This study is the first time for exploring the relationship between gut microbiota and its metabolite PAGln in stroke patients with T2D. Our data suggest that: (i) in the IS-T2D group, the gut microbiota was significantly imbalanced and the plasma PAGln levels increased partly caused by the disorder of gut microbiota, which was further confirmed by animal studies (Fig. 6); (ii) elevated PAGln levels were associated with poor functional outcomes and plasma NETs levels; (iii) PAGln levels, gut microbiota, and NETs levels could be used as combined diagnostic indicators for IS-T2D.
Previous research has established that PAGln was associated with CHD, peripheral artery disease, heart failure, and other cardiovascular diseases17, 38–40. In addition, PAGln was also closely related to obesity, diabetes, prediabetes, and other metabolic diseases16, 34, 41–43. Research shows that metabolic disorders such as insulin resistance, dyslipidemia, and fatty can increase platelet activity and aggregation via dysregulation of the NO-mediated signaling pathway, leading to thrombosis and atherosclerotic lesion formation44. And PAGln also has the effect of driving platelet invasiveness14. A metabolic disorder like diabetes mellitus may promote vascular injury via PAGln-mediated molecular mechanisms. Our results showed that IS-NT2D caused an increase in plasma PAGln levels. More importantly, PAGln levels were higher in IS patients with the complication of T2D. Although the plasma clearance of PAGln is closely related to renal function, that is, if renal function is impaired, the plasma PAGln clearance will be severely reduced, there was no significant difference in BUN or Scr levels between IS-T2D and IS-NT2D groups, indicating that the elevated PAGln levels in IS-T2D patients were mainly influenced by T2D38, 45. Except for stroke, Tang et al. found that plasma PAGln levels were significantly higher in patients with heart failure and diabetes mellitus than in those with heart failure alone, which suggested the elevation of plasma PAGln is associated with diabetes mellitus40. Our results also revealed that elevated PAGln was an independent risk factor for the patient with stroke and T2D, indicating that PAGln may be an important molecule of T2D in contributing to an exacerbation in stroke injury, of which the mechanism remains obscure and needs to be further explored.
PAGln is a metabolite of phenylalanine degradation by gut microbiota. Previous studies have revealed that, except for abnormal metabolism, people with diabetes mellitus had severely disturbed gut microbiota11, 46. Therefore, increased plasma PAGln levels in IS-T2D patients indicate that they might not only use more amino acids as energy sources, but have more gut microbiota to degrade phenylalanine compared with stroke patients, which was confirmed by our results. We found that the gut microbiota of IS-T2D patients was disordered, and there was more PAGln-related gut microbiota in these patients, including o_Enterobacterales, f_Enterobacteriaceae, p_Verrucomicrobiota, c_Verrucomicrobiota, s_Klebsiella_pneumoniae, and g_Klebsiella. Enterobacteriaceae is recognized as harmful gut microorganisms and showed a growth advantage in both stroke and T2D18, 47. A study has reported that Enterobacteriaceae was closely related to the mortality of stroke patients, increasing inflammatory factors such as TNF-α and IL-1β through the LPS-TLR4 pathway, thus accelerating systemic inflammation and exacerbating cerebral infarction18. In addition, studies have shown that Escherichia coli belonging to Enterobacteriaceae can catabolize aromatic compounds such as phenylalanine and phenylacetic acid48–50. The phenylacetic acid is the middle product that PHE is metabolized to PAGln. Our results showed that Enterobacteriaceae was one of the major differential bacteria in the IS-T2D group, which was positively correlated with PAGln, indicating that Enterobacteriaceae might cause the aggravation of brain injury by diabetes mellitus via metabolites PAGln. Consistent with Ottosson et al., a positive correlation was also found between plasma PAGln levels and Verrucomicrobiota in a CHD risk cohort17. In our study, c_Verrucomicrobiota was dominated by g_Akkermansia (99.4%). It is generally believed that Akkermansia is a kind of beneficial bacteria to maintain the health of the intestinal epithelium, which can enhance intestinal barrier function, produce short-chain fatty acids (SCFAs), improve metabolic disorders, and it has even been considered the new probiotics which could be developed and utilized51, 52. However, its role in the development of stroke and diabetes mellitus remains elusive. Studies have reported that Akkermansia was elevated in patients with prediabetes and diabetes as well as in diabetic animal models53, 54. In addition, it has been suggested that a significant increase in Akkermansia could contribute to the erosion of the mucous layer and enhancement of pathogen filtration in the intestinal epithelial layer when the dietary fiber was lacking in the host gut55. The latest study found that Akkermansia strains could be divided into five distinct candidate species with different host preferences and functions, of which only Akkermansia mucinophilus had the health-associated properties referring to being associated with low body mass index in the host56. In our study, Akkermansia was significantly decreased in IS-NT2D patients but increased in IS-T2D patients, and c_Verrucomicrobiota, dominated by Akkermansia, was significantly associated with PAGln. We speculated that T2D could increase a harmful species of Akkermansia, which is linked with PAGln metabolism. Therefore, it is necessary to further explore the relationship between each candidate species of Akkermansia and PAGln. Consistent with the study of Li et al., our results suggested that PAGln was significantly correlated with Klebsiella53. Klebsiella is widely believed to be a pathogenic bacterium and a strain with the TMA gene57–59. TMA is an essential precursor of the gut microbiota metabolite TMAO, which has been found to promote thrombosis via increasing platelet reactivity60. And TMAO is closely associated with atherosclerosis and poor prognosis in IS61, 62. Therefore, it is reasonable to hypothesize that Klebsiella is a strain with the characteristics of promoting thrombosis. In addition to the above bacteria, previous research also reported that PAGln was associated with Enterocoocus, Streptocoocus, Escherichia_ Shigella, G_f_o_SHA.98c_clostridia, and Eggerthella lenta17, 53, 63. And finally, we transplanted the fecal microbiota from patients with IS-T2D into rodents, ultimately resulting in increased plasma PAGln levels, which provided direct evidence that elevated PAGln levels in IS-T2D patients were partly caused by intestinal microorganism disorders. It is noteworthy that the PAGln level of rats receiving the fecal microbes from IS-T2D patients further increased after stroke. A possible explanation for this might be that under the acute stress of stroke, this part of microbiota further grew and multiplied, resulting in an increase of PAGln levels. Additionally, our study indicated that gut bacteria functions such as amino acid metabolism were enhanced and abnormal in patients with IS-T2D. Further work should be undertaken to search for strains related to PAGln metabolism and improve the prognosis of IS-T2D by targeting PAGln-producing bacteria.
The formation of NETs is originally identified as an important antimicrobial phenomenon, including resisting microbial invasion, stopping the microbial spread, and killing pathogens64. However, mounting evidence indicates that NETs play a negative role in many diseases such as endocrine diseases, nervous system diseases, and respiratory diseases65–67. Consistent with previous research, our results showed an increase in NETs levels in stroke patients and a further increase in stroke patients with T2D. Platelet-neutrophil interactions are a critical pathophysiological process of thromboinflammation68. In the arterial microenvironment of thrombosis, activated platelets trigger NE to form and release NETs containing thrombotic tissue factor (TF) through several mechanisms, including NE autophagy induced by the presentation of high mobility group frame 1 (HMGB1) protein, release of the NETs inducer–inorganic polyphosphate (polyP), and activation of the platelet Toll-like receptor 4(TLR 4)69–71. In addition, NETs, in turn, can further aggravate thromboinflammation via promoting thrombin and fibrin formation and binding to platelet-derived microparticles (PMPs) and coagulation factors72. Recently, a study had indicated the effects of the gut microbiota metabolite PAGln in driving platelet invasiveness and thrombosis via β adrenergic receptors14. According to our data, plasma PAGln levels were positively correlated with NETs, and NETs showed a dose-dependent increase according to PAGln levels. Therefore, we speculated that the gut microbiota metabolite PAGln might contribute to platelet-neutrophil interactions via enhancing platelet reactivity, which ultimately promotes the formation of NETs. Meanwhile, our data showed a positive correlation between PAGln and IS-T2D-associated gut microbiota, of which were positively correlated with NE. Dysregulation of gut microbiota in patients with arteriosclerotic cerebral small vessel disease has been reported to independently enhance the proinflammatory property of NE73. We thus inferred that IS-T2D-associated gut microbiota could exacerbate inflammation status of NE, and then combine with elevated PAGln in plasma to promote the formation of NETs, which may ultimately contribute to the aggravation of thromboinflammatory. In the meantime, our data suggest that elevated plasma PAGln levels were correlated with a 90-day poor prognosis in stroke patients, further supporting that the PAGln plays a negative role in the development of stroke. Above all, our study revealed that the presence of T2D could increase the inflammatory reaction in IS patients, mediated in part by the altered composition of gut microbiota and the effects of bacterial metabolite PAGln.
In this study, we not only confirmed the dysbiosis of gut microbiota and the increase of harmful molecular PAGln in IS-T2D patients, but also found that the gut microbiota dysbiosis, the increased plasma PAGln, and NETs could serve as important diagnostic markers for stroke with T2D.
Limitations of the study: Firstly, the stroke patients included in our study were within 2 weeks of onset, which was enough to cause significant differences in microbiota between samples in group18. In addition, the NIHSS scores of the patients included were low, limiting the representativeness of the study. Secondly, our sample size was relatively small, which needs to be further expanded. Meanwhile, we need to exclude confounding factors like dietary habits, lifestyle, and fecal status, and increase the T2D group in the future, to enhance persuasion of the experimental results. Finally, in the animal experiment, we only conducted a simple FMT experiment, and the transplanted fecal samples included bacteria, fungi, and other microorganisms, which is hard to determine which strain plays a role. Further experimental studies need to be carried out in the future.