Effects of hyperglycemia and ageing on the healing of colonic anastomoses in rats

Introduction: Despite the significant advances in the knowledge of the healing process, there is a limited number of studies demonstrating the relationships of this healing with ageing and elevated levels of glycemia. Purpose: To evaluate the effects of ageing and hyperglycemia on the healing of colonic anastomosis. Methods: 138 young and old male rats were utilized. Some of them were normoglycemic and others had hyperglycemia induced by streptozocin (50mg/kg). They were maintained under control for 90 days. They were then submitted to a termino-terminal anastomosis in the left colon. On the third, seventh and fourteenth days after surgery, their resistance was evaluated and a histopathological study of the anastomosis was carried out. Results: Gain in resistance was similar for both groups. The additive effect of age with hyperglycemia made a significant difference to the collagen I (p<0.001), III (p=0.022) and total (p<0.001). Among the old animals, the glycemia was a determining factor for the occurrence of a significant difference in total collagen (p=0.029) and collagen I (p=0.013). Among the normoglycemics, age determined a lower density of collagen I (p=0.002). Conclusion: There is delayed collagen synthesis and maturation of the scars of older animals, a situation that becomes more serious in older hyperglycemic animals, but insufficient to adversely affect the gain in resistance.


Introduction
Healing can be affected by negative influences of many factors such as: disturbances of coagulation, metabolic disturbances such as diabetes and hypothyroidism, vascular alterations such as atherosclerosis, radiotherapy, the chronic use of medicines such as steroidal and non-hormonal anti-inflammatory drugs, advanced age, infection and others 1 .
With the advance of age, it is common for the coexistence of diseases to affect the healing process.It has been shown that although there are differences in the healing process for colonic anastomoses in young and old rats, like the higher number of inflammatory cells and the higher deposition of collagen I in the post-surgery phase of anastomoses in young rats, age by and of itself did not cause the failure to heal 2 .Nevertheless, age did delay re-epithelization 3 .
One of the greatest physio-pathological consequences of hyperglycemia is excessive chemical interaction of glucose with proteins, leading to their connection without the aid of enzymes, non-enzymatic glycosylation.It has been seen that diabetics glycosylated seric protein levels that are four times higher than normal individuals 13 .Short half-life proteins have a reversible reaction, therefore damage is limited and reversible in long half-life proteins there is greater accumulation of end products of glycosylation to a point that it is irreversible, leading to structural and functional alterations 14 .
Oxygen has greater affinity for glycosylated hemoglobins, which is reflected in alterations in the hemoglobin dissociation curve, resulting in reduced release of oxygen to the tissues 15 .As collagen synthesis is dependent on oxygen, healing in diabetics may be compromised.
Diabetic rats have been described as showing deficiency in the closing of cutaneous injuries, with delayed re-epithelization, prolonged inflammatory response, impoverished angiogenesis, reduced matrix deposition and less contraction of the injury 16 .In the same way that it affects healing at the cutaneous level, diabetes can also interfere negatively in the healing of the gastrointestinal tract 17 .
Studies of the healing of injuries in the stomach and duodenum of diabetic rats have shown that even when late, the wounds of these animals are less resistant and have lower quantities of total collagen when compared to normal diabetic animals controlled by the use of insulin 18 .
Researchers admit that the lower resistance of the anastomosis is not due to the amount of collagen but possibly to the alteration in the quality of the collagen fibers 19,20 .These authors have shown that diabetes caused structural deficiencies in the healing of intestinal anastomoses, even when the glucose level returned to normal with the use of insulin 21 .
In diabetic rats with gastric injuries there is reduced fibroblast growth factor (FGF), a potent angiogenic factor which is capable of stimulating other cells, like the fibroblasts, the smooth muscular cells and the epithelial cells 22 .The resistance of the anastomoses in diabetic rats was found to have been compromised, with less type I collagen accumulation on the seventh day after surgery 23 .
Most studies on diabetes and healing research the precocious effects of hypergylcemia on the healing process.The maintenance of endothelium and the integrity of microcirculation is very important since it is responsible for the transport and offer of nutrients and the removal of metabolites from the tissues 15 .Damage caused by hyperglycemia leads to important implications for transcapillary transport and the cirulation of endothelial products like the Von Willebrand factor and the angiotensin converter enzymes that can lead to platelet aggregation abnormalities and interfere in the inflammatory process of the healing 15 .Many of the complications of diabetes are the result of damage caused to the microcirculation vessels 24 .
High glucose concentrations activate the aldose reductate which converts glucose to sorbitol.This accumulates within the cells and leads to an osmotic movement of water to their interior, producing edemas and damaging the tissues.These elevations of pressure aggravate the lesions of the endothelial cells.The endothelial damage, repaired with an excess protein deposit in the basal membranes, added to the effect of protein kinase C-ß, which stimulates the additional synthesis of these membranes and results in damage to the diapedesis of the leukocytes and the erythrocytes.These alterations account for the delayed inflammatory response 25,26 .Furthermore, it has been shown through experiments that phagocytic capacity and the production of hydrogen peroxide are reduced in the neutrophils of diabetics 27 .
The Brazilian Diabetic Society informs us that 12% of people over the age of forty are diabetics.If we consider diabetics and pre-diabetics, this number rises to 19.8% of people aged thirty to sixty-nine 28 .Moreover, Brazilians are living longer.The average lifespan in 2006 was 72.3 years 29 .
Colon cancer is the third most frequent in the world and the second in developed countries.In Brazil, the National Cancer Institute (INCA) calculates that there will be 12,490 new cases in men in 2008 and 14,500 in women 30 .
Therefore, if we consider the growing number of elderly people and the factors of diabetes and colon cancer, there is an evident need to study the healing process in these situations.
The aim of this study is to recognize the interferences caused to the collagen synthesis in colonic anastomoses by ageing and by hyperglycemia in animals.

Methods
The project was evaluated and approved by the Ethics in Research Committee at the PUC University (PUCPR), Report number 75/07, registration number 244.The study followed the guidelines for animal research as set out by the Brazilian College of Animal Experimentation (COBEA).
The sample was made up of 138 male Wistar rats (Rattus norvegicus albinus, Rodentia mammalia), supplied by the PUCPR vivarium.Of these, 69 were between 100 and 120 days old (average age 112 days) and 69 were aged between 650 and 710 days (average age of 672 days).The weight of the young rats was 228.17 ± 20.34g and the older ones weighed 364.48 ± 20.96g.They were kept in groups of five per standard cage for the species at the Central Vivarium where the light-dark cycle was 12 hours and humidity was the same as a normal environment without any artificial regulation.The temperature was a constant 20 ± 2 o C. The animals were fed standard commercial food and had free access to water throughout the experiment.
Glycemia was confirmed by a hemoglucotest.A drop of blood was collected from the end of the tail at the beginning of the study.The animals selected to belong to the hyperglycemic groups were weighed and received streptozocin (Zanosar ® ; Upjohn CO.), 50 mg/kg of weight by endovenous injection utilizing the dorsal vein of the penis and those selected for the normoglycemic groups received a corresponding volume.New glycemia checks were made after twenty-four hours and after ninety days, prior to surgery.An anesthetic by an intramuscular injection of 0.2ml/100g of weight from a mixture of 1ml of ketamine (50mg) and 1ml of xylazine (20mg) 31 .The trychotomy of the ventral abdominal wall was done and anti-sepsis with polyvinylpyrrolidone-iodine.
The surgical procedure consisted of a median laparotomy with four centimeters.The transverse colostomy was done at 2.5cm from the peritoneal reflection, followed by termino-terminal anastomosis on a single sero-muscular layer with eight interrupted 5.0 monofilament nylon stitches.We then moved on to the laparorraphy at two levels: the first, peritoneal-muscularaponeurotic; the second, of the skin, with continuous synthesis, utilizing 4.0 monofilament nylon.
After recuperating from the anesthetic, the rats were returned to their cages, where they had free access to standard commercial food and water.Immediately after surgery they were given an intramuscular dose of sodium diclofenac at a dosage of 10 mg/kg, as an analgesic and anti-inflammatory drug 31 .
The four groups were subdivided at random into three subgroups according to the date of euthanasia on the third, seventh and fourteenth days after surgery, for which a lethal dose of intraperitoneal sodium thiopental was used at 120mg/kg 32 .
After euthanasia, another laparotomy was performed, enabling us to observe the abdominal cavity, the existence of peritonitis or abscess, adherences and the integrity of the anastomoses.
A four-centimeter segment of the colon was dries, containing anastomosis in the mid portion and a study of the resistance of anastomoses to insuflation, obtaining the pressure measurements at the time of rupture in millimeters of mercury.
After this evaluation, the surgical parts were opened lengthwise and extended on filter paper and set in 10% formalin and forwarded for histopathological study.
Seriated cuts were made with four micrometers of thickness which, placed on blades and stained by the Sirus-Red technique enabled us to evaluate the density, quality and ordering of the collagen 33 .The cuts were evaluated without prior knowledge of the meaning of the codes utilized to identify the samples.
In the Sirus-Red analysis of the cuts, the thickest and most birefringent collagen fibers had a reddish orange color (collagen I); and the thinnest and most dispersed fibers, weakly birefringent, had a greenish coloring (collagen III) 33 .The images were captured by a Sony CCD101 camera and transmitted to a Trinitron Sony color monitor, frozen and digitalized through the oculus TCX board.The images were analyzed using Image-Plus ® 4.5 for Windows ® from MediaCybernetics on a Pentium microcomputer.Five fields enlarged 200x and located on the line of the scar were analyzed in each cut.In each of the fields the percentage of area occupied by the red and yellow fibers (collagen I) and green fibers (collagen III) was calculated.Considering that the other types of collagen constitute small fraction, for practical purposes, the sum of collagens I and III was considered as being the total collagen of the scar.
For comparison of the four groups in this study, the variance analysis was used with a factor (ANOVA) and the LSD test for multiple comparisons, or the non-parametric Kruskal-Wallis test, when appropriate.The condition of normality of the data was evaluated by the Shapiro-Wilks test and the homogeneity of variances by the Levène test.Values of p<0.05 indicated statistical significance.

Results
Eight animals from the YH group and 8 from the OH group died during the ninety days prior to the surgery.One hundred and twenty-two were operated on.In the immediate post-operative period, 2 from the YH group and 2 from the OH group died.Therefore, 118 animals were left for monitoring.
All the animals induced with streptozocin developed hyperglycemia.The levels were higher after 24 hours, although they remained high until the end of the experiment (Figure 1).No anastomotic dehiscence was registered and adherences were found, with different intensities, in the anastomoses of all groups.
The anastomoses gained resistance as time passed in the four groups, without any significant difference between them (Figure 2).The values of the total collagen density readings show that there was no difference between the groups in the evaluation on the third day (p=0.987),which was also the case on the seventh day (p=0.0439).However, there was a significant difference on the fourteenth day (p<0.001).The two by two comparison at this time, showed that there was more collagen in the amastomoses of the young normal animals than in the older hyperglycemic animals FIGURE 2 -Average of bursting pressure with gain at each time (p<0.001),demonstrating the interference of age and hyperglycemia.Colonic anastomoses in the older normal animals had more collagen than in the older hyperglycemic animals (p=0.029).In this case, as the two groups compared are of older animals, the factor of interference would be age.Comparing the collagen densities between young and old hyperglycemics, we observed that the young ones had more collagen (p<0.001)(Figure 3).Analysis of the collagen fractions enabled us to verify that at the three times studied, the collagen III density was higher than that of collagen I and that the proportion between them was continually modified with the evolving healing process.There was no significant difference in the collagen III density between the groups on the third day (p=0.191),nor on the seventh day (p=0.773).On the fourteenth day, there was a difference between the groups (p=0.038).Compared two by two, we could see that in the colonic anastomoses of the young animals there was a higher density of collagen III than in the older hyperglycemic animals (p=0.022),showing the interference of age and the levels of glycemia.Comparison of the densities of this fraction of collagen between the old groups showed that the noromglycemics were higher (p=0.013).In this case the interference of glycemia is evident (Figure 4).Analysis of the collagen I fraction was lower than that of collagen III at all times, although there was a gain with the progress of the healing process.At the first two times, the values were similar (third day -p=0.100and seventh day -p=0.264).On the fourteenth day, the collagen I density in the anastomoses was different (p<0.001).For the normoglycemic groups, the colonic anastomoses of the young had more collagen I than the old (p=0.002),thus the FIGURE 4 -Gain in collagen III in the four groups interference of age.Colonic anastomoses in young normal animals had more collagen I than young hyperglycemic animals (p<0.001).In this case, this was evidence of the influence of glycemia.Comparing the colonic anastomoses of young normal animals with the old hyperglycemic, the former had more collagen I (p<0.001),strengthening the hypothesis that age and glycemia would be factors of interference (Figure 5).

Discussion
Intolerance of glucose is the most important endocrinal alteration in the elderly population, being around 45% more frequent than in the younger population 34 .
One of the major consequences of hyperglycemia is nonenzymatic glycosylation of proteins 13 such as hemoglobin, a albumin, crystalline proteins, collagen, elastin, myelinic proteins and other 14 .There is an accumulation of end products of glycosylation until this deposition is irreversible, leading to structural and functional alterations 14,35 .When these proteins are exposed to high levels of glycemia for some days, the reactions of glycosylation of formed products are still reversible, which is not the case after months of exposure 14 .Hyperglycemia leads to end products of advanced glycosylation which are strengthened by aldolases by covalent reactions of animo-reactive groups.These end products of advanced glycosylation lead to increased oxidative stress and may activate the key to the transcription factor of the nuclear factor that is related to many cellular responses linked to cytokines.Thus these products appear to induce the production of PDGF, TNF-and IL-1-36 .These products may increase cross linking of collagen, inhibiting normal degradation.
A significant alteration has been shown in the degree of collagen glycosylation and its content in injuries after five days of hyperglycemia in rats 37 .This finding implies that recently formed glycosylation products, although reversible, are harmful to the healing process.In this study, after three months of exposure to hyperglycemia, the animals probably had the end products of advanced glycosylation, a fundamental condition for evaluating the effects of hyperglycemia on collagen synthesis and the related effects of healing intestinal anastomoses.
In literature we find less activity of leucocytes with a consequent lower inflammatory response, bactericide capacity and the formulation of granulation tissue 9,11,26 .Komesu et al. 26 described in rats with diabetes induced by alloxan an inflammatory response with a slower and more prolonged beginning.They also found lower density of neutrophils in the analysis of the third day.When they examined after a longer time, they noted that at the time when they were expecting higher density of macrophages, what they did find was a larger number of neutrophils, confirming a slow and drawn out beginning of the inflammatory phase.The glycoylated hemoglobin has reduced reactivity, with 2.3-dephosphoglycerate, which affects the transport of oxygen and results in tissue hypoxia 15 .If we consider that proline and lysine hydrolyxation depend on oxygen, its lack will result in lower collagen synthesis.
Insulin deficiency in diabetics leads to a reduction in the metabolism of carbohydrates, fats and proteins which are necessary for cellular activities and protein synthesis in the healing process.Insulin is necessary for the entry of glucose in the cells and to remove the energetic need for the synthesis of proteins and to inhibit lipolysis.Glucose is the main energy supply for the fibroblasts and for the polymorphonuclears.Thus, diabetics have reduced activities in these cells and therefore in healing 38 .There is a compromise of the chemotaxis, phagocytosis and bactericide capacity 9,27 , reduced expression of shock proteins 39 , lower antioxidant production and increased production of oxygen free radicals 40,41 during the initial phase of the process, which adversely affects healing.Furthermore, there is depletion of growth factors [5][6][7][8]11,12 , increased levels of glucocorticoids 42 , loss of cellular proliferation and increased regulation of apoptosis 10,11,[43][44][45] , characteristics of the later phase of healing in diabetics, which results in poor granulation tissue formation 11 .
Franzén and Roberg 46 , studying healing in diabetic rats, showed that fibroblasts have a lower number of cytoplasmatic protrusions with lower surface density of the plasmatic membrane, which indicates the reduced motility of these cells, a condition linked to reduced healing.Velander et al. 12 , working with pigs that had diabetes induced by streptozotocin, observed a significant fall of the levels of TGF-ß and IGF-1 on the seventh day, the latter being an important stimulator of the proliferation of keratinocytes and fibroblasts.
Minimum disturbances in the synthesis balance, deposition, degradation and remodeling of the collagen may result in a deficiency in the healing of the anastomosis 47 .The collagen fibers are responsible for the integrity of the intestinal wall 48 and the organization of fibers is essential to understanding the anastomotic healing process, although the quality, in comparison with the quantity of collagen fibers, is what plays the fundamental role in maintaining the integrity of the intestinal wall during this process 49 .
Several authors have reported in experimental studies that there is lower resistance and less collagen deposition in anastomoses done in the stomach, the ileum and the colon in animals with hyperglycemia [19][20][21][22][23]50 .
Despite significant advances in the knowledge of the healing process for colonic anastomoses, there is a limited number of studies demonstrating the relationship between this healing mechanism and ageing and high levels of glycemia.
Stoop et al. 51 reported that colonic anastomoses had equal resistance in young and old animals and that the collagen content was higher in the old.Seyer-Hansen et al. 52 , studied colon anastomoses in old and diabetic rats and concluded that resistance was not reduced with age and with diabetes.Meanwhile, Onodera et al. 23 tested the integrity of colon anastomoses in diabetic rats and stated that resistance was significantly compromised with less accumulation of collagen I in the evaluation after seven days and that these parameters were the same after fourteen days.
In a previous study in our line of research we observed that the inflammatory process was extended over a longer period of time in the anastomoses of older rats, that resistance was similar and that the collagen content up to the seventh day was also resistant, becoming greater in colonic anastomoses in young rats on the fourteenth day 2 .In another study, re-epithelization was seen to occur earlier in the anastomoses of young normoglycemic rats, as was the case with neovascularization 3 .According to Spravchikov et al. 43 in high concentrations of glucose, the effects of IGF-1 in the absorption of glucose by the keratinocytes and its proliferation and differentiation would be inhibited.
In this study the previous information was confirmed that resistance was not being affected, confirming the findings described by Seyer-Hansen et al. 52 .It was possible to see that colonic anastomoses in young and old normo or hyperglycemic rats had similar collagen density up to the seventh day.However, on the fourteenth day, the densities were different.There was less collagen in the anastomoses of the old and hyperglycemic animals (p<0.001).In this case, age and hyperglycemia influenced the result.It was also seen that the anastomoses in older normoglycemic animals had more collagen than the hyperglycemic animals (p=0.029) and in this case the factor responsible for this must have been the glycemia level.
By analyzing both young and old hyperglycemic rats, a lower density of collagen was found in the anastomoses of the latter and therefore age must be affecting the process.If we consider that the fraction of collagen I was higher in the anastomoses of the young normoglycemic rats than in the older normo and hyperglycemic rats, it is possible to suppose that not only was the collagen density affected but also the scar maturation.

Conclusions
1. Resistance of anastomosis is not affected by age or by hyperglycemia; 2. Collagen density is not affected in the early stages; 3. Collagen density is affected in later stages both by age and by high levels of glycemia; 4. Both fractions of collagen are affected, especially collagen I, which means that there is a delay in scar maturation.

FIGURE 1 -
FIGURE 1 -Variation of average levels of glucose in the four groups

FIGURE 3 -
FIGURE 3 -Collagen gain in the four groups at the three times

FIGURE 5 -
FIGURE 5 -Curve showing gain in collagen I density in the anastomoses of the four groups