| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Institute of Pathology, Case Western Reserve University, Cleveland, Ohio
b Department of Physiology, The University of Texas Health Science Center at San Antonio
David R. Sell, Institute of Pathology, Case Western Reserve University, 2085 Adelbert Road, Cleveland, OH 44106 E-mail: drs7{at}po.cwru.edu.
Decision Editor: John Faulkner, PhD
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
,ß-dicarbonyl compounds that can be trapped by aminoguanidine do not play a major role in collagen aging in the rat. Instead, post-Amadori pathways involving oxidative or nonoxidative fragmentation of the Amadori product emerge as the more likely mechanism of collagen cross-linking in aging. CERTAIN aspects of diabetic complications resemble accelerated aging, such as increased rate of cardiovascular disease, increased incidence of cataract formation, and limitations in joint mobility. Because these sequelae occur in the population as a whole during aging, a pragmatic interest is manifest in the discovery and development of pharmaceutical interventions for the prevention or treatment of age-related diseases and disabilities. One such intervention proposed and presently under scrutiny is the use of aminoguanidine (1).
Many studies have shown beneficial effects of aminoguanidine (AG) use in ameliorating or preventing complications caused by experimental diabetes and aging. At the organ level, these have included the inhibition of experimental diabetic retinopathy (2), nephropathy (3), and neuropathy (4). At the tissue level, ameliorations of diabetes-induced glomerular and retinal basement membrane thickening (5), motor nerve conduction velocity (6), retinal pericyte loss (2), cataract formation (7), and a host of vascular and collagen changes (8)(9)(10) have been observed. However, the potential usefulness of AG goes beyond the treatment of diabetes. In animal studies, AG use is able to either prevent or retard age-related diseases such as glomerular sclerosis (11) and cardiovascular disease, including arterial stiffening (12), atherogenesis (13), and cardiac hypertrophy (11)(12). It is also able to ameliorate cerebral damage from experimentally induced arterial occlusion as a model of stroke (14).
Many hypotheses have been proposed to explain the mechanisms and mode of action of AG. It has been found to inhibit various enzymes such as nitric oxide synthase (15), aldose reductase (16), diamine oxidase (17), and semicarbazide-sensitive amine oxidase (18). Alternatively, a hypothesis that bears significance for the present study has to do with its ability to block the Maillard reaction in the formation of advanced glycosylation end products (AGEs), thus preventing nitric oxide quenching (19) and collagen cross-linking by AGEs (1).
In the Maillard reaction (20), AG has been proposed to react with the Amadori product, thus blocking further reaction (21). However, this hypothesis has not been substantiated (5)(22)(23)(24). A more likely mechanism has to do with the ability of AG to trap highly reactive carbonyl intermediates of the Maillard pathways as triazine compounds (25)(26)(27). More recent in vitro studies have shown that AG inhibits only AGEs formed through ,ß-dicarbonyl fragments originating from these pathways, but it may have little effect on AGEs formed through intermediates of post-Amadori pathways where oxidative decomposition and fragmentation may not occur (26)(28). Examples of ,ß-dicarbonyl compounds trapped by AG include glyoxal, methylglyoxal, and 3-deoxyglucosone (26)(27)(29).
Glyoxal, methylglyoxal, and 3-deoxyglucosone (3-DG) are potent protein cross-linking agents (20)(26)(27)(30)(31). In addition, glyoxal and 3-DG are precursors of glycoxidation products, in particular carboxymethylysine (CML) and pentosidine, respectively (26)(32)(33). Their formation rate is accelerated in diabetes (34)(35) and is expected to increase with advancing age (34)(35)(36)(37) as a result of the age-related decrease in glucose tolerance that has been observed both in humans and rodents (38)(39). Conversely, because caloric restriction is associated with a decrease in mean glycemia (40)(41), one would expect their plasma levels and therefore the age-related tissue AGE product formation to decrease in calorically restricted rodents. The latter association was already reported for the glycoxidation products CML and pentosidine (37)(42).
Here, we have tested the hypothesis that chronic treatment with AG from 6 to 24 months of age should prevent the age-related collagen cross-linking and glycoxdiation as reflected by the tendon break time (TBT), pentosidine, and CML assays. If indeed the ,ß-dicarbonyl compounds that are being trapped by AG play an important role in collagen aging, one would then expect AG to inhibit collagen aging as previously reported in diabetic rats (43).
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
Tissue and Plasma Collection Procedures
When the rats were ages 10, 12, and 19 months, their blood was collected in heparinized tubes from a nick in their tail veins. As a way to avoid stress effects on glucose levels, samples were obtained within 1 minute of initial cage disturbance. Six rats from each group at ages 12 and 19 months, and the remaining rats at 24 months, were killed by CO2 narcosis. Tissues consisting of abdominal skin and whole tails were immediately excised. Skin was frozen in small plastic bags at -70°C. Tendon bundles were dissected from individual tails and similarly stored frozen in 1-ml microcentrifuge tubes containing saline.
Measurements of Plasma Glucose
Plasma was prepared by centrifugation of blood samples for 20 minutes at 4°C and then stored frozen at -70°C. Glucose levels were measured by the glucose oxidase method as previously described by Masoro and colleagues (40).
Measurement of TBT, Furosine, Pentosidine, and CML
Skin and tendon samples were shipped to Cleveland on dry ice and stored frozen at -70°C. TBT was determined as previously detailed (45). Samples of tendon and skin were also extracted and processed by methods described elsewhere (45). Pentosidine and furosine were determined in acid hydrolysates by high-performance liquid chromatography (HPLC) as described earlier (37)(45)(46). Similarly, CML was determined by HPLC with a two-step injection technique with postcolumn derivatization using phthalaldehyde reagent (OPA) as described elsewhere (47).
Statistics
Homogeneity of variance among treatments and normality within treatment groups were tested at = .05 by the Burr-Foster Q Test and the Wilk-Shapiro W Test adjusted for unequal sample sizes, respectively (48)(49). Data were transformed by either the square root or logarithmic transformation for statistical comparison of treatments (49). The outlier test that was conducted at = .05 and the test for differences in slopes were made by methods described by Snedecor and Cochran (50). Pearson correlations, regression analysis and equations, and analysis of variance (ANOVA) were computed by SPSS (Chicago, IL). The comparison of means in Fig. 2 was by the Newman-Keuls/SNK test (49).
|
|
|
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Because water consumption began to increase in both control and treated rats after the age of 20 months (Fig. 1), but food consumption showed opposing trends during this time (Fig. 1), the data of Fig. 1 and Fig. 1 were reexpressed as water consumed per food intake as shown in Fig. 1. This reexpression was necessary to show that the age-related increase in water intakes was not simply due to increased food intakes occurring in control rats during this time. Interestingly, differences between the control versus treated rats now appeared after the age of 18 months (Fig. 1). However, the results were still the same. An ANOVA showed significant (p < .0001) main effects for both age and treatment. Likewise, a slope analysis showed that the age-related increase occurred more rapidly (Fig. 1) and significantly (p < .0001) in the control versus treated rats (i.e., slope = 0.077 vs 0.064).
As a preliminary to the assessment of glycation and glycoxidation, glucose was measured at ages 10, 12, and 19 months in the plasma of these rats (Fig. 2). As previously explained, great care was taken not to stress the rats at the time of bleeding. Results showed significant effects for age (p < .0001) and treatment (p = .011). Results from a multiple comparison test (SNK) showed that glucose was moderately but significantly (p < .01) elevated in control versus AG-treated rats at ages 10 and 19 months, but not at 12 months (Fig. 2).
Collagen glycation (furosine), glycoxidation (pentosidine, CML), and cross-linking (TBT) were determined in skin and tendon (Fig. 3 and Fig. 4). Pearson correlation analyses (data not shown) showed that parameters measured within the same rat were highly correlated with each other. In this analysis, the correlation coefficients were found to be significant (p 
.002) for all comparisons (.45 < r < .86). TBT correlated most strongly with tendon CML (r = .75, p < .0001) and least with skin furosine (r = .47, p = .002). In turn, furosine was found to be most highly correlated with the pentosidine level of the same tissue. Thus, tendon furosine was strongly correlated with tendon pentosidine (r = .85, p < .0001), and skin furosine with skin pentosidine (r = .83, p < .0001). Skin and tendon pentosidine were highly correlated with skin (r = .84, p < .0001) and tendon CML (r = .79, p < .0001), respectively, as well as with each other (r = .84, p < .0001).
Parameters of Fig. 2 and Fig. 3 were separately analyzed by a two-way ANOVA consisting of age and AG treatment. As explained in the Methods, five data points have been excluded from these analyses because of nonnormality. The results showed that the effect caused by AG was not significant (p > .05) for all parameters, although that for TBT approached statistical significance (p = .077). However, as expected, all seven parameters were consistently and highly significantly (p < .0001) increased by age (Fig. 3 and Fig. 4).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Previously, it has been shown that AG is able to inhibit experimental diabetes-induced tail tendon cross-linking and autofluorescence of collagen in rodents (1)(2)(8)(9)(51). In the present research, we wanted to determine whether chronic AG treatment would affect the age-related increase of parameters previously measured in our laboratory for collagen cross-linking, glycation, and glycoxidation (37)(45)(46). A strong rationale for this study was the fact that we previously demonstrated an inverse relationship between longevity and pentosidine formation rate in skin of several mammalian species (37). Thus, the availability of a drug that would block its formation might be helpful in subsequent studies to investigate whether the processes determining longevity and tissue glycoxidation and cross-linking are linked.
The results of this study show that AG treatment of F344 rats from 6 to 24 months of age decreased the variation noted within parameters of untreated control rats somewhat. This has also been observed in another study investigating the effects of AG on glycation and glomerular capillary basement membrane thickness in birds (54). It was also found in the present study that AG had a marginal effect upon inhibiting the age-related increase of TBT, but without an effect on glycoxidative parameters measured in skin and tendon. Furthermore, even though there was a small but consistent difference between control and treated groups in TBT at each measured time point shown in Fig. 3, the inhibitory effect of AG upon TBT was not progressive (Fig. 3) and the difference as a whole only approached statistical significance by an ANOVA (p = .077).
As observed in a previous study with rodents (46), furosine progressively and dramatically increased with age at almost the same rate as that noted for pentosidine and CML (Fig. 4). Interestingly, furosine is a measure of the Amadori product specific for glucose (46), whereas both pentosidine and CML are measures of AGEs and can orginate from a variety of sugars (26)(37)(45)(46). In addition, CML can originate from dicarbonyl intermediates; that is, it can originate from glyoxal and glycolaldehyde, from the degradation of lipid (33), and the metabolism of serine by myeloperoxidase (55). Furthermore, the finding in the present study that furosine most highly correlated with the pentosidine level of the same tissue suggests that pentosidine originates from glycation.
Based on previous data, AG was not expected to have an effect on furosine levels because this drug reportedly does not affect the initial step of the Maillard reaction, namely glycation (5)(22)(23)(24). This was confirmed in Fig. 4. However, if AG was an efficacious anti-AGE agent, it would be expected to inhibit both tissue pentosidine and CML formation. As shown in Fig. 4, there was a trend for tendon CML to be progressively lower in the treated group at ages 12 and 19 months, but the trend did not hold true at 24 months. There was also no effect of AG treatment on skin CML (Fig. 4).
Among various reasons for the failure of AG to block cross-links and AGE formation is the administering mode and dosage of the drug. This has been a contentious issue in the human clinical trials. For example, in one study (56), a fairly high dose (1200 mg) administered orally to diabetic patients resulted in plasma levels of only 10 µg/ml. Thus, the efficacy of the drug to produce the desired effects has been questioned (57) and previously addressed in animal studies by several other investigators (10)(54)(58). However, the route and dosage used in the present study, that is, 1 g/l in drinking water, have proven adequate in many studies (8)(11)(12)(59). Regardless of the route of administration, the effect of AG on tissue glycoxidation has not always been consistent, especially in studies of experimental diabetes. For example, in the studies by Soulis-Liparota and colleagues (59), AG prevented the increase of diabetes-related collagen fluorescence in isolated glomeruli and renal tubules, but surprisingly not in the whole kidney of diabetic Sprague-Dawley (SG) rats. Likewise, in the study by Nyengaard and colleagues (10) using the same rat strain, it was found that the effects of AG on tissue glycoxidation and other diabetes-related parameters were strikingly discordant. On the one hand, it normalized aortic pentosidine, but on the other hand it surprisingly had no effect on collagen-linked fluorescence measured in aorta, kidney, and skin. Nyengaard and colleagues (10) measured AG concentrations in these tissues because of suspected dosage inadequacy. However, significant levels of this drug were detected in all tissues assayed. Thus, it may be concluded from the results of these investigators that the formation of these AGE fluorophores involved mechanisms not accessible to the action of AG, or do not involve dicarbonyl intermediates. Conversely, the ability of AG to block aortic pentosidine formation suggests that the latter may originate from a non-Amadori product precursor such as, for example, dehydroascorbic acid.
Evidence for the existence of mechanisms that are not blocked by AG in AGE formation was suggested initially by the work of Edelstein and Brownlee (22). AG was found to bind to fragments of glycated peptide, but surprisingly failed to bind to intact glycated peptide. In the one study by Glomb and Monnier (26), AG was totally ineffective in preventing the formation of CML when starting from a model Amadori product. In contrast, it was effective when glycation was started in the presence of the free sugar. Likewise, Booth and colleagues (28) reported that AG at different concentrations had negligible effects on the post-Amadori formation of AGEs from ribosylated serum albumin. Thus, these results suggest that AG is ineffective in the prevention of AGE formation in post-Amadori pathways not involving fragmentation products.
There is evidence that rodent strains may vary in their responsiveness to AG treatment. This is clearly shown in the results of Li and colleagues (11), in which AG provided to nondiabetic SG and F344 rats was able to ameliorate the age-related increases of albuminuria and proteinuria in both strains. However, it was found that F344 rats had less striking age-related structural changes in renal pathology and thus were in general less influenced by AG treatment compared with the SG strain. Interestingly, this study has comparative relevance to the present one in that both used F344 rats treated with AG at 1 g/l in their drinking water from ages 6 to 24 months. In the study by Li and colleagues (11), AG in F344 rats significantly reduced AGE formation measured by enzyme-linked immunosorbent assay in heart tissue, but had no effect on that of aorta and kidney during aging. In contrast, AG significantly reduced AGE formation in all three tissues in the SG rat. It was concluded that the SG rat was more responsive to the effects of AG because of this strain's known susceptibility to renal disease at old age. However, a more important factor in explaining strain differences may relate to differences in drug half-life, which were not studied.
In summary, we found that AG, when supplied to F344 rats in their drinking water from 6 to 24 months of age, had a modest effect on inhibiting the age-related increase in TBT, a parameter of cross-linking, and no effect on skin and tendon glycoxidation as measured by pentosidine and CML. These mechanisms most probably involve post-Amadori pathways where intermediates do not undergo decomposition and fragmentation side reactions. In support, Degenhardt and colleagues (60) reported no affect of AG's inhibiting the increase of pentosidine and CML in skin collagen of Lewis rats with streptozotocin-induced diabetes. Similarly, these investigators have also reached the same conclusion concerning the mechanism of formation of pentosidine and CML in skin. A secondary finding in the present study is that we observed an age-related increase in water intake in old rats greater than age 18 months. This also has been previously observed for F344 and SG strains of rats by Rowland and colleagues (61). The observation that AG was able to significantly ameliorate the increased water intake (Fig. 1 and Fig. 1), together with the noted age-related increase in albuminuria in this and other rat strains (11)(60), may reflect a beneficial effect of the drug on age-related nephrosis as reported by Li and colleagues (11).
Conclusions
From this study, one can conclude that AG is a poor inhibitor of the advanced Maillard reaction in collagen-rich tissues. It is likely that the strong effects of AG on diabetic nephropathy and renal immunoreactive AGE formation result from a biological reduction of AGE precursors rather than from its ability to chemically trap reactive intermediates of the Maillard reaction. Alternatively, the fact that AG did not have any effect on glycoxidation and cross-linking can be interpreted in terms of its inability to trap post-Amadori, nonoxidative intermediates of the Maillard reaction in vivo. In contrast, AG as an inhibitor of the reaction in vitro is well documented (1)(2)(20)(21)(22)(23)(24)(27)(28)(29). As previously mentioned, AG inhibits several enzymatic activities (15)(16)(17)(18), so the possibility exists that its main mechanism of action in vivo is indirectly related to the Maillard reaction.
|
Acknowledgments |
|---|
Received January 18, 2001
Accepted April 23, 2001
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
-oxoaldehydes glyoxal, methylglyoxal, and 3-deoxyglucosone under physiological conditions. Biochem Pharmacol. 60:55-65. [Medline]
-dicarbonyl compounds formed during Maillard reaction in vitro and in experimental diabetes. Diabetes. 47: (Suppl.1) A121
-(carboxymethyl)lysine, is a product of both lipid peroxidation and glycoxidation reactions. J Biol Chem. 271:9982-9986. -(carboxymethyl)lysine and N-(carboxymethyl)hydroxylysine in human skin collagen. Biochemistry. 30:1205-1210. [Medline]
-N-(2-furoylmethyl)-L-lysine (furosine) levels in rodent skin collagen. The relationship to dietary restriction and glycoxidation. Mech Ageing Dev. 95:81-99. [Medline]
This article has been cited by other articles:
 |
|
 |
|
  Z. Alikhani, M. Alikhani, C. M. Boyd, K. Nagao, P. C. Trackman, and D. T. Graves Advanced Glycation End Products Enhance Expression of Pro-apoptotic Genes and Stimulate Fibroblast Apoptosis through Cytoplasmic and Mitochondrial Pathways J. Biol. Chem., April 1, 2005; 280(13): 12087 - 12095. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. K. Bhattacharyya, M. Merz, and J. R. Thomas Modulation of Cutaneous Aging With Calorie Restriction in Fischer 344 Rats: A Histological Study Arch Facial Plast Surg, January 1, 2005; 7(1): 12 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Van Remmen, Y. Ikeno, M. Hamilton, M. Pahlavani, N. Wolf, S. R. Thorpe, N. L. Alderson, J. W. Baynes, C. J. Epstein, T.-T. Huang, et al. Life-long reduction in MnSOD activity results in increased DNA damage and higher incidence of cancer but does not accelerate aging Physiol Genomics, December 16, 2003; 16(1): 29 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Sell, M. A. Lane, M. E. Obrenovich, J. A. Mattison, A. Handy, D. K. Ingram, R. G. Cutler, G. S. Roth, and V. M. Monnier The Effect of Caloric Restriction on Glycation and Glycoxidation in Skin Collagen of Nonhuman Primates J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2003; 58(6): B508 - 516. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
, AG (error bar down).
