The primary finding of our study is that CD33 cell surface protein and mRNA expression levels are significantly reduced in monocytes from patients with type 2 diabetes. The analysis of the plasma cytokine profile of patients with type 2 diabetes showed that pro-inflammatory cytokine levels were increased, although only the levels of IL-8, IL-12, and TNF-α were significantly increased when compared to healthy volunteers. These results suggest that the elevated levels of pro-inflammatory cytokines in the serum of patients with type 2 diabetes could be related to the down-regulated expression of CD33. Increased levels of pro-inflammatory cytokines in the plasma of patients with type 2 diabetes have been widely reported [5, 9, 33]. On the other hand, spontaneous production of TNF-α, IL-6, and IL-8 was observed in human monocytes treated with anti-CD33 or by decreasing CD33 surface expression by RNA interference . Although diverse mechanisms have been proposed to explain the increase of inflammatory cytokines in patients with type 2 diabetes, to our knowledge, this is the first study to describe an association between CD33 and the inflammatory cytokine profile in type 2 diabetes patients.
This study also demonstrated that high glucose concentrations in vitro resulted in decreased expression of CD33 protein and mRNA in human monocytes from healthy donors.
In addition, we observed a significant increase in levels of TNF-α present in the supernatants of monocytes cultured under high glucose conditions (50 mmol/l D-glucose), although this increase was not observed for other cytokines. Increased levels of TNF-α mRNA from monocytes cultured with 33 mmol/l glucose have previously been described for healthy individuals . In addition, we did not detect an effect of high glucose concentrations on the production of IL-1β or IL-6 in monocytes. However, previous studies have reported inconsistent results; some have shown that hyperglycemia increases the production of IL-1β and IL-6 in the THP-1 cell line, whereas others have shown that hyperglycemia only induces the production of IL-6 in primary human monocytes [20, 34]. In addition, other studies have demonstrated reduced IL-1β levels in RAW264 murine macrophages exposed to 8-20 mmol/l D-glucose . These differences may be related to differences in the cell types, glucose concentrations or lengths of culture time used to measure cytokine levels. Other authors reported that high glucose concentration and LPS treatment act synergistically for stimulate the secretion of inflammatory cytokines in peripheral mononuclear cells from humans [36, 37]. Consistent with the formerly mentioned data from other authors, our results presented here serve to extend previous knowledge of the role of high glucose concentrations on the promotion of inflammation by demonstrating the in vitro effects of high glucose concentrations on TNF-α production by human monocytes.
We also presented evidence demonstrating that high glucose concentrations in vitro could increase the proportion of CD33low monocytes and reduce the proportion of CD33high monocytes, although TNF-α production was increased in both cell populations. These results support the idea that hyperglycemia leads to an increase in TNF-α production through a CD33-mediated mechanism, although there are likely additional mechanisms involved in production of TNF-α that are beyond the scope of this study. Furthermore, the spontaneous production of IL-6 by CD33-/lowplasmacytoid dendritic cells from patients with diabetes without atherosclerotic complications has been reported . These findings suggest that the increased production of pro-inflammatory cytokines in patients with type 2 diabetes may be partially associated with the subpopulation of CD33low monocytes.
The precise mechanisms by which hyperglycemia down-regulates CD33 expression have not yet been elucidated, although the generation of ROS by high glucose concentrations is believed to contribute to hyperglycemia-induced inflammatory responses [19, 29, 30]. Thus, we explored the association between ROS generation and CD33 expression in monocytes cultured under high glucose concentrations and treated with α-tocopherol. The results showed that α-tocopherol decreased ROS generation and prevented the effect of high glucose on CD33 expression. This result supports the idea that the oxidative stress generated by high glucose concentrations contributes to the down-regulation of CD33.
We observed an inhibition of TNF-α production in monocytes that were cultured under conditions of high glucose and were treated with α-tocopherol. This result indicates that ROS generation is involved in the TNF-α production by human monocytes cultured under high glucose conditions. Thus, the low expression of CD33 and the inhibition of TNF-α production in monocytes cultured under high glucose concentrations are primarily related to ROS generation. Therefore, we propose that ROS generation induced by high glucose conditions directly induces the down-regulation of CD33 expression. Alternatively, ROS generation could induce the production of pro-inflammatory cytokines that could then regulate the expression of CD33. A study by Shamsasenjan et al. postulated that IL-6 down-regulates CD33 expression in myeloma cells . However, in the current study, IL-6 production was not increased in the supernatants of monocytes cultured under high glucose conditions, and therefore, IL-6 is likely not the mechanism responsible for CD33 regulation under hyperglycemic conditions.
In this study, we also showed that high glucose concentrations could up-regulate the expression of SOCS3 mRNA in human monocytes, suggesting that this molecule may regulate the levels of monocyte CD33 expression. This hypothesis is consistent with results showing that SOCS3 could contribute to CD33 degradation in peripheral monocytes . Recently, it was reported that glucose ingestion induces the over-expression of SOCS3 in peripheral monocytes [31, 32, 39]. Interestingly, SOCS3 expression is induced by TNF-α and could therefore represent a feedback mechanism for inflammation associated with CD33 regulation [40, 41]. However, further studies are needed to assess whether TNF-α production regulates SOCS3 expression and its effect on CD33 expression.
Our study had limitations, the most critical of which was the limited ability of our in vitro model to recapitulate what occurs in patients with type 2 diabetes. Nonetheless, we demonstrated that a significant increase in TNF-α production and decrease in CD33 protein and mRNA expression were induced by high concentrations of glucose (30-50 mmol/l). A concentration of 50 mmol/l is equivalent to 900 mg/dl of blood glucose, which is a concentration that is rarely attained in type 2 diabetes patients is much greater than the mean value found in the diabetes patients included in this study (265.3 ± 79.71 mg/dl or 14.73 ± 4.42 mmol/l blood glucose). Hence, it is possible that the glucose sensitivity of TNF-α production associated with CD33 expression is greater in vivo than in vitro. However, this increased level of sensitivity may occur if other factors in vivo could potentiate glucose-induced ROS generation. Further studies are required to examine this possibility. The increase in TNF-α associated with the down-regulation of CD33 expression presented here constitutes an interesting in vitro model to further investigate the molecular processes involved in the modulation of inflammation by glucose.