In this study we show that daily feeding of a relatively small amount of sucrose in non-diabetic guinea pigs exacerbated Mtb infection-associated hyperglycemia, which significantly increased TB disease severity. Most notable was not only increased lung lesion and bacterial burden but also earlier hematogenous dissemination of bacilli in sucrose-fed animals, which resulted in earlier and more severe extrapulmonary TB. These studies were initiated following a fortuitous observation showing that the use of sucrose as an oral anti-TB drug carrier in guinea pigs resulted in more severe disease compared to non-sucrose fed controls. The importance of these data is that inducing post-prandial hyperglycemia in guinea pigs by supplementing a normal chow diet with sucrose had a significant negative impact on the severity of experimental Mtb infection. These data are significant since it has been shown that hyperglycemia is a TB risk factor and diabetics with uncontrolled hyperglycemia are at greater risk for active TB compared to diabetic patients who have successfully maintained near normal blood glucose levels 
. What remains unknown however, is whether hyperglycemia in non-diabetic humans also increases the risk and severity of TB as suggested by these data.
Invariably preceding the development of overt type 2 diabetes, there is a decrease in systemic insulin sensitivity referred to as insulin resistance, which is also a feature of chronic inflammatory conditions in non-diabetics. Our results show that hyperglycemia in Mtb infected guinea pigs fed sucrose was largely driven by Mtb infection associated inflammation and sucrose feeding contributed to hyperglycemia in the early stages of infection. Therefore, these data suggest that the persistent hyperglycemia was more likely due to insulin resistance associated with chronic Mtb infection. However, sucrose feeding lengthened the duration and severity of hyperglycemia as was evident by increased clinical disease severity and the formation and accumulation of tissue and serum AGEs.
One possible explanation for how chronic Mtb infection increases blood glucose levels is through the elevation of pro-inflammatory cytokines, specifically TNF-α. TNF-α rapidly induces insulin resistance by stimulating free fatty acid (FFA) synthesis and secretion, which interferes with post insulin receptor signaling via altered phosphorylation of insulin receptor substrates 
. Our data are consistent with FFA mediated insulin resistance in Mtb infected guinea pigs since FFA levels were elevated in both the water- and sucrose-fed animals, with similar elevations present at both days 30 and 60 of infection.
Although significant glucose intolerance was not demonstrated by OGTT, significant post-prandial hyperglycemia was confirmed 60 minutes after administration of 1 g/kg of glucose administered during the test. However, even a short interval of post-prandial hyperglycemia is known to result in significant oxidative stress in humans 
. Additionally, oral supplementation of sucrose has been shown to model insulin-resistance in rats as early as 2 weeks after initiation of treatment while persistent hyperglycemia does not manifest until 9 weeks after treatment 
. In this study, a comparatively low dose of 400 mg of sucrose daily was utilized to mimic the original observation made when sucrose was used as an anti-TB drug carrier. Compared to water-fed controls, this treatment induced more severe hyperglycemia by 30 days when combined with Mtb infection. We showed, however, that combining sucrose feeding and Mtb infection did not induce diabetes since oral glucose tolerance tests did not show any evidence of elevated fasting glucose or glucose intolerance typical of type 2 diabetes.
The degree of hyperglycemia, in particular non-diabetic hyperglycemia, has not been previously investigated as a risk factor for TB in humans and experimentally in animal models. In this study, non-diabetic guinea pigs with hyperglycemia had more severe and earlier onset of extrapulmonary TB as evident by higher lesion burden in the mediastinal lymph nodes at 30 days of infection followed by an increased bacterial burden and increased severity of spleen lesions on both days 30 and 60 of infection. Immunosuppression has been implicated as the mechanism by which diabetes and other TB risk factors increase the susceptibility to Mtb infection or exacerbate active TB disease 
. While suppression of the adaptive immune response potentially explains the increased numbers of viable bacilli isolated from both lung and spleen of sucrose-fed animals on day 60 of infection there are alternative explanations. The delay in the adaptive immune response as demonstrated in diabetic mouse models of Mtb infection combined with a pro-inflammatory innate response associated with hyperglycemia could also explain our findings 
. In addition, chronic hyperglycemia and free fatty acids may represent a readily available carbon source for rapidly replicating bacilli 
The extrapulmonary spread of bacilli to the regional lymph nodes is a consistent feature of the experimental Mtb infections in animal models including the guinea pig 
. The rate of dissemination and severity of extrapulmonary lesions is an indicator of increased susceptibility of the host or increased virulence of the challenge strain of Mtb 
. Studies performed in our laboratory and others have demonstrated an increased resistance to extrapulmonary TB in guinea pigs vaccinated with BCG prior to challenge, but few studies have demonstrated a decreased resistance in this model 
. Disseminated extrapulmonary TB in humans is associated with HIV infection and the risk increases proportionally with severity of immunosuppression 
. Similarly, diabetes, a condition known to alter immune function, is associated with increased risk of developing extrapulmonary dissemination of Mtb in humans 
. In this study, we show that non-diabetic hyperglycemia had a significant negative impact on disease progression resulting in earlier and more severe dissemination of bacilli. A potential mechanism leading to more extensive Mtb dissemination may be related to microvascular damage, a direct consequence of hyperglycemia, which is central to the pathogenesis of diabetic complications 
A potential link tying together high carbohydrate diet and a maladaptive pro-inflammatory state is the increased formation of circulating and tissue-associated AGEs. The presence of increased AGEs is tied to both a pro-oxidative and pro-inflammatory state where oxidative stress mediates the persistent inflammatory response 
. In patients with poorly controlled diabetes and increased glycated hemoglobin (HbA1c), peripheral blood leukocytes express increased innate and type 1 cytokines indicative of a pro-inflammatory phenotype 
. In this study, sucrose-fed guinea pigs developed more severe pulmonary and splenic inflammation, which corresponded to increased AGE formation in the TB lesions. The specific localization and accumulation of AGEs in TB lesions even in water-fed controls implicates AGEs as a potential driving force for the pro-inflammatory state in TB as well, which is exacerbated by diabetic or non-diabetic hyperglycemia. In part, this may be due to increased oxidative stress, which we have shown to be important in the pathogenesis of TB. Interestingly, this pathogenic mechanism is further supported by our findings that treatment of Mtb infected guinea pigs with the antioxidant drug N-acetyl cysteine significantly reduces disease burden including the extent of extrapulmonary dissemination 
There are multiple potential mechanisms for AGE formation in this study. Hyperglycemia may induce AGEs via direct interactions between glucose and serum or extracellular matrix proteins. Alternatively, increased intracellular glucose under hyperglycemic conditions promotes rapid formation of dicarbonyl intermediates of glycolysis including glyoxal, 3-deoxyglucasone and methylglyoxal, all potent inducers of AGE formation 
. Accumulation of AGEs associated with chronic inflammation has been shown to occur 
and methylglyoxal 
, as well as its precursor dihydroxyacetone 
, have been identified in association with TB lesions. This demonstrates the potential for at least localized AGE formation even in euglycemic animals infected with Mtb. AGE formation associated with Mtb infection was a prominent feature in this study where AGE accumulation increased in TB lesions between day 30 and 60 of infection even in water-fed guinea pigs. Additionally, elevated serum AGEs at day 60 of infection were equally elevated in both sucrose-fed and water-fed guinea pigs. Recently AGE formation and secretion by bacteria has been shown in vitro
and may represent a third source of AGEs in Mtb infection 
. However, the significant increase in tissue AGEs, which resulted from sucrose feeding, suggests that while bacteria and local tissue metabolism may be a source for AGEs during Mtb infection, the accumulation is greatly exacerbated by hyperglycemia.
The importance of understanding the impact diabetes and hyperglycemia have on TB pathogenesis and treatment has been highlighted as a global priority 
. Aside from overt clinical diabetes, the results herein indicate the potential for hyperglycemia in a non-diabetic state to significantly worsen active TB. These data emphasize the potential impact of stringent screening of glycemic responses in TB patients at the time of diagnosis even if diagnostic criteria for diabetes are not met 
. Our data indicate that inflammation-associated insulin resistance during Mtb infection may be an additional factor contributing to hyperglycemia. Based on our data, impaired control of Mtb infection, amplification of the inflammatory response and exacerbated extrapulmonary dissemination are evident in association with hyperglycemia-mediated AGE accumulation, which may represent a central mechanism in the pathobiology of diabetes-tuberculosis comorbidity.