Serum levels of 1,5-anhydroglucitol and 1,5-anhydrofructose-derived advanced glycation end products in patients undergoing hemodialysis

1,5-Anhydroglucitol (1,5-AG) is the exclusive metabolite of 1,5-anhydrofructose (1,5-AF) in humans. It is a reduction product of 1,5-AF, which is continuously produced by ligases derived from glycogen in the liver and other tissues, such as muscle and kidney [1]. The production rate of 1,5-AF is thought to be quite low because serum levels of 1,5-AG have been found to be as low as < 25 μg/mL in healthy individuals, which equates to about one-fortieth of the measured blood glucose level [2, 3]. Circulating 1,5-AG is usually excreted by the kidneys and is reabsorbed via sodium-glucose co-transporter 4 (SGLT4) in the renal proximal tubules [4, 5]. This cotransporter mainly reabsorbs glucose excreted by the glomeruli but can also reabsorb 1,5-AG. Thus, in people with urinary glucose levels exceeding the physiological range, SGLT4 is unable to reabsorb 1,5-AG, which results in decreased serum levels. Considering the physiological function of SGLT4, the serum level of 1,5-AG has been used as a predictive marker of the hyperglycemic state [6‐8].

Similarly to 1,5-AG, uric acid (UA) is mostly reabsorbed in the proximal tubules; albeit via another transporter, the urate transporter. A correlation between the serum levels of UA and 1,5-AG has been reported in diabetic patients with early-stage chronic kidney disease (CKD) [9, 10]. In patients with advanced CKD, serum UA levels occasionally increase with disease progression due to a decreased glomerular filtration rate (GFR). Accordingly, patients undergoing hemodialysis (HD) who have lost both excretory and resorptive renal functions may show accumulations of 1,5-AG and UA, since the dialysances of both is almost equivalent. Nevertheless, previous studies have reported that serum 1,5-AG levels are low in HD patients [11, 12]. The discrepancy between serum levels of UA and 1,5-AG is supposedly related to differences in oral intake and metabolic production during inter-dialysis days. However, the metabolic production of 1,5-AG has not yet been adequately studied. Furthermore, the mechanisms by which serum 1,5-AG levels are decreased in HD patients have not been fully studied. In addition, recent clinical studies have indicated a close association between low serum 1,5-AG levels and the development of cardiovascular disease in non-diabetic subjects [2, 13, 14].

Therefore, we measured serum 1,5-AG levels in 78 patients undergoing HD. In 29 HD patients, changes in serum 1,5-AG levels were also calculated during inter-dialysis days in order to determine the metabolic production of 1,5-AG. Furthermore, we measured serum levels of 1,5-AF-derived advanced glycation end products (AF-AGEs), which are novel, exclusive byproducts of 1,5-AF, such as 1,5-AG, using a specific antibody for AF-AGEs [1]. This study was the first to investigate AF-AGEs in a clinical setting in patients before and after HD, as well as in patients with diabetes mellitus.

This clinical study involved two protocols: one with HD on the first dialysis day of the week and another with HD on the second and third dialysis days of the week. The latter represents a protocol with an inter-dialysis day between two consecutive HD treatments.

1,5-AF and 1,5-AG are metabolites of glycogen in the AF pathway, which is a minor pathway compared with the Embden-Meyerhof pathway (Fig. 8). The serum level of 1,5-AG is a useful biomarker for the hyperglycemic state and has been reported to be low in diabetic patients, patients undergoing HD, and the elderly [6, 11‐13]. However, our study unexpectedly showed a nadir value for the serum 1,5-AG level that was clearly lower than what has previously been reported in patients undergoing HD [11, 12].

1,5-AG is considered an exclusive end product of the AF pathway in humans. Therefore, since renal tubular functions are completely disrupted in patients undergoing maintenance HD, the serum 1,5-AG level in HD patients should depend on the production of 1,5-AF, elimination by HD, and food intake. 1,5-AG can move freely through vascular and cellular walls and is distributed evenly, not only in the intravascular and extravascular spaces but also in the intracellular space [15]. Although the 1,5-AG reduction rate with HD has reportedly been as high as > 60% [12], we could not estimate the net amount of 1,5-AG eliminated by HD in this study because of undetectable post-HD serum values. However, in patients with consecutive HD treatments (the second and third dialysis days of the week), the pre-HD level of serum 1,5-AG was nearly the same at two consecutive HD sessions, suggesting that the 1,5-AG eliminated by HD was mostly recovered by metabolic production and oral intake over the interval days; that is, the inter-dialysis days. Assuming that is the case, we can estimate that the daily 1,5-AG production was 15–25 mg/day depending on the body fluid volume, which is much more than the 0.4–0.5 mg/day that was reported by Yamanouchi et al. [16].

Although many foods contain small amounts of 1,5-AG, the contribution of dietary 1,5-AG to the body pool of 1,5-AG is not more than 5 mg/day [16]. Therefore, the recovery of serum 1,5-AG levels during inter-dialysis days likely results from the flow of 1,5-AG from the extravascular pool into the vascular space rather than de novo metabolic production.

1,5-AF is converted exclusively to 1,5-AG by an AF-specific reductase, which requires the coenzyme NADPH, a derivative of niacin [17]. Serum levels of niacin in this study were maintained within the normal range (5.0 ± 0.7 μg/mL), despite abnormally low levels of 1,5-AG. Additionally, serum niacin levels were correlated with neither serum 1,5-AG nor AF-AGE levels. Thus, factors other than niacin depletion are likely to play a role in the 1,5-AG depletion in HD patients.

Recently, an AGE molecule derived from 1,5-AF has been observed in human serum [1]. AF-AGEs are thought to be novel metabolites derived from 1,5-AF, such as 1,5-AG. In body fluids, the enol and keto forms of 1,5-AF exist in equilibrium, and the latter is more likely to form glycation end products with proteins rather than glucose or fructose [1]. The 1,5-AF molecules involved in the formation of AGEs fail to undergo reduction in vivo. We found higher levels of serum AF-AGEs in HD patients than in healthy controls. Moreover, as shown in Additional File 1, the kinetic ratio of both reactions rapidly increased, and the serum 1,5-AG level decreased, which indicates accelerated production of AF-AGEs in HD patients.

Generally, early glycation products, such as Schiff and Amadori compounds, may be produced in amounts several times greater and more rapidly than AGEs. Accordingly, considerable amounts of 1,5-AF molecules are likely involved in the formation of AF-AGEs, avoiding reduction to 1,5-AG. In addition, the 1,5-AF reductase reaction may be slow because of a Km value as high as 1.02 mM [17], which is much greater than the serum 1,5-AG level in humans [18]. As shown in our previous report, 1,5-AF is either metabolized to 1,5-AG or undergoes glycation by intracellular proteins (Fig. 8) [1]. Thus, 1,5-AG production in the body may depend on the kinetics of both reactions, although we could not estimate Km values in this study because of the lack of serum concentrations of 1,5-AF. However, the increased level of AF-AGEs indicates an acceleration of the glycation pathway in HD patients, suggesting that a considerable amount of 1,5-AF is involved in glycation with cellular proteins, leading to decreased 1,5-AG production.

The significant correlation between 1,5-AG and AF-AGEs is a logical finding because both are end products derived from 1,5-AF (Fig. 4). However, the strong reciprocal correlation between the AF-AGEs/1,5-AG ratio and the 1,5-AG level indicates that the glycation process leading to AF-AGE formation results in a decrease in the reduction reaction to 1,5-AG. Therefore, this study strongly suggests that the glycation process of 1,5-AF is one mechanism by which the serum 1,5-AG level is decreased in this patient population. AF-AGEs are undialyzable proteins that elevate serum levels of AF-AGEs and augment the kinetic ratio of AF-AGEs/1,5-AG in HD patients (Fig. 5). However, it may also be worth calculating the AF-AGEs/1,5-AG ratio in other clinical settings in which patients have low serum levels of 1,5-AG.

In a previous study excluding patients with advanced renal dysfunction, it was reported that diagnosing diabetes mellitus based on serum 1,5-AG levels had an inferior ROC compared with fasting blood sugars and that age was associated with serum 1,5-AG levels [19]. In our study involving patients with end-stage renal disease, there were no correlations between the presence or absence of diabetes mellitus or age and the serum 1,5-AG level (P = 0.761, P = 0.610, respectively). In the dialysis patients, it was inferred that a large amount of 1,5-AG was removed, and the loss of 1,5-AG in the body pool influenced the lack of correlation.

Several recent clinical studies have demonstrated that low serum 1,5-AG levels are a strong indicator of cardiovascular disease in patients not undergoing HD, suggesting that 1,5-AG is involved in the development of this disorder [2, 13, 14, 20]. Nevertheless, the mechanisms by which serum 1,5-AG levels are decreased other than poor renal reabsorption have not been investigated.

In healthy individuals, the daily loss of body 1,5-AG by glycation can be recovered by renal reabsorption of 1,5-AG derived from foods; however, this mechanism is lost in HD patients. Accordingly, even if the daily loss of 1,5-AG in the body pool due to the glycation process is negligible, the amount lost over the years could be clinically non-negligible. Therefore, we believe that both prevention of excessive glycation and supplementation of 1,5-AG are important for maintaining normal serum 1,5-AG levels in HD patients.

Serum 1,5-AG levels were quite low in patients on regular HD, suggesting a significant reduction of 1,5-AG in the body serum pool of HD patients. On the other hand, serum levels of AF-AGEs, a glycation end product derived from 1,5-AF, were significantly increased in HD patients, suggesting a potential mechanism underlying the small pool size of 1,5-AG in HD patients. In addition, the serum AF-AGEs/1,5-AG ratio was extremely elevated in HD patients. Therefore, we believe that the AF-AGEs/1,5-AG ratio should be measured in other clinical settings in which patients have low serum levels of 1,5-AG.