Comprehensive antibody validation experiments were carried out using western blot and cell and tissue immunostaining. This approach allowed analysis of various available antibodies against DDAH1 and DDAH2 in order to identify the most specific antibodies for the final study (Supplementary Table S1-S2, Supplementary Figure S1-S2). We focused on the molecular mass of the protein band and its intensity in addition to the specificity of antibody binding with endogenous DDAH1 and DDAH2.
DDAH1 distribution in the mouse brain
To investigate DDAH1 protein distribution in the brain, we initially used anti-DDAH1 staining on mouse brain sections. As shown in Figure 1A, DDAH1 was widely distributed with an intense signal in the striatum, cortex and hippocampal formation (HPF). A moderate signal was observed in the thalamus and cerebellum (CB). In the striatal regions, cortex, HPF and CB the DDAH1 positive cells had an area of 12-16 µm2, star-shaped soma and many processes. In addition, a few cells with the same size but with a round soma and low expression level of DDAH1 were observed. Moreover, DDAH1 positive cells with star-shaped morphology were detected in white matter structures such as corpus callosum. In the thalamus and in some structures of the midbrain and hindbrain, low levels of DDAH1 expression were observed in round-shaped cells with 16-18 µm2 area, which were slightly bigger than the cells in the HPF and striatum.
Within cortical regions, DDAH1 immunoreactivity was intense in layers I – IV and isocortex (Figure 1A, box 1). In the striatum, cells with DDAH1 expression were evenly distributed. DDAH1 positive cells were detected in both the dorsal and ventral regions, in the lateral septal complex and, in a small amount, in the striatum-like amygdalar nuclei. The most intense DDAH1 expression, however, was observed in the dorsal region of the striatum (Figure 1A, box 2). In the thalamus, the DDAH1 signal was moderate and present only in a limited number of brain nuclei such as the paraventricular nucleus (PVT) (Figure 1A, box 3) and the reticular nucleus (RT). In HPF, few DDAH1 positive cells were present in Ammon’s horn (cornu Ammonis, CA) and the dentate gyrus (DG). In the CA, DDAH1 positive cells were observed in the stratum oriens (so), stratum radialis (sr), and stratum lacunosum-moleculare (slm) layers (Figure 1A, box 4). In the DG, strong DDAH1 expression was present in the polymorph and molecular layers. Interestingly, no DDAH1 positive cells were found in both the pyramidal layer (sp) of CA and the granule cell layer in the DG. However, we observed some processes of DDAH1 positive cells that were passing through these layers. DDAH1 immunoreactivity was strong in different olfactory areas. High expression levels were observed in the main olfactory bulb (MOB), anterior olfactory nucleus (AON) and in the layer II of the piriform cortex (Figure 1A, box 5). In CB, DDAH1 immunoreactivity was low in the Purkinje layer, without any detectable DDAH1 expression in other layers and cerebellar nuclei (Figure 1C). Additionally, some DDAH1 positive cells were observed in the dorsal nucleus Raphe (Figure 1B).
We compared our findings to publicly available mRNA expression datasets [41, 42] observing similar and intense expression profiles in the cortex, CB and MOB as well as low expression in the hypothalamus. Further high levels of DDAH1 protein in the striatum and HPF mapped to low Ddah1 expression [41]. On the contrary, low IHC reactivity reflected intense Ddah1 expression in the ventral tegmental area. Notably, the strongest Ddah1 mRNA signals were found in a wide number of thalamic nuclei [41] while we observed only moderate DDAH1 protein expression restricted to a small number of nuclei including the PVT and RT (see Supplementary Table S3).
DDAH1 is expressed in both neuronal and glial cells
Next, we determined specific cell type expression by co-labelling analysis of DDAH1 with neuronal marker (NeuN), glial fibrillary acidic protein (GFAP), oligodendrocyte transcription factor (Olig2), ionized calcium-binding adapter molecule 1 (Iba1) and platelet/endothelial cell adhesion molecule 1 (PECAM-1). (Figure 2H). On the one hand, we found overlapping signal between the neuronal marker NeuN and DDAH1 in RT and PVT (Figure 2A) but no overlap between DDAH1 and the astrocyte marker GFAP (Figure 2D). On the other hand, DDAH1 and GFAP were expressed by the same cells in all layers of the HPF (Figure 2C), but we did not observe any overlap between DDAH1 and NeuN (Figure 2B). Immunocytochemical analysis on primary cell culture of cortical and hippocampal origin confirmed the presence of DDAH1 in astrocytes and neurons (Supplementary Methods, Figure S3A-C). Additionally, we performed co-staining for DDAH1 and other cell type markers such as Olig2 for oligodendrocytes, Iba1 for microglia and PECAM-1 for endothelial cells. We observed partial co-labelling between DDAH1 and PECAM-1 throughout the brain, however the majority of PECAM-1 did not overlap with DDAH1 (Figure 2E). Despite prevalent Iba1 expression, e.g., within the striatum, we did not detect any co-labelling with DDAH1 (Figure 2F). Furthermore, no co-labelling between DDAH1 and Olig2 was found (Figure 2G). In summary, we found that DDAH1 is widely distributed in the rodent brain and expressed in a region specific manner in both neuronal and astrocyte cells as well as within the endothelium of the vascular structures.
DDAH2 protein is expressed in a limited number of brain regions
To build a map of DDAH2 protein expression, we firstly investigated its distribution on consecutive coronal mouse brain sections. As shown in Figure 3A, DDAH2 immunoreactivity was detected in a limited number of brain structures. We found high expression levels in the cortex, HPF, striatum, and pallidum. Low expression was observed in the cortical subplate. In all structures, DDAH2 positive cells had a round shaped soma, an area of 15-19 µm2 and stained processes.
Within the striatum, intense DDAH2 staining was observed in the ventral region, the lateral septal complex and the striatum-like amygdalar nuclei but not in the dorsal part. DDAH2 positive cells were observed in the lateral septal nucleus (LSN) (Figure 3D) of the lateral septal complex. Here, DDAH2 signal was detected in the caudal part of LSN, while no staining was apparent in the rostral and ventral parts. In the ventral striatum, DDAH2 positive cells were found in the fundus of the striatum (FS) (Figure 3E) and the olfactory tubercle (OT) (Figure 3B). DDAH2 immunoreactivity was also observed within the nucleus accumbens. In the striatum-like amygdalar nucleus, intense DDAH2 expression was found in the central amygdalar nuclei (CEA) (Figure 3A box 2), without staining in other nuclei such as the anterior or medial amygdalar nucleus. In the pallidum, DDAH2 staining was observed in the caudal part in the bed nuclei of the stria terminalis (BST) (Figure 3C). There was no DDAH2 fluorescence signal in other regions of the pallidum. Within the cortical regions, DDAH2 immunoreactivity was restricted to layer II of the piriform cortex (Figure 3A box 3) and layer II of the entorhinal area. Low DDAH2 expression was detected in the layer II of the auditory area, somatosensory, orbital, prelimbic, infralimbic and posterior parietal association areas in the cortex. Further, we observed region specific DDAH2 expression in the HPF restricted to the pyramidal layer of CA1 while no signal was observed in CA2 and CA3 (Figure 3A box 1). Additionally, intense DDAH2 expression was observed in the subiculum and the presubiculum. Finally, within the cortical subplate, DDAH2 positive cells were restricted to the endopiriform nucleus.
Again, we compared our findings to the publicly available mRNA expression datasets [41, 42]. Ddah2 mRNA signals in the cortex, in olfactory areas and in the HPF, matched our protein expression data (refer to Table S4). However, there were differences in Ddah2 mRNA distribution within the sub-regions of the HPF documented by the Allen Institute for Brain Science [41] and our data. According to the published data, Ddah2 mRNA can be identified in CA1, CA2 and CA3, whereas Ddah2 protein signal in our study was restricted to CA1. Striatal protein expression was found but could not be compared in detail due to the limited sagittal data availability on the Allen Mouse Brain Atlas [41]. Finally, Ddah2 mRNA expression in the HPF, striatum, and hypothalamus was confirmed in our protein assessment in accordance with the known Ddah2 mRNA distribution available at the Mouse Brain Atlas [42].
DDAH2 is expressed exclusively within neurons
Based on the already obtained cell size and shape information we performed co-labelling analysis with the neuronal marker NeuN. We found a complete overlap of DDAH2 positive cells with NeuN in all structures (Figure 4). DDAH2/NeuN expression was observed in CA1, while NeuN immunoreactivity continued in CA2 and CA3 (Figure 4A). The neuronal origin of DDAH2 positive cells was also confirmed for layer II of the piriform cortex while NeuN positive cells in both layers I and III did not express DDAH2 (Figure 4B). The CEA showed specific DDAH2 and NeuN positive staining, which was not observed in other amygdalar nuclei (Figure 4C). Intense DDAH2 and NeuN positive expression was seen in the caudal part of the LSN (Figure 4D). For validation, we also performed analysis of DDAH2 and GFAP double staining, which as expected did not yield any co-labelling (Figure 4E). Similarly, PECAM-1 positive endothelial did not display DDAH2 signal (Figure 4F). Results of immunocytochemical analysis on early postnatal primary cell culture of cortical and hippocampal origin confirmed the presence of DDAH2 in neurons; however, a minor fraction of DDAH2 positive cells also stained for GFAP (Figure S3D-F). In summary, within the adult mouse brain DDAH2 protein was found to be expressed exclusively in neuronal cells within a limited number of brain structures.
DDAH1 and DDAH2 are expressed by different cell populations
Next, we performed co-labelling experiments of DDAH1 and DDAH2. As shown in Figure 5, both DDAH1 and DDAH2 signal were observed in the HPF, cortex and striatum. However, DDAH1 and DDAH2 were consistently expressed by different cells. In the HPF, DDAH2 was present only in neurons of sp of CA1 whereas DDAH1 positive cells were observed in astrocytes in other layers (so, sr, slm) (Figure 5C). In the striatum, DDAH2 was expressed by neurons in a few striatal structures such as LSN, whereas DDAH1 was broadly present in astrocytes (Figure 5B). Additionally, DDAH2 and DDAH1 expression was observed in the BST (Figure 5B) and entorhinal cortex but did not show cellular overlap (Figure 5D). In summary, our results suggest that DDAH1 and DDAH2 protein are detected in the same regions of the brain but always in a cell type restricted manner.
Identical cell types in mouse and human brain tissue express DDAH1 and DDAH2
Finally, we performed comparative analysis in human and murine tissue. We established staining of DDAH1 and DDAH2 on human post-mortem tissue to compare its distribution with our murine data (Figure 6). At first, we performed DAB staining in tissue samples from the medial frontal gyrus, where both DDAH1 (Figure6A,B) and DDAH2 (Figure 6E,F) signals were observed. DDAH1 positive cells appeared small (app. 15 µm2) and star-shaped with multiple processes reminiscent of astrocytes, whilst DDAH2 positive cells had neuronal features and an area of app. 19 µm2. Comparable findings are derived from DDAH1 expression in mouse (Figure 6D) and human cortex (Figure 6C) as well as DDAH2 staining on mouse (Figure 6H) and human (Figure 6G) amygdala. In summary, similar cell types express DDAH1 and DDAH2 in both human and mouse CNS tissue.