidase-treated mouse DBP as DBPMAF preparation contained the active DBPMAF form in only 10% of the totalprotein. Although 100 pg/mouse of both GeMAF and DBPMAF wereused for each injection, the latter (DBPMAF) has 10 pg of activeMAF/mouse. In spite of the 10-fold difference in MAF activity, the

2191

Time (Days)

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

ANTITUMOR ACI1VITY OF MAP

DBPMAF treatment in the present study was more effective in dimmating Ehrlich aseites tumor cells than GeMAF therapy with theequivalent dosage. Such a finding may be explained by the possibilitythat mice can develop antibodies against human protein GcMAF,which would reduce its effectiveness with regimens involving multiplc administration of this agent during protracted time intervals.Commercial availability of interspecies antibodies against Ge proteinsupports the hypothesis that interspecies antibodies are raised againstnonhomologous amino acid sequences of DBP. Because GcMAF is apotent adjuvant for the antibody-producing capacity of mice (21),

administration of this heterologous protein to mice may produce atleast a low level of antibodies against GcMAF, in spite of highlyconserved DBPs. Because DBPMAF is of mouse origin, repeatedadministration of various doses of DBPMAF to Ehrlich aseites tumorbearing mice showed an excellent curative effect to the tumor, withoutadverse immunological effects.

The peptide of CdMAF contains 85 amino acid residues thatspan Ge protein amino acid residues 374—458 (COOH-terminal),as shown in Fig. 2. Because CdMAF is only 18.5% segment of theGcMAF peptide at the COOH-terminal, the possibility of developing antibodies against nonhomologous epitopes within the domain in mouse would be reduced more than 5-fold as comparedwith that of Ge protein. In fact, the data presented in this papersupport this hypothesis (see Fig. 4). Like GcMAF, CdMAF produces no side effects in humans. Of clinical significance is the factthat CdMAF is a nonhuman blood product.

In the present communication, our accumulated evidence enabledus to design the appropriate frequency (four times) of administrationof MAF (100 pg/mouse) to demonstrate differential efficacy ofDBPMAF, GeMAF, and CdMAF (Fig. 4). Because serumNaGalase activity is proportional to tumor burden in the hosts (22,34), serum NaGalase activity along with the total tumor cell countsin the peritoneal cavity seemed to be the most precise index forthe prognosis and curative state of the disease.

ACKNOWLEDGMENTS

We thank Dr. Sidney Weinhouse for his critical advice, guidance, andinterest in our work.

REFERENCES

1. Ngwenya, B. Z., and Yamamoto, N. Activation of peritoneal maerophages by lysophosphatidylcholine. Biochim. Biophys. Acts, 839: 9—15,1985.

2. Ngwenya, B. Z., and Yasnamoto, N. Effects of inflammation products on immunesystems: lysophosphatidyleholine stimulates macrophages. Cancer Immunol. Immunother.,21: 174—182,1986.

3. Ngwenya,B. Z., andYamamoto,N. Contributionof lysophosphatidyleholine-treatednonadherent cells to mechanism of macrophage activation. Proc. Soc. Exp. Biol.Med., 193: 118—124,1990.

4. Morton, D., Eibler, F. R., Malmgren, R. A., and Wood, W. C. Immunological factorswhich influence response to immunotherapy in malignant melanoma. Surgery (St.Louis), 68: 158—164,1970.

5. Thar, B., and Tanaka, T. Immunotherapy of cancer: regression of tumors afterintralesional injection of living Mycobacterium bovis. Science (Washington DC),172:271—273,1971.

6. Yamamoto, N., and Ngwenya, B. Z. Activation ofmacrophages by lysophospholipidsand ether derivatives of neutral lipids and phospholipids. Cancer Res., 47: 2008—2013,1987.

7. Yamamoto, N., Ngwenya, B. Z., Sery, T. W., and Pieringer, P. A. Activation ofmacrophages by ether analogues of lysophospholipids. Cancer Immunol. Immunother.,25:185—192,1987.

8. Yamamoto, N., St. Claire, D. A., Homma, S., and Ngwenya, B. Z. Activation ofmouse macrophages by alkylglycerols, inflammation products of cancerous tissues.Cancer Res., 48: 6044—6049,1988.

9. Yamamoto, N., Hoober, J. K., Yamamoto, S., and Yamamoto, N. Tumoricidal

capacities of macrophages photodynamically activated with hematoporphyrin denyative. Photochem. Photobiol., 56: 245—250,1992.

10. Yamamoto, N., Naraparaju, V. R., and Asbell, S. 0. Deglycosylation of senunvitamin D-binding protein and immunosuppression in cancer patients. Cancer Res.,56: 2827—2831,1996.

I 1. Homma, S., and Yamamoto, N. Activation process of macrophages after in vitrotreatment of mouse lymphocytes with dodecylglycerol. CIin. Exp. Immunol., 79:307—313,1990.

12. Yamamoto, N., Homma, S., and Miliman, I. Identification of the serum factorrequired for in vitro activation of macrophages: role of yitamin D-binding protein(group-specific component, Ge) in lysophospholipid activation of mouse penitonealmacrophages. J. Immunol., 147: 273—280,1991.

13. Yamamoto, N., Homma, S., Haddad, J. G., and Kowalski, M. N. Vitamin D3-bindingprotein required for in vitro activation of macnophages after dodecyiglycerol treatment of mouse penitoneal cells. Immunology, 74: 420—424,1991.

14. Homma, S., Yamamoto, M., and Yamamoto, N. Vitamin D-binding protein (groupspecific component, Ge) is the sole serum protein required for macrophage activationafter treatment of peritoneal cells with lysophosphatidylcholine. Immunol. Cell Biol..71: 249—257,1993.

15. Yamamoto, N., and Homma, S. Vitamin D3-binding protein (group-specific component) is a precursor for the macrophage activating signal from lysophosphatidylcholine-treated lymphocytes. Proc. Nail. Acad. Sci. USA, 88: 8539—8543,1991.

16. Yamamoto, N., and Kumashiro, R. Conversion of vitamin D3-binding protein (groupspecific component) to a macrophage-activatingfactor by the stepwise action of a-galactosidase of B cells and sialidase of T cells. J. ImmunoL, 151: 2794-2902, 1993.

17. Naraparaju, V. R., and Yamamoto, N. Roles of@-galactosidase ofB lymphocytes andsialidase of T lymphocytes in inflammation-primed activation of macrophages. hnmunol. Lets., 43: 143—148,1994.

18. Yamamoto, N., and Naraparaju, V. R. Role of mouse vitamin D3-binding protein inactivation of macrophages. J. Immunol., 157: 1744—1751,1996.

19. Yamamoto, N., and Naraparaju, V. R. A defect in @-galactosidaseof B lymphocytesin the osteopetrotic (op/op) mouse. Immunol. Lett., 50: 35—40,1996.

20. Yamamoto, N., and Naraparaju, V. R. A defect in inducible @-galactosidase of Blymphocytes in the osteopetrotic (mi/mi) mouse. Immunology. 88: 604—610, 1996.

21. Yamamoto, N. Structural definition of a potent macrophage activating factor derivedfrom vitamin D3-bindingprotein with adjuvant activity for antibody production. Mol.Immunol.,33: 1157—1164,1996.

22. Yamamoto, N., Naraparaju, V. R., and Urade, M. Prognostic utility of serum a-N-acetylgalactosaminidase and immunosuppression resulted from deglycosylation ofserum Ge protein in oral cancer patients. Cancer Res., 57: 295—299,1997.

23. Link, R. P., Perlman, K. L, Pierce, E. A., Schnoes, H. K., and DeLuca, H. F.Purification of human serum vitamin D-binding protein by 25-hydroxyvitamin D3-Sepharose chromatography. Anal. Biochem., 157: 262—269,1986.

24. Yamamoto, N., Naraparaju, V. R., and Snnivasula, S. M. Structural modification ofserum vitamin D3-binding protein and immunosuppression in fflV-infected patients.AIDS Res. Hum. Retroviruses, 11: 1373—1378,1995.

25. Yang, F., Bergeron, J. M.. Linehan, L. A., Lalley, P. A., Sakaguchi, A. Y., andBowman. B. H. Mapping and conservation of the group-specific component gene inmouse. Genomics, 7: 509—516,1990.

26. Yamamoto, N., Lindsay, D. D., Naraparaju, V. R., Ireland, R. A., and Popoff, S. N.A defect in the inflammation-primed macrophage activation cascade in osteopetrotic(op) rats. J. Immunol., 152: 5100—5107, 1994.

27. Weinhold, K. J., Goldstein, L. T., and Wheelock, E. F. The tumor-dormant state:quantitation of L5178Y cells and host immune responses during the establishment andcourse of dormancy in syngeneic DBA/2 mice. J. Exp. Med., 149: 732—744,1979.

28. Weinhold, K. J., Goldstein, L. T., and Wheelock, E. F. The tumor-dormant stateestablished with L5178Y lymphoma cells in immunized syngeneic munine hosts.Nature (Lond.), 270: 59—61,1977.

29. Robinson, M. K., and Wheelock, E. F. Identification of macrophage-mediated cytolytic activity as a tumor-suppressive mechanism during maintenance of the L5178Ytumor-dormant state in DBA/2 mice. J. Immunol., 126: 673—679,1981.

30. Weiner, L. M., Moldofsky, P. J., Gatenby, R. A., O'Dwyer, J., O'Brien, J., Litwin, S.,and Comis, R. L. Antibody delivery and effector cell activation in Phase II trial ofrecombinant y-interferon and munine monoclonal antibody CO17—lA in advancedcolorectal carcinoma. Cancer Res., 48: 2568—2573,1988.

31. Houghton, A. L., Mintzer, D., Cordon-Cardo, C., Welt, S., Fliegel, S., Vadhan, S.,Carswell, E., Melamed, M., Oettgen, H. F., and Old, L. J. Mouse monoclonal 1g03antibody detecting GD3 ganglioside: a Phase I trial in patient with malignant melanoma. Proc. Nail. Acad. Sci. USA, 82: 1242—1246,1985.

32. Naraparaju, V. R., Wimmers, R. S., Neil, R. N., Orchard, P. J., and Yamamoto, N.Originof immunosuppressioninjuvenileleukemiaandtherapeuticefficacyof vitamm D3-binding protein-derived macrophage-activating factor. Proc. Am. Assoc.Cancer Res., 37: 213, 1996.

33. Yamamoto, N., Naraparaju, V. R., Neil, R. N., Suyama, H., and Nakazato, H.Therapeutic efficacy of vitamin D3-binding protein-derived macnophage-activatingfactor for prostate, breast, and colon cancers. Proc. Am. Assoc. Cancer Res., 38: 31,1997.

34. Koga, Y.. Naraparaju, V. R., and Yamamoto, N. Antitumor effects of vitaminD@-binding protein-derived macrophage-activating factor on Ehrlich tumor-bearingmice. Proc. Am. Assoc. Cancer Res., 37: 481, 1996.

2192

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

1997;57:2187-2192. Cancer Res   Nobuto Yamamoto and Venkateswara R. Naraparaju  Factorwith Vitamin D-binding Protein-derived Macrophage Activating Immunotherapy of BALB/c Mice Bearing Ehrlich Ascites Tumor

  Updated version

  http://cancerres.aacrjournals.org/content/57/11/2187

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on December 12, 2015. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from