Spatial learning and memory of young and aging rats following injection with human Wharton's jelly-mesenchymal stem cells

Human Wharton’s jelly‐mesenchymal stem cells (hWJ‐MSC) are an emerging potential source of stem cells derived from the umbilical cord. Previous studies have shown their potential as treatment for traumatic brain injury and Parkinson’s disease. However, no study has yet investigated the effect of hWJ‐MSC injections in countering spatial learning and memory impairment in aging rats. The effect of hWJ‐MSC injection on young rats is also unknown. The objective of this research was to analyze the effect of an hWJ‐MSC injection on spatial learning, memory, density of putative neural progenitor cells (pNPC), and neuronal apoptosis in the dentate gyrus (DG) of young and aging rats. Injection of hWJ‐MSC did not change spatial learning and memory in young rats until two months post‐injection. This might be due to retained pNPC density and neuronal apoptosis in the DG of young rats after injection of hWJ‐MSC. In contrast, injection of hWJ‐MSC promoted both spatial learning and memory in aging rats, a finding that might be attributable to the increased pNPC density and attenuated neuronal apoptosis in DG of aging rats during the two months post‐injection. Our study suggests that a single injection of hWJ‐MSC might be sufficient to promote improvement in long‐term learning and memory in aging rats.


Introduction
Aging is process of gradual deterioration across multiple organs which is attributed to decreasing regenerative abil ity (Lazarov et al. 2010). The brain is one of the organs affected by the aging process. Brain aging can cause spa tial learning and memory impairment (Drapeau et al. 2003; Terry et al. 2011. Spatial learning is the process of encod ing information regarding position based on environmen tal cues into the cognitive map to guide navigational pro cess (Floresco 2010); and memory is encoded and stored information which can be recalled when needed (Smith 1980). Spatial learning and memory are regulated by neuro genesis and apoptosis in the hippocampus (Dupret et al. 2007). Neurogenesis is the formation of newborn neurons by neural progenitor cells (Deng et al. 2010). Newborn neurons are important to encoding new memories in the learning process (Deng et al. 2010; Kempermann 2002. Neurons in the hippocampus also undergo apoptosis, a process of programmed cell death (Elmore 2007). Pro gressive cell death in the hippocampus can cause impair ment in spatial learning ability.
Previous studies have shown many attempts to counter neuronal damage caused by aging in the hippocampus such as dietary regulations, exercise, and injection of young blood (Galli et al. 2002; Speisman et al. 2013; Villeda et al. 2014. Another alternative to prevent neuronal degrada tion is stem cells injection. Stem cells are undifferentiated cells which are able to selfrenew and differentiate into many types of cells (Pera et al. 2000). Stem cells could be a potential antiaging agent because they could directly re pair damaged tissue by differentiating into many cell types (Boyette and Tuan 2014). Stem cells also produce growth factor, which could inhibit apoptosis and induce prolifer ation of progenitor cells (Bali et al. 2016).
Wharton's jellymesenchymal stem cells (WJMSC) are a new emerging potential source of stem cells. Pre vious studies have showed that WJMSC can be isolated robustly, have a higher proliferation rate than other mes enchymal stem cells, and multipotent ability to differen tiate into chondrocyte adipocyte, osteocyte, as well as transdifferentiate into hepatocyte, cardiomyocyte, glia, and neurons (Mitchell et al. 2003; Nekanti et al. 2010; Wang et al. 2004; Weiss et al. 2006. WJMSC also have immunosuppressive ability, which can reduce immune re jection upon injection (Weiss et al. 2008; Zhou et al. 2011. Moreover, the isolation process is ethically unchallenged because the umbilical cord is often regarded as postbirth waste (Wang et al. 2004; Weiss et al. 2006. Human Wharton's jellymesenchymal stem cells (hWJMSC) has been used in preclinical tests as treat ment for traumatic brain injury and Parkinson's disease model (Cheng et al. 2015; Weiss et al. 2006. However, to our best knowledge, no study has shown the effect of hWJMSC injection to counter spatial learning and mem ory impairment in aging rats. The effect of hWJMSC on young rats is also yet to be known. Moreover, the effect of hWJMSC injection on the cellular conditions in den tate gyrus of young and aging rats also unknown. There fore, this study aimed to analyze the effect of hWJMSC injection to: spatial learning, memory, density of putative neural progenitor cells (pNPC), and neuronal apoptosis in dentate gyrus (DG) of young and aging rats.

Experimental procedures
This research has been approved by Animal Care and Use Committee (ACUC), Educational Animal Hospital, Fac ulty of Veterinary Medicine, Bogor Agricultural Univer sity. All rats used in this research were from Unit Pen gelola Hewan Laboratorium (UPHL), Faculty of Veteri nary Medicine, Bogor Agricultural University. This re search used 9 young rats (± 5 month old) and 9 aging rats (± 24 month old) (Rattus norvegicus) strain Sprague Dawley, each divided into 3 groups: 1 control group and 2 treatment groups. Each group consisted of 3 rats. Con trol group was given 0.5 mL saline (vehicle). Treatment groups were given 1.5×106 hWJMSC cells/rat in vehicle. hWJMSC was prepared in Stem Cell and Cancer Institute (Indonesia). Isolation and characterization of hWJMSC used in this study has been described in Antoninus et al. (2012). All treatments were given via tail intravenous in jection. Rats were anesthetized with ketamine 10 mg/kg and xylazine 1 mg/kg of body weight. Treatment groups were divided based on euthanasia period: 1 month (SC1) and 2 month (SC2) postinjection to assess longterm ef fect of hWJMSC injection.

Spatial learning test
Spatial learning ability was measured using Ymaze al ternation test. Test protocol was based on Juliandi et al. (2015). Spatial learning ability was measured in 3 time points: before treatment (7 d before treatment), 1month posttreatment (M1), and 2 month posttreatment (M2). Y maze alternation test was done using Yarm maze appara tus which consisted of 3 identical arms (40×10×30 (H) cm) at 120°away from each other. The center platform had 10 cm sidelength. Ymaze alternation test was performed by counting rat's alternation in Yarm maze for 5 min. Alter nations were counted through video analysis. One alter nation was defined as 3 consecutive arms entered by rats. Alternation was considered correct if rats entered 3 dif ferent arms continuously which is based on rats' instinct to explore new environment. The apparatus was cleaned with ethanol 70% after each session to minimize possible olfactory cue. Spatial learning ability was measured by an alyzing correct alternation percentage, which is percentage of ratio between total correct alternation and total alterna tion.

Object recognition memory test
Rat's memory was measured by novel object recognition test (NORT). NORT protocol was based on Bevins and Besheer (2006). This test was based on rats preference to novel object. NORT was done in an acrylic compart ment (30×30×32 (H) cm). NORT consisted of two phase: acquisition and test phase. Acquisition phase was done 1 d before treatment. In acquisition phase, each rat freely explored the apparatus which had two identical objects for 10 min. The identical objects used in this experiment are two plastic yellow balls with 10 cm diameter. The pur pose of this acquisition phase was to familiarize each rat with those two identical objects which referred as famil iar objects. After acquisition phase, rat was put in their house cage and given 3 delay timepoints to test the mem ory retention. Test phase in this research was done in 1 d (D1) posttreatment to assess shortterm memory, 1 month (M1), and 2 month (M2) posttreatment to measure long term memory. In test phase, each rat explored the appa ratus which has one of the familiar object replaced with novel object for 3 min. Novel object used in this research was a stuffed animal (chicken) (10×8×9 (H) cm). Appa ratus and 2 objects were cleaned after each session with ethanol 70% to decrease olfactory cue. Rat's memory was measured using discrimination ratio which is percentage of ratio between time to explore novel object with total to explore both novel and familiar object. Time to explore was measured using video analysis. Rat was considered exploring one object if rat's snout was within 3 cm from the object.

Hematoxylin-Eosin staining
Rat's brain was isolated using perfusion method. Rat's brain was fixed in buffered neutral formalin (BNF) 10% for 4 d. Coronal trimming was done in 3 mm poste rior bregma. Brain section containing hippocampus was made into 5 µm paraffin sections. Paraffin sections were stained with hematoxylineosin and observed with light microscope which connected to camera. Photomicrograph of dentate gyrus sections were captured using Indomi croView software (Indomicro, Indonesia) for further anal ysis. Putative neural progenitor cells (pNPC) and apop totic neuron in dentate gyrus (DG) were identified based on Han et al. (2008) and Hashem et al. (2010). Counting of pNPC and DG apoptotic neurons density were done using ImageJ 1.50i software (NIH, USA).

Immunohistochemistry
Antigen retrieval was conducted either by incubation in L.A.B. Solution (Polysciences) for 15 min at room temper ature (RT) or by autoclaving in Target Retrieval Solution (Dako) for 15 min at 105°C. After 3 washes with PBS, the sections were incubated for 1 h at RT in blocking solution (PBS containing 3% FBS and 0.1% Triton X100). They were then incubated overnight at 4°C with the appropri ate primary antibodies. The following primary antibodies were used: mouse antihuman Ki67 (1:500, BD Pharmin gen) and goat antiactive caspase3 (1:500, R&D Systems). After 3 washes with PBS, the sections were incubated for 2 h at RT with the appropriate secondary antibodies. The fol lowing secondary antibodies were used: Cy3conjugated donkey antimouse and Cy5conjugated donkey antigoat (1:500, Jackson ImmunoResearch). After 3 washes with PBS, nuclei were stained for 15 min at RT with Hoechst 33258 (Nacalai Tesque). Sections were mounted on cover slips with ImmuMount (Thermo Scientific). Cell count ing was conducted manually and photographed using LSM 710 confocal microscope (Zeiss) equipped with a camera and appropriate epifluorescence filters. The total num ber of positive cells was counted in every twelve section (480 µm apart). Positive cells were counted throughout the rostrocaudal extent of the granule cell layer (GCL), and the derived numbers were multiplied by 12 (slice series) to obtain total cell number per GCL.

Data analysis
Spatial learning and memory were analyzed with one way ANOVA, twoway ANOVA and independent ttest. Analysis on pNPC density and neuronal apoptosis in DG were done using oneway ANOVA with posthoc test us ing Tukey honest significant difference (HSD). Statistical analysis was performed in SPSS 22 (IBM, USA) with min imum significance level p<0.05.

Spatial learning ability
Young rats in control group had an increased spatial learn ing ability from before treatment up to 2 month post treatment (p=0.046, oneway ANOVA) (Figure 1a). In comparison, aging rats in control group had a consistent spatial learning ability throughout 2 month (p=0.858, one way ANOVA) (Figure 1b). This result shows that young rats were able to learn the Ymaze apparatus. In compari son, aging rats did not learn the apparatus throughout test period.
Injection of hWJMSC did not alter spatial learning of young rats compared to control group (p=0.118, twoway ANOVA) (Figure 1a). Shift in spatial learning ability was not observed in young rats injected with hWJMSC from before treatment to 2 month postinjection (p=0.063, one way ANOVA). Conversely, injection of hWJMSC pro moted spatial learning ability of aging rats (p=0.013, two way ANOVA) (Figure 1b). Increased spatial learning abil ity was observed in aging rats injected with hWJMSC (p=0.0052, oneway ANOVA). Rats injected with hWJ MSC had higher spatial learning ability in 1 month post injection (M1) (p=0.0098, independent ttest) and 2 month postinjection (M2) (p=0.021, independent ttest), com pared with control group. This result suggests that hWJ MSC had a longterm effect on promoting spatial learning of aging rats.

Object recognition memory
All rats had 50% discrimination ratio in the acquisi tion phase (Figure 2). This result indicates that all rats were exploring both familiar objects on equal proportion. Injection of hWJMSC did not alter object recognition memory of young rats (p=0.108, twoway ANOVA) (Fig  ure 2a). Retained object recognition memory were ob served on both young rats in control group (p=0.788, one way ANOVA) and young rats injected with hWJMSC (p=0.110, oneway ANOVA).
Treatment with hWJMSC promoted object recogni tion memory of aging rats (p=0.0013, twoway ANOVA) (Figure 2b). Compared to control group, object recogni tion memory of aging rats in hWJMSC group was signif icantly higher in 1 d (D1) (p=0.042, independent ttest) and 1 month (M1) (p=0.0097, independent ttest) post injection. However, no significant differences were ob served in object recognition memory between aging rats in control group and hWJMSC group in 2 month (M2) posttreatment (p=0.110, independent ttest).
No significant changes were observed in the con trol group of aging rats (p=0.345, oneway ANOVA). In contrary, significant increase was observed in aging rats injected with hWJMSC (p=0.0026, oneway ANOVA). Object recognition memory of aging rats in hWJMSC group was increased and peaked in 1 d postinjection (D1) (p=0.042, Tukey HSD) compared to acquisition phase. This result shows that hWJMSC promoted shortterm memory of aging rats. Significant increase also observed in 1 month post injection compared to acquisition phase (p=0.029, Tukey HSD). Object recognition memory in 1 month postinjection (M1) was not significantly differ ent compared to 1 d postinjection (D1) (p=0.252, Tukey HSD). This result indicates memory retention from short term to longterm. However, significant decrease in object recognition was observed in 2 month postinjection (M2) in comparison to 1 d postinjection (D1) (p=0.252, Tukey HSD). This result shows that injection of hWJMSC did not produce permanent effect on object recognition of ag ing rats.

Putative neural progenitor cells density and neuronal apoptosis of young rats
Photomicrograph of rat's brain coronal section showed dentate gyrus as part of hippocampus (Figure 3a). Pho tomicrograph of dentate gyrus showed putative neural pro genitor cells (pNPC) which identified by the nongranural morphology, dark cytoplasm, and resides in the deepest part of granular cell layer (GCL) or the subgranular zone (SGZ) (Figure 3b). Apoptotic neurons in dentate gyrus were identified by vacuolation, shrunken cytoplasm, and pyknotic nucleus (Figure 3c).
Neuronal apoptosis in dentate gyrus of young rats was consistent despite treatment with hWJMSC (p=0.522, oneway ANOVA) (Figure 4b). Increased pNPC density without significant change on neuronal apoptosis showed that injection of hWJMSC induce proliferation of neural progenitor cells in dentate gyrus.

Discussion
To our knowledge, this is the first report on the effect of human Wharton's jellymesenchymal stem cells (hWJ MSC) xenogeneic transplantation in vivo to both aging and young rats. Injection of hWJMSC produced age specific effect. Treatment with hWJMSC to young rats did not change the spatial learning ( Figure 1a) and object recognition memory (Figure 1a). In comparison, signifi cant change was observed in spatial learning (Figure 1b) as well as object recognition memory (Figure 2b) of ag ing rats. This agespecific effect might be attributable to difference in proliferation rate of young and aging rats. Previous studies have presumed that brain might has certain neurogenesis threshold to retain spatial learning ability and memory (Drapeau et al. 2003; Juliandi et al. 2015. Rats with NPC proliferation above threshold, such as in young rats, would not produce higher spatial learn ing and memory despite induction of NPC proliferation. This research supports this hypothesis. Injection of hWJ MSC to young rats did increase pNPC density (Figure 4a), which reflect NPC proliferation. However, this induction was not followed by change in spatial learning and object recognition memory. We presumed that formation of new born neurons which stores new memories are already suf ficient or above the threshold, so induction of NPC prolif eration after hWJMSC injection did not change the spatial learning and object recognition memory.
Rats with neurogenesis below threshold, such as caused by aging, showed spatial learning impairments (Drapeau et al. 2003). Aging process is known to de crease NPC proliferation due to increasing quiescence of NPC (Heine et al. 2004; Hattiangady andShetty 2008). This agerelated decline could cause reduction in newborn neurons formation which is vital in memory and learn ing process (Bernal and Peterson 2004). Indeed, in this study, aging rats in control group had lower pNPC density, spatial learning ability, as well as memory. In line, ag ing rats injected with hWJMSC had higher pNPC density which showed induction of proliferation as well as spa tial learning and memory. Therefore, we presumed that hWJMSC injection might be able to induce NPC prolifer ation to above threshold which resulted in increased spatial learning and memory.
We also found that hWJMSC could attenuate neu ronal apoptosis in DG. Inhibition of neuronal apoptosis in DG is important to retain stored memories and spa tial learning ability (Ramírez et al. 2005). Inhibition of neuronal apoptosis might also reflect increased newborn neurons survival. Approximately 9,000 newborn neurons were produced daily in young rats (Heine et al. 2004). In tegration of newborn neurons in massive numbers could cause disruption in hippocampal circuitry, which leads to seizures and forgetting (Meltzer et al. 2005). Therefore, 8090% of newborn neurons produced were removed by apoptosis (li Ming and Song 2011; Sierra et al. 2010; Kem permann et al. 2003. Injection of hWJMSC in this study increased prolifer ation of NPC. Induced NPC proliferation should have been followed by increased neuronal apoptosis. However, rats injected with hWJMSC showed reduced neuronal apopto sis in DG of aging rats, which suggests increased survival of newborn neurons. Similar result was found in previous studies in which aging rats were given exercise treatment (Kim et al. 2010; Lee et al. 2005. Combined, this results also suggest that acquisition of newborn neurons is more selective in aging rats compared to young rats. Increased survival of newborn neurons in aging rats also could pro mote spatial learning and memory (Drapeau et al. 2003).
The precise mechanism on how hWJMSC could af fect cellular condition in dentate gyrus and learning abil ity is unknown. Previous study has shown that porcine WJMSC could integrate and proliferate in substantia ni gra and ventral tegmental area until 8weeks postinjection in Parkinsonian rat model (Medicetty et al. 2004). Porcine WJMSC injected directly into rat's brain also migrate into region in ventral of corpus callosum on 6weeks post injection . In general, WJMSC could survive in rats even long after injection period. However, none of the above uses systemic injection as we did in this study.
Indeed, the main challenge in systemic injection of stem cells is whether they could penetrate blood brain bar rier (BBB) which tightly regulates substances going into the brain (Goncharova et al. 2014). Elahy et al. (2015) showed that increased capillary permeability happens due to BBB breakdown in normal aged mice (24 month) which is the same age as rats used in this study. Furthermore, Montagne et al. (2015) showed that aging can cause BBB breakdown specifically in hippocampus. This might open up possibility of hWJMSC integration into aging hip pocampus and attenuates neuronal degeneration either via celltocell contact or cytokine secretion. This study can not provide data to show whether hWJMSC is indeed could migrate into dentate gyrus. Therefore, further stud ies using specific marker such as human nuclei antibody are needed to confirm the integration and survival of hWJ MSC injected in aging rats.
Regardless, we presumed that induction of NPC pro liferation and inhibition of neuronal apoptosis in DG might be attributable to secretion of growth factor by hWJMSC. Yang et al. (2008) showed that hWJMSC could produce vascular endothelial growth factor (VEGF) and basic fi broblast growth factor (bFGF) which are wellknown to increase NPC proliferation (Jin et al. 2002; Lin et al. 2015.
Neuroprotective effect of hWJMSC might due to secre tion of VEGF and granulocytecolony stimulating factor (GCSF) (Koh et al. 2008; Yang et al. 2008). Both G CSF and VEGF are known to inhibit neuronal apoptosis (Yata et al. 2007; Solaroglu et al. 2009; Sun et al. 2003. However, as aforementioned, further research is needed, especially to confirm whether the effect is indeed caused solely by growth factor secreted by hWJMSC, or by pro motion of endogenous growth factor produced in SGZ, or even both

Conclusions
In conclusion, this study has provided information regard ing the effect of hWJMSC injection to spatial learning and memory of young and aging rats. Intravenous injection of hWJMSC did not alter both spatial learning and mem ory of youth rats. In contrast, hWJMSC promoted spatial learning as well as memory of aging rats. Increased pNPC density and attenuation of neuronal apoptosis in DG might be responsible for the promotion of both spatial learning and memory of aging rats.