Combination of vasostatin and cyclophosphamide in the therapy of

Growth of tumors is strongly dependent upon supply of nutrients and oxygen by de novo formed blood vessels. Inhibiting angiogenesis suppresses growth of primary tumors as well and affects development of metastases. We demonstrate that recombinant MBP/vasostatin fusion protein inhibits proliferation of endothelial cells in vitro. The therapeutic usefulness of such intratumorally delivered recombinant protein was then assessed by investigating its ability to inhibit growth of experimental murine melanomas. In the model of B16-F10 melanoma the MBP/vasostatin construct significantly delayed tumor growth and prolonged survival of treated mice. A combination therapy involving MBP/vasostatin construct and cyclophosphamide was even more effective and led to further inhibition of the tumor growth and extended survival. We show that such combination might be useful in the clinical setting, especially to treat tumors which have already formed microvessel networks.


INTRODUCTION
The development of both primary and metastatic tumors is dependent on angiogenesis, i.e. formation of new blood capillaries from preexisting vessels (Folkman, 1971).It is widely assumed that, without accompanying angiogenesis, nests of cancer cells cannot exceed 2-3 mm 3 in size (Folkman, 1971).The progress of angiogenesis appears to be controlled by an equilibrium of proangiogenic (stimulating) and antiangiogenic (inhibiting) factors.Over a dozen angiogenesis-promoting molecules have been identified, mainly growth factors (e.g., VEGF or bFGF), whereas the inhibitor group includes a number of endogenous proteins (e.g., angiostatin or endostatin).Angiogenesis is either induced or repressed, depending on the equilibrium switch that reflects shifts between both types of factors involved (Folkman, 2003).Ongoing angiogenesis thus requires either increased levels of stimulating factors and/or decreased levels of their natural inhibitors (Hanahan & Folkman, 1996).
Both angiogenesis-stimulating factors and inhibitors targeting angiogenesis have been used in attempts to gain control over tumor growth.Repressing the formation of tumor blood vessels leads to inhibited growth of primary tumors as well as of metastases (Folkman, 2003).
Our previous investigation concerned the benefits of gene therapy based on the use of a vasostatin gene construct (Jazowiecka-Rakus et al., 2006).However, in terms of therapeutic agent dose flex-ibility, the protein approach is certainly superior to gene therapy.The study presented herein sought to demonstrate the therapeutic advantage of a composite therapy involving administration of recombinant vasostatin protein and a well-known chemotherapeutic, cyclophosphamide.

Vasostatin gene cloning.
In order to clone the vasostatin gene, the pCARL plasmid-containing calreticulin sequence (obtained from Dr. M. Michalak, University of Alberta, Edmonton, AB, Canada), was used as a template.The following primers were used: (+) 5' aaaaaggatccgagcccgccgtctacttc 3' and (-) 5' aaaaaaagcttcattccaaggagccggactc 3'.The underlined sequences denote the BamHI restriction site in the (+) primer and the HindIII restriction site in the (-) starter.The sequence in bold denotes the termination codon (TGA).
The amplified product was cloned into pMal-c2x, a bacterial expression vector (New England, Biolabs) using the malE gene reading frame.This gene encodes a maltose binding protein (MBP) which has a strong affinity for maltose.The pMal-c2x plasmid contains a deletion of the malE signal sequence, allowing this protein to remain in the cytoplasm.
Isolation and purification of recombinant proteins.pMal-c2x/vasostatin and pMal-c2x (control) plasmids were introduced into Escherichia coli TB1 strain.The bacteria were cultured at 37°C, with intensive shaking, in LB supplemented with 0.2% dextrose and 100 µg/ml ampicillin, until OD 600 about 0.5.After about 2½ h IPTG was added (final concentration 0.3 mM).Following a 3-h induction, the cultures were centrifuged for 20 min/6 000 rpm.The supernatant was discarded and the pellet placed on ice and resuspended in CB buffer (200 mM NaCl, 1 mM EDTA, 20 mM Tris/HCl, pH 7.4)) at 10 ml of buffer per 1 g of pellet.Bacterial lysates were sonicated using twelve 10-s impulses of medium amplitude (Branson sonifier).Then, the lysates were centrifuged for 30 min at 4°C/9 000 rpm and the supernatant (containing soluble protein fraction) was collected.
The MBP/vasostatin fusion protein was affinity-chromatography purified using an XK 16/20 column (Amersham, Biosciences) with a 15-ml amylose resin bed (New England, Biolabs).The column was first prewashed with eight volumes of CB buffer and then the clarified supernatant fraction of solu-Vasostatin and CTX in murine melanoma tumor suppression ble proteins (2.5 µg/µl) was loaded on top of the column.Fractions containing unbound proteins were eluted with 12 vol. of CB buffer whereas the purified MBP/vasostatin was eluted using CB buffer supplemented with 10 mM maltose.Protein concentration in the eluate was determined using Bradford assay.The eluate was then dialyzed (18-20 h at 4°C, 15 000 MWCO Spectrum bags) against 1.5 l (triple exchange) of buffer (25 mM NaCl, 20 mM Tris/HCl, pH 8.0).The dialysate was further purified on a Source Q-type ion-exchange chromatography column (Amersham Biosciences) using 20 mM Tris/HCl (pH 8.0) buffer.The protein was eluted from the column using 18-ml aliquots of 20 mM Tris/HCl (pH 8.0) with an increasing NaCl gradient (25-400 mM).Based on protein quantitation and analyses of polyacrylamide gel separations, the collected fractions were concentrated down to 1 ml using Ultrafree-4 devices (10 000 MWCO for MBP or 30 000 MWCO for MBP/vasostatin; Millipore).
The MBP/vasostatin fusion protein was further purified from endotoxins on agarose bed columns containing immobilized polymyxin B (Detoxi-Gel, Pierce Biotechnology).Before loading protein preparations, the column was first prewashed with 5 vol.
(5 ml) of 1% sodium deoxycholate, 5 vol. of endotoxin-free water and 5 vol. of buffer (0.1 M NaCl, 20 mM Tris/HCl, pH 7.5).Purified proteins were eluted using the same buffer.Eluates containing the highest concentrations of the MBP/vasostatin fusion protein or MBP were collected.In order to maximize endotoxin elimination, the clean-up procedure was repeated twice.Endotoxin content in the final protein preparations was determined using a Limulus Amebocyte Lysate QCL-1000 kit (Cambrex).
Immunohistochemical identification of vasostatin (Western blot).To identify MBP/vasostatin we used a polyclonal rabbit antibody recognizing MBP (New England, Biolabs) or a polyclonal goat antibody recognizing calreticulin (Santa Cruz) as well as anti-rabbit or anti-goat immunoglobulins (Vector).
Protein preparations were first separated electrophoretically on polyacrylamide gels and electrotransferred (4°C, 90 min, 300 mA) onto nitrocellulose filters (Schleicher & Schuell).The filters were then rinsed (1 h at room temp.) with Tris-buffered saline (TBS) containing 3% milk, placed in primary antibody solution (rabbit IgG recognizing MBP or goat IgG recognizing calreticulin, cross-reactive with vasostatin) in TBS with 3% milk and incubated overnight at 4°C.The filters were subsequently washed four times (5 min each) in TBST (TBS supplemented with 0.1% Tween-20), once in TBS and then placed in the solution of a second antibody (recognizing either rabbit IgG or goat IgG, and biotin-conjugated) in TBS with 1% milk and incubated for 1.5 h at room temp.The filters were again washed for 5 min in TBST (four times), once in TBS and were further incubated for 1 h at room temp.with diluted (1:1 000) streptavidin-biotinylated horseradish peroxidase complex (Amersham Biosciences).Following a TBS wash, the filters were incubated for about 1 min in a peroxidase substrate solution containing 0.2% 3,3'diaminebenzidine (DAB), 0.5 M Tris/HCl (pH 7.4) and 0.3% H 2 O 2 .After bands had appeared the filters were washed in deionized water.
Influence of MBP/vasostatin on endothelial cells growth (MTT test).Bovine aortic endothelial cells (BAEC) were seeded at 4 × 10 3 cells/well in a 96-well plate in 100 µl RPMI supplemented with 10% fetal bovine serum (FBS).After 2-h incubation at 37°C, the medium was replaced with 75 µl RPMI containing 10% FBS and MBP (control) or MBP/vasostatin.Three protein concentrations were used (0.1, 1 and 10 µg/ml), and experiments were done in triplicate.After 1-h incubation, 75 µl bFGF (20 ng/ ml) diluted in RPMI with 10% FBS was added to the wells and incubation was continued for 24 h at 37°C.At the end of the incubation, medium was removed from the wells and 100 µl of MTT (0.5 mg MTT/1 ml PBS‾) solution was added.Plates were further incubated for 3 h at 37°C.In order to dissolve formazan formed, an equal volume of 0.04 N HCl in isopropanol was added to the MTT solution.The absorbance of the samples was measured at 570 nm using an ELISA EL x 800 reader (BioTek Instruments).The percentage of live cells was estimated as: Animals.C57BL/6 mice used throughout the study were from an on-site Animal Facility.The experiments were approved by the Ethics Committee at the Medical University in Warsaw.
In vivo test of angiogenesis inhibition.Inhibition of angiogenesis by the recombinant protein was carried out according to Chen et al. (1999) using 4-week-old C57BL6 female mice.Briefly, animals were anesthetized with 200 µl 2.5% avertin (about 15 μl/g body mass) and 300 µl of Matrigel (Becton Dickinson) infused with fibroblast growth factor (bFGF) and the recombinant MBP/vasostatin protein (50 µg/ml) was implanted intradermally 1 cm below the sternum.After 7 days the gel was removed, placed in Drabkin's solution and incubated for 6 h at room temp.with intermittent mixing.The samples were then centrifuged for 5 min/2 500 rpm Supernatant was collected and absorbance measured at 540 nm.The hemoglobin concentration (g/dl) was calculated as: where 13.8 denotes the concentration of the hemoglobin standard used.
Cyclophosphamide (170 mg/kg body mass) was injected intraperitoneally starting on the 6th or 7th day of experiment (4 administrations every 6 days were given).For details see Mitrus et al. (2006).
Tumors were measured every 2-3 days using calipers and their volume calculated as in O' Reilly et al. (1997): Statistics. Results were compared using a non-parametric Mann-Whitney U-test.Animal survival (Kaplan-Meier plot) was analyzed using logrank test.Intergroup differences were considered as statistically significant at P < 0.05.Calculations were done using Statistica 5.0 software.

Cloning of vasostatin gene
A 540-bp DNA fragment encoding human vasostatin was obtained using PCR.The amplified product was cloned into a bacterial expression vec-tor pMal-c2x (New England, Biolabs) within the reading frame of the malE gene.The fusion gene (combining malE and vasostatin coding sequence in the pMal-c2x/vasostatin plasmid) was sequenced.The obtained malE sequence was compared to data supplied by New England Biolabs, whereas the vasostatin sequence was compared to that appearing in the GenBank database (GI: 5921996).

Isolation and purification of proteins
The MBP/vasostatin fusion protein was isolated from cultures of bacteria harboring the pMal-c2x/vasostatin plasmid construct (Fig. 1A).Yield of the recombinant fusion protein reached 30-50 mg of protein/liter of bacterial culture.MBP protein (control) was obtained (from pMal-c2x plasmid) and purified as described (Fig. 1B).The MBP protein (about 51 kDa) carries additionally the α-fragment of β-galactosidase whereas MBP forming part of the MBP/vasostatin fusion protein does not include the α-fragment of β-galactosidase and therefrore its mass is lower (42 kDa).Insertion of the vasostatin gene into the pMal-c2x vector inhibits expression of the LacZ gene encoding the α-fragment of β-galactosidase.Translation is terminated at the stop codon inserted into the vasostatin sequence using the (-) starter (see Materials and Methods).The whole MBP/vasostatin fusion protein has a molecular mass of 66 kDa.
The proteins were identified using Western blotting.Polyclonal antibodies recognizing either MBP (Fig. 2A) or (to identify vasostatin) calreticulin (Fig. 2B) were used.The mass (66 kDa) of the identified protein corresponded to that of the MBP/vasostatin fusion protein.
The endotoxin level of the purified proteins used subsequently for in vitro and in vivo experiments was less than < 0.1 EU/μg protein.

Determination of the antiangiogenic properties of the MBP/vasostatin fusion protein
In order to determine the effect of the fusion protein on the proliferation of endothelial cells in vitro, the MTT test was performed.The MBP/vasostatin fusion protein inhibited proliferation of endothelial cells (BAEC) stimulated by fibroblast growth factor (46% at 10 µg/ml, 37% at 1 µg/ml and 33% at 0.1 µg/ ml).MBP used as a control produced no inhibition of BAEC proliferation (Fig. 3).
An angiogenesis-inhibition test using Matrigel was performed to determine the hemoglobin content in intradermally implanted gel plugs supplemented with bFGF and the recombinant proteins.The test represents an indirect functional measure of the density of vessels formed within the Matrigel plug.A decreased hemoglobin concentration reflects inhibited angiogenesis in vivo.The test showed that the examined fusion protein inhibited (by about 48%) the formation of new vessels within Matrigel.MBP did not affect the angiogenesis (Fig. 4).

MBP/vasostatin-mediated in vivo therapy of mice bearing B16(F10) tumors
In subsequent therapeutic experiments a dose of MBP/vasostatin fusion protein was established that retarded the growth of experimental tumors by at least 80%.The vasostatin construct was administered to mice daily, starting on the 6th day from in-oculation, and continuing for 11 days.Under these circumstances the minimum dose that resulted in so defined tumor growth inhibition was established as 60 μg of MBP/vasostatin per animal and such amounts were used in further studies (Fig. 5).
In a subsequent experiment animals were administered with 60 μg of the fusion protein/100 μl FB or 40 μg of MBP/100 μl FB or 100 μl FB (the two latter being controls).The 40 μg MBP/100 μl FB per mouse dose was calculated from the ratio of molecular masses of MBP and the fusion protein (MBP constitutes about 64% of MBP/vasostatin).A similar line of reasoning was previously followed by Pike et al. (1998).The therapy was started on the 6th day of the experiment.Both proteins were administered Proteins separated on 10% (A) or 12% (B) polyacrylamide gel were transferred onto nitrocellulose filter and placed in a solution of MBP-recognizing rabbit antibodies (A) or of goat antibodies recognizing vasostatin (B).The filters were incubated with anti-rabbit (A) or anti-goat (B) immunoglobulin conjugated with biotin and a complex of streptavidin and biotinylated horseradish peroxidase.In order to visualize immunocomplexes the filters were placed in a solution of DAB, a peroxidase substrate.Total protein before induction (lane 1).Fraction 3 h after IPTG induction (0.3 mM) (lane 2).Insoluble fraction (lane 3).Soluble fraction (lane 4).

Figure 3. In vitro inhibition of proliferation of endothelial cells.
Endothelial cells (BAEC, 4 × 10 3 /well) were incubated with bFGF (25 ng/ml) and various concentrations of proteins: (0.1-10 µg/ml) at 37 o C for 24 h.Following addition of MTT the percentage of live cells was calculated (see Materials and Methods).The result obtained for 10 µg/ml MBP/ vasostatin group (P = 0.049535) is statistically significant as compared to bFGF and 10 µg/ml MBP control groups (Mann-Whitney U-test).
J. Jazowiecka-Rakus and others intratumorally for the following 14 days.The results show that MBP/vasostatin markedly inhibits tumor growth and prolongs survival of treated animals.On the other hand, administration of either MBP or FB solution did not show any therapeutic effect (Fig. 6).
Finally, the MBP/vasostatin therapy was combined with CTX administration (Fig. 7).The chemotherapeutic (170 mg/kg body mass) was injected intraperitoneally, every 6 days starting on the 7th day from inoculation, as described by Browder et al. (2000).The fusion protein (60 μg in 100 μl FB/ mouse) was administered intratumorally, starting from the 5th day from inoculation, 5 times a week for 21 days.On the day of CTX administration the mice did not receive injections of the recombinant protein.Inhibition of tumor growth was observed in mice treated with cyclophosphamide, with the fusion protein and with both agents combined.However, the survival of the treated animals was prolonged the most in the combined therapy, as compared to CTX alone (P = 0.01939).

DISCUSSION
Several observations indicate a dependence of invasive tumor growth upon formation of its own network of blood vessels (Folkman, 1971).By inhibiting the process of angiogenesis, also growth suppression of primary tumors as well as metastases is possible (Folkman, 2000).This finding has led to new possibilities in cancer therapy.
This study was aimed at investigating the therapeutic usefulness of a recombinant protein inhibitor of angiogenesis in preventing growth of B16(F10) murine melanoma experimental tumors.In our previous study (Jazowiecka-Rakus et al., 2006), we tested a related therapeutic approach, based on the use of vasostatin gene.The protein approach tested herein has an important advantage of permitting modifications of the therapeutic agent's doses.Since, despite numerous attempts, we were unable to isolate vasostatin devoid of the MBP tag (not shown), we decided to verify the therapeutic usefulness of the MBP/vasostatin recombinant protein, with MBP alone used as a control.
The recombinant fusion protein obtained here indeed shows antiangiogenic properties.We found that proliferation of endothelial cells (BAEC) was inhibited about 50% by the fusion protein at 10 μg/ml (Fig. 3).The in vivo test of angiogenesis inhibition using Matrigel established that the studied protein did inhibit formation of new blood vessels stimulated by fibroblast growth factor bFGF present in the Matrigel plug (Fig. 4).
We also determined the therapeutic dose (60 μg of MBP/vasostatin per mouse) of the antiangiogenic protein causing suppression of B16(F10) murine melanoma tumor growth by at least 80% (Fig. 5).Repeated administrations (14 times) led to a marked improvement of animal survival (Fig. 6).
The use of the MBP/vasostatin construct did not lead, however, to complete curing of the animals (Fig. 6).Cessation of drug administration led to regrowth of the tumors.Chronic administration of the antiangiogenic agent could probably arrest the tu-  Mice were inoculated intradermally with 2 × 10 5 B16(F10) cells/animal in 100 µl PBSˉ and divided into groups of six.Starting from the 6th day from inoculation and continuing daily for 11 days, various doses of MBP/vasostatin protein: (1-100 μg, see graph) in 100 μl FB/animal were administered intratumorally.Each datapoint represents average tumor volume ± S.D.For example, on day 16 statistically significant differences were observed between mice from groups receiving 60 μg (P = 0.046) and 100 μg (P=0.01171) of the protein, as compared to the control group (Mann-Whitney U-test).
mor growth for the duration of the treatment (Klement et al., 2000).
Further experiments were performed to see if a combined therapy involving administration of MBP/vasostatin and CTX would lead to improved therapeutic results.According to Browder et al. (2000), CTX administered at an appropriate dose (170 mg/kg body mass, every 6 days) induces, among other effects, apoptosis of endothelial cells lining tumor blood vessels.It turns out that not only low doses of CTX induce an increased expression of thrombospondin-1, an inhibitor of angiogenesis (Hamano et al., 2004), but that CTX has immunostimulatory properties (Loeffler et al., 2005).
In the combined therapy approach, administration of MBP/vasostatin (60 μg/mouse) was started five days after inoculation of the animals with cancer cells.The protein was applied five times a week for a total of 21 days.Cyclophosphamide (170 mg/ kg body mass), on the other hand, was administered intraperitoneally, every six days starting on the 7th day after inoculation.Inhibition of tumor growth and extended survival were indeed clearly observable in mice treated with the combination therapy (Fig. 7).
Application of the recombinant therapeutic protein was started on the 6th day following inoculation, when tumors were 20-30 mm 3 in volume Figure 6.Therapy of mice bearing B16(F10) tumors.Animals were inoculated intradermally with 2 × 10 5 B16(F10) cells/mouse in 100 µl PBS‾ and divided into groups of six.The following controls were used: mice receiving MBP only (40 μg in 100 μl FB/mouse), mice inoculated with B16(F10) cells only and mice that received FB only (100 μl/mouse).The treated group included mice that were receiving MBP/vasostatin in 100 µl FB/mouse.Starting on the 6th day after inoculation, mice from the treated group received recombinant protein daily for subsequent 14 days.The results of survival analysis (logrank test) for the animals treated with the recombinant MBP/vasostatin protein are statistically significant, as compared to control groups: B16(F10) (P = 0.03734), FB (P = 0.01599) and MBP (P = 0.03552).Animals were inoculated intradermally with 2 × 10 5 B16(F10) cells/mouse in 100 µl PBS‾ and divided into groups of five.The control group included animals that were only inoculated with neoplastic cells.The following study groups were established: mice receiving MBP/vasostatin (60 μg in 100 μl FB/mouse); mice receiving cyclophosphamide alone, and mice undergoing combined therapy.Starting on the 5th day after inoculation, mice received intratumorally the recombinant protein for 21 days (on day 5, 6, 8-12, 14-18, 20-24, 26-29), with the exception of days in which cyclophosphamide was administered.CTX (170 mg/ kg body mass) was administered on the 7th day after inoculation, and then every six days, i.e. on days 7, 13, 19 and 25.Each datapoint in graph A shows average tumor volume ± S.D.For example, on day 16 statistically significant differences are seen between the group receiving MBP/vasostatin plus cyclophosphamide and the controls, including mice inoculated with neoplastic cells only (P = 0.0139), mice treated with cyclophosphamide only (P = 0.0135), and mice treated with MBP/vasostatin only (P = 0.0139) (Mann-Whitney U-test).The results of survival analysis (log-rank test) for the animals treated with recombinant MBP/vasostatin and CTX are statistically significant as compared to control groups: mice inoculated with neoplastic cells only (P = 0.02119), mice receiving MBP/vasostatin only (P = 0.01815), and mice receiving CTX only (P = 0.01939).

B
and must had already developed their microvessel networks.Therefore, our therapeutic approach was aimed at inhibiting the formation of new vessels from the already existing ones.
One drawback of the protein-based antiangiogenic strategy, besides the necessity of continued drug administration, is its high cost, labor-intensive preparation and purification of the protein, as well as its immunogenicity (Schellekens, 2002).Immunogenic forms of the protein drug may arise during the lengthy production steps and/or during storage as a result of an altered tertiary protein structure caused by aggregation (Meritet et al., 2001).
The therapeutic strategy tested in this study, based on the use of a recombinant antiangiogenic protein combined with cyclophosphamide, markedly inhibited growth of B16(F10) murine melanoma experimental tumors in mice.Such an approach to cancer therapy might be useful in the clinic, especially when attempting to gain growth control of tumors which have already formed microvessel networks.

Figure 2 .
Figure 2. Identification of MBP/vasostatin fusion protein (Western blot).Proteins separated on 10% (A) or 12% (B) polyacrylamide gel were transferred onto nitrocellulose filter and placed in a solution of MBP-recognizing rabbit antibodies (A) or of goat antibodies recognizing vasostatin (B).The filters were incubated with anti-rabbit (A) or anti-goat (B) immunoglobulin conjugated with biotin and a complex of streptavidin and biotinylated horseradish peroxidase.In order to visualize immunocomplexes the filters were placed in a solution of DAB, a peroxidase substrate.Total protein before induction (lane 1).Fraction 3 h after IPTG induction (0.3 mM) (lane 2).Insoluble fraction (lane 3).Soluble fraction (lane 4).Proteins not bound to the bed (lane 5).First wash (fraction 6).Next washes (lane 7).Protein eluate (lane 8).Prestained Protein Ladder size standard (Fermentas) (lane M).

Figure 4 .
Figure 4.In vivo test of angiogenesis inhibition using Matrigel.C57Bl6 female mice (4-week-old) were used.Each group counted 4 animals.Matrigel with bFGF (150 ng/ml) and proteins (50 µg/ml) added was implanted (see Materials and Methods).Hemoglobin content was determined after 7 days.The result obtained for the MBP/vasostatin group is statistically significant as compared to bFGF (P = 0.03389) and MBP (P = 0.03389) control groups (Mann-Whitney Utest).The experiment was done in duplicate with similar results.

Figure 5 .
Figure 5. Determining therapeutic dose of MBP/ vasostatin protein that inhibits murine melanoma tumor growth.Mice were inoculated intradermally with 2 × 10 5 B16(F10) cells/animal in 100 µl PBSˉ and divided into groups of six.Starting from the 6th day from inoculation and continuing daily for 11 days, various doses of MBP/vasostatin protein: (1-100 μg, see graph) in 100 μl FB/animal were administered intratumorally.Each datapoint represents average tumor volume ± S.D.For example, on day 16 statistically significant differences were observed between mice from groups receiving 60 μg (P = 0.046) and 100 μg (P=0.01171) of the protein, as compared to the control group (Mann-Whitney U-test).

Figure 7 .
Figure 7. Inhibition of murine melanoma tumor growth by treatment with MBP/vasostatin fusion protein and cyclophosphamide.A. Inhibition of tumor growth.B. Survival extension.Animals were inoculated intradermally with 2 × 10 5 B16(F10) cells/mouse in 100 µl PBS‾ and divided into groups of five.The control group included animals that were only inoculated with neoplastic cells.The following study groups were established: mice receiving MBP/vasostatin (60 μg in 100 μl FB/mouse); mice receiving cyclophosphamide alone, and mice undergoing combined therapy.Starting on the 5th day after inoculation, mice received intratumorally the recombinant protein for 21 days (on day 5, 6, 8-12, 14-18, 20-24, 26-29), with the exception of days in which cyclophosphamide was administered.CTX (170 mg/ kg body mass) was administered on the 7th day after inoculation, and then every six days, i.e. on days 7, 13, 19 and 25.Each datapoint in graph A shows average tumor volume ± S.D.For example, on day 16 statistically significant differences are seen between the group receiving MBP/vasostatin plus cyclophosphamide and the controls, including mice inoculated with neoplastic cells only (P = 0.0139), mice treated with cyclophosphamide only (P = 0.0135), and mice treated with MBP/vasostatin only (P = 0.0139) (Mann-Whitney U-test).The results of survival analysis (log-rank test) for the animals treated with recombinant MBP/vasostatin and CTX are statistically significant as compared to control groups: mice inoculated with neoplastic cells only (P = 0.02119), mice receiving MBP/vasostatin only (P = 0.01815), and mice receiving CTX only (P = 0.01939).