Oral Pyruvate Protection of Dorsal Root Ganglia in Simulated Weightlessness Rats

Background: Previous studies demonstrate that long-tern microgravity induces multi-organ injury and dysfunction, including the dorsal root ganglia (DRG) damage. This study investigated oral pyruvate protective effects on lumbar 5 (L5) DRG nerve tissues in rats subjected to hindlimb unweighting (HU). Results: Male Sprague-Dawley rats were randomly assigned to four groups (n=10): control group (Group CON), suspension group (Group SUS), normal saline group (Group SAL) and sodium pyruvate group (Group PYR), respectively. The rats of SUS, SAL and PYR groups were simulated with microgravity by tail suspension of HU for 8 weeks. Rats in Groups SAL and PYR fed with normal saline and pyruvate saline, respectively. Histopathological and immunouorescence examinations were conducted and levels of glial cell line-derived neurotrophic factor (GDNF), glial brillary acidic protein (GFAP), ATP and ATPase were measured in DRG tissues; L5 spinal cord scans were also carried out in rats following HU. Results showed that the HU resulted in signicant alterations in DRG nerve tissues’ structure and function in Groups SUS and SAL, whereas morphological changes were not signicantly distinguished between Group PYR and Group CON; GDNF, GFAP, ATP and ATPase levels were mostly preserved in Group PYR, but still worse than in Group CON. The signicance of oral pyruvate protection against DRG injury following HU and the dose and formula of oral pyruvate solutions were discussed for use in astronauts’ spaceight. Conclusions: oral pyruvate effectively protected L5 DRG from pathological alterations and dysfunction induced by the HU in rats. Further studies and clinical trials are warranted.


Background
The status of astronauts' health and safety in space exploration missions is a very important issue. When they enter the space environment, astronauts' physiological functions are affected in various facets, such as negative calcium balance, muscle atrophy and bone loss, cardiovascular and immune system disorders, brain and spinal cord injury (SCI) and so on [1][2][3][4][5] And several interventions for these disturbances were proposed [6,7].
Dorsal root ganglion (DRG) is a critical anatomical structure involved in stimulating sensation of pain injury. It is also an important synthetic site of neurotrophic factors and neurotransmitters, which play a pivotal role in the nutrition and injury repair of nerves. Under simulated weightlessness, the expression of neurotrophic factors and genomics is altered and the degenerative alteration of myelin sheath appears in the fth lumbar spine (L5) DRG in rats. These aberrances re ect its damage from the microscopic point of view [5,8].
Pyruvate is the intermediate product of glycolysis at the junction of glucose anaerobic and aerobic metabolism. It is the metabolic hub of the three major nutrients, not only an energy substrate, but also an anti-oxidative stress and anti-in ammatory agent. Pyruvate holds several superior physiological and pharmacological characteristics in comprehensive improvement of glucometabolic disturbances and protection of mitochondrial function, so that it is multi-organ protective against noxious stimuli [9][10][11].
Several reports have evidenced that pyruvate has protective function against damage of brain and nerve tissues caused by various pathological insults [8,9,12]. Further, recent ndings indicated that enteral pyruvate in a pyruvate-enriched oral rehydration solution (Pyr-ORS) profoundly preserved neurons structure and function following ashyxial cardiac arrest in rats [13].
Based on our prior observation [8], the present study aims to investigate the effect of oral sodium pyruvate solution on the DRG injury, by highlight of the changes in morphological structure and metabolic function of DRG nervous cells, in rats subjected with simulated weightlessness and provides the rst experimental evidence for oral pyruvate in prevention and treatment of the back pain and beyond in astronauts' space ight.

Materials And Methods
Animal and tissue treatment Forty 3-month old, male, Sprague-Dawley (SD) rats ( ), purchased from the Animal Experimental Center of the Academy of Military Medical Sciences, Beijing, were divided into 4 groups ( ): control group (Group CON), suspension group (Group SUS), normal saline group (Group SAL) and sodium pyruvate saline group (Group PYR). The design procedures were following previous ndings [8]. The rats of Groups SUS, SAL and PYR were simulated with microgravity effects by the tail suspension of hindlimb unweighting (HU) model [8]. The hind limbs (legs) of the rats were held off the horizontal plane approximately at a 30°a ngle. Rats in the CON and SUS groups were fed with conventional laboratory water. The rats of Group SAL and Group PYR were orally given with 0.9% sodium chloride solution (normal saline, NS, [Na + ] 154 mmol/L, [Cl -] ) and sodium pyruvate saline daily to replace the conventional laboratory water, respectively.
Pyruvate saline was prepared in the laboratory on the basis of the patent (Patent No.: US 8,835,508 B2) formula: ([Na + ] 154 mmol/L, [Cl -] ) and [Pyr -] ). The total volume of drinking water fed daily was about in each rat.
The rats were housed in cages (2-3 in each) with experimental wood chip bedding and placed in a 12hour light/dark cycle in the similar conditions during experimentation with standard laboratory food and light, as previously conducted [8]. After 1 week of adaptive suspension, the experimental rats with hindlimb unweighting lasted 8 weeks. At the end of the experiment, the rats were euthanized by intraperitoneal injection of sodium pentobarbital with a lethal dose ( ) and dissected. The rats' lumbar 5 dorsal root ganglia (L5 DRG in both sides each rat) were quickly excised, frozen in the isopentane cooled with liquid nitrogen [8]. Normal saline (Cisen Pharmaceutical Co., China) was obtained from the Animal Laboratory of PLA General Hospital and sodium pyruvate was purchased from Sigma Co. (St Louis, MO).
Hematoxylin-Eosin staining (HE) and toluidine blue staining As described before [8], the DRG tissue samples were xed in formalin for 24 hours, washed and dehydrated by ethanol gradient. Then, tissue blocks were embedded in conventional para n, treated with Page 4/25 dimethyl benzene and nally sliced in sections. The sections were incubated at 60℃ for 1 hour and then dewaxed with xylene, dehydrated with ethanol and stained with hematoxylin and eosin.
For toluidine blue staining, the sections of DRG were placed on the coverslips, washed by phosphate buffered saline (PBS) and xed in 4% paraformaldehyde for 30 minutes and stained with 1% toluidine blue at room temperature for 30 minutes. The sections then were washed with absolute ethanol, dried, cleared and nally coverslipped. Tissue sections were analyzed by a light microscopy (400X, OLYMPUS BX46). The researchers were blinded to observe the experimental samples. The quantity of Nissl-stained neurons per 1 mm 2 region of the ve random areas was counted in average by using an optical microscopy and the ImageJ tool software (Nikon NIS-Elements, USA).

Western blot analysis
As previously described, the DRG tissues were prepared for Western Blotting [8,14]. Brie y, total proteins of DRG were extracted by standard methods and the protein content was determined by the Bradford method. The cell lysate separated by using SDS-PAGE and then transferred to the PVDF membrane. All membranes were blocked in TBST contained 5% skim milk for 30 minutes at room temperature, followed by incubation with the primary antibody (1:1000, rabbit anti-rat GDNF [ab64337, Abcam]; 1:1000, rabbit anti-rat GFAP [ab138701, Abcam] and 1:5000, anti-β-actin [ab95437, Abcam]) overnight at 4℃, respectively. Blots were washed with TBST and incubated in secondary antibodies (goat anti-rabbit IgG [1:2000, Abcam]) for 1 h at 37℃. The protein expression was visualized with enhanced chemiluminescence (ECL) reagents (Amersham, Cleveland, OH, USA) and the membranes were exposed to Hyper lm (Amersham). The intensity values of glial cell-derived neurotrophic factor (GDNF) and glial brillary acidic protein (GFAP) were determined with the Quantity-One software (BIO-RAD, USA) and each group was compared with β-chain. The results for the two proteins were expressed in relative ratios, respectively.

Spinal Cord Scan
All SD rats, before sacri ces, were pre-anesthetized in the pre-anesthetic box (ABM, Yuyan Instruments Co. Ltd., China) with an anesthetic dose of 5cps and oxygen of 2lps. After rats were deeply anesthetized (both eyes closed without reaction by shaking the anesthesia box) for 2 minutes, rats were moved into the magnet and xed head to a magnetic resonance imaging (MRI) lift bed with a prostrate position, as previously reported [15]. With continuous anesthesia (2% iso urane gas), all rats were xed with the head advanced and prone position. After the xation was completed, continuous anesthesia was performed in a group sequence. Ensured the connection was correct to keep rats' condition monitored during the inspection process in L5, a fast spin echo (FSE) sequence PDWI is used to scan the region of interest (ROI) across sections, sagittal surfaces and coronal surfaces. Scan parameters were repetition time: 2000 ms; echo time: 28 ms, scan layer: 24 layers, eld of view: ; number of signal average: 15; scan time: 5 min 50 ms, using the small animal NMR instrument GE Signa Excite 0.35T scanner. Using the ImageJ tool software (Nikon NIS-Elements, USA) to calculate the relative area ratio of the gray matter in four groups with Group CON as 1.0.

Tissue ATP content
The L5 DRG tissue in each rat was quickly homogenized to prepare the tissue lysate and the protein concentration was determined. ATP levels in DRG tissues were measured with absorbance values, as previously reported [16]. The tissue ATP content was determined stringently following the procedures provided by the test kit purchased from Nanjing Jiancheng Bioengineering Institute, China. Finally, the absorbance values at 636 nm were read with each tube by HPLC (high performance liquid chromatograph, Agilent CO., USA). The ATP content of the test sample was calculated, according to the following formula: ATP content (nmol/gprot) = (test tube-control tube) / (standard tube-blank tube) × standard content × dilution times ÷ protein (gprot/L)

Tissue ATPase activity
In strict accordance with the ATPase test kit instructions (SuZhou Comin Biotechnology Co., Ltd, China), L5 DRG tissues were accurately weighed and the weight over volume ratio of 1:9 was obtained by adding 0.9% sodium chloride solution to make enzyme solution samples. Samples were homogenized and centrifuged at 2000rpm at 4℃for 10 minutes. The supernatant 0.2 mL were diluted to homogenate with 0.9% sodium chloride 0.8 ml. Then, enzymatic and phosphorylation reactions were performed according to the operation table in the test kit. Finally, the amount of inorganic phosphoric acid was calculated by colorimetric determination at 660 nm (Agilent CO., USA). The ATPase activity was calculated according to the following formula: ATPase activity (μmol/mgprot/h) = (test tube-control tube) / (standard tube-blank tube) × standard content × dilution times × 6 ÷ protein (mgprot/ml) Myelin sheath immuno uorescence As previously described [17] for immuno uorescence detection to identify the expression of myelin sheath, a continuous cross section of a 20-μm thick L5 DRG tissue was produced by cryostat (Microm, Heidelberg, Germany). The sections were thawed at room temperature for 30 minutes for staining. The tissues were washed with PBS and then incubated with 3% hydrogen peroxide for 10 minutes. Next, sections were rinsed by normal goat serum at room temperature for 2 hours and incubated with polyclonal degenerated myelin basic protein (dgen-MBP) primary antibody (1:1000, Chemicon Inc., Temecula, USA) overnight at 4℃. After rinsing, the sections were incubated at 4℃ overnight with a secondary antibody (1:200, donkey anti-rabbit IgG-FITC; Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA). Then, sections were washed three times with PBS for 10 minutes each time. Finally, the sections were mounted by using Floursave mounting reagent (Cal Biochem, La Jolla, USA). Comparisons among groups were made based on stained area and mean optical density (OD). Images were collected and the mean OD value was quanti ed using ImageJ software (National institute of Health, USA).

Data analysis
The SPSS 17.0 software was used for statistical analyses of data expressed as mean ± SD. Comparison of data among four groups were analyzed, using the Student's t test. One-way analysis of variance test was used to calculate signi cance among the various groups. Statistically signi cant was de ned as P <0.05.

Results
Effects of pyruvate on cell morphology and Nissl bodies in L5 DRG tissues HE staining In the PYR group, the DRG nervous cells showed only mild swelling and a small amount of cytoplasm vacuoles. In contrast, in the SUS and SAL groups, the DRG cells presented quite swelling, gap widened in varying degrees and some cytoplasmic vacuoles and numerous ne granular basophilic materials among cells.
These changes appeared more signi cant in the SUS and SAL groups than in the CON and PYR groups.
The cell morphological appearance in Group PYR was slightly changed in comparison with that in Group CON (Fig. 1).
In the PYR group, the DRG neurons showed only mild swelling and a small amount of cytoplasm vacuoles. In the SUS and SAL groups, the neurons presented quite swelling, gap widened in varying degrees and some cytoplasmic vacuoles and there were numerous ne granular basophilic materials among the cells. The cell morphological appearance in Group PYR was slightly changed in comparison with that in Group CON (scale bars 25μm).

Nissl body in neuron
In the CON and PYR groups, Nissl basophilic granules were abundant with near normal appearances in neurons. However, in the SUS and SAL groups, the morphological size and the distribution pattern of Nissl bodies were smaller and scattered, which were lightly stained. In some DRG cells, the Nissl bodies also appeared disassociated. In addition, the observed and quanti ed changes were more signi cant in the SUS and SAL groups than in the CON and PYR groups and there was no signi cant difference in numbers of Nissl body stained neurons between Groups CON and PYR (Fig. 2 A and B).
In the CON and PYR groups, Nissl basophilic granules were abundant with near normal appearance. However, in the SUS and SAL groups, the Nissl bodies were lightly stained with smaller morphological sizes and scattered distribution. In some DRG neurons, the Nissl bodies also appeared disassociated in Groups SUS and SAL (scale bars 100μm).
The amounts of Nissl-stained neurons/1.0 mm 2 in ve random elds were quanti ed by using a light microscope (X400; ve section slices were evaluated in average in each rat; n = 10). In addition, the observed and quanti ed changes were more obvious in the SUS and SAL groups than in the CON and PYR groups and there was no signi cant difference in numbers of Nissl body staining neurons between Groups CON and PYR. There were statistically signi cant differences between Groups SUS and SAL from Group PYR, respectively (**P<0.01).

Degenerated-MBP expression
In the SUS and SAL groups, the cell number of dgen-MBP was increased; the cells were swollen and their boundaries were unclear with heavy uorescent in the immuno uorescence detection. These changes did not obviously appear in the CON and PYR groups (Fig. 3 A). The mean OD value of dgen-MBP was signi cantly increased in Groups SUS and SAL, compared with Groups CON and PYR. As shown in Fig. 3 B, there were statistically signi cant differences between Groups SUS and SAL from Groups CON and PYR (P<0.01), however, there was no difference between both later groups (P>0.05) in the mean OD value.
The SUS and SAL groups showed more expressions of dgen-MBP than the CON and PYR groups (scale bars 50μm).
Mean OD values of dgen-MBP immuno uorescence staining in L5 DRG were higher in Groups SUS and SAL than in Groups CON and PYR. There were statistically signi cant differences between Groups SUS and SAL from Group PYR, respectively (**P<0.01).

dgen-MBP: degenerated myelin basic protein
Spinal cord scan In the CON groups, the gray matter and white matter areas in L5 were normal. In the SUS and SAL groups, the areas of gray matter and white matter were atrophy. These alterations appeared more signi cant in the SUS and SAL groups than in the PYR group that was close to Group CON. The relative area ratios of gray matter in the HU groups to the CON group also illustrated a signi cant decrease in Groups SUS and SAL, but a comparable outcome in Groups PYR and CON (Fig. 4 A and B).
In Group CON and PYR, the posterior part of the gray matter dorsal horn was slightly thicker and the outer layer was gentle; the edge of the anterior horn of the gray matter was enlarged and smooth. In Group SUS and SAL, the middle and lateral parts of the gray matter dorsal horn were slightly curved and the out layer was steep. In Group SUS, the medial margin of the anterior horn was thinner; in Group SAL, the anterior horn of the gray matter was signi cantly reduced (scale bars 1μm).
The gray matter area was bigger in the PYR group than in the SUS and SAL groups. The alterations appeared more signi cant in the SUS and SAL groups than in the CON and PYR groups and they were close to each other in Groups CON and PYR. The gray matter areas were statistically signi cant differences between SUS and SAL from PYR groups, respectively (**P<0.01).
Effects of pyruvate on GDNF and GFAP expressions in L5 DRG tissues GDNF protein As shown in Fig. 5, 8 weeks of simulated microgravity induced a signi cant reduction in GDNF expressions. Compared with the CON group, the relative GDNF protein content in the DRG was reduced by 21% and 20% in the SUS and SAL groups, respectively. In contrast, the GDNF content was 123% and 112% higher in the PYR group than in the SUS and SAL groups, respectively. There were statistically signi cant differences between SUA and SAL groups from CON and PYR groups (P<0.01). However, the protein content was signi cantly higher in group PYR than in group CON (p<0.01).
Simulated microgravity induced signi cant reduction in GDNF expressions. There were statistically signi cant differences between SUA and SAL groups from CON and PYR groups. However, the protein content was signi cantly higher in Group PYR than in Group CON. There were statistically signi cant differences between Groups CON, SUS and SAL from Group PYR, respectively (**P<0.01).

GFAP expression
The relative GFAP protein expression was signi cantly increased in all three suspension groups, compared with Group CON, as shown in Fig. 6. GFAP contents were 2.5 times and 2.3 times in Groups SUS and SAL over Group CON, respectively. However, the GFAP protein in Group PYR was reduced by 28% and 24%, compared to Groups SUS and SAL, respectively. There were statistically signi cant differences between Groups SUA and SAL from Groups CON and PYR (P<0.01). And the GFAP level was still signi cantly higher in Group PYR than in Group CON (P< 0.01).
The GFAP protein expressions were signi cantly increased in all three suspension groups. There were statistically signi cant differences between Groups SUA and SAL from Groups CON and PYR. And the GFAP level was still signi cantly higher in Group PYR than in Group CON. There were statistically signi cant differences between CON, SUS and SAL from Group PYR, respectively (**P<0.01).
Effects of pyruvate on ATP and ATPase expressions in L5 DRG tissues ATP content The adenosine triphosphate (ATP) content was decreased in all three suspension groups, compared to Group CON. ATP contents (nmol/gprot) were reduced by 60% and 61% in Groups SUS and SAL from Group CON, respectively. However, they were 35% and 38% higher in the PYR group than in the SUS and SAL groups, respectively. There were statistically signi cant differences between Groups SUS and SAL from Groups CON and PYR (P<0.05; P<0.01; Fig. 7 A). However, the ATP level was still signi cantly lower in Group PYR, compared with Group CON (P<0.01).
The ATP contents were decreased in all three suspension groups. There were statistically signi cant differences between Groups SUA and SAL from Groups CON and PYR. However, the ATP level was still signi cantly lower in Group PYR compared with Group CON. There was a statistically signi cant difference between Group SUS and Group PYR (P*<0.05) and signi cant different between Groups CON and SAL from Group PYR, respectively (**P<0.01).

ATPase activity
The adenosine triphosphatase (ATPase) activity was reduced in all three suspension groups relative to Group CON. ATPase activities (μmol/mgprot/h) in Groups SUS and SAL were decreased by 57% and 59% from Group CON, respectively. However, they were 42% and 51% higher in the PYR group than in the SUS and SAL groups, respectively. There were statistically signi cant differences between the SUS and SAL groups from the CON and PYR groups (P<0.01; Fig. 7 B). But, the ATPase activity in Group PYR was still signi cantly reduced relative to Group CON (P<0.01).
The ATPase activity was reduced in all three suspension groups. There were statistically signi cant differences between the SUS and SAL groups from the CON and PYR groups. But, the ATPase activity in Group PYR was still signi cantly reduced relative to Group CON. There were statistically signi cant differences between Groups CON, SUS and SAL from Group PYR, respectively (**P<0.01).
The rats' diet, defecation and general activity in the experiment were normal, no tail injury and death occurred.

Major ndings and interpretation
Morphological changes As demonstrated previously [3,8], present results further showed that the microgravity in Groups SUS and SAL led to signi cant morphological changes in L5 DRG neurons in rats, while the cell features of the PYR group were almost undistinguishable from those in the CON group without HU experienced in HE staining (Fig. 1). The Nissl body is composed of neurons' rough endoplasmic reticulum (ER) and free polyribosomes. Its alterations of functional status and morphology can re ect neuron's activities in synthesis of proteins and relevant nutritional factors and its survival [18,19]. Results discovered that the impairment of Nissl stained neurons in Groups SUS and SAL was serious, while changes in Group PYR were slight, compared to Group CON under a light microscope (Fig. 2 A and B). Further, alterations in immuno uorescence of dgen-MBP in L5 DRG and L5 spinal cord scan were in parallel with changes above in various groups following the 8-week HU in rats (Fig. 3 A and B; Fig. 4 A and B).
These data rst evidenced that HU-induced microstructure damage in DRG could be signi cantly alleviated by oral pyruvate. Oral pyruvate protected Nissl body from HU injury was supported by recent ndings that pyruvate inhibited ER stress in high glucose, in vitro (Accepted, FEBS Open Bio, 2020). Notably, an analogical protection of Nissl staining in motor neurons by ethyl pyruvate (EP, a derivate of sodium pyruvate) in spinal cord ischemia reperfusion injury was also reported recently [20].
Functional changes GDNF has a strong neurotrophic effect on spinal motor neurons in prevention of their degeneration and can rescue cells from apoptosis as well as their death following injury. When nerve damage occurs, the GDNF content can be altered with early stimulated increase following by subsequent decline [21,22]. As previously reported that DRG injury induced GDNF reduction [8,23,24], the results here showed that relative GDNF proteins in Groups SUS and SAL were signi cantly decreased, compared with those in Groups CON and PYR following an 8-week HU. In contrary, the GDNF expression in Group PYR was accumulated surprisingly to over 2-fold of those in Groups SUS and SAL, even much more than that in Group CON. The results of GDNF mRNA detection in the groups were also comparable with changes above (data not shown) [25]. The results evidently showed that the DRG injury at the early stage had a prominent feedback increase of GDNF protein with pyruvate supplementation daily (Fig. 5). The underlying causes of profoundly enhanced GDNF protein by pyruvate are unknown at present. To our best knowledge, there has been no report regarding the pyruvate stimulation of GDNF expression, in vivo, though EP also showed a mimic effect in primary astrocyte cultures, in vitro [26]. Interestingly, it has been known that histone deacetylase inhibitors (HDACi), like sodium butyrate, up-regulate GDNF gene transcription in astrocytes to protect dopaminergic neurons, via HDAC inhibition [27], and pyruvate is one of HDACis [28,29]. Nevertheless, it is rst revealed that oral pyruvate can robustly promote the secretion of GDNF that attenuates DRG damage during HU. Currently, exogenous GDNF is suitable for SCI treatment [30,31]. Thus, it is assumed that more bene ts may be targeted with pyruvate addition in a novel GDNF combination intervention for SCI treatment.
GFAP is a unique cytoskeleton protein in astrocytes of DRG, the main component of intermediate lament network that warrants cell integrity and resilience. It is considered a sign of the maturation of astrocytes and mainly involved in the structure and metabolism of neurons, but rarely expressed in physiological conditions [32]. Results here showed that relative GFAP expressions were signi cantly increased in Groups SUS and SAL (Fig. 6), indicating that astrocytes were activated in DRG injury and the predominant expression of GFAP might be associated with the repairment of DRG damage. It was signi cantly declined in Group PYR than in both Groups SUS and SAL, but still raised relative to Group CON. These data demonstrated that oral pyruvate could provide the neuroprotective effect on damaged DRG in the early HU injury, as previously indicated that pyruvate could limit or better control the reactive astrocyte proliferation (gliosis) and sustain the morphological integrity of injured astrocytes [9,33]. Also, GFAP mRNA detection showed changes in the groups in a similar manner (data not shown). However, current pyruvate dosage could not fully prevent the damage. Again, EP was reported to attenuate its expression in astrocytes in rats with ischemic spinal cord [34].
Intriguingly, GDNF is reductive of astrogliosis, as early found that GDNF promoted the spinal cord graft survival and growth, but reduced GFAP immunoreactivity [35,36]. Present results rst demonstrated that oral pyruvate upregulated the former and attenuated the later in DRG nerves during HU. These results suggest that GDNF and GFAP as well as pyruvate play a pivotal role in maintaining environmental stability, neuronal survival and plasticity repair in the nervous system [32].
The levels of ATP and ATPase in Groups SUS and SAL were signi cantly decreased relative to those in Groups CON and PYR (Fig. 7 A and B). Results evidenced that cellular energy metabolism was declined by HU injury and oral pyruvate in Group PYR profoundly preserved energy metabolism in DRG nerves during the period of microgravity, so that nerve tissues' structure and function were mostly protected. Although ATP and ATPase levels in Group PYR were relatively preserved, they were still signi cantly inferior to those in Group CON. Metabolic studies reveal that glycolysis is usually stimulated with inhibition of glucose oxidative phosphorylation in microgravity conditions [37]. Blood pyruvate, as pyruvate dehydrogenase kinase (PDK) inhibitor like dichloroacetate (DCA), can easily enter mitochondria of nervous cells, activate the depressed pyruvate dehydrogenase (PDH) via PDK inhibition to even full context and promote the tricarboxylic acid (TCA) cycle metabolism [35,38,39], though the PDH activity was not detected here (see below). Further, its anti-oxidative stress and anti-in ammation may simultaneously bene t in TCA cycle function in addition as an important energetic substrate [10,13,40]. As a result, ATP production and ATPase activity were signi cantly preserved in Group PYR.
In all, these results discovered that Group PYR showed morphological alterations close to Group CON, while the energy metabolic disorder and cell dysfunction were strikingly attenuated by oral pyruvate, but not fully recovered with current pyruvate dosage from the HU injury. The reason of divergence between histological and functional changes was not clear at present. The dose-effect curve remains explored.

Oral pyruvate implication in space ight
Present results provide additional evidence with 0.55% pyruvate that a low oral pyruvate (0.33-4% in drink water) does protect the structure and function of nerve systems [13,[41][42][43][44]. Thus, oral pyruvate may be useful in attenuation or prevention of DRG, SCI and other disorders for astronauts in space missions.
The dose of oral pyruvate (0.55%) in the experiment was empirically chosen and the ideal concentration and optimal dosage of oral pyruvate solution for astronauts remain explored. Notably, oral or intravenous (IV) pyruvate in large dosages has been reported in several preliminary clinical trials with safety and effectiveness in absence of clinical toxicity [45][46][47][48][49]. However, oral ingestion of small doses of pyruvate alone does not elevate plasma pyruvate, but a large dosage is of gastrointestinal irritation in humans 45,50]. On the other hand, the novel pyruvate solution (Pyr-ORS): equimolar pyruvate replacement of bicarbonate or citrate in the World Health Organization-guided ORS (WHO-ORS) [51], containing 0.35% pyruvate and 1.35% glucose as an experimental drink water, effectively preserved neurons structure and function in hippocampal CA1 regions following asphyxial cardiac arrest in rats [13]. Further, Pyr-ORS and 1% pyruvate as drink water comparably showed a robust PDH activity renovated and blood sugar and kidney function preserved in diabetic db/db mice [52].
Several animal studies strongly indicate that around 10.0g of oral sodium pyruvate daily as a functional drink in a formula as Pyr-ORS may be bene cial and feasible for astronauts in space ights due to Na + -Glucose co-transporter existed in intestinal epithelium in mammalian [13,42,51]. As to metabolic pro le of post-pyruvate, it was recently illustrated [53,54]. Further studies are required and the clinical trials with Pyr-ORS formula, modi ed if needed, are warranted in astronauts' training ground and in general SCI treatment.
Finally, it has to be pointed out that although EP showed the protection of nervous system [20,26,34], there is a distinct difference between them: EP does not work in humans [54], leading to its clinical perspective hopelessly. Limitations 1) Present data focused on a principal proof, in vivo, for oral pyruvate protection in rats subjected to HU injury. Nevertheless, cell experiments, in vitro, to demonstrate the direct effect of pyruvate on DRG nerve cells following weightlessness would further support the hypothesis and explore the underlying molecular mechanisms of the pyruvate action. 2) If double immuno uorescence with anti-neuro lament antibody in addition to dgen-MBP expression was performed, the MBP pictures would be more visible with pyruvate treatment. 3) Oral pyruvate in Pyr-ORS (0.35%) was not compared in the experiment; to establish a doseresponse manner would further illustrate oral pyruvate effects on DRG in rats subjected to microgravity.

Conclusion
Hindlimb unweighting resulted in signi cant L5 DRG damage in both tissue morphology and cell function in rats. Present results rst demonstrate that oral pyruvate can effectively protect the DRG tissues against HU injury. The optimal dosage of oral pyruvate and pyruvate formula remain explored for astronauts' use in space missions. Hematoxylin-Eosin (HE) staining of L5 DRG tissues in various treatment groups        Relative GFAP expressions by Western blotting in L5 DRG tissues in various treatment groups