Mast cells increase adult neural precursor proliferation and differentiation but this potential is not realized in vivo under physiological conditions

There is growing evidence that both peripheral and resident immune cells play an important part in regulating adult neural stem cell proliferation and neurogenesis, although the contribution of the various immune cell types is still unclear. Mast cells, a population of immune cells known for their role in the allergic response, have been implicated in the regulation of adult hippocampal neurogenesis. Mast cell-deficient c-kitW-sh/W-sh mice have previously been shown to exhibit significantly decreased adult hippocampal neurogenesis and associated learning and memory deficits. However, given that numerous other cell types also express high levels of c-kit, the utility of these mice as a reliable model of mast cell-specific depletion is questionable. We show here, using a different model of mast cell deficiency (Mcpt5CreR26DTA/DTA), that precursor proliferation and adult neurogenesis are not influenced by mast cells in vivo. Interestingly, when applied at supraphysiological doses, mast cells can activate latent hippocampal precursor cells and increase subventricular zone precursor proliferation in vitro, an effect that can be blocked with specific histamine-receptor antagonists. Thus, we conclude that while both mast cells and their major chemical mediator histamine have the potential to affect neural precursor proliferation and neurogenesis, this is unlikely to occur under physiological conditions.


MC are significantly depleted in the Mcpt5CreR26 DTA/DTA transgenic model. Given that MC have
been previously shown to reside in small numbers in the brain, particularly in the regions close to the hippocampus 6 , we first aimed to quantify the MC number and localization within the brain of WT mice. We observed limited numbers of avidin + /c-kit + MC in the meninges (166.8 ± 16.3) and hippocampal fissure (130.8 ± 59.1), but not in the SVZ (1.2 ± 1.2; Fig. 1A-C) of WT mice. The Mcpt5CreR26 DTA/DTA transgenic mouse model has been shown to result in dramatically fewer MC compared to WT mice in the peritoneal cavity (1.99% of WT), abdominal skin (11.1%) back skin (11.2%) and ear skin (3.5%) 20 . To confirm that the brains of Mcpt5CreR26 DTA/ DTA knock-out mice are also depleted of MC, we analyzed MC numbers within the brain of WT and MC-KO mice. In the brains of the MC-KO mice the total number of brain-resident MC was significantly decreased to 10.1% of those found in WT littermates (WT: 339.5 ± 48.1 vs MC-KO: 34.5 ± 10.3 vs C57BL/6: 332.8 ± 60.5 MC per brain, p =< 0.01; Fig. 1D). The c-kit + avidin − cells in the DG are neurons. We identified a small population of cells positive for c-kit, but unlike MC, negative for avidin (c-kit + avidin − ) in the DG of adult WT mice. While very few c-kit + avidin − cells could be found in the cortex, olfactory bulb, and hippocampus, no c-kit + avidin − cells at all could be identified in the SVZ. On average 190 c-kit + avidin − cells were found per DG of all three genotypes, indicating that in our MC-deficient model the number of c-kit + avidin − cells is not affected (C57BL/6: 189 ± 22.2 vs WT: 190.5 ± 21.6 vs MC-KO: 189 ± 38.7 c-kit + avidin − cells per hippocampus, n = 4 mice per group). To determine the identity of these c-kit + avidin − cells we phenotyped them with a panel of cell-type specific markers. We could show that the c-kit + avidin − cells were not GFAP + astrocytes (Fig. 3A), nestin + precursor cells (Fig. 3B), DCX + cells (Fig. 3C), or Iba1 + microglia (Fig. 3D). Interestingly, the c-kit + avidin − cells expressed the neuronal marker NeuN (Fig. 3E). Because they were DCX-negative they could not be the immature, early postmitotic new neurons, which are DCX/NeuN double-positive. Their morphology and location also argued against their nature as granule cells. It has been previously suggested that c-kit expression in the adult mouse brain was restricted to GAD67-expressing GABAergic interneurons 21 . Further phenotying of the c-kit + avidin − cells with a number of interneuron-specific markers indeed revealed that the majority of the c-kit + avidin − cells in the hippocampus as well as the cortex were GAD67 + (Fig. 3F,G), GABAergic (Fig. 3H) interneurons.

MC increase proliferative potential of SVZ neural precursor cells in vitro.
We have shown that in vivo, under physiological conditions MC depletion does not affect neural precursor proliferation. We next wanted to determine whether MC, when co-cultured in very high numbers with neural precursor cells, are able to release factors that can influence neural stem cell proliferation and differentiation. To do this we co-cultured 2 × 10 5 PCMCs with primary cells isolated from the two neurogenic niches using the neurosphere assay.
Firstly, the purity of MC in in vitro cultures was determined. Strong c-kit immunoreactivity is characteristic of MC as well as other cell types including neurons, astrocytes and microglia, melanoblasts, germ cells and hematopoietic stem cells, but its expression is lost in mature immune cells. Given the isolation procedure, although some limited contamination with hematopoietic stem cells is possible, hematopoietic stem cells are found at only very low numbers in the peripheral blood. The c-kit expression in cultured cells was assessed by immunofluorescence and quantified by flow cytometry. MC isolated from four C57BL/6 mice were expanded in separate cultures  and analyzed. On average 99.1% of cultured cells were c-kit + (Fig. 4A), indicating very high homogeneity of PCMCs. When cultured in the presence of MC (2 × 10 5 ), we observed a small but not statistically significant decrease in the number of SVZ neurospheres generated (−MC: 273 ± 71; +MC: 219 ± 84, p = 0.059; Fig. 4B), suggesting that MC derived factors do not initiate proliferation of quiescent neural stem cells isolated from the SVZ. Interestingly, in contrast to this finding although MC are not found in the SVZ under physiological conditions, when co-cultured in vitro they significantly increased SVZ neurosphere diameter ( Fig. 4C-E; −MC: 92.9 ± 2.2 μm vs MC + : 144.7 ± 2.6 μm, p < 0.0001). We also assessed neuronal differentiation in MC co-cultured SVZ neurospheres by quantifying the percentages of astrocytes (GFAP + ) and immature neurons (β-III-tubulin + ). Here, we observed a significant increase in the percentage of neurons in the neurospheres cultured in the presence of MC (−MC: 2.9 ± 0.46 vs +MC: 12.0 ± 2.6%, p = 0.0026; Fig. 4F

In vitro histamine treatment increases proliferation of SVZ but not DG progenitor cells. Having
shown that MC-released factors can significantly increase precursor proliferation and neuronal differentiation, we next investigated whether this effect was mediated by histamine, one of the most prominent mediators released by MC. To determine the potential effect of histamine on DG and SVZ precursor proliferation and differentiation, primary SVZ and DG cells were cultured in different concentrations of histamine (1 μM and 1 mM) using the in vitro neurosphere assay. In addition, to determine which receptor mediates the potential histamine effect, DG and SVZ cells were cultured with the antagonists for each histamine receptor.

Discussion
In contrast to the published experiments in which baseline hippocampal neurogenesis is decreased in adult MC-deficient c-kit W-sh/W-sh mice 6 , our results from the Mcpt5CreR26 DTA/DTA model suggest that MC deficiency does not have an effect on baseline adult hippocampal proliferation and neurogenesis in vivo. We cannot however conclusively rule out that the possibility that the approximately 90% reduction in MC in our model is not sufficient and that the few remaining MC (by our calculation, approximately eight per whole hippocampus) can functionally compensate for this depletion. Likewise, using the c-kit W-sh/W-sh mice, although MC are fully depleted, the authors cannot definitively confirm that loss of these cells is responsible for the concomitant decrease in hippocampal neurogenesis, as altered c-kit signaling and/or suppression of c-kit expression in cell types other than MC might have confounded this finding. Nautiyal and colleagues previously reported that c-kit W-sh/W-sh depletion the survival of proliferating cells was observed in the DG of MC KO mice (n = 7 mice) compared to WT mice (n = 6 mice). (F) MC depletion had no effect on the percentage of surviving new-born neurons, 28 d following BrdU labeling. (G) Representative image of BrdU and NeuN (Fox3) staining. Scale bars, 100 μm. has no effect on SVZ neurogenesis 6 . This, together with the fact that we also observed no change in SVZ precursor proliferation in the Mcpt5CreR26 DTA/DTA MC depletion model, makes it unlikely that MC depletion would result in a decrease in neurogenesis in this niche.
In the adult mouse nervous system, SCF and c-kit expression have been detected in multiple cell types including neurons in the OB, neocortex, archicortex, cerebellum, brain stem and spinal cord 19 , glial cells in the neocortex, hypothalamus, cerebellum and brainstem, and precursor cells in the SVZ 18 . A number of studies have shown that c-kit is expressed by the precursor cells of the adult rodent SVZ 18,22 . These c-kit-positive precursor cells were responsive to SCF signaling, increasing their proliferation following subcutaneous SCF injection 22 . Other studies have identified c-kit-positive progenitor cells in the adult olfactory mucosa, and selective depletion of these cells resulted in in decreased olfactory neurogenesis 23,24 .
High expression of c-kit in rat hippocampal CA1-CA4 regions and impaired spatial learning and memory in homozygous c-kit mutant rats has also been reported 25 . Similarly, SCF-mutant mice (Steel-mice) demonstrated spatial learning deficits 6,26 . These findings suggest a potential involvement of SCF/c-kit signaling in learning and memory. This was further supported by the report that a selective knockout of huntingtin-associated protein 1 (Hap1) expression led to reduced early postnatal neurogenesis in the murine hippocampus 21 . In that study, c-kit was identified as direct molecular target of Hap1, and loss of Hap1 caused downregulation of c-kit expression, resulting in decreased postnatal hippocampal neurogenesis. In our study, however, we found that microglia, astrocytes and precursor cells in the adult mouse DG were in fact not c-kit + .
The situation in adulthood might substantially differ from younger mice. Expression of c-kit and SCF in cultured neurons, astroglia, and microglia isolated from early postnatal mice is high 19 , whereas we found that c-kit + cells in the adult brain are GFAP-negative and Iba1-negative. It has been shown that during early postnatal development (P7-P15), c-kit is expressed in hippocampal NPCs (nestin + /c-kit + ), immature neurons (DCX + /c-kit + ), and mature neurons (NeuN + c-kit + ), whereas in 4-week-old mice c-kit expression was restricted to GAD67-expressing GABAergic interneurons even though overall expression in the hippocampus was high compared to the cortex, striatum and hypothalamus at this age 21 .
In our study, we identified a population of c-kit + avidin − NeuN + cells in the dentate gyrus of 2-month-old mice, confirming that c-kit + neurons remain present in the mature hippocampus. Phenotyping of these c-kit + neurons with GAD67 and GABA confirmed that they are indeed a small subpopulation of c-kit + GABAergic interneurons. This is in agreement with other studies that report c-kit expression in basket, stellate, and Golgi cells in the cerebellar cortex 19,27-29 and suggests a function for c-kit in inhibitory interneurons during postnatal cerebellar development 30 . Interestingly, the c-kit-expressing neurons in the cerebral cortex and olfactory bulb are also mainly interneurons 19 . In summary, these data indicate that c-kit may have a widespread function during embryonic and early postnatal brain development that becomes more restricted during adulthood. Taken together, the above studies suggest that c-kit depletion must have wider implications, also for the hippocampus, than the depletion of MC.
We could identify MC in the meninges and the hippocampal formation of C57BL/6 mice, which is in line with previous reports of MC in the hippocampal formation of B10.PL mice 31 . To further analyze a potential interaction of MC and their released factors with neural stem/progenitor cells in the neurogenic niches, we co-cultured peritoneal MC with SVZ-and DG-derived cells in vitro. Our results suggest that MC-derived factors activate proliferation of quiescent DG neural stem cells. Notably, our study reveals a significant boost in proliferative potential of cycling SVZ progenitors when co-cultured with MC, indicated by a dramatic increase in neurosphere diameter. This is particularly interesting, as the presence of MC in the SVZ niche in vivo was never observed. These results indicate that, although MC can influence SVZ precursor proliferation in vitro, under physiological conditions in vivo this interaction is unlikely.
Since MC account for 90% of the hippocampal, and up to 50% of total brain histamine and are the main source of this neuromodulator in peripheral tissues 8,32 , we next verified the effects of histamine on SVZ-and DG-derived  (H1R, H2R, H3R), which are widely distributed throughout the CNS 8 , we also determined which of these receptors histamine is potentially acting through in both neurogenic niches. All three histamine receptors are expressed by SVZ-derived neural stem cells 33,34 .
We observed a significant increase in SVZ-derived neurosphere number and size when treated with 1 mM histamine. This result is not consistent with previously reported effects of histamine on neonatal SVZ stem/ progenitor cells where a strong pro-neurogenic effect of histamine on neonatal SVZ and embryonic cortical cell cultures was observed, without changes in proliferation 34,35 . Another report has demonstrated that chronic histamine administration in the adult SVZ in vivo does not induce an overall increase in cell proliferation but instead may trigger neuronal commitment of SVZ cells, and identified histamine as a crucial modulator of neuronal differentiation in the SVZ-OB axis 36 . Nevertheless, in published in vitro studies 500 μM was the highest histamine concentration tested, possibly indicating that elevated concentrations of histamine (1 mM) may be needed to activate SVZ cell proliferation. Furthermore, the effect of endogenous histamine was abolished when SVZ-derived cells were cultured in the presence of H1R, H2R and H3R antagonists, confirming previously published results showing that histamine actions in the SVZ may be mediated by the activation of all three histaminergic receptors. Importantly, several studies have identified histamine as a potent pro-neurogenic mediator, responsible for priming of NSC in the SVZ toward the neuronal phenotype [34][35][36][37] . This is in line with the results from our study, which showed a trend towards increased neuronal differentiation in the SVZ cells treated with 1 mM histamine and a significant decrease in those treated with the H1R and H2R antagonists. We found all three histamine receptors to be expressed in the DG (our unpublished results). In addition, a recent study demonstrated expression of the H3R in the hippocampus and showed that S38093, a novel histamine H3R antagonist promoted hippocampal neurogenesis in 3-month-old mice and improved context discrimination in aged mice 38 . In accordance with this study, we found a small but non-significant reduction in DG neurosphere number following histamine treatment. Similar to the SVZ however, we observed a significant decrease in DG-derived neurosphere number when cells were cultured in the presence of the H1R and H2R antagonists. Surprisingly however, we found no effect of the H3R antagonist.
Despite a lack of in vitro data for the hippocampus, previous in vivo experiments revealed impaired spatial learning and reduced adult hippocampal neurogenesis in H1R-knockout mice 39 . Both the aforementioned study and our own identify a potential role of endogenous histamine in adult hippocampal neurogenesis that might be executed by H1R activation. However, the exact mechanisms responsible for the reduced number of newborn cells in the hippocampus of H1R-KO mice, as well as in DG-derived cell cultures after chlorpheniramine treatment, have not been yet been sufficiently explored.
It is worth to note that the neurosphere assays described above were performed in the absence of exogenous histamine indicating that endogenous histamine may be present in our cultures. It is possible that mast cells, or other histamine releasing cells such as basophils and lymphocytes 40,41 , may be present at small numbers in our cultures. Non-mast cell derived histamine may be found in several tissues, including the brain, where it functions as a neurotransmitter 42 . In addition, MC granular contents have also been shown to diffuse a great distance, and their granules may be seen up to 500 μm away from the cell of origin 43,44 , potentially explaining why histamine may be present in regions of the brain that are devoid of MC (i.e. the SVZ).
In summary, we are just beginning to understand the role that immune cells play in regulating adult neurogenesis under physiological conditions. Here, we show using the Mcpt5CreR26 DTA/DTA MC-deficient mouse line that MC are not required for the maintenance of normal baseline adult neurogenesis, even though the main factor they release potentially influences adult neurogenesis. Based on our results, together with previously published data, we hypothesize that the observed decrease in hippocampal neurogenesis and deficits in hippocampus-dependent spatial learning and memory in c-kit W-sh/W-sh mice 6 may be, at least partly, the result of disrupted SCF/c-kit signaling in the neurogenic niche and may not be directly connected with MC-deficiency. While both MC and their chemical mediator histamine have the potential to influence adult neural stem cell proliferation and differentiation when supplied at supraphysiological conditions ex vivo, in the steady-state adult brain MC are unlikely to affect neural stem cell proliferation and adult neurogenesis. BrdU injection and tissue preparation. At 8 weeks of age, MC-KO mice and WT mice received a single intraperitoneal injection of 50 mg/kg 5-bromo-2-deoxyuridine (BrdU). To measure precursor proliferation, mice were sacrificed 12 h later and to measure neuronal survival mice were sacrificed after 28 d. Mice were deeply anesthetized and perfused transcardially with 0.9% NaCl, followed by 4% PBS buffered paraformaldehyde (PFA). Afterwards, brains were fixed for 24 h in 4% PFA at 4 °C, washed in PBS and incubated in 30% sucrose dissolved in PBS. Coronal 40 μm thick sections were cut throughout the entire brain on a microtome (Leica).

Staining, quantification and image analysis.
For a list of all antibodies used in this study see Table S1.
For diaminobenzadine (DAB) detection, sections were first incubated in 0.6% hydrogen peroxide diluted in PBS for 30 min at room temperature (RT) then in 2.5 M HCl for 30 min at 37 °C (BrdU staining only). After rinsing in PBS, sections were blocked in PBS++ (0.2% triton-X100 and 10% donkey serum (DS)) for 60 min at RT. Sections were incubated overnight (ON) at 4 °C in primary antibodies diluted in PBS+ (0.2% triton-X100 and 3% DS), washed extensively with PBS and incubated for 3 h at RT in secondary antibodies diluted in PBS+ . For the fluorescently-labeled sections, a 10 min incubation with 4′,6-diaminidono-2-phenylindole (DAPI; 1:1000 in PBS) was included prior to the final PBS wash. After rinsing in PBS, fluorescently-labeled sections were mounted with Aqua-Poly/Mount (Polysciences Inc). The DAB-labelled sections were incubated for 1 h at RT in Vectastain ABC-Elite (Vector Laboratories), washed, incubated for 5 min at RT with DAB (Vector Laboratories), cleared with Neo-Clear (Merck) and mounted in Neo-Mount (Merck). DAB-labelled cells were counted in the subgranular zone (SGZ) of the hippocampus and in the SVZ using an upright light microscope (DM750; Leica). Fluorescence images were acquired on the Zeiss Apotome Axio Imager Z1 microscope using either the Plan-APOCHROMAT 20x/0,75 or Plan-APOCHROMAT 40x/0,95 objectives and AxioVision (V 4.8.2.0) software. For cell phenotyping optical Z-sectioning was performed at 1 μm intervals. Images were analysed on merged z-stacks in single optical sections and maximum z-projected images were created using Fiji software (http://fiji.sc). All microscopic quantification was performed blinded to the experimental conditions. Isolation and culture of peritoneal-cell derived mast cells (PCMCs). Mice were anaesthetized small incision was made below the abdomen of the mouse and 5 ml of ice cold sterile PBS was injected into the SCIENTIFIC RepoRts | (2017) 7:17859 | DOI:10.1038/s41598-017-18184-2 peritoneum. The peritoneal liquid containing PCMCs was collected and centrifuged. The pellet was re-suspended in RPMI medium containing 10% fetal calf serum, 50 units/ml penicillin/streptomycin, 2 mM L-Ala L-glutamine (Merck), 1 mM sodium pyruvate (Merck) and 50 μM 2-mercapto-ethanol supplemented with interleukin-3 (10 ng/ml) and stem cell factor (30 ng/ml). Cells were transferred into a T25 culture flask and cultured at 37 °C, 5% CO 2 . At day 2 the non-adherent cells were removed and 5 ml of fresh, supplemented RPMI medium was added to the culture flask. At day 6 an additional 5 ml of supplemented RPMI medium was added. At day 10, non-adherent PCMCs were obtained from culture medium for further experiments.
Assessment of cultured MC purity by flow cytometry analysis. Cultured PCMCs were fixed and permeabilized with 4% PFA, 0.1% triton X-100 in PBS for 15 min at RT. Subsequently, cells were washed and incubated for 30 min at RT in rat anti-c-kit diluted in 1% bovine serum albumin and 0.1% triton X-100 in PBS. Afterwards, cells were washed, incubated with secondary donkey anti-rat Cy3 and re-suspended in PBS. Unstained and secondary only controls were also prepared. Samples were analyzed by flow cytometry (FACS AriaIII; BD Biosciences) using FACSDiva Software. Dead cells and cell debris were gated out on the basis of their forward and side scatter profile. c-kit + cells were designated as MC.
MC and neural precursor cell co-culture. The SVZ and the DG were dissected from brains of 8-week-old C57BL/6 mice as described previously 45 . Isolated DG and SVZ neural precursor cells were resuspended in neurosphere growth medium (neural basal medium supplemented with 1x B27, 1x GlutaMAX, 2 µg/ml heparin, 50 units/ml penicillin/streptomycin) supplemented with 20 ng/ml purified mouse receptor-grade epidermal growth factor (ProSpec), and 20 ng/ml recombinant bovine fibroblast growth factor (ProSepc)) and seeded into a 24 well plate (500 μl/well). Hippocampal adherent monolayer cultures were initiated as described previously 45 and used between passages 7-9. Transwell inserts (Sigma-Aldrich) were placed in every second well and 2 × 10 5 PCMCs were seeded per transwell. After 2 days the transwells were removed. At day 7 (SVZ) and day 14 (DG) neurospheres were quantified, sized and transferred to a 24-well plate containing the neurosphere growth medium without growth factors and a poly-D-lysine/laminin coated coverslip. To label proliferating adherent precursor cells, a pulse of BrdU (10 μg/ml) was added to the wells and BrdU + cells stained as described above. Neurospheres were differentiated for 7 d and adherent cultures for 5 d at 37 °C with 5% CO 2 , in growth medium devoid of growth factors, then fixed with 4% PFA for 15 min at 37 °C.
Statistical significance was depicted as (*)p < 0.05, (**)p < 0.01 and (***)p < 0.001. Data availability. All data generated and analysed during this study are included in this published article