Effects of hydrogen-rich saline in neuroinflammation and mitochondrial dysfunction in rat model of sepsis-associated encephalopathy

Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection [1, 2]. Severe infection in children leads to cardiovascular and/or noncardiovascular organ dysfunction referred to as “sepsis-associated organ dysfunction” [3]. Sepsis-induced inflammatory response results in disseminated cerebral dysfunction termed “sepsis-associated encephalopathy, (SAE)” [4], which is characterized by various clinical or laboratory abnormalities in the absence of central nervous system infection, such as abnormal cerebral anatomical structures, haemorrhage or embolism. The incidence of SAE in paediatrics has not been demonstrated [5]. SAE has a long-term impairment, such as behavioral changes, memory and cognitive dysfunction, with socioeconomic burdens on patients/parents and society [6]. In children, the short-term mortality reported are cognitive impairment and poor academic performance [7]; the long-term include delayed neurodevelopment, low verbal IQ, a decline in school performance and low intelligence at short-term follow-up [8].

Manifestations of SAE include seizures, delirium, focal cognitive deficits, hallucinations, abnormal sleep rhythms, personality changes, lack of concentration, and depressive symptoms [9‐11]. At present, there are no diagnostic criteria and risk factor stratification for suspected patients with SAE. In other words, direct CNS infection or other types of encephalopathy have to be excluded from diagnosing SAE [12, 13], which hinders its early detection and appropriate implementation of management protocols, thus associated with high mortality rate. This scenario becomes more evident in paediatric patients with few cases being reported in the literature. Clinical assessment, electrophysiological, neurological imaging and biomarkers are implemented to aid diagnosis and direct therapeutic strategies; though most of these diagnostic tools are potentially hampered by sedation, and mechanical ventilation [14‐16].

SAE treatment is mainly focused on managing the underlying conditions, as there is no specific treatment protocol [11]. Antibiotics and supportive therapy are the mainstays of treatment, while the sedative medication is used to treat those showing agitation features [10]. Dexmedetomidine [9, 13], therapeutic plasma exchange (TPE) [12, 17], activated protein C [18], etc. have been used to treat patients with suspected SAE. Despite these signs of progress being made, effective treatments are lacking in suspected SAE individuals, especially in children. Thus, other agents are being explored both in vivo and ex vivo which are showing positive and promising results.

The hydrogen molecule is a small (molecular weight 2 Da), electrically neutral, and nonpolar, which is colorless, odorless, diatomic gas produced by intestinal bacteria in mammals. Molecular hydrogen (H2) can easily penetrate several barriers, such as the blood–brain barrier, the placental barrier, and the testis barrier [19‐21]. H2 is permeable to biomembranes and can easily target subcellular organelles, especially the mitochondria that constitute the biggest source of cellular ROS [22]; it is non-toxic, balances the pH of body fluids and does not affect the normal metabolic redox reaction due to its small molecular weight and antioxidative activity which selectively affects only the strongest oxidant [23]. It is neither inflammable nor explosive at low concentrations (< 4.6% in air and 4.1% in pure oxygen) [24]. Molecular hydrogen promotes cell detoxification, increases cell hydration, and strengthens the host immune system [25]. H2 consumption can improve biomarkers of inflammation and redox homeostasis [19]. H2 also attenuates injury and dysfunction of important organs (heart, liver, lung, kidneys, and brain) and physiological barriers (epithelial cell barrier, vascular endothelial cell barrier) by suppressing oxidative stress and inflammation as well as reducing apoptosis and regulating sepsis-induced autophagy [25].

Extensive studies both in vivo and ex vivo have shown that H2 play an effective therapeutic role in many disease entities such as sepsis, ischaemia–reperfusion injury, hypoxic-ischemic brain damage (HIBD), organ transplantation, stroke, MODS, type 2 diabetes, atherosclerosis, bronchopulmonary dysplasia (BPD), neurodegenerative diseases, skin diseases, cancers, ionising radiation and oxygen toxicity [20, 25‐27]. From the aforementioned, it is perceivable that H2 has multiple roles not only in the CNS but also in other systems by regulating a myriad of mediators and factors activated following brain injuries, such as attenuating neuroinflammation [22, 28, 29], decreasing oxidative stress parameters [30, 31], ameliorating inflammatory response and neuronal apoptosis [32, 33], regulation of signaling pathways [34, 35], improving mitochondrial dysfunction [36, 37], regulating astrocyte and microglial activation [38]. However, there is limited experimental evidence of hydrogen-rich saline (HRS) regulating inflammatory cytokines, neuronal injury, apoptotic cascades and mitochondrial dysfunction in paediatric patients with SAE.

In this study, we investigate the effect of HRS in regulating neuronal apoptosis, mitochondrial dysfunction and neuroinflammation in a juvenile SAE rat model, and the possible underlying mechanism(s). We hypothesized that the administration of HRS following LPS-induced sepsis may attenuate neuroinflammation, astrocyte and microglia activation, mitochondrial dysfunction and neuronal apoptosis, potentially improving overall neurological outcomes.

In this study, we reported the neuroprotective effects of hydrogen-rich saline (HRS) in attenuating neuroinflammation, neuronal injury, improved survival rate, and improving mitochondrial dysfunction in the juvenile SAE rat model. This is also the first study to demonstrate the effects of HRS in the paediatric SAE model. SAE represents a challenge in the ICU with high mortality and morbidity but with few available treatment options other than systemic support and antibiotics that are sometimes associated with brain dysfunction in critically ill patients as a side effect [41], especially in children due to the complexity of their brain development and disease course.

The pathophysiology of SAE is multifactorial involving diffuse neuroinflammation, disrupted BBB, mitochondrial dysfunction, oxidative stress, excitotoxicity, and cerebral autoregulation impairment [11, 42, 43]. These pathogenetic mechanisms have similar characterisation both in adults and children, though their pathogenesis and clinical presentation might differ due to the evolving brain. For instance, systemic adaptive and immune responses following infections, the resistance of the immature brain to injury [44, 45], and functional BBB response after brain injury insults [46]. Most of these factors are said to be regulated by HRS, in part via neuroinflammatory attenuation, oxidative stress modulation, maintaining BBB integrity, ameliorating mitochondrial dysfunction, regulating astrocyte and microglia activation and suppressing NLRP3 inflammasome activation in adults SAE models [19, 25, 47].

Neuroinflammation results from the upregulation of pro-inflammatory cytokines (PICs) that are involved in microcirculatory dysfunction by potentially altering blood flow [11]. Upregulated PICs mediate SAE occurrence due to their direct correlation with BBB disruption, brain edema, neutrophil infiltration, astrocytosis, and apoptosis of brain cells [48]. The mRNA expression of TNF-α and its receptor, TNFR1 is also upregulated following LPS induction in the septic encephalopathy model [49]. Similarly, stimulated microglia and astrocytes by cytokines further produce other cytokines, chemokines, nitric oxide, excitatory amino acids, COX2, and reactive oxygen species (ROS), which are detrimental to the immature brain due to the enhanced vulnerability of maturing cells [50]. Our results showed that PICs, e.g., TNF-α and IL-1β were upregulated while anti-inflammatory cytokines, e.g., IL-10 was downregulated in septic rats induced by LPS thereby exacerbating neuroinflammation. However, HRS treatment attenuated neuroinflammation by reducing TNF-α and IL-1β expression levels and subsequently increasing IL-10 levels both in serum and brain tissues of septic rats. Previous studies have reported similar effects [47, 51]. A recent control trial also demonstrated the protective effect of HRS in healthy individuals by decreasing inflammatory response and increasing antioxidant capacity [52]. In vivo studies showed that reactive astrocytes (A1) phenotype is also induced by TNF-α and IL-1β after being stimulated by LPS [53], the current study also reported similar results in cultured cells, thus it is conceivable that there exist a synergy between astrocytes activation and upregulation of pro-inflammatory mediators (TNF-α and IL-1β) during sepsis. In essence, decreasing A1 reactive astrocytes is crucial in curtailing neuroinflammation and astrocytes activation. It is therefore plausible to state that the mechanistic effects of HRS is through balancing and maintaining activation of inflammatory mediators as demonstrated above where TNF-α and IL-1β were downregulated and IL-10 upregulated, suppressing A1 reactive astrocytes and microglia activation in septic rat model. Additionally, we demonstrated that HRS given at early onset of sepsis induction is more effective in reducing TNF-α and IL-1β than when given at later time point, though more studies are needed to validate this effect.

Neuronal sensitivity from increased levels of nitric oxide (NO) produced by activated microglia can exacerbate neuronal apoptosis [9]. Certain chemokines also promote neuronal apoptosis, as evidenced by increased upregulation of Ccl2 or Cxcl2 protein levels in the PLS model, resulting in hippocampal neuronal apoptosis, thus supporting a direct role of these chemokines in neuronal death [54]. Another study reported that postnatal day 1 (p1) is more vulnerable to neurotoxicity than postnatal day 2 (p2) in septic rat model induced either induced by hypoxia–ischemia, LPS or both, due to the immaturity of neuronal cells [44]. In our present study, both neuronal injury (evidenced by increased number of damaged neurons, dissolved cellular structure and dark nuclei, shrunken cytoplasm and perineuronal vacuolisation) and neuronal apoptotic cell death were increased after LPS administration in SAE rats, while HRS treatment abrogated neuronal injury and apoptotic neuronal cell death both in the cerebral cortex and hippocampus. HRS enhanced the efficiency of the defense response evidenced by improved resolution of neuroinflammation in sickness behavior induced by LPS [29], as well as attenuation of neuronal apoptosis in early brain injury of SAH rat model [55]. We also reported improved neurological function at 48 h following LPS induction, as well as an improved survival rate after HRS administration. Other studies have also demonstrated the effect of H2 in attenuating neurobehavioral deficits and neuronal apoptosis in a neonatal hypoxia–ischemia rat model [38, 56]. Based on these results and ours, it is conceivable that HRS protective mechanism in septic models is probably via counteracting pro-apoptotic mediators’ activation and ameliorating neuronal injury.

Mitochondria as the powerhouse of cellular metabolism, is responsible for over 90% of the total body oxygen consumption and ATP generation via oxidative phosphorylation [57]. Mitochondria play a vital role in neuronal functions, and alteration of mitochondrial dynamics including fission and fusion can have deleterious effects. In an SAE model, LPS stimulation led to a loss of mitochondrial membrane potential, propagation of dynamin-related protein 1 (Drp1) and p53 recruitment with subsequent initiation of cell death pathways [58]. Decreased mitochondrial ATP generation is caused by increased release of ROS and NO, which in turn induce neuronal apoptosis by releasing cytochrome C [59]. In addition, activation of reactive nitrogen species (RNS), NO and ROS can inhibit complexes I and IV of the electron transport chain (ETC), which disrupt mitochondrial function that is implicated in SAE pathogenesis [9, 60]. Excessive production of ROS during sepsis exacerbates mitochondrial damage and decrease in ATP concentration, and subsequent depletion of cellular/energy metabolism, resulting in bioenergetic failure. Our present study demonstrated that the mitochondrial membrane potential (MMP) and ATP content were decreased in septic rats while ROS production was increased. Increased release of ROS after LPS induction resulted in mitochondrial dysfunction and directly or indirectly inhibiting mitochondrial respiratory chain complexes vital for normal cellular metabolism. Damaged mitochondria further impaired it ability to generate ATP, thus altering the bioenergetic status of cellular function. However, this excess release of ROS was abrogated after HRS treatment, thus improving mitochondrial dysfunction. ATP content is vital for maintaining constant supply of energy needed by the body. ATP concentration was reported to decrease in non-survivors of septic patient attributed to increase rise of AMP resulting in imbalance in ATP turnover, thus decreased ATP production [57]. Our present study demonstrated similar results where ATP content was decreased in septic rats that were conversely enhanced after HRS administration. Similarly, HRS improved MMP, further highlighting it effect in improving mitochondrial dysfunction. Interestingly, reactive astrocytes (A1) result in less formation of synapses and that even those that have formed are weaker compared to those produced by healthy resting astrocytes [53]. A recent study has demonstrated that single or multiple systemic injection of LPS result in global loss of synapses [40]. Our results further showed that the number of mitochondria and synapses following the LPS challenge were decreased. Thus, inhibiting A1 reactive astrocytes can have positive effect in increasing synaptic formation as well as maintaining strong and healthy synapses, which our results has demonstrated after HRS treatment. Similarly, mitochondrial biogenesis is also disrupted due to exacerbated PICs activation; and our results showed that the mitochondrial membrane was disrupted after LPS administration, resulting in ultrastructural damage to mitochondrial biogenesis. Other studies have reported similar results [4]. Furthermore, mitochondrial dysfunction due to uncontrolled leakage of electrons from the electron transfer chain can be mitigated after HRS administration, thus improving mitochondrial energy metabolism [61].

Recent studies showed that astrocytes and microglial cells are involved across SAE pathomechanistic spectrum, with inflammatory activation occurring mainly in microglial cells [62]. Microglia activation involved two phenotypes, M1 cells that produce PICs and ROS and M2 cells that produce anti-inflammatory mediators that play neuroprotective effect and tissue repair ability [62, 63]. Activated microglia deteriorates BBB integrity subsequently enhancing ROS release, which leads to brain dysfunction [48]. Studies have shown that microglia depletion during severe sepsis development is associated with early exacerbation of the brain and systemic inflammation in the sepsis model [64]. Similarly, astrocytes, which control homeostasis and catabolism, also have two forms; reactive astrogliosis triggers nervous tissue damage by attracting immune cells specifically to the injured region and facilitates their extravasation and tissue infiltration [65]. In addition, astrocytes have been reported to restore neuronal mitochodrial membrane potential that result in increased mitochondrial content [66]. in vivo studies have shown that reactive microglia are needed to induce A1 reactive astrocytes in LPS-treated cells [53]. Thus, astrocytes activation is exacerbated by reactive microglia, both are implicated in the pathogenesis of sepsis. Our present study demonstrated that LPS induced increased activation of astrocyte and microglia in septic rats as shown by elevated expressions of GFAP and Iba-1 in both cerebral cortex and hippocampal neurons, causing neurotoxicity. HRS administration decreased activation of these markers, thus ameliorating astrogliosis and microglial activation. Further analysis by immunofluorescence staining showed increased immunoreactivity of GFAP and Iba-1 in the cerebral cortex and hippocampus that was reverted after HRS treatment, highlighting the possible mechanistic action of HRS in modulating astrocytes and microglial activation implicated in the SAE pathogenesis. Another study reported similar findings where HRS treatment decreases M1 polarization thereby regulating microglia activation in a septic rat model [47]. At protein level, both GFAP and Iba-1 were downregulated in the HRS group when determined by western blot, suggesting it molecular actions in regulating these proteins. Though, further studies are needed to explore its targeted molecular function, our results have shown that HRS is crucial in maintaining mitochondrial function, modulating inflammatory mediators and regulating astrocytes and microglia activation.

A recent study showed that early fluid resuscitation and inhalation of 2% hydrogen could decrease oxidative damage and inhibit the overexpression of inflammatory factors, as well as downregulating the expression of proapoptotic protein Fas, and up-regulate the expression of anti-apoptotic protein Bcl2 in septic shock rats [67]. Our results also showed the effect of HRS in regulating apoptotic mediators, where Bcl-2 was downregulated and upregulated Bax expression levels in septic rats, while HRS treatment exerted the opposite effect in regulating these mediators at molecular level.

We demonstrated in this study that early HRS administration after the onset of sepsis is more favourable in exerting its effectiveness in regulating inflammatory response and improving mitochondrial dysfunction than when given at a later time point. This decrease in HRS effectiveness is attributed to delayed administration of HRS compounded with increased SAE severity. However, this trend of effectiveness warrants more studies to explore this phenomenon. Secondly, the present study did not assess the long-term effect of HRS in improving neurological function and memory impairment, as reported in adult SAE models [68, 69]. Although our current study does not provide the exact mode of action of HRS, a possible sequence of events from exposure to HRS to protection of sepsis-associated encephalopathy could be postulated as follow: HRS could exert it’s action on mitochondria first through its anti-oxidative stress function, followed by anti-apoptotic activity through inhibition of the intrinsic apoptosis pathways such as inhibition of the release of cytochrome C and the related cascade, resulting in increase of pro-survival molecules such as AkT and phosphorylated AkT (p-AkT) and activation of mTOR pathway leading to inhibition of microglia activation and attenuation of inflammation. Naturally, the mechanism(s) underlying HRS neuroprotective effects in attenuating neuroinflammation and modulating astrocyte and microglia activation need further elucidation in paediatric SAE models. Nevertheless, our present study has shown the potential mechanistic effects of HRS in ameliorating inflammatory response; inhibit astrocyte and microglia activation, neuronal injury, neuronal apoptosis, and mitochondrial dysfunction, thus postulating its application in paediatric SAE as a promising therapeutic agent.