Ageing and longevity are one of the main concerns all over the world. In recent years, proteomics technology was widely used to identify age related plasma proteins, which can provide new clues about the mechanisms of aging process. Previous proteomic studies showed that most of age-relevant proteins were enriched in insulin-like growth factor (IGF) signaling, mitogen-activated protein kinases (MAPK), hypoxia-inducible factor 1 (HIF1), cytokine signaling, Forkhead Box O (FOXO) metabolic pathways, folate metabolism, advance glycation end products (AGE), and receptor AGE (RAGE) metabolic pathway [13]. Till now, few studies have focused on the relationship between blood glucose regulation and longevity. Regulation of blood glucose balance is part of the regulation of life activities. It is an important condition for maintaining homeostasis. When the body's blood glucose regulation is out of balance many diseases such as diabetes mellitus can be caused. According to the previous studies, we speculate that the longevity population may have some advantages in blood glucose regulation. In this study, we utilized TMT-based proteomics method to analyze the differences of plasma proteomics profiles between non-longevity area participants (with exceeding standard FUN level) and offsprings of longevous families (with normal FUN level). In total, we identified 155 DEPs (non-longevity area participants vs. offsprings of longevous families, 132 down-regulated and 23 up-regulated). According to bioinformatics analysis, several DEPs were enriched in glycolysis/gluconeogenesis, pyruvate metabolism, propanoate metabolism, fructose and mannose metabolism, pentose phosphate pathway, glucagon signaling pathway, PI3K-Akt signaling pathway, etc. (Additional file 8: Table S7). These are involved in processes of metabolism of carbohydrate and regulation of blood glucose concentration.

Carbohydrate metabolism can be divided into catabolism and anabolism, mainly including aerobic oxidation, glycolysis, pentose phosphate pathway, gluconeogenesis andetc. It is well known that many enzymes play important roles in carbohydrate metabolism processes, such as aerobic oxidation enzymes in glycolysis and pentose phosphate pathway (e.g., triosephosphate isomerase (TPI), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), enolase (ENO), phosphoglycerate kinase (PGK), pyruvate kinase (PK), fructose-bisphos-phate aldolase (ALDO), L-lactate dehydrogenase (LDH), malate dehydrogenase (MDH), glucose phosphate isomerase (GPI), transaldolase (TALDO), and enzymes in gluconeogenesis (e.g., MDH, PK, LDH) [14,15,16]. In the present study, both aerobic oxidation enzymes (ALDOA, TPI1, GAPDH, PGK1, ENO1, GPI and TALDO1) and gluconeogenesis-related enzymes (MDH1, PKM, LDHA, and LDHB) were up-regulated in the samples from Bama longevity hotspot. These results suggested that compared with the high FUN population in non-longevity areas, the offsprings of longevous families in Bama improved both the catabolism of carbohydrate (gluconeogenesis) and the anabolism of carbohydrate (aerobic oxidation, glycolysis, pentose phosphate pathways and etc.), thus promoting the metabolic process of glucose.

The production and utilization of blood glucose are regulated by hormones such as insulin and glucagon. Insulin can suppress blood glucose levels by promoting the transformation of bloodstream glucose into glycogen, fat and other non-sugars, and inhibiting glucose production from the liver [17]. Glucagon is the counter-regulatory hormone to the hypoglycemic effects of insulin, and thus the increased plasma glucagon levels will result in increased hepatic glucose production by suppression of glycogenesis and glycolysis, and stimulating of glycogenolysis and gluconeogenesis [18].

The metabolic functions of insulin are mainly exerted by PI3K- Akt pathway in insulin cell signaling. The PI3K- Akt pathway mediates many of the metabolic actions of insulin via phosphorylation of key metabolic substrates such as glycogen synthase kinase-3 for glycogen synthesis [19,20,21]. Several researches have demonstrated that 14-3-3 isoforms (i.e., 14-3-3 Ɛ/YWHAE, 14-3-3 β/YWHAB, 14-3-3 γ/YWHAG, 14-3-3 η/YWHAH, 14-3-3 θ /YWHAQ, 14-3-3 ζ/YWHAZ, and 14-3-3 σ/SFN) interact with effectors (e.g., IRS-1, Raf-1, AS160/TBD1C4, and FOXO1) in the insulin signaling pathway and in glucose metabolism [20, 22,23,24,25]. Lim GE et al. found insulin sensitivity decreased in an YWHAZ gene knockout mice model [26, 27]. In present study, the result showed that the most conspicuously enriched protein domain was 14-3-3 domain, YWHAZ, YWHAB, YWHAG and YWHAE were higher in samples of offsprings from Bama longevity hotspot (with normal FUN level) than in non-longevity area participants (with exceeding standard FUN level), which indicated that compared with non-longevity area participants (with exceeding standard FUN level), insulin plays a stronger role in Bama participants (with normal blood glucose).

As a major Ca2+ binding protein in non-muscle cells, calmodulin (CaM) is activated by Ca2+ and then undergoes a conformational change which allowing it to activate numerous downstream targets [28]. In humans, CaM is encoded by three genes (CALM1, CALM2, and CALM3) [29]. Glucagon is secreted from pancreaticα-cells in response to low levels of blood glucose, and intracellular Ca2+ activity is required for glucagon secretion [30]. Many studies support the hypothesis that the glucagon receptor type 1 (GR1)/phospholipase C (PLC)/inositol-3, 4, 5-triphosphate (IP3)/Ca2+/CaM pathway is the predominant or exclusive signal for glucagon in vivo most of the time [31, 32]. Epstein et al. found glucagon was expressed increasingly in islet cells in a mouse model of islet β-cell CaM overexpression [33]. Similarly, in the present study, we found that CALM3 was up-regulated in the samples from Bama longevity hotspot (Fig. 5), which suggests the regulation of glucagon in Bama participants (with normal blood glucose) is stronger than that in non-longevity area participants (with exceeding standard FUN level).

Type 2 diabetes was an inflammatory condition, which associated with increasing levels of acute phase inflammatory reactants in serum [34,35,36,37,38]. In our study, several inflammatory reactant proteins, e.g., C—reactive protein (CRP), serum amyloid A-2 protein (SAA2), complement factor H (CFH), scavenger receptor cysteine-rich type 1 protein M130 (CD163), and lipopolysaccharide-binding protein (LBP) were down-regulated in the samples from Bama longevity hotspot, while serum amyloid A-4 protein (SAA4) and plasma serine protease inhibitor (SERPINA5) were up-regulated. Among these proteins, CRP and serum amyloid A (SAA) are important representative acute phase inflammatory proteins. It is reported that CRP significantly increased in the presence of inflammation and the elevated CRP level was associated with insulin resistance and an increased risk of diabetes [39,40,41]. As another important acute inflammation protein, SAA helps to link the complex network of cells and proteins mediating inflammation. The SAA gene family contains four genes, namely SAA1, SAA2, SAA3 and SAA4. SAA1 and SAA2 are acute-phase proteins, while SAA3 is a non-translated pseudogene and SAA4 protein is not induced during the acute phase response of inflammation [42,43,44]. Our proteomic result revealed that CRP and SAA2 are higher in non-longevity areas participants (with exceeding standard FUN level) than in offsprings of longevous families (with normal FUN level), which is consistent with the aforementioned reports.

Besides the aforementioned proteins, some other proteins including SPARC, PARK 7, and IGFBP-1 were significantly down-regulated, whereas pantetheinase (VNN1) was significantly up-regulated in non-longevity area participants (with exceeding standard FUN level) (Figs. 5, 6, and Supporting Additional file 3: Table S2). Consistent with our result, it is reported that the SPARC levels were decreased in islets with diabetes and SPARC deficiency could lead to DM in SPARC null mice [45, 46]. On the contrary, Wu et al. reported that plasma SPARC levels were significantly increased in T2DM patients, and Xu et al. found that increased plasma SPARC levels were relevant to insulin resistance and dyslipidemia in gestational diabetes patients [47, 48]. Therefore, more studies are still needed for further validation of the mechanism of SPARC on glycemic control. Furthermore, DJ-1 gene (PARK7) was found to be down-regulated in pancreatic islets of patients with type 2 diabetes mellitus (T2DM) [49], and a low serum concentration of IGFBP-1 is associated with gestational diabetes mellitus (GDM), unfavorable metabolic profile, glucose intolerance and risk of diabetes mellitus [50, 51], which are similar to our results. In addition, consistent with our results, some studies reported blood levels of VNN1 were increased in diabetic patients, and VNN1 increased the expression of gluconeogenic genes and hepatic glucose output, which led to hyperglycemia in a diabetic mice model [52, 53]. Further research is needed to reveal the glycemic control mechanism of these DEPs.