Modulation of transforming growth factor beta 2 ( TGF-beta 2 ) by inositol hexaphosphate in colon carcinogenesis in rats 1 Modulação do fator transformador de crescimento beta 2 ( TGF beta 2 ) pelo inositol hexafosfato na carcinogênese colônica em ratos

Purpose: To evaluate modulation in the expression of Transforming growth factor beta2 (TGF-beta2) in short-term colon carcinogenesis. Methods: 64 male rats was used, comprising 4 groups of 16 animals each: group 1 received Inositol hexaphosphate (IP6) and azoxymethane (AOM); group 2, AOM alone; group 3, IP6 alone; group 4 was used as control. Groups 1 and 3 were given 1% IP6 in drinking water for 6 weeks. AOM was administered subcutaneously at weeks 3 and 4 of the experiment at 20 mg/kg of body weight each week. Immunohistochemical processing was performed with the use of anti-TGF-beta2 primary antibodies in right colon samples and quantitation of TGF-beta2 as percentage of expression, through computer-assisted image processing. Results: mean values of TGF-beta2 expression were 9.0 ± 3.9% for group 4 (control), 12.7 ± 4.0% for group 3 (IP6), 19.3 ± 6.2% for group 2 (AOM), and 13.1 ± 5.3% for group 1 (IP6+AOM). The value of p was calculated as 0.0001 for a 5% or lower significance level. Conclusion: the experiment revealed a significant increase in TGF-beta2 expression in right colon with the administration of AOM, and a significant decrease in TGF-beta2 expression when IP6 was administered with AOM.


Introduction
Transforming growth factor beta (TGF-beta) is a member of a family of proteins that regulate numerous biological activities, including cell proliferation, differentiation, adhesion, and apoptosis, extracellular matrix (ECM) production, and early embryo development and immunity.The involvement of members of this family in carcinogenesis is a complex one, but a predominantly suppressing activity is observed in normal tissues.In tumorigenesis, however, changes in cell expression and response favor oncogenic activity 1,2 .TGF-beta is secreted by subepithelial myofibroblasts 3 and is stored in the extracellular matrix as a latent complex, which can be cleaved by the following alterations: (a) physicochemical changes (microalterations in cell acidity, extreme pH values, gamma radiation, reactive oxygen species-ROS); (b) enzymatic changes and with the interaction of nonspecific proteins (proteases, metalloproteinases type 2 and 9, glycosidases, trombospondin, and mediators of integrin); (c) induction by drugs (antiestrogens, retinoids, vitamin D3 derivatives, and glucocorticoids) 4 .Cleavage of the latent complex promotes TGF-beta2 activation, generating strong cell signals after bonding with its cell membrane receptor-TGFBR2 5 .Carcinogenesis is composed of a sequence of steps-initiation, promotion, and tumoral progression-and TGF-beta action has been identified during cancer progression in the following functions: (a) limiting normal epithelial growth and early tumors; (b) causing loss of growthinhibitory response by the loss of TGF-beta receptors or SMAD proteins, or by the specific loss of response of cytostatic genes, selected by more aggressive tumoral growth, facilitated by the additional acquisition of oncogenic mutation; (c) altering tumoral cells that display impaired cytostatic response, while preserving components of the TGF-beta signaling pathway, making them more invasive; (d) developing an environment of immunosupression for the function of T cells, evading cytotoxic removal mediated by T lymphocytes-an event that takes place when a tumor is already established; (e) mediating induction of angiogenic response, facilitating the recruitment of new tumor-nourishing blood vessels; (f) promoting tumoral cell adhesion to the endothelium and/or extravasation of tumoral cells at methastatic sites; and (g) stimulating gene expression of osteoclast differentiation factor (interleukin 11) and connective tissue growth factor 6 .A number of animal studies have demonstrated that inositol hexaphosphate (IP6) inhibits tumoral growth in diverse types of cancer, including that of colon.Proposed action mechanisms are: increased activity of natural killer cells, changes in signal transduction with PI3-K blockade, stimulation of genes responsible for cell differentiation, apoptosis, and antioxidant activity 7 .Because changes in tissue oxygenation levels accompany several pathological states, including cancer, a mechanism must exist that activates latent TGF-beta, whose redox sensitivity is within the range of that displayed by the precursor peptide LTGFbeta.LTGF-beta redox sensitivity is presumably directed toward recovery of homeostasis; however, oxidation may also be a mechanism of LTGF-beta activation that can be deleterious during disease mechanisms involving chronic ROS production 8 .These aspects led us to raise doubts on the role and ability of IP6 in modulating the expression of TGF-beta2 upon administration of a free-radical-generating substance that is carcinogenic to colon.In order to clarify these aspects, we developed the present study, whose purpose was to investigate experimentally the modulation of TGF-beta2 expression with the administration of IP6 in short-term colon carcinogenesis in rats.

Methods
Sixty four male Wistar rats were used.They were bred in the Central Animal Facilities of UFMS and had a mean weight of 152 ± 26 g.The study was evaluated by the Animal Ethics Committee of UFMS, approved, and certified by Protocol 01/ 2001 of April 25, 2001.The experiment was carried out in the Laboratory of Experimental Carcinogenesis, an addition to the Central Animal Facilities of UFMS.Polypropylene cages (standard size for 5 rats) with galvanized wire lids were used, each housing four animals randomly selected by draw.The animals were then acclimated to the housing conditions for 7 days under artificial light (130 to 325 lux) with light/dark cycles of 12 h each, mean temperature of 24 ± 2 °C, and mean humidity of 52 ± 8%.They were fed ad libitum on rat chow (Nuvilab CR1, Nuvital Nutrientes e Produtos Veterinários Ltda, Curitiba, Brazil) and filtered water.The 64 rats were distributed into four groups by draw.Tail tattoos, made with an indelible black ink pen, were used to identify groups (1 through 4) and individual animals (A through Q) within each group.Groups distribution was as follows: group l (IP6+AOM): 16(n) animals receiving IP6 and AOM; group 2 (AOM): 16(n), receiving AOM alone; group 3 (IP6): 16(n), receiving IP6 alone; group 4 (control): 16(n) untreated animals.

Procedures
Administration of substances IP6 (anti-tumoral substance, C 6 H 6 O 24 P 6 Na 12 .Sigma, cat.P3168) was administered as a 1% solution in drinking water ad libitum to groups 1 and 3, for six weeks.AOM (carcinogenic substance, C 2 H 6 N 2 O. Sigma, cat.A9517, lot 70K0847) was administered subcutaneously at 20 mg/kg of body weight (prepared as 100 mg AOM/0.1 ml + 49.9 ml 0.9% physiological solution, resulting in 2 mg AOM/ml) to groups 1 and 2 in the beginning of the third and fourth weeks of the six-week period, at the same times when groups 3 and 4 were given 0.9% saline solution subcutaneously at 10 ml/kg of body weight-in volumes equivalent to those of diluted AOM administered to groups 1 and 2 (Figure 1).

Euthanasia
At the end of the six-week period, the animals were identified, weighed, and subjected to euthanasia by intraperitoneal thiopental at 150 mg/kg of body weight.

Harvesting of material for study
Each rat was positioned in dorsal decubitus and submitted to midline laparotomy.The ileocecal region was then identified and the terminal ileum and right colon were harvested en bloc.

Preparation of material for study
Each colon specimen was opened along the antimesenteric border and the mucosa was rinsed with Ringer's solution.A 1cm long segment was then resected from the right colon, distally from the ileocolic region.The segment of colon wall was sandwiched between the plates of a hinged perforated double holder in order to be maintained straightened, and the resulting set was immersed in 10% buffered formol solution for 24 hours.

Preparation of blocks and slides
Each sample was embedded in paraffin (with sample identification being preserved) and sectioned with a microtome (4-micrometer-thick sagittal sections of colon wall).The resulting sections were mounted on silanized glass slides (first cut for histology; second cut for immunohistochemistry).

Histological analysis
The tissue sections were stained with hematoxylin-eosin and viewed under an optical microscope at 400X magnifications for identification of crypts and aberrant crypt foci (ACF).The following histological abnormalities of ACF were taken into account: increased nucleus-cytoplasm ratio, nuclear stratification, loss of nuclear polarity, structural abnormality, and number of crypts in the ACF (Figure 2).

Microscopy and image capture
Optical microscopy was performed with a microscope (Nikon, mod.Elipse E200) equipped with 10X binocular lenses and a 40X/0.65 objective.The microscope was coupled to a video camera (Samsung, mod.SCC131) connected to a microcomputer (Duron 750-MHz processor, 128-MB RAM, 20-GB hard disk, Microsoft Windows 98 SE) through a videoboard (Pinnacle Studio PCTV, USB).

Immunohistochemistry quantitation
TGF-beta2 was quantified by computer-assisted image processing with ImageLab software.One single image from each sample was captured, which was centered either on the crypt or the ACF.For quantitation purposes, the ACF were examined and selected so that when more than one focus was available, a focus containing two crypts was chosen.In the absence of ACF with two, three, or four crypts, a random selection of crypts or ACF with one single crypt was made.Each image captured was saved (for documentation purposes) and codified according to a previously designed scheme.The quantified brown coloration density corresponds to TGF-beta2 expression in area percentage (using a RGB filter, blue background, 0-to-147 color interval) in the selected crypt or ACF (Figure 3).

Statistical analysis
A 95% confidence interval and a 5% significance level (p ≤ 0.05) were adopted.ANOVA was used for comparing: (a) the mean weights of animals per group across all groups at the beginning of the experiment; (b) the mean differences between the initial and final weights across groups (Table 3); (c) the mean values, in percentage, of TGF-beta2 expression as revealed by immunohistochemistry processing.Paired comparison of the mean values obtained through ANOVA was then performed by applying the Fisher's Least Significant Difference (LSD) test.

Statistical calculation
The analysis of variance among groups calculated for TGF-beta2 expression yielded p = 0.0001 (significance: p £ 0.05).Results of the paired comparison of mean values are shown in Table 2.

Effect on TGF-beta2 with administration of AOM
A significant increase in TGF-beta2 expression was found at the end of the fourth week after administration of AOM (group 2).A significant decrease in TGF-beta2 expression was found when IP6 was supplied.The administration of IP6 was started two weeks before that of the carcinogenic substance AOM (Figure 4).

Effect on TGF-beta2 with administration of IP6
No significant difference was found for TGF-beta2 expression in group 3, which received IP6 alone, as compared with group 4 (control).A significant decrease in TGF-beta2 expression (group 1) was found when the antitumoral substance IP6 was supplied in conjunction with AOM, as compared with group 2 (AOM alone).

General observations
No significant differences were found when the mean weight gains were compared across groups (p = 0.2274).No deaths occurred in the experimental period.Three samples failed to be collected because of losses during microtomy or because sectioning made them unusable.

Discussion
In this investigation, rats were chosen as experimental animals because ACF have been studied in them.Also, rat gene TGFBR2 9 has been identified as a tumoral suppressor.These facts further increased the reliability in the carcinogenesis model already standardized and recommended by the International Agency for Research on Cancer.The adoption of a short-term colon carcinogenesis model, which uses ACF identification as an endpoint structure, addresses the need for investigating a phenomenon related to primary prevention of cancer, at a stage when this disturb is not yet phenotypically characterized.When the genetic bases of colorectal cancer and critical signaling pathways of tumorigenesis (APC/beta-catenin, DNA mismatch repair (MMR), and TGF-beta/SMAD pathways) are considered, a higher prevalence is found for colorectal cancer mutations with genetic instability of gene TGFBR2 in culture cell lines.In sporadic cancer, a prevalence of gene TGFBR2 has been identified in 82% to 90% of cases; in familiar cancer, the prevalence ranged from 78% to 83% 10 .Theses levels point to the need to investigate its potential importance in cancer prevention, and for implementing procedures aimed at modulating TGF-beta2 expression in the tumorigenesis process.Because apoptosis is a relevant process-a physiologic condition necessary for the elimination of defective cells-the choice of substances that participate in it and in antioxidation defined the criteria that led to the adoption of TGF-beta2 and IP6 in the experiment.Also, IP6 is a natural substance, being available from daily-use foodstuffs, a feature that favors, in economic terms, its wide-scale consumption in nonpurified form.Both the well-establish link between diet and cancer and the new genomic technologies have made it possible to investigate the nutritional modulation of carcinogenic pathways with the use of nutrients, micronutrients, and phytochemicals 11 .In rats, the administration of 1% IP6 promotes a decrease in the appearance of ACF 12 , whereas at a concentration of 2% this substance inhibits tumor establishment 13 .Chemoprevention of colon cancer has come to be regarded as an important procedure, so that known substances can be ranked in terms of the potency that leads the a decrease in the frequency of ACF establishment-of 31% with the use of 1% IP6 and of 70% for 2% IP6.It is worth mentioning that 2% IP6 is an alkaline solution, and the choice of IP6 at a 1% concentration and neutral pH favors the oral use of the solution, with preserved effectiveness 14 .The technique for keeping the colon wall segments straightened prevented the retraction caused by immersion in formol, which results in shrinkage of muscle fibers.This procedure made tissue samples more suitable for sagittal microtomy and subsequent optical microscopy and viewing of crypts and ACF.Computer-assisted image analysis was chosen to quantify the brown-tinted areas corresponding to TGF-beta2 expression and made visible by immunohistochemical processing.This type of analysis was needed because TGF-beta2, having low molecular weight, has intra-and extracellular distribution.With this method, it became possible to quantify TGF-beta2 expression in an objective and reproducible manner.This is in contrast with most investigations reported in the literature, which utilize semiquantitative immunohistochemistry, whereas in the present study a quantitative immunohistochemistry methodology was required 15 .An algorithm was developed to deconvolve the color information acquired with red-green-blue (RGB) cameras and to calculate the contribution of each of the applied stains based on stain-specific RGB absorption.The algorithm was tested using different combinations of diaminobenzidine, hematoxylin and eosin at different staining levels.Quantification of the different stains was not significantly influenced by the combination of multiple stains in a single sample.The color deconvolution algorithm resulted in comparable quantification independent of the stain combinations as long as the histochemical procedures did not influence the amount of stain in the sample due to bleaching because of stain solubility and saturation of staining was prevented.This image analysis algorithm provides a robust and flexible method for objective immunohistochemical analysis of samples stained with up to three different stains using a laboratory microscope, standard RGB camera setup and the public domain program NIH Image 16 .The absence of significant differences in weight gain across groups (p = 0.2274) corroborates results from other studies where no weight changes were observed when IP6 was supplied 17 even when the dietary and proteic quality of the animal chow was evaluated.In the present study, food was supplied ad libitum, since a calorie-restricted diet has been shown elsewhere to decrease significantly the expression of the proteins TGF-beta1/ beta2 and COX-2 in the mucosa of obese rats 18 .Apoptosis is significantly triggered upon the administration of AOM, to eliminate cells damaged by the administration of this substance 19 .This was evident in group 2, where a significant increase (p = 0.0000) of TGF-beta2 expression occurred, as compared with group 4 (control).The administration of IP6 alone promoted an increase in TGF-beta2 expression, though not in a significant manner, as compared with the control group.This finding, however, can be regarded as relevant, as it contradicts the expectation that IP6 might normalize cellular physiology through one of its actions and promote normal or hypoexpression of TGF-beta2.Such increase might be explained by demonstrating that apoptosis is increased by 60% at the lower portion of the crypt when a diet containing IP6 is supplied 20 .As many as 82.6% of mucinous goblet cells present in ACF having 1 to 3 crypts contain labile class III mucin 21 .In the present experiment, no TGF-beta2 expression was observed in the mucin of goblet cells in those groups receiving AOM (Fig. 3, groups 1 and 2).Evidence for a significant increase in TGF-beta2 expression at the end of the fourth week after the initial administration of AOM, in ACFtype lesions regarded as nonadenomatous and premalignant, has not been reported in the literature.This allows the occurrence of early alterations in TGF-beta2 to be regarded as preceding the phase of tumorigenesis progression; this finding differs from the previously accepted view on the genetic events related to colon cancer tumorigenesis, where TGF-beta2 is increased from the adenoma to the carcinoma phase (Chung  DC, 2000).The present study has demonstrated that in group 1 (IP6+AOM) the administration of IP6 provides a modulating action, significantly decreasing TGF-beta2 expression, as compared with group 2 (AOM) (p = 0.0009).The experiment reveals that the administration of the carcinogenic substance stimulates an increase in TGF-beta2, which is maintained at the end of the fourth week following the administration of AOM.It can be inferred that a mechanism must be at play, other than the generation of free radicals and DNA damage, as the latter undergoes repair within 72 hours 22 .The existence of a stimulus promoting sustained high levels of TGF-beta2 at the end of the fourth week deserves investigation.The modulation of TGF-beta2 expression promoted by the administration of IP6 reinforces the recommended chemopreventive use of selected foods containing IP6, and/or the addition of IP6 to diets.These measures are expected to reduce the incidence of colon cancer, as they should help to

FIGURE 1 -
FIGURE 1 -Protocol for the administration of substances.

FIGURE 3 -
FIGURE 3 -Computer-assisted image processing.TGF-beta2 was quantified by computer-assisted image processing with ImageLab software.One single image from each sample was captured, which was centered either on the crypt or the ACF.The quantified brown coloration density corresponds to TGF-beta2 expression in area percentage (using a RGB filter, blue background, 0to-147 color interval) in the selected crypt or ACF.

TABLE 1 -
Immunohistochemistry results.Percentages of the captured areas of TGF-beta2 expression in aberrant crypt foci in colon, as quantified through computerassisted image processing (ImageLab software) from immunohistochemistry-processed slides.Campo Grande, Brazil 2004.

TABLE 2 -
Paired comparison of TGFbeta2 expression in the groups.Campo Grande, Brazil 2004.

TABLE 3 -
Weight gain during the experiment.Mean and standart deviation of the difference between final and initial weights.Campo Grande, Brazil, 2004.FIGURE 4 -Schematical representation of significant influence of inositol hexaphosphate on TGFbeta2 expression in aberrant crypt foci in colon (Group 1) at the end of the sixthweek experiment, as compared with group 2, in azoxymethaneinduced colon carcinogenesis model in rats.*p=0.0001