Deficits in hippocampal neurogenesis in obesity-dependent and -independent type-2 diabetes mellitus mouse models

Hippocampal neurogenesis plays an important role in learning and memory function throughout life. Declines in this process have been observed in both aging and Alzheimer’s disease (AD). Type 2 Diabetes mellitus (T2DM) is a disorder characterized by insulin resistance and impaired glucose metabolism. T2DM often results in cognitive decline in adults, and significantly increases the risk of AD development. The pathways underlying T2DM-induced cognitive deficits are not known. Some studies suggest that alterations in hippocampal neurogenesis may contribute to cognitive deterioration, however, the fate of neurogenesis in these studies is highly controversial. To address this problem, we utilized two models of T2DM: (1) obesity-independent MKR transgenic mice expressing a mutated form of the human insulin-like growth factor 1 receptor (IGF-1R) in skeletal muscle, and (2) Obesity-dependent db/db mice harboring a mutation in the leptin receptor. Our results show that both models of T2DM display compromised hippocampal neurogenesis. We show that the number of new neurons in the hippocampus of these mice is reduced. Clone formation capacity of neural progenitor cells isolated from the db/db mice is deficient. Expression of insulin receptor and epidermal growth factor receptor was reduced in hippocampal neurospheres isolated from db/db mice. Results from this study warrant further investigation into the mechanisms underlying decreased neurogenesis in T2DM and its link to the cognitive decline observed in this disorder.


Scientific RepoRtS
| (2020) 10:16368 | https://doi.org/10.1038/s41598-020-73401-9 www.nature.com/scientificreports/ Hippocampal neurogenesis plays important roles in the regulation of hippocampal function, learning and memory 11 . In addition, impaired hippocampal neurogenesis is linked to AD [12][13][14] . Several reports suggest that hippocampal neurogenesis is altered in mouse models of T2DM. However, the information is often contradictory or inconclusive. For example, the proliferation of bromodeoxyuridine positive (BrdU +) cells was increased in the subventricular zone (SVZ) and subgranular zone (SGZ) of the DG of Goto-Kakizaki (GK) rat, while their survival significantly decreased. These alterations took place in parallel to increased blood glucose level 15 . However, the identity of these BrdU + cells was not specified. Stranahan and colleagues (2008) report that the numbers of BrdU + cells and BrdU + Tuj1 + are reduced in the SGZ of db/db mice compared to wild type 16 . In contrast, another study reports increased number of BrdU + or DCX + cells in the SVZ, SGZ and cortex of db/ db mice. Here too, the identity of BrdU + cells was not studied. In addition, it is not clear whether the number of cells was adjusted to brain volume, given the progressive atrophy 8 . In the context of obesity-related T2DM, a high-fat diet was reported to impair the number of proliferating cells in the SGZ of the hippocampus in male, but not female rats 17 . In contrast, a recent study shows that high-fat diet compromises neurogenesis in C57BL/6 J females but not in males 18 .
Interestingly, several studies suggest that metformin, a drug used to manage T2DM, has pro-neurogenic characteristics 19 . For example, Wang and colleagues observed that metformin enhances the differentiation of embryonic cortical NPCs, as well as of adult-born NPCs in the SVZ and the SGZ in mice 20 . Chronic administration of metformin facilitates cell proliferation and neuronal differentiation and inhibits diabetes-related neuroinflammation in the brains of DIO mice 21 . Treatment of the neonatal mice with metformin following hypoxia-induced brain injury facilitated neurogenesis and led to functional recovery 22 . This effect has been proposed to be mediated via insulin receptor substrate 1 21 , while another study suggests that the effect on selfrenewal and proliferation occurs via the p53 family member and transcription factor TAp73, and it promotes neuronal differentiation of these cells by activating the 5′ adenosine monophosphate-activated protein kinase/ protein kinase C/transcriptional coactivator CREB-binding protein pathway AMPK-aPKC-CBP pathway 20,23 .
Another study suggests that GLP-1 mimetic Exendin-4, which is used for the treatment of T2DM, enhances hippocampal neurogenesis in three mouse models of T2DM: ob/ob, db/db and high-fat-diet-fed mice 24 . Taken together, these studies suggest that while there is a possible connection between hippocampal neurogenesis and cognitive deficits in T2DM, the fate of hippocampal neurogenesis in T2DM is yet to be established.
To address the current controversy, we analyzed two mouse models of T2DM that may represent different disease mechanisms: obesity-dependent db/db mice, expressing a spontaneous mutation in the leptin receptor (Lepr db ), and obesity-independent MKR mice, expressing the human insulin-like growth factor I receptor (IGF-1R) gene containing the K1003R mutation driven by the murine creatine kinase, muscle (Ckm) promoter. The extent of neural stem cells and new neurons in the hippocampus of these mice was examined using neurogenic proxies. Consistently, in both models we observed deficits in the number of new neurons as well as deficits in neural progenitor cell proliferation.

Results
Alterations in the proliferative and differentiation capacity of neural progenitor cells in db/ db model of T2DM. To determine the extent of hippocampal neurogenesis in T2DM, we first quantified the number of cells expressing nestin and/or doublecortin (DCX) in the subgranular zone (SGZ) in brain sections of 12-week-old male db/db mice using unbiased stereology ( Fig. 1A-D). We observed that the number of neural progenitor cells (NPCs, nestin + DCX−) was similar in db/db and wild type mice (Fig. 1A). The number of neuroblasts (nestin + DCX +) was reduced in db/db, albeit not statistically significant (P = 0.0668, N = 4, Fig. 1B,D). Interestingly, the number of immature neurons (Nestin−DCX +) was significantly lower in db/db mice compared to wild type (P = 0.0490, N = 4, Fig. 1C,D), suggesting less new neurons in db/db mice, particularly as age and disease progress. To further examine this, NPCs were isolated from the DG of db/db and wild type mice and grown in culture to form neurospheres. Next, neurospheres were singly dissociated and subjected to a clone-forming assay. Quantification revealed that the number and size of neurospheres and the total number of cells were reduced in the db/db cultures compared to wild type neurospheres (N = 3, Fig. 1E-H). This may suggest that proliferation of NPCs derived from the db/db mice is reduced compared to wild type mouse in the subgranular zone of the dentate gyrus of db/db and wild type mice using unbiased stereology. The number of neural progenitor cells was similar between genotypes. There was a trending decrease in the number of proliferating neuroblasts (P = 0.0668, N = 4). A significant decrease was observed in the number of immature neurons (*P < 0.05) in the db/db mice (Unpaired t-test, 12 weeks of age, N = 4). (D) Representative images of wild type and db/db brain sections immunolabeled with antibodies raised against doublecortin (DCX) or nestin and counterstained with DAPI. Scale bar = 75 μm. (E-G) Clonogenic assay of primary neural progenitor cells isolated from the dentate gyrus of db/db and wild type mice. Analysis of the number (E) and diameter (F) of neurospheres formed, as well as the total number of cells (G) show a significant decrease in all parameters in the db/db compared to wild type. (H) Representative images of neurospheres cultured from wild type and db/db mice (N = 3, unpaired t-test, *P < 0.05). Inserts show high power images of a single neurosphere. (I) Real time PCR analysis of mRNA expression of epidermal growth factor receptor (EGFR) in primary neurospheres shows a significant decrease in the db/db genotype (N = 3, unpaired t-test, *P < 0.05). (J,K) Western blot analysis (J) and semi-quantification using image J (K) of the expression of insulin receptor 1α (IR1α) in protein lysate of neurospheres isolated from db/db and wild type mice. Expression of IR1α is severely compromised in the db/db genotype (N = 3, unpaired t-test, *P < 0.05).

Scientific RepoRtS
| (2020) 10:16368 | https://doi.org/10.1038/s41598-020-73401-9 www.nature.com/scientificreports/ NPCs. Another possibility is that the survival of NPCs or neuroblasts derived from db/db mice is compromised. To start to explore these possibilities, we examined the expression levels of epidermal growth factor receptor (EGFR) and insulin receptor 1α (IR-1α), respectively. EGF is a critical proliferation factor of NPCs. Insulin plays a major role in neuronal differentiation and is likely to be compromised in the T2DM brain. In the healthy brain, IR-1α is highly abundant in the hippocampus and in NPCs and regulates hippocampus-dependent learning and memory 25 . The expression of EGFR and IR-1α were examined in RNA extracts and protein lysates, respectively, of neurosphere cultures derived from the brains of db/db and wild type mice. We observed a significant reduction in the expression of EGFR in db/db NPCs compared to wild type mouse NPCs (Fig. 1I, N = 3). In addition, we observed a marked reduction in the expression level of IR-1α in db/db NPCs compared to wild type NPCs (Fig. 1J,K; N = 3). Taken together, these results suggest impairments in the proliferation and differentiation of new neurons in the hippocampus of db/db mice at 12 weeks of age.
Impairments in hippocampal neurogenesis in the MKR mouse recapitulate impairments in the db/db. Given that there are several mechanisms leading to T2DM, we chose to investigate an obesityindependent model (the MKR mouse) to determine if the impairments observed in the db/db mouse represent the state of neurogenesis in T2DM. Similar to findings in the db/db brains, results show that levels of NPCs (Nestin + DCX−) were comparable in the hippocampi of MKR and wild type mice ( Fig. 2A, N = 6). Interestingly, we observed a nearly significant decrease in the number of proliferating neuroblasts (Nestin + DCX +) in the hippocampi of MKR compared to wild type, which could indicate deficits in neuroblast proliferation or neuronal maturation (Fig. 2B, N = 6). In agreement with reduced number of neuroblasts, a significant decrease was observed in the number of immature neurons (Nestin−DCX +) in the hippocampus of MKR mice (Fig. 2C,D, N = 6).

Differential alterations in neural stem cell population in hippocampus of db/db and MKR mice.
To determine if impairments in neurogenesis involve alterations in the neural stem cell (NSC) pool, we next quantified the number of Nestin + GFAP + in the hippocampus of db/db and wild type mice. Intriguingly, there was a reduction in the number of Nestin + GFAP + cells in the hippocampi of db/db mice compared to the wild type, albeit not statistically significant (P = 0.0524, N = 3, Fig. 3A,D). There was no change in the level of Nestin + GFAP−, which designates several populations of NPCs at different stages of development (Fig. 3B,  N = 3). Interestingly, we further observed an increase in the number of mature astrocytes (Nestin−GFAP +) in the db/db compared to the wild type mice, however it was not significant (Fig. 3C, N = 3). Examination of these populations in the MKR mice revealed that the number of NSCs (Nestin + GFAP + , N = 6, Fig. 4A) in the hippocampus was lower in db/db compared to the wild type mice, albeit not statistically significant (P = 0.202). No change was observed in the total number of NPCs (Nestin + GFAP−, N = 6, Fig. 4B). In contrast to the db/db, there was a trending decrease in the number of mature astrocytes (P = 0.052, N = 6, Fig. 4C). However, alterations in the number of mature astrocytes did not reach significance in neither genotype, and the lack of consistency warrants further investigation.
In summary, based on the results of this study, we conclude that hippocampal neurogenesis is deficient in both T2DM mouse models used here, manifested by reduced number of new neurons. We further observed reduced levels of EGFR and IR in NPCs derived from the hippocampi of db/db mice, which may underlie some of these deficits.

Discussion
This study addresses discrepancies in the literature concerning the fate of neurogenesis in T2DM mouse models. These discrepancies are likely the result of lack of use of appropriate neurogenic proxies, and inadequate or lack of quantification. Nevertheless, determining the fate of hippocampal neurogenesis in the brains of T2DM patients and mouse models is critical for understanding the mechanism of cognitive deterioration. Recent intriguing studies show that hippocampal neurogenesis is impaired in mild cognitive impairment and AD, as well as in mouse models of AD 12,13 . This may imply that dysfunction of hippocampal neurogenesis in T2DM may contribute to the development of AD. To this end, in this study we show that the number of new neurons is reduced in the SGZ of the hippocampus of two mouse models of T2DM, suggesting that less new neurons are available for the recruitment and incorporation in the granular layer of the DG. This is significant because it may lead to hippocampal dysfunction and cognitive deficits, and may provide a putative mechanism underlying the substantial portion of T2DM patients who deteriorate cognitively. Observing reduced number of new neurons in the hippocampus of two different mouse models of T2DM validates this result as physiologically-relevant and excludes the possibility that it is a transgene-related off-target effect. Notably, MKR mice express a mutated form of the human insulin-like growth factor 1 receptor (IGF-1R) in skeletal muscle solely, and yet, pathology develops in the brain and in neural progenitor cells and new neurons specifically and recapitulates the impairments observed in the db/db mouse.
Second, our observations showed trending reductions in the numbers of Nestin + DCX + cells in the SGZ of both the db/db and MKR mice. This result was strengthened by the observed reduction in cell proliferation in neurospheres isolated from db/db mice. Nevertheless, it should be noted that the population of cells positive for Nestin + DCX + , as well as the population of cells encompassing neurospheres in vitro, each includes several subpopulations that proliferate at different rates. In addition, a difference in clone-formation capacity in culture could be due to mechanisms other than proliferation, for example, cell adhesion. Further investigation is needed in order to clarify the specific mechanisms underlying the reduced numbers of proliferating NPCs and neuroblasts in these mouse models. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Third, we observed reduced expression of IR-1α in NPCs derived from the db/db mouse. We cannot exclude the possibility that this may be due to the mutation in leptin receptor expressed in the db/db mice. Specifically, previous studies suggest that both insulin and leptin circulate at levels that vary in proportion to body fat stores and both enter the CNS in proportion to their plasma level, where they act on their respective receptors 26 . Deficiency of either hormone or their respective receptors in the CNS results in insulin resistance 27,28 . In light of the tight functional association, it is expected that the mutation in leptin receptor would affect insulin receptor expression and signaling. Thus, levels of IR-1α in NPCs should be validated in other mouse models of T2DM. www.nature.com/scientificreports/ Another mechanism by which insulin resistance or deprivation can compromise neurogenesis is via the Wnt3-NeuroD1 pathway. Previous studies have shown that Wnt3 promotes neuronal differentiation by the activation of the transcription factor NeuroD1, which in turn, induces the expression of insulin [29][30][31] . Notably, deficiency of NeuroD1 causes malformation of the DG and sever diabetes [32][33][34] . Lastly, it should be noted that insulin, IGF-1, www.nature.com/scientificreports/ and IGF-2 also play a role in oligodendrocyte differentiation and neuronal survival [32][33][34] . Thus, dysfunctional insulin signaling would have major implications for neurogenesis and gliogenesis. Fourth, in addition to IR-1α, we observed that levels of EGFR are reduced in NPCs derived from the db/db mouse. EGF is a critical regulator of NPC proliferation. Interestingly, EGFR can be compromised directly or indirectly by lack of leptin receptor signaling. Evidently, a leptin-EGFR-Notch 1 axis has been described under pathological conditions 35 . Interestingly, a recent study described EGF-like metabolic hormones that bind EGFR as critical modulators of neural activity, coupling insulin secretion to the nutritional status in Drosophila 36 . Future studies should address the mechanism by which deficits in EGFR and IR-1α in NPCs play a role in altered hippocampal neurogenesis in T2DM.
Fifth, inconsistencies were found between db/db and MKR as to the number of NSCs and mature astrocytes as compared to their respective wild type strains. This may be either genotype related or a representation of differential effects of diabetic mechanisms on the neurogenic population in the two mouse models. These observations warrant further investigations addressing the effect of distinct molecular signaling mechanisms on hippocampal neurogenesis. In that regard, it should be noted that this study examined levels of hippocampal neurogenesis in male mice. Sex differences in levels of neurogenesis have been reported in T2DM related mouse models 17,18 . Thus, future studies should include a comparative analysis of neurogenesis in both sexes.
In summary, the results of this study suggest that hippocampal neurogenesis is deficient in mouse models of T2DM. Deficient neurogenesis may contribute to the development of cognitive impairment in T2DM patients. Understanding the mechanism underlying defective neurogenesis may help alleviate T2DM-linked memory impairments.

Mice. All animal experiments were approved by the University of Illinois at Chicago Institutional Animal
Care and Use Committee, ACC Protocol # 17-123 (Lazarov) and # . All experiments were performed in accordance with institutional guidelines and regulations.
Immunohistochemistry. Following perfusion with cold PBS, brains were post-fixed in 4% PFA for 24 h.
Unbiased stereology. Cell counts were performed using design-based stereology (StereoInvestigator, MBF Biosciences). For the analysis, every sixth section of brain tissue was quantified using the optical fractionator workflow of StereoInvestigator. Regions of interest were traced under 5 × magnification with counting performed under 63 × magnification with a counting frame of 225 µm × 145 µm and sampling grid of 100 µm × 100 µm with 12.5 µm top and bottom guard zones. Primary neurosphere culture. Isolation. 8-12 weeks old mice were euthanized via isoflurane overdose and brains dissected into HBSS stored on ice followed by dissection of the hippocampus. Tissue was minced with a sterile scalpel and transferred to 3 mL warm media (DMEM/F-12, 20 mM KCl, 2 μg/mL heparin, 1% penicillin-streptomycin, 20% B27 supplement, 10% N2 supplement). After tissue was settled, media was removed, and tissue was dissociated with 0.1% trypsin-EDTA and incubated with agitation at 37 °C for 7 min. Trypsin inhibitor (139 μg/mL plus 10 μg/mL DNase I in HBSS -/-) was added and samples spun at 700 rpm for 5 min. Tissue pellet was resuspended in 1 mL warm media supplemented with 20 ng/mL EGF and 10 ng/mL bFGF and pipetted 20 × with a P1000 pipet and cultured at 37 °C, 5% CO 2 and passaged every 7 days.
Passaging. Media is collected and spun at 1000 rpm for 5 min. Neurospheres are dissociated with 0.05% trypsin-EDTA and incubated in a 37 °C water bath for 7 min. Trypsin inhibitor was added, and cells spun at 1000 rpm for 5 min. Cells are counted and plated at 10,000 cells/cm 2 .