The Anticancer Activity of Alpha-tomatine against Mammary Adenocarcinoma in Mice

a Aim. To evaluate the anticancer effect of alpha-tomatine (i.p.) either alone or in combination with doxorubicin (i.v.) in a mouse tumour model. Methods. We studied the effect of repeated alpha-tomatine (0.1 – 9 mg/kg) and/or doxorubicin (2 mg/kg) on the growth and mitotic activity of the solid Ehrlich tumour in vivo, as well as on the survival of the tumour-bearing mice. Results. Monotherapy with alpha-tomatine had a significant dose-dependent anticancer effect which peaked at 1 mg/kg. This was shown by both slowed tumour growth and reduced tumour cell proliferation. We also provide the first evidence that the combination alpha-tomatine (1 mg/kg) and doxorubicin (2 mg/kg) had a synergistic effect and significantly prolonged the survival of the mice. Neither alpha-tomatine nor doxorubicin influenced the infiltration of tumours with CD3+ lymphocytes; nor were we able to find an in vivo modulation of the key molecules of two regulatory pathways reported in vitro as the principal anti-cancer mechanisms of alpha-tomatine, i.e. iNOS and phosphorylated ERK2. However, alpha-tomatine still led to intracellular DNA inhibition and protein synthesis in Ehrlich tumour cells in a short-term culture ex vivo with IC 50 values of 8.7 and 6.6 µM. Conclusions. The results suggest that ΤΟΜ, especially in combination with doxorubicin, may be a promising agent for the treatment of malignant solid tumours. Despite growing knowledge of the mechanisms of ΤΟΜ action in cancer cells, most aspects remain unclear. Parallel organ toxicity, especially potential liver effects, requires careful attention when performing in vivo studies in the future.


INTRODUCTION
α−Tomatine (TOM) is the major glycoalkaloid of tomato (Solanum lycopersicum L., syn.: Lycopersicon esculentum Mill.).The molecule is composed of an aglycone core, tomatidine, (22S,25S)-5α-spirosolan-3β-ol, which is attached with its 3-OH group to the saccharide moiety called lycotetraose (Fig. 1).In plants, ΤΟΜ may provide defence against pathogenic fungi, bacteria, viruses and herbivores 1 .This effect corresponds with the content of the agent and changes throughout the lifespan of the tomato plant; e.g. during tomato fruit maturation, the levels of ΤΟΜ decrease considerably, which, among other factors, brings about a reduction in the bitter flavour 2 .
In mammals, a wide variety of health-promoting properties of ΤΟΜ have been reported, including the lowering of plasma concentrations of LDL and triacylglycerols 3 and cardiotonic 1 and antiviral activity 4 .ΤΟΜ may also contribute to the stimulation of antigen-specific humoral and cellular immune response, including the augmenta-tion of anticancer defence 5 .Friedman et al. 6 reported the chemopreventive effect of ΤΟΜ against liver and stomach cancer in rainbow trout (Oncorhynchus mykiss Walb.). Lee et al. 7 showed the antiproliferative effect of ΤΟΜ (c ~ 1 µg/L) against human colon and liver cell lines, which was more powerful than that of some classical anticancer agents such as doxorubicin and camptothecin.Similar results were obtained with high-tomatine extracts from green tomatoes 8 .
On investigating the mechanism of action, it was found that ΤΟΜ affects important signaling pathways responsible for the regulation of cell proliferation, differentiation, and migration 9,10 .Shih et al. 10 found ΤΟΜ (c ≤ 2 µmol/L) to suppress invasion and migration in a human lung adenocarcinoma cell line.It also inhibited the phosphorylation of Akt (protein kinase B), a serine-threonine kinase which regulates the function of many cellular proteins involved in metastasis and the proliferation of cancer cells 11 , and the phosphorylation of extracellular signal-regulated kinase 1 and 2 (ERK 1,2).The ERK (extracellular regulated kinase) pathway often plays a role in oncogenic transformation and in the regulation of metalloproteinases and urokinase expression, i.e. of the enzymes which degrade the extracellular matrix and help cancer cells invade tissues and metastasize 12,13 .Shih et al. 10 detected lower activities of metalloproteinases and urokinase in cells incubated with ΤΟΜ.ΤΟΜ also inhibited the activation of focal adhesion kinase, the phosphatidylinositol 3-kinase (PI3K)/Akt signal pathway, and NF-κB (nuclear factor kappa B) in cancer cells 14,15 .In leukemic cells, TOM induced apoptosis, which was caspase-independent and associated with the down-regulation of survivin expression and the translocation of AIF (apoptosis inducing factor) into the nucleus 16 .In our recent study on MOLT-4 cells, too, TOM induced caspase-independent cell death associated with an increase in p53 and in the mitochondrial protein PUMA.The inhibition of proliferation by alphatomatine was linked with an increase in p21 WAF1/ CIP1 level and activation of the checkpoint kinase 2 (Chk2) (ref. 17).
Despite the very encouraging data from in vitro experiments, to our knowledge only one other very recently published study has investigated the anticancer effect of ΤΟΜ in a mammal model in vivo.It found that TOM at a dose of 5 mg/kg (administered i.p. every other day) inhibits the growth of HL-60 leukaemia xenografts in immunodeficient mice 16 .However, the effect of the therapy on mice survival or its dose dependence was not followed.In the present study, we evaluated in detail the effect of the repeated intraperitoneal administration of ΤΟΜ at doses from 0.1 to 9 mg/kg, either alone or in combination with doxorubicin, on tumour growth, mitotic activity, and the survival of immunocompetent mice bearing the solid Ehrlich tumour.We also tested the effect of ΤΟΜ on the DNA and protein synthesis of Ehrlich tumour cells in short-term culture ex vivo.As a surrogate for potential toxicity, serum bilirubin and liver enzymes were measured.

MATERIALS AND METHODS
Design of the in vivo study.Two hundred and eight NMRI female mice weighing on average 27.2±1.50gobtained from BioTest s.r.o.(Konárovice, Czech Republic − CR) were fed a standard diet and water ad libitum.The solid Ehrlich tumour was purchased from the Research Institute for Pharmacy and Biochemistry (VUFB), Prague, CR, and then maintained in NMRI mice by periodical transplantations in the laboratory.The homogenized tumour tissue was inoculated subcutaneously into all but 16 mice (controls without tumour) on day 0, using 0.2 mL of 1/1 (v/v) homogenate freshly prepared in isotonic glucose solution.The tumour-bearing mice were then divided into 12 groups as follows: saline treated controls with the tumour; five groups of tumour-bearing animals treated with 0.1, 0.3, 1, 3, or 9 g TOM/kg; a further five groups of tumour-bearing mice receiving 0.1, 0.3, 1, 3, or 9 g ΤΟΜ/kg, but in combination with 2 mg/kg of doxorubicin (DOX).Each group consisted of 16 animals.ΤΟΜ (as hydrate) was purchased from Sigma-Aldrich Co. (CR); DOX was of pharmaceutical grade (Doxorubicin Teva 0.2%, 50 mg/25 mL, Pharmachemie B.V., Haarlem, Netherlands).The doses of ΤΟΜ were selected according to reported LD 50 for i.p. administration 18 .The i.v.dose for DOX was chosen as a suboptimal dose (our data not shown 19 ) representing approximately 40% of the LD10 (ref. 20).Both ΤΟΜ (i.p.) and DOX (i.v.) were administered on days 1, 4 and 7, in volumes of 0.2 mL per 20 g body weight (DOX as solutions in 0.9% saline; ΤΟΜ in 0.9% saline acidified with HCl).In the groups treated with both substances, ΤΟΜ was administered 30 min after DOX.The reason for this was the suspected vasoconstriction activity of TOM observed in our preliminary experiments (data not shown).Mice were weighed on the first, fourth, sixth and seventh days.On the eighth day, half of the mice were sacrificed; the remaining animals were left in order to observe their survival.One hour before the sacrifice of the first half of the animals, a solution of 5-bromo-2'deoxy-uridine (BrdU; 100 mg/kg, Sigma-Aldrich, CR) was injected i.p. to assess the DNA synthesis of cells (particularly tumorous).In addition to the total body weight, tumours, livers, lungs, hearts, spleens and blood were also weighed.The organs were examined macroscopically, the tumours also histologically.Biochemical analysis of the plasma samples − assessment of the concentrations of total and conjugated bilirubin, as well as the activities of ALT (alanine aminotransferase) and AST (aspartate aminotransferase) -were performed on Modular PP (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions.
The care and handling of mice conformed to European Union recommendations on handling experimental animals and was approved by the Animal Ethics Commission of the Medical Faculty in Hradec Králové, CR (Nr.30790/2010-30).

Histological evaluation.
One hour before the animals were sacrificed, BrdU (Sigma-Aldrich, CR) was injected intraperitoneally (100 mg/kg of body weight) to assess the DNA synthesis of cells.Samples of prepared tumours were fixed by immersion in 4% neutral formaldehyde for 3 days.Paraffin sections (7 μm thick) were prepared by conventional histological processing.The indirect peroxidase-anti-peroxidase method was used for immunohistochemical analysis.The hydrolysis of DNA with 2N HCl was necessary prior to the incubation of slices with mouse monoclonal anti-BrdU (Dako, CR; dilution 1:100).Biotinylated anti-mouse secondary antibody (Jackson ImmunoResearch Lab., USA; dilution 1:500) and streptavidin conjugate of peroxidase (Dako, CR; dilution 1:300) were subsequently used.Visualization was performed conventionally by DAB (Sigma-Aldrich, CR), and Gill's haematoxylin was used for counterstaining.Photomicrographs were made using an Olympus BX 51 microscope equipped with a DP 70 digital camera (Olympus, Prague, CR) and Quick Photo Camera 2.3 software (Promicra, Prague, CR).
Ex vivo DNA and protein synthesis assay.The above mentioned Ehrlich tumour was homogenized and inoculated i.p. (10 6 tumour cells to female NMRI mice (see above)).The cells for this experiment were obtained from the peritoneal cavity on day 7 after i.p. inoculation.Cell suspensions were packed by low-speed centrifugation (600 x g for 10 min at 4 °C).The cells were resuspended in Krebs III phosphate buffer (all components obtained from Sigma-Aldrich Co.), pH 7.4, without calcium but enriched with D-glucose (540 mg/L) and ascitic plasma (2.5 %, v/v).The cell suspensions were incubated for 150 min at 37 °C with the addition of [6-3 H]-thymidine (73 kBq/mL of incubated suspension) and L-[U-14 C]-amino acid mixture (5 kBq/mL of incubated suspension) obtained from Lacomed s.r.o., CR, and TOM (1 -9 µM) or DOX (4.05 -258.6 μM).The cell density was 1.5x10 6 cells/mL incubated suspension.Samples were taken at 30 min intervals from 0 to 150 min and processed by the method described by Mattern using filter paper discs 21 .The activities of 5% trichloracetic acid-insoluble fractions were measured by liquid scintillation spectrometry (Triton scintillation fluid (Chemopetrol, CR), Tri-Carb 2900 TR liquid scintillation counter (PerkinElmer, Inc., U.S.A.).The IC 50 was calculated as the concentration at which the initial velocity of precursor incorporation was halved compared to the controls.The mathematical model used is described in our previous paper 22 .
Statistical Analysis.Data are presented as means and standard deviations.Two-Way Analysis of Variance with post-hoc multiple comparisons by Fisher's LSD test was used to test for differences.Survival times in groups are expressed as geometric mean with a 95% confidence interval.Kaplan-Meier curves and logrank tests were used to compare survival times in groups.The level of significance was α=0.05.MS Excel 2003 and NCSS software was used for the calculations and statistical evaluations.

Effect on the survival of tumour-bearing mice
The mean overall survival of tumour bearing mice without therapy was 14.5 days.The dose of ΤΟΜ with the greatest effect on the survival rate was 1 mg/kg, both in the monotherapy and in combination with DOX (Fig. 2).However, only the combination of 1 mg/kg of ΤΟΜ with DOX prolonged the survival time significantly compared with untreated tumour-bearing control mice (P<0.01) and with doxorubicin monotherapy (P<0.05).On the other hand, doxorubicin alone had no significant effect on the survival time.

Effect on tumour growth
Fig. 3 shows the weight of tumours in mice treated with saline, ΤΟΜ, DOX or the TOM/DOX combination measured on day 8 and significantly documents differ-Fig.2. Kaplan-Meier analysis of survival.The administration of α-tomatine 1 mg/kg i.p. (T1) with doxorubicin 2 mg/kg i.v.(D) significantly prolonged the survival of tumour-bearing mice when compared with tumour-bearing controls treated with saline i.p. (Ctrl).The monotherapy with α-tomatine 1 mg/kg i.p. had the second largest effect, though this was not statistically significant.This plot shows only some selected groups.ences in the effect of the used drugs.Both ΤΟΜ and DOX alone as well as their combination had a significant inhibitory effect on tumour growth.The greatest effect was noted after the administration of ΤΟΜ in doses of 1, 3 and 9 mg/kg in combination with DOX.Fisher's LSD test revealed that the tumour weights in groups treated with ΤΟΜ alone (0.1 -3 mg/kg) did not differ significantly from those treated with its combination with DOX (P>0.05).The groups subjected to ex vivo proteomic assays as well as to histological assessment were therefore chosen with respect to the optimal effect on survival and tumour growth.

Effect on mitotic activity in vivo
The microscopic structure of the tumours stained for mitotic activity (incorporation of BrdU) is depicted in Fig. 4. A slightly decreased density of newly generated (BrdUpositive) tumour cells was noted compared with untreated control mice (Fig. 4a) and groups treated with either the combination of 1 mg/kg of ΤΟΜ and DOX (Fig. 4c) or 1 mg/kg of ΤΟΜ alone (Fig. 4d).In group treated only with DOX (Fig. 4b), the decrease in mitotic activity was less apparent.However, in all evaluated tumours, clusters of intensely proliferating cells were separated by strips or fields of currently quiescent (non-dividing) cells.In mice which underwent any of the treatment methods, these areas of BrdU-negative cells increased (frequently markedly) in size (Figs.4b-d).Precise quantitative analysis was, however, not possible due to the clonal origin of tumour cells resulting in the formation of typical clusters with a very irregular distribution of proliferating (BrdU-positive) cells (Figs. 4 a-d).Thus, the decrease in proliferating fraction was measured by quantification of PCNA using Western blot, which was verified in groups treated by TOM at 1 mg/kg either alone or with doxorubicin (Fig. 5).

Western blot
To assess the major pathways that may be involved in anticancer process, we evaluated the presence of the CD3 antigen -the indicator of T-lymphocytes; phosphorylated ERK2 -a key active regulatory molecule; and iNOS (inducible nitric oxide synthase) -a target downstream of NF-κB as an indicator of its activation.We did not detect any significant changes in the expression of these proteins after ΤΟΜ monotherapy (Fig. 6).The induction of iNOS was only observed in mice treated with both agents.

Pathology and biochemical analysis of the plasma
No macroscopic metastases were found in the organs.The body weights of mice treated with ΤΟΜ 1 -9 mg/kg or DOX, either alone or in combination with TOM 0.3 -9 mg/kg, were significantly higher than those of untreated mice, both the tumour-free and the tumour-bearing controls (Fig. 3).Hepatomegaly, which was distinct in the tumour-bearing mice, was reduced in groups receiving 1 and 3 mg/kg of ΤΟΜ as well as in the combination groups with TOM 1 -9 mg/kg (but not in the group treated only with DOX) when compared to the tumour-bearing controls (Fig. 3).Similar results were obtained in relation to the weight of the spleen, where -in addition to the groups treated with only 1 and 3 mg/kg of ΤΟΜ -all groups receiving DOX, including all TOM/DOX combination groups, displayed reduced or no splenomegaly (Fig. 3).Interestingly, the activities of ALT and AST in blood serum were increased only in mice treated with the monotherapy of ΤΟΜ 1 -9 mg/kg.On the other hand, in groups receiving combinations of TOM and DOX, lower concentrations of total bilirubin were observed (Fig. 7).The levels of conjugated bilirubin did not differ between the groups (data not shown).

Effect on DNA and protein synthesis ex vivo
ΤΟΜ inhibited the incorporation of the [6-3 H]thymidine and L-[U-14 C]-amino acid mixture (IC 50 values were 8.7 and 6.6 μM, respectively), which can be interpreted as the inhibition of both DNA and protein synthesis; however, the effect on protein synthesis appears to be even stronger.DOX inhibited only DNA synthesis (IC 50 = 7.7 μM); its effect on protein synthesis could not be detected.

DISCUSSION
In the present study, we show for the first time the effect of i.p. injected TOM against a solid tumour in vivo.We also show the beneficial effect of the concurrent intravenous administration of DOX.ΤΟΜ reduced the growth of the solid Ehrlich tumour, which resulted in a prolongation of the survival of tumour-bearing animals receiving a combination therapy of TOM and DOX.Our results suggest that the optimal dose of TOM is close to 1 mg/ kg.Lower doses were not effective, as they were incapable of reducing tumour growth.Higher doses, on the other hand, did not offer any additional activity.These findings are unique, because with the exception of one very recent study on leukaemia xenograft in mouse models 16 , no similar data have been reported yet for the effects of TOM or related glycoalkaloids from the Solanaceae family, in comparable tumour-bearing mammal models.
The use of the Ehrlich tumour, a poorly differentiated malignant tumour derived from a spontaneous adenocarcinoma of the mouse mammary gland in immunocompetent mice, has the advantage of involving the immune system in the inhibition of tumour progression [23][24][25] .In order to avoid excessive curability arising from the host's immune response 26 , we started the treatment one day after the inoculation of tumour cells.Moreover, we attempted to detect the CD3 surface antigen of T-cells, as the hallmark of the activation of immune anti-cancer defence in this animal model 27 , and saw no influence of either ΤΟΜ (1 mg/kg) or DOX on tumour infiltration by CD3+ cells.This suggests that the mechanisms of activated innate immunity may not be crucial for the observed in vivo anticancer effect of ΤΟΜ.The recent findings in immunodeficient mice 16 support this conclusion.
In line with previous reports 7,8 , in our study ΤΟΜ was also able to inhibit the growth of cancer cells ex vivo.In this case, ΤΟΜ blocked both DNA and protein synthesis, whereas doxorubicin exerted only a weak inhibitory action against protein synthesis, which is in accordance with previous findings 28,29 .The stronger effect of TOM on protein than on DNA synthesis might indicate an effect on energy metabolism 30 rather than the primary induction of DNA damage or inhibition of replicative DNA synthesis.In order to investigate the global effect of TOM on cell functions, we analysed the expression of proteins described formerly as potential targets for ΤΟΜ action in ex vivo samples of tumours: iNOS as a NF-κB downstream target, and phosphorylated ERK2.The negative findings in this area contradict previous in vitro studies and suggest another mechanism, which is probably different from that of doxorubicin, because ΤΟΜ did not interfere with anthracycline and potentiated its anticancer activity.This synergism may also be the result of P-gp (permeability glycoprotein) efflux transporter blockade, which was detected previously for tomatidine, the aglycon of ΤΟΜ 31 .The impact on cell membrane structures would be even more important for the overall action of ΤΟΜ because it interacts strongly with cholesterol and may cause cell membrane disruption and intracellular stress [32][33][34] .Whether TOM can also induce apoptosis in solid tumours such as mammary adenocarcinoma in the manner described by Chao et al. 16 will need to be clarified by further investigation.Overall,, the mechanism of ΤΟΜ anticancer activity appears to be complex, not well understood and may serve as a very interesting focus for further research.
Splenomegaly, caused by the accumulation of lymphocytic elements, is typically associated with progression of the Ehrlich tumour, along with thymus atrophy [35][36][37][38] .The weight of the tumour-free liver was also found to correspond with the growth of subcutaneous murine adenocarcinoma 39 .Since this effect on both organs was attenuated in the presence of agents in concentrations which also effectively reduced cancer growth, we speculate that it may be attributable to the anticancer activity of both drugs rather than to their direct effect in the organs.Although we did not evaluate whether ΤΟΜ at the doses used had an effect on liver weight in healthy mice, Friedman et al. 40 demonstrated no significant effect of feeding healthy mice ΤΟΜ (2.4 mmol/kg for 7 d) on liver or body weight, in contrast to the related potato glycoalkaloids.In that study, however, the different bioavailability after oral administration may play a role.We detected an increase in the activities of ALT and AST in ΤΟΜ i.p. treated groups.Similar results have also been reported by Dalvi 41 after i.p. injection of a close congener of ΤΟΜ, α−solanine (20 mg/kg) to rats.Thus, ΤΟΜ seems to be hepatotoxic too, because several previous studies on a similar cancer model reported a decrease in ALS/AST levels back to the levels of tumour free mice after treatment with liver-friendly anticancer agents 42,43 .However, the lack of an increase in ALT/AST upon injection with the same doses of ΤΟΜ 30 min after DOX had been administered -as observed in our study -requires further investigation.One possible explanation may be that DOX interferes with the binding of ΤΟΜ to the cytoplasmatic membrane of hepatocytes.
The overall toxicity of ΤΟΜ depends on the route of administration.Whereas oral ΤΟΜ shows apparently lower toxicity than the potato glycoalkaloids α−chaconine and α−solanine 1,40 , presumably due to the poor absorption of ΤΟΜ complexes with cholesterol in the gut, the lethal i.p. doses of ΤΟΜ and potato glycoalkaloids are similar.In mice, the parenteral toxicity of ΤΟΜ is characterized by an LD 50 of 32.4 mg/kg for i.p. administration 17 and of 18 mg/kg for intravenous injection 44 .We observed peripheral necrosis (tail tips, auricles) at doses as high as 5 mg/kg (i.v.) and 30 mg/kg (i.p.) (data not shown).Such a narrow therapeutic window for the intravenous treatment was one of the reasons we chose the intraperitoneal route of administration at the selected dose spectrum.Death after administration of a lethal dose of ΤΟΜ, if it occurs, is sudden.The symptoms following parenteral injection include short-lived hypotension without a great effect on heart rate, an increased respiratory rate, and haemolysis 44 .Although haemolysis is reported, we demonstrated reduced concentrations of total bilirubin in serum, especially in groups where ΤΟΜ was combined with DOX.The levels of conjugated bilirubin in serum were unaffected either by the tumour, or by ΤΟΜ alone.One of the possible underlying mechanisms of the total hypobilirubinaemia observed might be the interaction of the compounds with the binding of unconjugated bilirubin to plasma protein and the potentiation of its elimination by the liver.Another possible mechanism might be the induction of UDP-glucuronyltransferase, as DOX was reported to be its strong inducer 45 .In conclusion, our results suggest that ΤΟΜ, especially in combination with doxorubicin, may be a promising agent for the treatment of malignant solid tumours.Despite growing knowledge of the mechanisms of ΤΟΜ action in cancer cells, most aspects still remain unclear.Parallel organ toxicity, especially potential liver effects of the agent, requires careful attention when performing in vivo studies in the future.

Fig. 4 .
Fig.4.Mitotic activity was demonstrated in areas of approximately similar location -next to the outer surface of the tumour taken from the solid nodous part.Clusters of intensely proliferating cells were separated by strips or fields (x) of currently quiescent (BrdU-negative) cells, which subsequently increased (frequently markedly) in size in all treated groups.Some parts of the tumour (in all groups) were richly vascularized (V) but also haemorrhagic (H) and necrotic areas of different size were frequently present.The host tissue (skeletal muscles -M, fat cells -F or connective tissue) was incorporated into the tumour by its infiltrative and invasive growth.

Fig. 6 .
Fig. 6.The presence of CD3 antigen -the indicator of T-lymphocytes, phosphorylated ERK2 and iNOS in tumour homogenates from mice treated with saline (Ctrl), α-tomatine 1 mg/kg i.p. (T1) and the combination of α-tomatine 1 mg/kg i.p. with doxorubicin 2 mg/kg i.v.(T1 + D).No significant changes in expression of these proteins after α-tomatine monotherapy were detected.The induction of iNOS was only observed in mice treated with both agents.Values are the means ± SEM.Significantly different from the controls ( * P < 0.05).