Air pollution is an important environmental risk factor that influences people’s health. WHO indicated that 92% of the world’s population were breathing air which does not meet minimum standards [16]. Earlier studies demonstrated that the exposure of PM2.5 could lead to increased incidence and mortality from cardiovascular and respiratory diseases [17, 18]. Moreover, epidemiological studies have shown that exposure to PM2.5 may be related to obesity and type 2 diabetes [19, 20]. Children who are chronically exposed to PM2.5 above the standards would suffer from obesity and cognitive deficits [21]. Recently, we have found that chronic PM2.5 exposure for 9 months could led to neuroinflammatory and cognitive decline in mice [5]. However, PM2.5 exposure induced neuroinflammatory reactions in hippocampus and cortex could be found only after 12-month exposure, which occurred after cognitive impairment, implying that neuroinflammtion may be the results but did not trigger the neuronal dysfunction [5]. The underlying molecular mechanism of early events in brain induced by PM2.5 exposure remains unclear. In the current study, neuronal apoptosis were observed earlier by 24-week PM2.5 exposure, in parallel with aggravated cognitive function in mice. Besides, we also found AMD1, as one of most important enzymes responsible for polyamine synthesis, together with its product spermidine, participated in PM2.5 exposure triggered neuronal apoptosis.
During whole-body exposure, the concentration and composition of PM2.5 were analyzed, as it has been reported before [5]. As we all know, PM2.5 is an important pollutant composed of a complex mixture of organic and inorganic components. Previous evidence showed that inorganic elements such as sulfur, potassium, chlorine, calcium were the major composition of PM2.5 and the PAHs contained benzo(b)fluoranthene were detected in PM2.5 as well, which has carcinogenic and neurotoxic properties [22]. In the current study, mice were exposed to ambient PM2.5 in a whole-body manner for 24 weeks. In our expectation, in MWM test, mice from PM group manifested longer latency time to reach the hidden platform and reduced frequency and time spent in the target quadrant (Fig. 1B). Therefore, these results again demonstrated that PM2.5 was complex and harmful mixture which could do harm to central nervous system.
We then put efforts on to explore the molecular mechanism triggering the neuronal dysfunction. Though neuroinflammation is the common pathway of neurological diseases [23], much evidence has shown neuronal apoptosis as important event in the pathologic process of cognitive dysfunction [24, 25]. Concerning to about PM2.5 exposure induced neuronal apoptosis, evidence from mouse model of gestational PM2.5 exposure showed neuronal apoptosis in hippocampus of mice offspring [24]. In vitro data also showed that PM2.5 exposure season-dependently induced neuronal apoptosis and synaptic injuries [9]. From our results, TUNEL positive cells were obviously increased in CA1 area of the hippocampus from mice exposed to PM2.5 for 24 weeks, indicating elevated neuronal apoptosis rate by PM2.5 exposure (Fig. 2A). Besides, we found the upregulated levels of pro-apoptotic protein Bax and cleaved caspase-3, and the downregulated level of anti-apoptotic protein Bcl-2 from both in vivo and in vitro experiments (Fig. 2B, Fig. 3B, C), which was consistent with evidence reported before. Therefore, in the next step, we will focus on exploring the specific molecular mechanism of PM2.5 inducing neuronal apoptosis.
In order to explore novel target that may contribute to PM2.5 induced neuronal apoptosis, total RNA from brain tissues homogenate was screened for transcriptome sequencing analysis. The results showed that genes with upregulated transcription levels included Pgrmc1, Tceal6, Syt11, Pea15a, AMD1 et al, while with downregulated transcription levels included Tubb2b, mt-Nd5, Tubb4b, Shisa9 et al. By literature research, AMD1, as one of the speed-limiting enzymes in polyamine metabolism in eukaryotic cells [26], became one of our interests. The reduced relative transcriptional level and protein expression level of AMD1 by PM2.5 exposure in mice were further verified by qPCR and immunoblotting (Fig. 4A-B), which was confirmed as dose-dependent in vitro (Fig. 4C). AMD1 silencing or inhibition was reported to regulate apoptosis in neuroblastoma [11], but little evidence was provided from neurons. By our results, we found that reduced cell viability and increased neuronal apoptosis could be triggered not only by PM2.5 exposure, but also by AMD1 inhibitor in PC12 cells and primary neurons (Fig. 5), demonstrating the possible participation of AMD1 in PM2.5 induced neuronal dysfunction.
AMD1 is believed to play an important role in cell growth and proliferation, which may be dependent on metabolic process it participates in. AMD1 decarboxylated S-adenosylmethionine, which provide aminopropyl for putrescine conversion to spermidine and spermine [27]. Spermidine, one of natural polyamines, is not only a major inducer of growth and proliferation in eukaryotic cell, but also has significant mitigative and protective effects against the development of neurodegenerative diseases and impaired cognitive ability [28]. Numerous studies have showed that spermidine treatment could attenuate oxidative stress, neuroinflammation and possess promising neuroprotective effect against degenerative changes [29, 30]. Reduced level of spermidine was reported concomitant with declining memory abilities in aging fruit flies [31]. Another study showed that spermidine promoted cell liability in retinal ganglion neuronal cells [32]. By our work, intracellular concentration of spermidine was evaluated by HPLC, and data showed that spermidine concentration in PM group was significantly reduced (Fig. 6A). By contrast, exogenous spermidine supplementation could efficiently rescue the impaired cell viability, increased apoptotic rate and corresponding apoptosis related protein expression triggered by PM2.5 treatment (Fig. 6B-C). Piled up all the results, we may speculate that AMD1 participate in PM2.5 induced neuronal apoptosis, while its product spermidine was the key target.
As is known to all, mitochondria are one of the most important organelle involved in apoptosis [33, 34]. Numerous studies have shown that mitochondrial dysfunction may eventually lead to apoptosis [35, 36]. Besides, spermidine was proven to ameliorate neuronal aging by improving mitochondrial function in vitro [37]. Based on these evidence, mitochondrial membrane potential was examined in PC12 cells by JC-1 staining, showing that exogenous application of spermidine could rescue PM2.5 treatment induced the mitochondrial membrane potential depolarization (Fig. 7A). Mitochondrial membrane potential depolarization may then switch on the intrinsic apoptosis pathway. Once activated, mitochondrial intermembrane space proteins, notably cytochrome C, are released into the cytosol whereupon they activate caspases. Executioner caspase, such as caspase 3, effectively kills the cell within minutes by cleavage of hundreds of different substrates in parallel [38–40]. By our results, the expression levels of mitochondrial function closely related protein Bax/Bcl-2, cytochrome C, cleaved caspase-9 and cleaved caspase-3 were up-regulated by PM2.5 treatment, which could be recovered by spermidine pretreatment (Fig. 7B). Therefore, our results suggested that PM2.5 aggravated PC12 cells via activating mitochondrial-mediated apoptosis pathway, which could be alleviated by exogenous spermidine. All these data suggested that AMD1 and its product spermidine played crucial role in PM2.5 induced neuronal injury.