Molecular imaging of nestin in neuroinflammatory conditions reveals marked signal induction in activated microglia

To visualize the spatial and temporal dynamics of NPG cellular responses/nestin signal in real time and from the brains of live animals, we generated transgenic mice bearing a dual bicistronic reporter system [firefly luciferase (luc) and GFP] under the transcriptional control of the nestin gene promoter. The 5.1-kb fragment of the nestin gene promoter was amplified from the recombinant plasmid pNERV and cloned into a TOPO vector then sequenced and inserted into the recombinant plasmid pIRES-Luc2-AcGFP (Fig. 1a). The feasibility of this approach was clearly demonstrated in our previous studies [7, 8]. Namely, the advantage of the dual reporter system emerged from the fact that fluorescence signals can be used to achieve microscopic resolution while bioluminescence, due to favorable emission spectra of luciferase, is optimized for in vivo imaging. Furthermore, previous studies demonstrated that the expression of nestin in the neuroepithelial cells is dependent on the presence of the transcriptional enhancer present in the second intron of the gene; therefore, to preferentially direct the transgene expression into CNS stem cells, a 654-bp fragment containing the second intron of the rat nestin gene was also amplified from and inserted between the Cla1 restriction sites at the end of the final construct [1]. Nine transgenic founders were obtained using this construct, and following validation analysis, we focused on the transgenic reporter line with the strongest signal induction and appropriate nestin-driven transgene expression. The mice developed normally and, as expected, did not develop any overt phenotype. As presented in Fig. 1b, the transgenic mice were genotyped using PCR with primers aiming at the segment of the luc transgene (Fig. 1b). To determine the functionality of the transgene, the animals were screened for in vivo bioluminescence signal in control, baseline conditions, following focal brain ischemia and in the context of acute and chronic neuroinflammatory conditions. We first assessed the distribution of the nestin signal in the baseline, control conditions in adult mice. As shown in Fig. 1c, in physiological conditions, the nestin bioluminescence signal was restricted to the olfactory bulb region and areas reflecting subventricular zones and hippocampus. To further validate our transgenic model, we next investigated the expression patterns of the nestin-driven transgenes luciferase and GFP. The analysis of the nestin signal by in vivo bioluminescence was followed by double-immunofluorescence analysis confirming the nestin/GFP co-localization. Moreover, the appropriate negative control experiments have been performed to additionally validate our transgenic model and in vivo imaging approaches (Additional file 1: Figure S1A–C) as well as the used immunofluorescence protocols (Additional file 1: Figure S1D–G). As shown in Additional file 1: Figure S1A–C, the in vivo imaging negative controls in wild-type (WT) mice were performed in baseline conditions and 24 h after MCAO and LPS injection. Importantly, in any of the tested conditions, we did not observe a non-specific bioluminescence signal. In addition, we also analyzed brain sections of WT mice for non-specific immunofluorescence signals in control conditions as well as following MCAO and LPS challenge. Importantly, as shown in Additional file 1: Figure S1D–G, in the tested conditions, we did not observe a non-specific immunofluorescence staining, thus further confirming the validity of our transgenic model system.

In the adult brain, neurogenesis persists in two brain regions referred to as neurogenic niches, the subgranular zone of the dentate gyrus (DG) in the hippocampus and the subependyma of the lateral ventricles called the subventricular zone (SVZ). While SVZ progenitors give rise to rostral migratory stream and incorporate into the olfactory bulb as new olfactory neurons [16], the subgranular zone progenitors migrate to the granule cell layer of the dentate gyrus and differentiate primarily into granule cells and/or interneurons [16, 17]. Importantly, the glial and neuronal progenitors are marked by expression of the intermediate filament nestin. Previous studies have demonstrated that in normal condition, there is a low level of NPG proliferation and nestin expression [16, 17]. In keeping with previous findings, analysis of the adult brains in normal conditions revealed a low basal level of the endogenous nestin protein presented in the ventricular zone (Fig. 1e–g), dentate gyrus (Fig. 1i–k), and olfactory bulb (Fig. 1m–o) and low levels of the nestin-driven GFP transgene immunoreactivity. As further revealed by double-immunofluorescence analysis, the majority of nestin-positive cells co-localize with the GFP, thus suggesting an adequate transgene expression pattern (Fig. 1e–o). Hence, in the adult mouse brain and in physiological conditions, the nestin signal expression is indeed restricted to a neurogenic niche region comprising the SVZ and DG as well as the olfactory bulb (OB).

To date, it has been widely established that ischemic injury in the brain triggers adult neurogenesis in the SVZ of the lateral ventricles and the SGZ of the dentate gyrus [18‐21]. Furthermore, cerebral ischemia is associated with a marked acute and chronic inflammatory response, events that may significantly affect nestin signal and/or affect its cellular expression patterns. To visualize the spatial and temporal dynamics of the nestin signal after stroke, the transgenic nestin-luc/GFP mice were subjected to 90-min transient MCAO followed by different reperfusion periods. The imaging protocol was initiated shortly before stroke (baseline level measurements), and the same animals were longitudinally imaged starting at 6 h, 24 h, and 3, 5, 7, 10, and up to 14 days after the initial ischemic injury (Fig. 2a–f). As expected, analysis of the baseline nestin signal showed low bioluminescence baseline levels. The ischemic injury induced a marked and progressive increase of the nestin signal, reaching its peak at 24–72 h and 1 week after ischemic lesion. After day 7, the intensity of the nestin signal slowly decreased reaching the baseline value at 10 days and stayed at that level till 2 weeks following stroke (Fig. 2g). As previously described, to determine whether photons detected by charge-coupled device camera after cerebral ischemia were emitted from the appropriate brain regions, we performed spectral imaging and a 3D reconstruction of the recorded signals using DLIT algorithms (see the “Methods” section) 7 days after stroke [13]. As revealed in Fig. 2h, the 3D signal reconstruction of the bioluminescent sources revealed that the nestin signal was indeed arising from the areas surrounding the brain structures normally affected by MCAO and involved in neurogenesis, including the SVZ and striatum as well as the hippocampus. These findings were further validated by immunohistochemistry analysis (Fig. 3a–j). Importantly, recorded signal distribution was consistent with our previous results obtained by in vivo imaging of different reporter mice following transient MCAO [7, 10].

Previous evidence suggests that the nestin-positive progenitors may also give rise and differentiate into GFAP-positive astrocytes; we next investigated the cellular expression/induction patterns of nestin after stroke. Based on the results obtained by in vivo imaging, i.e., peak signal induction of nestin (see Fig. 2g), the ischemic brains were collected at 3–7 days after ischemia and stained for nestin and GFP. Because previous evidence suggests a marked up-regulation of nestin in reactive astrocyte surrounding the lesion site, we first analyzed the pattern of the nestin-driven GFP expression in the peri-infarct/stroked zone. The specificity of transgene expression was then assessed by analysis of GFP expression in the cell types known to be highly responsive to brain damage. Immunofluorescence analysis of the peri-infarct region showed co-localization of the nestin-driven transgene GFP with endogenous protein or with the known astrocyte marker GFAP (Fig. 3a–g). Next, the cellular expression pattern of the GFP transgene was also assessed in the hippocampus (Fig. 4h–j) and in the SVZ (Fig. 3k–m), neuroanatomical regions known to be involved in the adult brain neurogenesis. We observed a marked up-regulation of nestin in the ipsilateral dentate gyrus at 1-week post-injury (Fig. 3n). Namely, the quantitative analysis revealed a significant increase in the numbers of the GFP and nestin-positive cells at 7 days after initial stroke. The similar pattern of the GFP and nestin immunoreactivities was observed at the SVZ region (Fig. 3o). Taken together, quantitative analysis of the GFP, GFAP, and nestin-positive cells revealed a significant increase in cell numbers starting at 3 days and peaking 7 days after stroke (Fig. 3n–q). Importantly, and as revealed by combined in vivo imaging and double-immunofluorescence data, the early peak of the nestin signal observed by in vivo imaging (biophotonic/bioluminescence imaging protocol) (24–72 h after stroke) is caused, in part, by a marked increase in the nestin signal arising from the activated astrocytes located at the peri-infarct region.

Contrary to brain response to injuries, including stroke, previous evidence suggests that the acute and chronic inflammation may have a negative impact on neurogenesis [22‐27]. To assess the spatial and temporal dynamics of the nestin signal in the context of neuroinflammatory conditions, we used a well-established model of the LPS-induced acute and chronic inflammation. We used two different approaches. The nestin-luc/GFP transgenic mice were injected by single dose of LPS or treated chronically over 14 days (see the “Methods” section) (Fig. 4a–l). The in vivo analysis of bioluminescence revealed that, when compared to baseline levels, a single LPS injection induced a transient but robust increase of the nestin signal peaking 24 h after initial stimuli (Fig. 4a–d). Interestingly, after the single LPS injection, the signal tends to decline after 3 days reaching the baseline levels at 5–7 days (Fig. 4m). In the chronic inflammatory model, the nestin signal was up-regulated after 24 h of LPS and tends to increase after every 3 days (periodical injection) (Fig. 4e–l). However, at the end of the 14 days, the recorded signals were below the baseline levels (Fig. 4m). The cellular expression patterns of nestin were further confirmed by immunofluorescence analysis. The quantitative analysis of the nestin and GFAP expression reveals a marked increase in nestin (Fig. 5a–c, g–i) and GFAP (Fig. 5d, f, j–l) immunoreactivities in the dentate gyrus and in the SVZ 1 day after LPS injection. This increase in immunoreactivities correlates with the increase in the numbers of nestin and GFAP-positive cells in the two regions analyzed. The number of nestin and GFAP-positive cells is significantly increased at day 1 and starts to decrease in the DG and in the SVZ as early as 4 days after LPS injection to reach lower than the basal level at 14 days (Fig. 5m, n). In order to determine whether the observed increase in the number of the GFAP-positive cells is due to an augmentation in proliferation, we performed double immunofluorescence for Ki67 (marker of proliferation) and GFAP 1 day after LPS injection. In fact, very few GFAP cells expressed the proliferation marker Ki67 in the SVZ (Fig. 5o–q) or DG (Fig. 5r–t). Thus, our results suggest that the observed increase in the number of GFAP-positive cells after LPS, at the tested time point, is not due to proliferation [28].

While nestin have been widely considered as a putative neuronal stem cell marker, in neuroinflammatory and in post-injury conditions, nestin could also be observed in activated and GFAP-positive astrocytes (see Figs. 3 and 4) [29, 30]. Importantly, the immunofluorescence analysis of the brain sections after stroke revealed several GFP-positive cells with glial morphology that were not positive for the astrocyte marker GFAP. We next asked whether resident glial cells including microglia may express and/or up-regulate nestin in neuroinflammatory conditions. We analyzed activated microglial cells for nestin expression at different time points following LPS challenge and after MCAO. Our areas of interest were the DG and SVZ regions, and since MCAO injury affects the area close to the SVZ, this region also included a peri-infarct zone/stroke area. As expected, the LPS challenge and the inflammatory response following MCAO were characterized by a marked increase in Iba-1 staining (Fig. 6a–r). Interestingly, a double-immunofluorescence analysis revealed that a significant number of activated microglia/macrophages were indeed positive for nestin. As shown in Fig. 6s–z, quantitative analysis of the Iba-1 and the Iba-1/nestin-positive cells revealed that following the LPS challenge, a number of Iba-1/nestin-positive cells showed the highest increase 24 h after the LPS challenge, while following MCAO, the number of nestin-positive microglia/macrophages peaked 7 days after stroke followed by decline at 14 days following initial ischemic injury (Fig. 6u, v, y, z). Note that the SVZ region in the stroke-affected mice also comprises an ischemic area affected by MCAO (Fig. 6p–r). Here it is important to mention that, in the brain, the systemic LPS-induced microglial activation does not lead to infiltration of the peripheral cells; thus, Iba-1 and Iba-1/nestin-positive cells are indeed resident microglia [31]. In the stroke model, however, we cannot exclude that the subset of the nestin/Iba-1-positive cells may be of peripheral origin, thus representing the microglia/macrophage population.

One of the important features of microglial activation, especially in the context of ischemic injury, is a robust proliferative response [11, 32]. Thus, we next investigated whether nestin/Iba-1+ cells are indeed dividing microglia. We performed immunofluorescence analysis for Ki67 (marker of proliferation), nestin, and Iba-1. Given the temporal differences in the peak numbers of the Iba-1-positive cells following LPS challenge and MCAO (see Fig. 6), identification analysis of the Iba-1/nestin/Ki67-positive cells was performed either 24 h (LPS) or 72 h (MCAO) following initial brain challenge. As revealed in Fig. 7, following stroke, only few Iba-1/nestin/Ki67-positive cells were detected in the stroke area (SA), DG, and SVZ 72 h after MCAO (Fig. 7a–p). The similar findings were obtained following LPS challenge (Fig.7m–t). Namely, we identified only few Iba-1/nestin/Ki67-positive cells in the DG and SVZ. To further confirm that microglial cells up-regulate nestin in inflammatory conditions, we performed additional series of in vitro experiments using primary glial cultures, enriched in microglia. Indeed as revealed in Fig. 7u–bb, the LPS treatment was associated with a robust induction of nestin in CD11b-positive cells. In addition, the triple Iba-1/nestin/Ki67 staining revealed that some of the nestin-positive microglial cells were dividing (Fig. 7bb).

Taken together, our results strongly suggest that following injuries and in the context of acute innate immune challenge, in addition to reactive astrocytes, activated Iba-1 microglia/macrophages strongly up-regulate nestin. This suggests that nestin may serve as a context-dependent marker of activated microglia/macrophages.