Adult neurogenic process in the subventricular zone‐olfactory bulb system is regulated by Tau protein under prolonged stress

Abstract Objectives The area of the subventricular zone (SVZ) in the adult brain exhibits the highest number of proliferative cells, which, together with the olfactory bulb (OB), maintains constant brain plasticity through the generation, migration and integration of newly born neurons. Despite Tau and its malfunction is increasingly related to deficits of adult hippocampal neurogenesis and brain plasticity under pathological conditions [e.g. in Alzheimer's disease (AD)], it remains unknown whether Tau plays a role in the neurogenic process of the SVZ and OB system under conditions of chronic stress, a well‐known sculptor of brain and risk factor for AD. Materials and methods Different types of newly born cells in SVZ and OB were analysed in animals that lack Tau gene (Tau‐KO) and their wild‐type littermates (WT) under control or chronic stress conditions. Results We demonstrate that chronic stress reduced the number of proliferating cells and neuroblasts in the SVZ leading to decreased number of newborn neurons in the OB of adult WT, but not Tau‐KO, mice. Interestingly, while stress‐evoked changes were not detected in OB granular cell layer, Tau‐KO exhibited increased number of mature neurons in this layer indicating altered neuronal migration due to Tau loss. Conclusions Our findings suggest the critical involvement of Tau in the neurogenesis suppression of SVZ and OB neurogenic niche under stressful conditions highlighting the role of Tau protein as an essential regulator of stress‐driven plasticity deficits.


| INTRODUC TI ON
The brain is the most adaptive of all organs due to its continuous plasticity in response to a variety of internal and environmental stimuli. A dynamic form of neuronal plasticity in the adult brain is neurogenesis, the process of generation of new, functional neurons from neural stem cells (NSCs) and progenitor cells, 1 which enables the brain to adapt to the constantly evolving interaction between environmental signals and the brain's internal reaction to these stimuli. 2 The two main neurogenic niches of the adult brain are the hippocampal dentate gyrus (DG) and the subventricular zone (SVZ) of lateral ventricles 3 ; the latter exhibits the biggest amount of proliferative cells in the brain. 4,5 Newly born neurons generated in the SVZ of the adult brain migrate through the rostral migratory stream to the olfactory bulb (OB) where they differentiate into local circuit interneurons that are implicated in learning and memory processes related to smell sensation in rodents. 6,7 In brief, type B cells are quiescent neural stem cells (NSCs) expressing glial fibrillary acidic protein (GFAP) that give rise to type C cells (also known as transientamplifying progenitors); type C cells give rise to type A cells, which are neuroblasts expressing doublecortin and migrate to the OB. 8,9 In the OB, these neuroblasts differentiate into interneurons and migrate radially to the outer cell layers, namely granular cell layer (GCL), mitral cell layer (MCL) and glomerular cell layer (GL). Specifically, they differentiate in the GCL and MCL into granule cells (GC) and in the GL into periglomerular cells (PGC). 10,11 Additionally, it has also been described that type B cells generate oligodendrocytes-see also Figure 1. 12,13 Although the extent and relevance of adult neurogenesis in humans are currently debated, 14,15 accumulating evidence suggests that neurogenesis persists in the adult brain of both humans and rodent animals during the entire lifespan while it drops in Alzheimer's disease (AD) [16][17][18][19][20] and other pathological conditions causally related to AD, such as depression and stress. [21][22][23] Chronic stress, a major precipitant of depression and AD [24][25][26][27] is known to impair brain plasticity, including suppression of neurogenesis. 23,[28][29][30][31][32] Recent evidence about stress-driven neurogenic deficits highlights the critical role for the cytoskeletal Tau protein 22,33 a prominent stabilizer of microtubules (MT), 34 which promotes co-organization of MT and actin networks. [35][36][37][38] However, our knowledge related to the impact of chronic stress on adult neurogenesis is mainly based on the hippocampus, as the vast majority of studies have neglected the other main neurogenic niche of the adult brain, the SVZ. As a matter of fact, the SVZ area exhibits the highest number of proliferative cells in the adult brain 4,5 and constitutes the origin of the newly born cells/neurons that migrate into the OB under control conditions 39 and to other neocortical sites under injury (eg. after stroke or trauma). 40 Also, hippocampal and SVZ-OB cytogenic areas exhibit essential differences in their anatomical/layer organization and input received from other brain areas as well as the type of newborn cells generated in each of these two neurogenic niches. 41 Moreover, their vulnerability to stress or pharmacological/irradiation treatment may be different, as previous studies suggested. 42,43 Despite that the above findings point towards essential differences between these two neurogenic niches in the adult brain, our knowledge about the cell-type specific impact on chronic stress on SVZ-OB system and the underlying mechanisms remain poor.
In light of the limited and conflicting evidence about whether exposure to stressful conditions affects (or not) the SVZ and OB cytogenesis in the adult brain 42,43 and the selective involvement of Tau in specific types of newborn cells (eg DG newborn neurons, but not glial cells), the current study aims to clarify the effect of stress on different populations of newborn cells in the SVZ-OB system. For that purpose, we have exposed animals lacking Tau protein (Tau-KO) and their wild type (WT) littermates to a chronic unpredictable stress (CUS) paradigm and evaluated differences in the cell population resident in the SVZ and OB neurogenic niches. Our findings suggest that exposure to chronic stress suppresses proliferation as well as neuronal differentiation and maturation in the SVZ and OB of the adult brain while the absence of Tau protein diminishes these neuroplastic effects of stress highlighting an essential role for Tau in the mechanisms through which prolonged stressful conditions damage brain plasticity.

| Animals
Twenty-eight male mice lacking Tau protein (Tau-KO) and their wild-type littermates (6-7 months old; C57BL/6J background) were used in this study divided into stressed and control groups (7 ani-

| Chronic stress paradigm
Animals were exposed to a 9-week chronic unpredictable stress (CUS) paradigm during the daily period of light, while control (nonstressed; CON) mice remained undisturbed in their home cages.
The CUS protocol included 4 different stressors: restraint, vibrating platform, overcrowding and exposure to a hot air stream. Animals were exposed to one stressor per day for 3 hours (restraint, vibrating platform, overcrowding) or 30 minutes (hot air stream). The order of stressors and the time of the day at which the stressor was applied were randomly chosen and varied among weeks to promote unpredictability, as previously described. 22,25,28,44 At the end of the CUS protocol, mice body weight was measured and blood was collected from all animals. Blood serum was isolated after centrifugation and corticosterone (CORT) levels were measured using a radioimmunoassay kit (R&D Systems) according to the manufacturer's instructions.

| BrdU treatment
For assessment of cell proliferation, a set of control and stressed WT and Tau-KO animals (four animals per group) were injected with 5-bromo-2′-deoxyuridine (BrdU; 50 mg/kg per day) for 3 consecutive days before killing. For cell survival monitoring, another set of control and stressed animals of both genotypes (three animals per group) were injected with BrdU (50 mg/kg per day) for 3 consecutive days, 4 weeks before killing-see also Figure 1A.
F I G U R E 1 Chronic stress suppresses the number of proliferating cells and neuroblasts in the adult subventricular zone of WT, but not Tau-KO, animals. A, Schematic representation of the experimental design where wild-type (WT) and Tau-knockout (Tau-KO) mice were divided into control (CON) and chronic stress (STR) groups. Animals of all groups were randomly divided into two groups receiving 5-bromo-2′-deoxyuridine (BrdU) injections before sacrifice (left panel) and 4 weeks before sacrifice (right panel). B,C, Schematic illustration of the mouse brain (B) highlighting the neurogenic areas of the subventricular zone (SVZ) and olfactory bulb (OB) as well as different types of newly born SVZ cells analysed and the markers used for their monitoring. D-F, Representative microphotograph of BrdU/DCX double-labelled cells (arrow head) in the SVZ (D). Chronic stress evoked a decrease in BrdU-positive cell density (reflecting proliferating cells) in WT, but not Tau-KO, animals. Note that stressed Tau-KO animals present higher number of proliferating cells when compared to stressed WT animals (E). Similarly, stress reduced the percentage of DCX/BrdU double-labelled cells (reflecting neuroblasts) only in WT animals (F). All numerical data are shown as mean ± s.e.m (*P < .05). CON, control-non-stressed; STR, stressed; BrdU, 5-bromo-2′-deoxyuridine; DCX, doublecortin; WT, wild type; Tau-KO, Tau-knockout

| Tissue preparation
At the end of the CUS protocol, animals were deeply anesthetized (ketamine hydrochloride [150 mg/kg] plus medetomidine [0.3 mg/kg]) and transcardially perfused with saline followed by ice-cold 4% paraformaldehyde perfusion. Brains were removed, post-fixed in 4% paraformaldehyde for 2 hours and then transferred to a 30% sucrose solution until they sunk. Then, brains were included in optimal cutting temperature compound (OCT; Tissue Tek, Sakura FineTek), snap-frozen in liquid nitrogen with 2-methylbutane and sectioned in a cryostat (Leica CM1900) into 20 μm sections.

| Immunofluorescence staining
Coronal brain sections of SVZ and OB were double-stained for BrdU

| Open field
We used an open-field square arena (43.2 cm × 43.2 cm) surrounded by tall Perspex walls (Med Associates Inc). Each mouse was placed in the centre and allowed to explore the arena for 10 minutes. Infrared beams and manufacturer's software were used to automatically register animals' movements.

| Ultrasonic vocalizations
Measurement of ultrasonic vocalizations (USVs) was performed as previously described with some modifications. 45 Briefly, each animal was placed in a cage for 24 hours. Then, the animal was in close proximity with a female animal, and USVs were recorded for 15 minutes using the Avisoft-Recorder (version 5.1.04) and manually analysed with AvisoftSAS Lab Pro (version 5.1.22, Avisoft Bioacoustics).

| Novel object recognition test
The test arena consisted of a white rectangular box (33 cm × 33 cm × 33 cm). Mice were placed for 20 minutes during 3 consecutive days inside the test arena (habituation). On the following day, mice were placed in the test arena, which contained two identical objects equally distant, and returned to their home cage after 10 minutes of exploration. The next day, the animals were presented a novel object (NO) and one old, familiar object (FO) for 10 minutes; both objects were generally similar regarding height and volume but they were different in shape, colour and texture. Animal's behaviour was recorded, and the time spent

| Statistical analysis
Data were analysed using two-way analysis of variance (ANOVA) before the application of appropriate post hoc pair-wise comparisons (GraphPad Prism v.6.01; GraphPad Software). Differences were considered statistically significant when P <.05. Results are presented as mean ± SEM

| Exposure to chronic stress reduces proliferating cells in the subventricular zone of the adult brain while Tau ablation blocks this stress effect
For clarifying the impact of prolonged stress exposure on the neurogenic niche of the subventricular zone (SVZ) -olfactory bulb (OB) system and monitoring the potential role of Tau in the stress-driven regulation of cytogenesis, we exposed wild-type (WT) mice and their littermates lacking Tau protein (Tau-KO) to a chronic unpredictable stress (CUS) paradigm for 9 weeks (see Figure 1). For detection of newly generated cells in the SVZ-OB system, we followed the widely used approach of administration of the synthetic nucleotide bromodeoxyuridine (BrdU), which is incorporated into the newly synthesized DNA during the S phase of the cell cycle. To evaluate proliferation in the SVZ, animals were injected with BrdU for three consecutive days before killing ( Figure 1A

| Chronic stress affects neuroblasts in the SVZ in a Tau-dependent manner
Neural stem cells (NSCs) in the SVZ may give rise to neuronal and oligodendrocytes precursors ( Figure 1C). To monitor the impact of chronic stress on SVZ neuroblasts, brain sections from mice injected with BrdU before sacrifice ( Figure 1D) were double-stained with antibodies against BrdU and DCX; the latter is a cytoskeletal protein  Figure 1C). To monitor NSCs, we performed staining with antibodies against BrdU and GFAP, a cytoplasmic marker that identifies NSCs in the SVZ (Figure 2A,Bsee also Figure S2). As shown in Figure 2C, we found no differences in the percentage of GFAP/BrdU double-labelled cells in the SVZ among groups, suggesting that chronic stress exposure does not affect this cell population in animals of both genotypes.
Furthermore, we have also monitored oligodendrocyte precursors by double labelling with BrdU and Olig2, a nuclear marker that identifies oligodendrocyte progenitors ( Figure 2D and Figure S2).
Here, we detected no significant differences in Olig2/BrdUlabelled cells among all groups, indicative of an absence of any significant effect of stress or Tau deletion in this cell population of the SVZ ( Figure 2E).

| Newly born neurons are differentially regulated by Tau deletion and chronic stress in the different sublayers of the olfactory bulb
As the olfactory bulb (OB) is the brain region where the newly born neurons generated within the SVZ migrate, we next analysed the effect of Tau deletion and chronic stress on another set of animals. In this case, animals had been injected with BrdU 4 weeks before the sacrifice, to give enough time for the newly gener- Tau-KO (P = .265) but was significantly higher when compared to stressed WT (P = .036). Altogether, the above data suggest that exposure to chronic stress reduced the newly born neurons found in the MCL and GL of OB, and this decrease was blocked in Tau-KO animals indicating that Tau protein is critical in stressinduced neurogenic suppression.

| Tau deletion does not interfere with the endocrine response to stress but blocks the related behavioural changes
Exposure to chronic stress is known to impact organism's ho-   Figure 4C). Additionally, we monitored ultrasonic vocalizations (USVs) of the animals as an index of their emotional status. 45 Our results showed a Stress × Genotype interaction in the number of USVs emitted by each animal (two-way ANOVA F 1, 24 = 5.578, F I G U R E 3 Impact of chronic stress and Tau on newborn neurons in differential sublayers of the olfactory bulb (OB). A, For olfactory bulb (OB) analysis, mice were injected with BrdU 4 weeks before sacrifice. B-C, Schematic illustration of the mouse brain highlighting the olfactory bulb (OB) level of analysis followed by an olfactory bulb coronal section (C) and the different sublayers analysed; granular cell layer (GCL), mitral cell layer  The current study focused on the analysis of neurogenesis in the SVZ-OB system of the adult rodent brain after prolonged exposure to environmental stress. Clinical and experimental evidence has long shown that exposure to stressful conditions is a strong precipitant of depressive pathology while a cardinal feature of the response to chronic stress is the atrophy of specific brain regions, as detected by both brain imaging and stereological techniques. 57,58 These plastic changes of the brain include dendritic atrophy and synaptic loss accompanied by the suppressed generation of newly born cells in specific areas of the adult brain. 28,29,48,57,59 Whereas the hippocampus has been the main focus of a plethora of clinical and experimental studies of depressed and/or stress-exposed human subjects and related animal models, clinical studies have also reported that adults with a history of early life stress or major depressive disorder present F I G U R E 4 Tau deletion does not interfere with endocrine response to stress but attenuates stress-induced behavioural impairment. A-B, Stressed animals of both WT and Tau-KO genotype exhibited reduced body weight (A) and increased levels of corticosterone, the main stress hormones (B), when compared with their corresponding control animals. C, Total distance travelled in the open-field arena was not different among groups. D, Chronic stress decreased the number of ultrasonic vocalizations (USVs) in WT animals indicating deficits of emotional status; no effect of stress was found in Tau-KO animals. E, Preference index in the Novel Object Recognition test was reduced in stressed WTs when compared to WT controls; this stress effect was not found in Tau-KOs suggesting a Tau-dependent cognitive impairment caused by chronic stress. All numerical data are shown as mean ± SEM (*P <.05). USVs, ultrasonic vocalizations; WT, wild type; Tau-KO, Tauknockout reduced OB volume and odorant detection impairment 60-63 indicating the potential impact of chronic stress on the OB. In line with that clinical evidence, few experimental studies on rodents exposed to chronic stress and/or to high levels of stress hormones, glucocorticoids, have shown reduced SVZ neurogenesis and olfactory deficits along with depressive-like and anxiety symptoms. 43,64,65 Different stress paradigms, such as maternal separation, repeated exposure to forced swim stress and chronic administration of corticosterone resulted in reduced BrdU-labelled proliferating cells in the SVZ. 43,[64][65][66] In line with these reports, we hereby demonstrate that 9-week exposure to a CUS protocol reduced both proliferation and neuronal differentiation of newly born cells in the SVZ, as assessed by the reduced number of BrdU-labelled cells as well as DCX/BrdU-labelled neuronal precursors and immature neurons in SVZ ( Figure 5). On the other hand, exposure to short stress periods (e.g. 2-day repeated foot shock stress paradigm 67 ) or a milder stress paradigm (e.g. chronic mild stress 42 or corticosterone 64 ) do not seem to impact SVZ neurogenesis; nevertheless, the above stress paradigms reduced neurogenesis in the hippocampus indicating a differential vulnerability of proliferating cells to stress and stress hormones between the dentate gyrus (DG) and SVZ neurogenic niches (see also 64 ). This is of great importance as different areas of the adult brain are shown to exhibit different vulnerability to the detrimental effects of chronic stress on their plasticity and function (e.g. hippocampus vs. frontal cortex 68 ) while other brain areas respond with opposite effects to chronic stress; for instance, stress causes atrophy to adult hippocampus whereas hypertrophy to amygdala and nucleus accumbens. 48,69 Following a 4-week period of newly born cells survival and migration into the OB, we found that the 9-week CUS protocol also suppress the BrdU-labelled cell population in the MCL and GL of the OB;

| D ISCUSS I ON
This was accompanied by a reduced number of NeuN/BrdU-labelled newborn neurons in OB of the adult brain ( Figure 5). These findings indicate that chronic stress diminishes neuronal maturation as well as the survival rate of newly born neurons in the OB that may contribute to the previously described deficits of olfactory memory induced by prolong stress and/or corticosterone exposure. 43,64 The current findings on both SVZ and OB brain areas suggest that 9 weeks of Chronic stress suppresses proliferation, neuronal, but not oligodendrocytic, differentiation and maturation of newly born cells in the subventricular zone (SVZ) and olfactory bulb (OB) of the adult brain as assessed by decreased levels of proliferating cells and neuroblasts in SVZ and reduced newborn neurons in the mitral cell and glomerular cell lB neurogenic niche of the adult brainayers (MCL and GL, respectively) of OB. However, the stress impact on the above cell populations was blocked in animals lacking Tau (Tau-KO). These findings suggest that Tau protein is essentially involved in the neurogenesis-suppressing role of chronic stress on the SVZ-OB neurogenic niche of the adult brain in line with previously reported reduction of the hippocampal neurogenesis by stress. 22,33 Interestingly, compared to wild type (WT), newborn neurons of Tau-KO animals seem to be accumulated in the first layer of the OB, the granular cell layer (GCL), indicating delayed migration of newborn neurons in the other OB cell layers and subsequently, decreased number of these cells in the mitral cell layer (MCL). GCL, granular cell layer; MCL, mitral cell layer; GL, glomerular cell layer; NSCs, neural stem cells; OPCs, oligodendrocytes progenitor cells; WT; wild-type, Tau-KO, Tau-knockout SVZ-OB neurogenic system in the network of brain areas damaged by chronic stress. Through an integrated manner, chronic stress damages various domains of behavioural performance such as different types of memory (i.e. associative, spatial and odour memories) and emotional status (e.g. anxiety levels, depressive behaviour).
Tau is an important protein involved in the regulation of cytoskeletal assembly and different cellular processes, such as axonal branching and transport, as well as in neuronal polarity, migration and differentiation. 22,[72][73][74] Despite the compensatory mechanisms (e.g. increased expression of other cytoskeletal proteins) that have been suggested to attribute to the lack of gross behavioural and neurostructural abnormalities in animals that lack Tau protein (Tau-KO), the absence of Tau appears to cause a transient delay in the dendritic maturation of new-born neurons 33 as well as a delay in their migration. 73,75 Moreover, it was recently shown that 14-month-old Tau-KO mice exhibit increased BrdU-labelled proliferating cells in the SVZ. 76 In contrast, our analysis of BrdU-labelled proliferating cell in SVZ did not detect a significant difference between WT and Tau-KO animals; this difference may be attributed to the age difference between our (6-7-month-old mice) and Criado-Marrero study (14 months old). 76 However, our findings demonstrate a significant increase in the number of newborn neurons in the OB, a region that was not monitored in Criado-Marrero's study. 76 Moreover, this increase in the number of newborn neurons was detected in the first layer of the OB (granular cell layer; GCL) of Tau-KO animals accompanied by a tendency for a decrease in the next OB layers (mitral and glomerular cell layers-see Figure 3 and Suppl. Figure 3), indicative of a delay of neuronal migration in OB led by the absence of Tau.
Although the precise mechanisms through which lack of Tau may induce neuronal migration deficits in the OB are still under investigation, inhibition of the Rho-ROCK signalling pathway by Tau absence in glioblastoma cells was recently shown to induce the remodelling of the actin cytoskeleton leading to delayed cell migration. 77 Alternatively, the interplay between Tau protein and the transduction of reelin, a protein that is crucial in neuronal migration and to the formation of synaptic connections in the brain, may have an equally important role here. [78][79][80] Moreover, the malfunction of Tau is causally related to cytoskeletal dysregulation, neuronal malfunction and atrophy under different pathological conditions including AD as well as stroke and brain trauma. 48,70,[81][82][83] As SVZ neurogenesis is suggested to participate in the endogenous regenerative response of the brain to stroke or trauma, 40,55,56 future studies should clarify the potential role of Tau in the adult neurogenic process under these pathological conditions, too.
Recent animal studies from our team and others have proposed the involvement of Tau in the regulation of adult neurogenesis of the hippocampus after exposure to acute or prolonged stress. 22,33 Specifically, exposure to stress leads to Tau hyperphosphorylation and accumulation in newly born neurons of the adult brain. 2 It is known that dephosphorylated Tau binds more stable to microtubules while increased phosphorylation of Tau is shown to reduce its microtubulebinding capacity leading to microtubule instability. 84 Through a constant regulation of its phosphorylation-dephosphorylation equilibrium, Tau protein is involved in many cellular functions as it regulates the cytoskeletal stability influencing morphogenesis of neurons. 85 Thus, stress-evoked alterations of the tight control of Tau phosphorylation could impair the complex and tight regulation of Tau, diminishing the cellular control over the cytoskeletal dynamics and network essential for proliferating cells and neuroblasts. Interestingly, animals lacking Tau were spared from the neurogenesis-damaging effects of chronic stress as, in contrast to WTs, stressed Tau-KO presented no reduction of proliferating cells and neuroblasts in the SVZ, followed by lack of decreased newborn neurons in the OB ( Figure 5). Extending previous evidence about Tau-dependent suppression of the hippocampal neurogenesis under stress conditions, 22,33 the current study suggests the essential mediation of Tau in the stress-driven neurogenic deficits in the SVZ-OB system of the adult brain. Moreover, progenitor cells in the hippocampus and SVZ are also able to differentiate into non-neuronal cell types (e.g. astrocytes, oligodendrocytes); however, our knowledge about whether and how chronic stress impacts other types of newborn cells (non-neuronal ones) in the adult brain remains limited. Our findings suggest that the population of oligodendrocyte progenitor cells in the SVZ is not affected by stress indicating that the detrimental impact of chronic stress on the SVZ-OB niche is mainly neuronal with an essential mediating role for Tau protein. Indeed, the neuroprotective role of Tau reduction against chronic stress is extended beyond newly born cells (and neurogenesis) as it is also evident in old, pre-existing neurons of the hippocampus as well as of other brain areas (e.g. prefrontal cortex). 48,70 Moreover, emerging evidence from animal models of diverse brain pathologies (e.g. Alzheimer's disease, epilepsy, stroke, traumatic brain injury) 81,83,86 suggests Tau as a converging protein of neuronal damage between different insults and disorders highlighting its broad neuroplastic and neuropathological role. 85 In summary, the current study provides novel insights about the involvement of Tau protein in the mechanisms that reduce cell proliferation, neuronal differentiation and migration within the SVZ-OB neurogenic niche under prolonged stressful conditions. Together with previous work suggesting the role of Tau in stress-evoked hippocampal plasticity changes, 22,33,48 these findings bring further information about the biological underpinnings of the stress-driven deficits on the adult brain circuits regulating mood and cognition. A better understanding of the mechanisms underlying the neuroplastic effect of chronic stress on the adult brain may help the development of targeted therapies for stress-related disorders, such as depression and Alzheimer's disease, which are characterized by deficits of neuronal plasticity. Agreement, through the European Regional Development Fund (ERDF).

ACK N OWLED G M ENTS
Additionally, this work has been funded by ICVS Scientific Microscopy Platform, a member of the national infrastructure PPBI -Portuguese Platform of Bioimaging (PPBI-POCI-01-0145-FEDER-022122.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
CD was involved in all stages of the experimental procedures, data collection and analysis, interpretation and manuscript prepa-

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.