Characterization of CamKII-transgenic mice: neuron-restricted reprogramming in young mice

Since the transgenic model we used is entirely novel, we aimed to characterize it initially and verify that indeed Yamanaka factors (YF) were being effectively expressed. To do this, we began by studying the first offspring whose genotyping confirmed the presence of both transgenes. We maintained continuous activation of the doxycycline-inducible system from birth until the mice reached three months of age (when the nervous system should be fully formed), time when we collected samples.

To assure that exogenous YF were effectively expressed, quantitative PCR (qPCR) was performed against the sequence E2A-cMyc of the transgenes OSKM using hippocampus, neocortex and cerebellum samples. The results revealed a clearly higher expression in -CaMKII-OSKM compared to transgenic control mice in hippocampus and neocortex, but no difference was observed in the cerebellum between transgenic and control mice, being almost undetectable (Supplementary Fig.1a).

Attending the expression of one of the YFs (Klf4) in cerebral histological regions, we were able to determine more precisely the specific areas in which the promoters were active throughout the experiment (Supplementary Fig.1b). It is important to note that due to the nature of the transgene, the expression of Klf4 (located in the third position within the gene order of the construct, after Oct4 and Sox2) implies that the other genes should have also been expressed in the neuron, so the expression pattern observed with Klf4 would correspond to that of the transgene as a whole.

The highest expression occurs in different neuronal layers of the hippocampal formation (DG, CA3, CA1, Subiculum), dorsal and ventral striatum, and neocortex (somatosensory, somatomotor, visual or orbital areas). Some expression was also observed in the thalamus but at a much lower level. No expression was found in other diencephalic regions, such as the globus pallidus, or in brainstem regions such as the substantia nigra, superior or inferior colliculus, cerebellum, or medulla. The varied pattern of -CaMKII promoter expression throughout the different cerebral regions in these mice is in line with that described previously20.

In order to better understand the molecular process behind, we have focused first on the hippocampal formation, which not only have exhibited one of the highest Klf4 protein expressions in this model but also because this region is involved in different key brain processes such as memory functioning or adult neurogenesis, both processes affected by ageing21. We obtained bulk transcriptomic data from the hippocampi of 8 different animals to which doxycycline was not administered since birth, allowing neuronal cells from double transgenic mice to express YF continuously since birth. RNA-seq analysis (ENA accession number is PRJEB56610), by using Deseq2722 R software package, detected 1419 differentially expressed genes (DEG) (q value<0.05) in neuronal-reprogrammed animals regarding control mice. From all of them, 604 DEG resulted in decreased and 815 in increased expression. The MA-plot shows the log2 fold changes (M) between two conditions over the mean of normalised counts (A) for all samples (Supplementary Fig.1c). In the heatmap (Supplementary Fig.1d), is also shown how the expression of YF in a subpopulation of neurons, has led to significant changes in the transcriptome of a relevant group of genes. According to the Gene Ontology (GO) knowledgebase, which is the largest source of information on the functions of genes and proteins, and to the Kyoto Encyclopedia of Genes and Genomes (KEGG), we have been able to track down groups of differentially expressed genes (DEG) involved in certain cellular functions (Supplementary Fig.1e-f; Supplementary Data1). Data obtained revealed changes in the expression of genes related with regulation of nervous system development (46 DEG; 1.68E-07 p.adj), stem cell differentiation (21 DEG; 0.027 p.adj) or maintenance (17 DEG; 0.02 p.adj), central nervous system neuron differentiation (23 DEG; 1.74E-03 p.adj) or specifically in regulation of neuron differentiation (31 DEG; 1.91E-05 p.adj), would confirm that processes related with reprogramming in -CaMKII-OSKM mice are taken place. In addition, transcription changes were found in genes related with: extracellular matrix organization (48 DEG: 5.09E-10 p.adj), structure (29 DEG; 1.29E-08 p.adj) or in the regulation of cell-cell adhesion (62 DEG; 2.86E-10 p.adj); in the regulation of synaptic organization (38 DEG; 1.68E-07 p.adj) and of synaptic structure or activity (39 DEG; 1.20E-07 p.adj); learning or memory (39 DEG; 4.30E-06 p.adj) and cognition (42 DEG; 3.85E-06 p.adj); dendrite morphology (26 DEG; 3.26E-05 p.adj) or development (34 DEG; 0.00019 p.adj), axogenesis (61 DEG; 3.07E-09 p.adj) as well as in regulation of neurogenesis (57 DEG; 6.93E-09 p.adj). It is important to note that all these functions are ultimately affected with ageing.

Furthermore, using a bioinformatics tool such as Ingenuity Pathway Analysis (IPA)23, we were able to identify a common upstream regulator of downstream genes. IPA revealed several different types of upstream molecules (~400), including transcription regulators, transporters, cytokines, growth factors, kinases or various enzymes. Among all upstream regulators, HMG20A, IL4, FGFR, C1QA, KTMT2D, KAT2A or CREBBP are within the top 20 most significantly activated regulators (P<10-05 and Z-score 2; among the top 30 upstream regulator). These regulators have been associated with epigenetic modifications (KTMT2D, KAT2A), neuronal differentiation (HMG20A, FGFR, KTMT2D), neurodevelopment (C1QA, FGFR, CREBBP), the ageing process (IL4, FGFR) or memory functioning (KAT2A). On the other hand, APOE, TP63, Ptprd or PSEN1 were among the most significantly inhibited regulators (P<10-04 and Z-score 2; among the top 30 upstream regulator), and they are also involved in ageing or ageing-associated diseases (APOE, TP63, Ptprd or PSEN1), cell adhesion (Ptprd) or neural differentiation (PSEN1).

During these experiments the impact of continuous neuron-restricted reprogramming on mortality rate was analysed. Thus, in the case of these young mice that underwent continuous induction of transgenes from birth, a high mortality rate (around 60%) was observed, along with the presence of hydrocephalus in some cases. However, for the mice in which the transgene system was only activated in adulthood at 6 months old, whose study will be described below, the mortality rate decreased to zero. Neither in the case of continuous induction nor in the case of cyclic induction did we find teratomas, confirming that the expression of YF in neurons did not lead to cellular dedifferentiation in vivo. This result is in line with the findings of Kim and colleagues17.

Next, we proceed to study the impact of neuron-restricted reprogramming on adult mice during long-term treatment, initiating the treatment in mice aged 6 months to nearly one-year-old, a protocol similar to that described previously13. Some mice were subjected to continuous factor expression by the continuous withdrawal of doxycycline over the 4-month treatment period, while another group of mice underwent cyclic doxycycline administration. This cyclic protocol involved administering water for 3 days per week and doxycycline for the remaining 4 days (Fig.1). The cyclical expression of Yamanaka factors was confirmed by the histological study of Klf4 protein expression at different time intervals during a week cyclical protocol (Supplementary Fig.2a-b). The analysis of immunofluorescence obtained from Klf4 Yamanaka factor expression have shown that indeed, after 3 days of induction, there is a significant increase (p value=0.0043) in the expression of Klf4 Yamanaka factor (Supplementary Fig.2c). This induction returns to day 0 levels after 4 days of continuous doxycycline administration (p value=0.0146).

a Crossbreeding was conducted between -CaMKII-tTA and TetO-OSKM transgenic mice to generate the -CaMKII-OSKM mice. b Schematic representation of the OSKM transgene showing the location of DNA sequences encoding 2A peptides that separate each Yamanaka factor. c Temporal representation of the three OSKM treatments applied to the murine model in the present study. Doxycycline administration prevents binding between the transactivator tTA and the TetO promoter and thus inhibits transcription of Yamanaka factors.

To characterize the histological expression of YF in adult -CaMKII mice, we employed antibodies targeting KLF4, just as in the previously shown study with young mice (Fig.2a, b; Supplementary Fig.3a). Quantitative analysis of immunohistofluorescence in adult -CaMKII mice subjected to cyclic transgene induction from 6 to 11 months, demonstrated significantly higher YF expression compared to transgenic control mice (Fig.2c, d). This difference was more pronounced in the deeper layers of the somatosensory neocortex (Fig.2c). In the control group, YF expression was nearly absent in all regions of the cerebral cortex, as shown in Fig.2a. Additionally, continuous transgene expression resulted in higher YF expression levels compared to transgenic control mice and adult -CaMKII mice with cyclic YF expression. This difference was more evident in the hippocampal region (Supplementary Fig.3b, c).

a, b Representative microphotographs of the Klf4 Yamanaka Factor immunoreactivity (in green) obtained in sagittal brain sections from control 11 months old adult mice (n=5) (a), -CaMKII-OSKM mice with cyclic induction (n=5) of transgene system from 6 to 11 months of age (b). Schematic outlines approximating the boundaries of some of the most relevant areas of the central nervous system have been overlaid on the microphotographs. Scale bar shown in a and b indicates 1000 m. c, d Graphical representation of the meanSEM of the percentage of occupied area by Klf4-immunoreactivity in somatosensory neocortex, distinguishing deep layers from superficial (c) and different regions from hippocampal region, including CA1, CA3 and dentate gyrus (d). *p<0.05 and ** p<0.01 by Students paired t-test.

With respect to the regional distribution of Klf4 expression, adult -CaMKII mice with cyclic and continuous expression of YF exhibited an immunofluorescence pattern of Klf4 protein expression similar to that of mice with continuous expression from birth at 3 months of age (Fig.2b; Supplementary Fig.3a). The highest expression was found throughout all the neocortex, mainly in deeper layers. Expression was also found in the hippocampus, subiculum, caudate putamen, piriform cortex and thalamus, but cerebellum or other medullary nucleus seemed to lack YF expression.

We first analysed the animals anxiety levels by observing their exploration behaviour within the central area of the box. Typically, rodents tend to remain close to the walls and avoid open spaces, a behaviour known as thigmotaxis24. As rodents age, it appears that they tend to spend less time in the central zone of the open field, which would translate to higher levels of anxiety25,26. The results showed statistically significant increase (p value=0.0003) in the time spent by the -CaMKII-OSKM mouse in the central zone of the tray when they had cyclic treatment in comparison with the control group (Fig.3a, b). As already described (see Materials and Methods), a short-term (2hours) novel object recognition test was used to assess memory performance. The results of the test (Fig.3c, d) showed a higher memory index in terms of time (P value=0,0149) and entries (P value=0,0204) in -CaMKII-OSKM regarding the control group. Spatial memory was also evaluated through the Y-maze test (Fig.3e, f), which showed improvements in the -CaMKII-OSKM mice with cyclical administration of doxycycline with respect to the control group (P value=0,0422). Therefore, cyclic activation of YF expression restricted to a subpopulation of neurons was enough to improve different types of memory in middle-age mice. Contrary to the cognitive effects found in -CaMKII-OSKM adult mice with the cyclical induction of YF, continuous induction did not result in significant changes, either in the open field test or in the other tests conducted.

a, c, e Schematic representation of the organisation of the space where the behavioural tests were performed, a Open field test, b Novel Object Recognition test and c Y maze test. Representative tracking maps obtained with Any Maze software, showing the trajectory of the centre of the rodent during the behavioural test. b, d, f Graphical representation of the meanSEM of the time (s) between the number of entries in the area analyzed (see Material & Methods) in different experimental groups. Control 11 months old adult mice (n=8) and -CaMKII-OSKM mice with cyclic induction (n=9); control 11 months old adult mice (n=8) and -CaMKII-OSKM mice with continuous induction (n=7). *p<0.05, ** p<0.01, *** p<0.001.

Given that only the cyclical induction of YF in adult neurons yielded noteworthy improvements in cognition compared to continuous induction, we opted to focus on this approach, which would entail expression of YF within the -CamKII promoter subpopulation of neurons.

We obtained transcriptomic data from both hippocampal and neocortical tissues of adult animals. Bulk RNA-seq analysis (ENA accession number is PRJEB65922), by using Deseq2722 R software package, detected in total 94 differentially expressed genes (DEG) (q value<0.05) in neuronal-reprogrammed animals regarding control mice (Supplementary Data2). Out of all of them, the majority (~75%) showed decreased expression (70 DEG), while around 25% their expression was found to have increased. The gene expression data can be visualized in Fig.4 where the MA-plot is shown for all samples in the neocortex and in the hippocampus (Fig.4a, b; Supplementary Data2). In the heatmaps (Fig.4c, d), it is also shown how the expression of YF in a subpopulation of neurons, has led to significant changes in the transcriptome of a relevant group of genes.

a, b MA-plot from neocortex (a) and hippocampus (b) samples which represent genes coloured in blue that have q values less than 0.05. Points which fall out of the window are plotted as open triangles pointing either up or down. Heatmap from neocortex (c) and hippocampus (d) samples. Data are displayed in a grid where each row corresponds to a gene and each column to a sample (from two different conditions). The colour and intensity in the heatmap represent changes of gene expression from the list of genes with q value in the Principal Component Analysis (PCA). Emapplot from neocortex (e) and hippocampal (f) samples, showing an enrichment Map for enrichment result of over-representation test or gene set enrichment analysis. g GO term analysis (cellular component) of altered genes.

According to the Gene Ontology (GO) knowledgebase, we have been able to track down groups of differentially expressed genes (DEG) involved in certain cellular functions (Fig.4e, f; raw data in Supplementary Data2). Firstly, it is important to note that, when examining the cellular component of ORA analysis (considering the both hippocampus and neocortex), the transcriptomic data revealed that reported changes were mainly located in specific neuron compartments, with the dendritic component standing out prominently (Fig.4g). This finding confirms the specificity of YF expression solely in neurons. Data obtained revealed that genes with altered expression (qvalue<0.05) were included in different biological processes, as regulation of nervous system development (Nectin3/Chrna4/Lrrtm4/Gabra5; 3.7E-04 padj) and neuron differentiation (Dab1/Trpc6/Brinp3/Neurog2; 5.5E-04 padj) that indicate functions compatible with a reprogramming process. Moreover, in general many of these differentially expressed genes (DEG), could be primarily grouped into alterations in processes related to the extracellular matrix (e.g. ECM organization Itga8/Col15a1/Adamts16/Fbln2/Grem1/Col22a1; 5.45E-04 Padj) and cell adhesion (e.g. regulation of cell substrate-adhesion; Col26a1/Pcsk5/Thy1/Ajap1/Fbln2/Ppm1f/Grem1, 7.27E-06), neuronal activity involving different classes of neurotransmitters (13 different processes, e.g., neurotransmitter receptor activity; Chrna4/Chrnb3/Gabra5/Htr2a/Hrh3, 2.96E-05 padj), cognition (Itga8/Chrna4/Gabra5/Htr2a/Hrh3/Jun, 9.4E-04), and processes always associated with neuronal structures and functions, with a particular emphasis on postsynaptic processes (e.g, postsynaptic specialization; Nectin3/Itga8/Chrna4/Dab1/Lrrtm4/Gabra5/Als2/Lzts3, 9.53E-05 padj).

Taking into account the hippocampus and the neocortex separately, the expression of these factors seems to have led to a somewhat more intense reprogramming process in the second region compared to the hippocampal region, considering the number of genes with altered expression (43 DEGs in the hippocampus vs. 59 DEGs in the neocortex). Among the genes with the most statistically significant differential expression, notable examples relate to the extracellular matrix, with collagen alpha 1 type XXVI (Col26a1; 7.5831E-11 padj) in the hippocampus and collagen alpha 1 type XXII (Col22a1; 1.63E-4 padj) in the neocortex. Furthermore, zinc finger proteins, like Zfp804b (3.6816E-06 padj) in the neocortex and Zfp386 (4.94E-09 padj) in the hippocampus, are noteworthy. Given their capacity to bind to chromatin, these proteins are believed to play a central role in neuronal reprogramming processes27 and the latter have been involved in silencing LINE-1 elements28,29. In addition, there have been genes whose expression has been found to be altered in both cortical areas, such as Glis3 (another zinc finger protein) and Fbln2. The expression of the Glis3 gene, functionally involved in reprogramming processes30 and known to increase with aging31,32, was significantly reduced in both hippocampal and neocortical regions in this study. Fbln2 is an extracellular matrix protein with roles in tissue remodelling and embryonic development33,34.

Since transcriptomic studies have revealed significant effects of partial reprogramming on regulatory neuronal activity genes we aimed to investigate, in a more specific manner, how these changes in neuronal activity may have contributed to the cognitive improvement observed in these animals during partial reprogramming. For this purpose, we conducted histological analyses on tissue samples using an immediate early gene c-Fos marker. In neurons, c-Fos expression is induced under conditions of neuronal plasticity, including learning and memory35. It has been widely used as a neuronal activity marker since they are rapidly and transiently induced by neuronal stimuli in the brain36. In this study, the mice were immediately perfused upon completion of the memory test. In this way, we were able to study the activity levels of the memory circuits during the execution of these memory tests. The analysis of c-Fos-immunoreactive cells was focused on the hippocampus (Fig.5a), due to its essential role on recognition/spatial memory performance. We found a higher number of neurons active (c-Fos-immunoreactive) just after memory test performance in -CaMKII-OSKM transgenic mice regarding transgenic control mice, in both granular cell layer of dentate gyrus and in the pyramidal cell layer of CA1 (Fig.5b). This result potentially indicates more active circuits during memory testing due to YF expression restricted to neurons.

a Representative microphotographs of immunoreactivity obtained for c-Fos protein expression (in red) in the hippocampus region from control (n=8) and -CaMKII-OSKM mice (n=10). Scale bar indicates 200 m. b Graphical representation of the meanSEM of c-Fos-immunoreactive cells per mm3 at different neuronal layers of the hippocampal region (Dentate gyrus, DG; CA1; CA3). *p<0.05.

Additionally, we have found in these mice a significant inverse correlation (R=0.964) between levels of Klf4 expression and density of c-Fos-immunoreactive cells (Supplementary Fig.4). Excessive expression of Klf4 led to lower increase of c-Fos-immunoreactive cell density in the hippocampus of -CaMKII-OSKM transgenic mice. These results are consistent with those found in -CaMKII-OSKM mice with continuous induction of the YF. In these mice, where Klf4-YF expression is continuous, we observed worse cognitive performance compared to those with cyclical induction. All these results underscore once again the importance of the level of induction of the YF. Moderate rather than excessive induction is what achieves beneficial effects on the cognition of aged mice.

Considering significant changes previously identified in the extracellular matrix (ECM) as a result of reprogramming in young -CaMKII-OSKM mice (Supplementary Fig.1, Supplementary Fig.5), which led to an overall reduction in its structure, in addition to transcriptomic data obtained from -CaMKII-OSKM adult mice showing significant alterations in genes related to ECM, we aimed to investigate whether partial reprogramming in adult mice would result in youthful ECM reorganization. Thus, we studied the expression of the cartilage-specific core protein proteoglycan (aggrecan), which binds to specific proteoglycans and allows visualisation of the so-called perineuronal extracellular matrix networks (Fig.6a and d from the overall panoramic view). Immunoreactivity analysis for this protein was carried out in both the neocortex and hippocampal formation at both experimental groups (control and -CamK-OSKM adult mice with cyclic overexpression of YF). Figure6a shows representative microphotographs of the neocortex using an antibody against aggrecan protein, where, to facilitate the analysis, the supragranular layers (layers I-IV) have been distinguished from the infragranular layers (layers V and VI). The results have shown a prominent inclination towards an overall reduction in the percentage of area occupied by the aggrecan protein across the entire neocortex (p value=0.0518), attaining statistical significance within the deeper neocortical layers (V and VI) among -CaMKII-OSKM mice in comparison to the control group (p value=0.0258; Fig.6b).

a Representative microphotographs of immunoreactivity obtained for Aggrecan protein expression (in red) in the somatosensorial neocortex from control (n=8) and -CaMKII-OSKM mice (n=9). On the right side of the panel are enlargements of the panoramic view displaying the structure of perineuronal networks formed by the extracellular matrix. Scale bar shown in A indicates 200 m and 15 m in the magnification. b Graphical representation of the meanSEM of the percentage of area occupied by Aggrecan-immunoreactive cells per mm3 in total volume of the somatosensory neocortical region and at different cortical layers. *p<0.05. c Density of perineuronal net units (PNNs) in the somatosensorial neocortex (number of Aggrecan-immunoreactive PNNs in each brain slice by the volume of the somatosensory neocortical region, *p<0.05). d Representative microphotographs of immunoreactivity obtained for Aggrecan protein expression (in red) in the hippocampus of the different murine models. On the right side of the panel are enlargements of the panoramic view displaying the structure of perineuronal networks formed by the extracellular matrix. Scale bar shown in (a) indicates 200 m and 15 m in the magnification. e Representation of the meanSEM of the percentage of area occupied by Aggrecan immunoreactive per mm3 at different neuronal layers of the hippocampal region (Dentate gyrus, DG; CA1; CA3). f Density of perineuronal net units (PNNs) in total hippocampus (number of Aggrecan-immunoreactive PNNs in each brain slice by the volume of the area analysed).

In line with these results, the analysis of the density of perineuronal net units also shows a significant reduction following the induction of partial reprogramming in -CaMKII-OSKM mice (Fig.6c). In contrast to the neocortical areas, noteworthy statistical differences were not found in the hippocampus between both experimental groups for the area occupied by the immunoreactive aggrecan matrix (Fig.6e), nor regarding the density of PNN units (Fig.6f). This general reduction found by immunofluorescence detection in aggrecan-immunoreactive extracellular matrix has been corroborated by Western blot technique (Supplementary Fig.6a, b). We observed a highly significant decrease (P=0.0001) in its expression following the cyclical induction of the Yamanaka factors. These data confirm the significant role of extracellular matrix reorganization during cyclical reprogramming processes in the brain.

In vitro studies have demonstrated that YF alone is not sufficient to induce neuronal dedifferentiation17. In this study, we decided to try to validate these findings in vivo, determining whether, under YF expression, neurons could undergo dedifferentiation. For this analysis, we employed a distinct set of antibodies. Doublecortin protein (Dcx) is expressed in migrating neuroblasts and immature neurons, making it a reliable marker for adult neurogenesis. In general, in adult rodents, it is only possible to find immature neurons in regions where neurogenesis occurs, which are typically only two, one of which is the subgranular zone (SGZ) of the hippocampal dentate gyrus. Thus, we tried to detect doublecortin labelling outside of the subgranular zone in -CaMKII-OSKM mice. This would indicate that processes of dedifferentiation owing to YF expression from mature neurons to a previous state of maturation have occurred. In these animals, throughout the cerebral cortex we only observed doublecortin labelling in the subgranular zone, similar to what was seen in the control group. We did not find these cells in the rest of the hippocampus or the neocortex. Moreover, we observed no differences in the density of Dcx-immunoreactive cells between the control and -CamKII-OSKM mice in SGZ (Supplementary Fig.7a, d). Furthermore, we aimed to study other markers of earlier neuronal development such as the intermediate progenitor marker T-box brain gene 2 (Tbr2) to ensure that partial reprogramming was not regressing to even earlier stages than those identified by the doublecortin marker. The results in the SGZ revealed no differences in Tbr2 expression between the transgenic and control groups (Supplementary Fig.7b, e). Additionally, 5-chloro-2-deoxyuridine (CldU) was administered three weeks before perfusion for each mouse in both experimental groups to identify cells that were newly generated at that time. Results have shown how the number of three weeks-old cells labelled with CldU was not significantly changed in the DG of the -CaMKII-OSKM, indicating that partial reprogramming by YF not only does it not appear to influence adult neurogenesis itself, but it also would not affect the proliferation of new cells in adult mice (Supplementary Fig.7c). Differences in the density of CldU-labelled cells between control and -CaMKII-OSKM adult mice in the somatosensory neocortex were not found either (Supplementary Fig.7c, f).

Considering that the reprogramming induced by Yamanaka factors correlates with epigenetic changes7,37, in this study we aimed to verify whether the selective partial reprogramming of neurons led them to more youthful epigenetic states. Early studies in rats showed that methylation of histones H3 and H4 changes gradually with increasing age38. Moreover, a systematic study of posttranslational modifications of histones in the brain of senescence-accelerated prone mouse 8 (SAMP8) model revealed a significant decrease of H4K20me3 marker during ageing39. Here, we have found a remethylation of this epigenetic marker at H4 after cyclic neuronal induction of YF expression in adult mice (Fig.7). The results showed that H4K20me3 (histone 4 lysine 20 trimethylation) marker increases in -CamKII-OSKM adult mice overexpressing YF in a cyclic manner regarding control mice throughout all the neocortex layers (p value=0.0124; Fig.7a, b), as well as in CA1 (p value=0.0372) and CA3 (p value=0.0288) pyramidal cell layers at the hippocampal region (Fig.7c, d), but not in the granular cell layer of DG. According to previous studies these epigenetic changes in areas where partial reprogramming is taking out would lead to a more youthful epigenetic pattern in those reprogrammed cortical neurons39.

a Representative microphotographs of immunoreactivity obtained for H4K20me3 expression marker (in green) in the somatosensory neocortex from control (n=8) and -CaMKII-OSKM mice (n=9). On the right side of the panel, enlargements are displayed, showing H4K20me3-immunoreactivity expression in a group of cells present in the deeper neocortical layers. Scale bar shown indicates 200 m and 15 m in the magnification. b Representative percentage of mean intensity (arbitrary units) obtained from H4K20me3 immunoreactivity in neocortex, distinguishing supragranular layer (I-IV) and infragranular layers (V-VI). *p<0.05. c Representative microphotographs of immunoreactivity obtained for H4K20me3 expression marker (in green) in the hippocampal region of the different murine models. On the right side of the panel, enlargements are displayed, showing H4K20me3-immunoreactivity expression in a group of cells present in the pyramidal neuronal layer of the CA1 region. Scale bar shown indicates 200 m and 15 m in the magnification. d Percentage of mean intensity (arbitrary units) obtained from H4K20me3 immunoreactivity in hippocampus, distinguishing CA1, CA3 and dentate gyrus (DG). *p<0.05.

More here:
In vivo cyclic overexpression of Yamanaka factors restricted to neurons reverses age-associated phenotypes and ... - Nature.com