Organization of Neuropeptide Y-Immunoreactive Cells in the Mongolian gerbil (Meriones unguiculatus) Visual Cortex

Neuropeptide Y (NPY) is found throughout the central nervous system where it appears to be involved in the regulation of a wide range of physiological effects. The Mongolian gerbil, a member of the rodent family Muridae, is a diurnal animal and has been widely used in various aspects of biomedical research. This study was conducted to investigate the organization of NPY-immunoreactive (IR) neurons in the gerbil visual cortex using NPY immunocytochemistry. The highest density of NPY-IR neurons was located in layer V (50.58%). The major type of NPY-IR neuron was a multipolar round/oval cell type (44.57%). Double-color immunofluorescence revealed that 89.55% and 89.95% of NPY-IR neurons contained gamma-aminobutyric acid (GABA) or somatostatin, respectively. Several processes of the NPY-IR neurons surrounded GABAergic interneurons. Although 30.81% of the NPY-IR neurons contained calretinin, NPY and calbindin-D28K-IR neurons were co-expressed rarely (3.75%) and NPY did not co-express parvalbumin. Triple-color immunofluorescence with anti-GluR2 or CaMKII antibodies suggested that some non-GABAergic NPY-IR neurons may make excitatory synaptic contacts. This study indicates that NPY-IR neurons have a notable architecture and are unique subpopulations of the interneurons of the gerbil visual cortex, which could provide additional valuable data for elucidating the role of NPY in the visual process in diurnal animals.


Quantitative Analysis
For the quantitative analysis of laminar distribution, the NPY-IR neurons were photographed using a Zeiss Axioplan microscope with a 20× objective. A total of 30 sections were sampled, each with a width of 2000 µm, from each of three animals (10 tissue sections from each animal). The number of labeled neurons was expressed as an average of cell numbers and as a percentage of the total population of labeled neurons. The percentage frequency was calculated as follows: NPY-IR neurons at each layer/NPY-IR neurons in total layers. In three animals, the morphological types and the average diameter and area of NPY-IR neurons were analyzed. The morphological types were determined for 175 neurons analyzed from 59 sections in three gerbils, and the average diameter and area of NPY-IR neurons were determined for 111 neurons analyzed from 22 sections in three gerbils. All analyses were conducted using a 40× Zeiss Plan-Apochromat objective. To obtain the best images, we analyzed the cells under DIC optics. Only the cell profiles containing a nucleus and at least one faintly visible nucleolus were included in the analysis. The average diameter and area of labeled neurons were computed using a digital camera (Carl Zeiss Meditec Inc.). A cursor was moved manually around the outer contour of each cell using the Zeiss AxioVison system. Images were adjusted according to brightness and contrast using the Adobe Photoshop CS software (Adobe Systems Inc., San Jose, CA, USA). Double-labeled neurons were counted from a total of 12 (except calbindin-D28K: total 18 sections) different tissue sections from each of the three animals, each 2000 µm in width, across all layers. Double-and triple-labeled images were obtained and viewed under a Zeiss LSM800 laser scanning confocal microscope (Carl Zeiss Meditec Inc.) with 40× and 100× objectives.
Drawings of the NPY-IR neurons were produced using a Zeiss Axioplan microscope (Carl Zeiss Meditec Inc.) with a 40× and 63× Zeiss Plan-Apochromat objective (Carl Zeiss Meditec Inc.). NPY-IR neurons were imaged on a computer monitor, and cells were drawn on acetate sheets. The final images were drawn using Adobe Photoshop CS (Adobe Systems Inc.).

Laminar Distribution of NPY-IR Neurons
The size of the Mongolian gerbil brain was approximately 2 cm in length (anterior to posterior) ( Figure 1A). Figure 1B shows a low magnification image of the gerbil visual cortex in the coronal plane with thionin staining. NPY-IR neurons were sparsely and selectively distributed in the visual cortex of the gerbil. Figure 1C shows a thionin-stained section which reveals the cortical layers and laminar distribution of NPY-IR neurons ( Figure  1D). NPY-IR neurons were located throughout all layers of the gerbil visual cortex except layer I. NPY-IR fibers were distributed throughout all layers of the visual cortex forming a plexus of labeled fibers with differential densities in different layers. Figure 1E, a dark field figure, is a representation of the distribution of the NPY-IR fibers in the gerbil visual cortex. The highest density of NPY-IR neurons was distributed in layer V. Layers II-IV showed considerably fewer numbers of NPY-IR neurons. Quantitative maps of the cells revealed the density of NPY-IR neurons in each layer ( Figure 1F). Regarding the proportion of the total population of labeled neurons, 0% of NPY-IR neurons were found in layer I, 4.28% were found in layer II, 12.06% were found in layer III, 7.00% were found in layer IV, 50.58% were found in layer V, and 26.07% of NPY-IR neurons were found in layer VI.

Morphology of NPY-IR Neurons
At least six types of NPY-IR neurons were found in the gerbil visual cortex as follows: multipolar round/oval, multipolar stellate, vertical fusiform, horizontal, pyriform, and Martinotti cells. Figure Figure 2A,G. The cells had a round/oval-shaped cell body and multiple dendrites coursing in all directions. The next most common NPY-IR neurons were stellate cells, as shown in Figure 2B,H. Stellate cells had a polygonal-shaped cell body and multiple dendrites coursing in all directions. The multipolar round/oval and stellate cells had medium dendritic fields (200-300 μm in diameter). These cell types typically had 3-6 primary processes with sparsely branched processes. In general, the cell bodies of the round/oval cells were smaller than those of the stellate cells.

Morphology of NPY-IR Neurons
At least six types of NPY-IR neurons were found in the gerbil visual cortex as follows: multipolar round/oval, multipolar stellate, vertical fusiform, horizontal, pyriform, and Martinotti cells. Figure  The cells had a round/oval-shaped cell body and multiple dendrites coursing in all directions. The next most common NPY-IR neurons were stellate cells, as shown in Figure  2B,H. Stellate cells had a polygonal-shaped cell body and multiple dendrites coursing in all directions. The multipolar round/oval and stellate cells had medium dendritic fields (200-300 µm in diameter). These cell types typically had 3-6 primary processes with sparsely branched processes. In general, the cell bodies of the round/oval cells were smaller than those of the stellate cells. Figure 2C,K show a vertical fusiform cell with a vertical fusiform cell body, a primary long process ascending toward the pial surface, and a descending process. Figure 2D,I show a horizontal cell with a horizontal fusiform cell body and horizontally oriented processes. The fusiform cells had medium-to-large dendritic fields (300-400 µm in diameter) with two processes. In general, cell bodies of fusiform cells were also relatively smaller compared with those of the stellate cells. In the present study, the processes of the round/oval, stellate, and fusiform cells were aspinous.   These cells showed very few short descending processes, and one apical process which extended toward the pial surface in layer I. Neuroglialform interneurons which have distinctive short multiple dendrites that spread in all directions, from small, round cell bodies, were not identified as NPY-IR neurons in the present study. Figure 2M shows a histogram of the percentage of each cell type. Quantitatively, 44.57% ± 4.95% (mean ± S.D.) (78 of 175 cells) of NPY-IR neurons were round/oval, 21.14% ± 3.05% (37 of 175 cells) were stellate, 14.29% ± 1.29% (25 of 175 cells) were horizontal, 13.14% ± 3.04% (23 of 175 cells) were vertical fusiform, 4.57% ± 1.78% (8 of 175 cells) were pyriform, and 2.29% ± 0.89% (4 of 175 cells) were Martinotti cells. Figure 2N shows a histogram of the frequency distributions of each cell type in each layer. The highest number of round/oval (56.16%) and stellate (34.15%) neurons were located in layer V. Vertical fusiform neurons were mostly distributed in layer III (65.38%), whereas horizontal neurons were mostly distributed in layer VI (60.61%). Pyriform neurons were distributed in layers II-VI and Martinotti cells were distributed in layers III-V in the present study. Pyriform and Martinotti cells were not present in sufficient numbers to draw a meaningful conclusion.
neurons were found in layer VI. There was no layer-specific distribution of NPY-IR neurons that were double-labeled with calbindin-D28K. Most of NPY-IR neurons were labeled with somatostatin (arrowheads in Figure 3I3,J3), but other cells were labeled with only one of the antibodies ( Figure 3J3). No obvious relationship was detected between cell morphology and whether the cell was single-, double-, or triple-labeled. To estimate the percentage of double-labeled cells, we assessed each of the four sections from each of the three animals and counted the number of NPY-IR neurons and double-labeled cells across the layers of the visual cortex of the gerbil. As the number of double-labeled cells of NPY with calbindin-D28K was low, we assessed six sections from each of the three animals to increase the accuracy of statistical analyses. Quantitatively, 89.55% ± 8.58% (180 of 201 cells) of NPY-IR neurons were double-labeled with GABA, 3.75% ± 1.85% (10 of 267 cells) with calbindin-D28K, 30.81% ± 4.44% (61 of 198 cells) with calretinin, 0% (0 of 176 cells) with parvalbumin, and 89.95 ± 5.09% (224 of 249 cells) with somatostatin (Table 1).    Figure 4A-H shows the GABAergic interneurons ( Figure 4B,F) that were surrounded by NPY-IR fibers ( Figure 4A,C,E,G). Some NPY-IR fibers densely surrounded the GABAergic interneurons ( Figure 4D,H). Figure 4I-T show triple-labeling with antibodies against NPY ( Figure 4I,O), GABA ( Figure 4J,P), and GluR2 ( Figure 4L) or CaMKII ( Figure 4R). Figure 4M,N,S,T show the processes of NPY-IR neurons that were not labeled with GABA ( Figure 4K,Q) making contact with immunopuncta of GluR2 or CaMKII. When NPY and GluR2 or CaMKII were colocalized in the same focal plane, they were regarded as contacts. (arrowheads in Figure 4M,N,S,T).

Discussion
The present study showed that NPY-IR neurons in the gerbil visual cortex are distributed throughout layers II-VI, with the highest density detected in layer V. The

Discussion
The present study showed that NPY-IR neurons in the gerbil visual cortex are distributed throughout layers II-VI, with the highest density detected in layer V. The labeled cells had various morphologies. NPY-IR neurons contained GABA, calbindin-D28K or calretinin at distinctively different ratios, whereas none of the NPY-IR neurons contained parvalbumin.
The distribution of NPY-IR neurons was primarily concentrated in the deep cortical layers of the gerbil visual cortex. The infragranular layers, layers V and VI, were the prominent location of NPY-IR neurons with the highest density in layer V. On the other hand, a relatively small number of NPY-IR neurons was distributed in other layers. Notably, there were no NPY-IR neurons in layer I of the gerbil visual cortex. Similar to the results of the present study, in humans [30] and monkeys [76], NPY-IR neurons have been found to be most frequently distributed in layer V or VI and more sparsely in supragranular layers. In cats, NPY-IR neurons have also been shown to be primarily distributed in infragranular layers and few in layers II/III [36]. In rats, the majority of NPY-IR neurons have been reported to be in deep cortical layers such as layer VI [39,77,78], similar to the results of the present study. Moreover, in cats [35,36] and rats [39,77,78], NPY-IR neurons were absent in layer I in adults, except some transient expression of NPY in layer I during development. However, there are subtle species differences in NPY expression. Berman and Fredrickson [29] demonstrated the highest number of NPY-IR neurons in layers II-III in the human visual cortex among the six cortical layers. Moreover, there were some NPY-IR neurons that were distributed in layer I in the visual cortices of humans [29,30,79] and macaques [33,34]. It is also notable that some differences in NPY distribution have been reported even within the same members of Muridae family. In contrast to the gerbil visual cortex in the present study, the mouse visual cortex was found to have a high distribution of NPY-IR neurons in layers II-III in addition to layer VI [40]. Some NPY-IR neurons were also found in layer I of the mouse visual cortex [40]. Thus, with respect to the distribution within the same members of the Muridae family, at least one solid conclusion can be drawn by comparing nocturnal mouse and diurnal gerbil data. Differences in the pattern of distribution of NPY-IR neurons in the two species are clearly seen. The functional significance of such species differences in NPY expression patterns remains to be elucidated. However, these differences in expression in layers may contribute to slight functional variations in different species.
The NPY-IR neurons in the gerbil visual cortex were morphologically diverse subpopulations of non-pyramidal neurons. Most were multipolar round/oval cells. Some vertical fusiform, multipolar stellate, pyriform, and horizontal neurons also expressed NPY. Consistent with the present result, NPY-IR neurons that were almost exclusively non-pyramidal cells have been found in humans [29,30,79], monkeys [31,33,34,76,80], cats [36,49], and mice [40], and a large proportion of the labeled cells were round/oval in shape. Oval cells were also the predominant type of NPY-IR neurons even in the invertebrate octopus optic lobe [46,81]. The combined results of the present and previous studies indicate that NPY-IR neurons are interneurons in the visual cortex. However, a few studies have demonstrated the existence of a few NPY-IR pyramidal cells in the human visual cortex [30,79]. These results suggest that some NPY-IR neurons may be projection neurons. A recent study demonstrated that NPY-IR labeled projection neurons in the inferior colliculus [82].
In the present study, we estimated the cell size of the NPY-IR neurons and found that the majority of NPY-IR neurons in the gerbil visual cortex were small to medium in size. However, in studies of other species, the size of NPY-IR neurons in the visual cortex has been shown to vary. Most of the NPY-IR neurons in the adult cat were not small in size [35][36][37], and large NPY-IR neurons measuring >20 µm in size were also found in cats [49]. In the human visual cortex, the NPY-IR neurons were highly varied in size (7-23 µm), with larger cells reported as well [29]. The invertebrate octopus also had a highly variable size (9-27 µm) of NPY-IR neurons in the optic lobe [46]. Although Gonchar et al. [40] did not measure NPY-IR neuron size, all the NPY-IR neurons in their study (Figures 1, 4 and 5) were small in size.
There has been no report on the quantitative analysis of the cell size of NPY-IR neurons in the visual cortex of rodents to compare with the gerbils in the present study. Considering the overall morphologically distinct subpopulations of NPY-IR neurons with variable sizes among different species, NPY may serve varied biological functions in different species.
With respect to the double labeling data of NPY-IR neurons with GABA-IR neurons, we can also draw at least one solid conclusion within the same members of Muridae family. Differences in double labeling between mice and gerbils can be seen clearly. In the present study, we found that approximately 10% of NPY-IR neurons did not express GABA or somatostatin. In contrast to the gerbil visual cortex, almost all the NPY-IR neurons expressed GABA in the visual cortices of mice [40] and rats [39]. The NPY-IR neurons are supposed to be mostly GABAergic in the cerebral cortex [48,80,83,84], hippocampus [85], and hypothalamus [86] but not in the caudate-putamen nuclei [47]. A previous study also revealed that 97.7% of NPY-IR neurons expressed somatostatin in the human cortex [87]. These results suggest that a discrepancy exists in the GABA-and somatostatin-containing NPY-IR neurons among different areas and species. In our previous study of melanopsinexpressing intrinsically photosensitive retinal ganglion cells (ipRGCs), the M2 cells that have been frequently found in other rodents, we surprisingly did not find them in the Mongolian gerbil [69]. The present study on NPY and GABA and previous studies on ipRGCs indicate species-specific variation of neuronal subpopulations in mice and gerbils, both of which belong to the Muridae family.
The present study showed that NPY-IR neurons in the gerbil visual cortex were labeled by calretinin (30.81%) and calbindin-D28K (3.75%). However, none of the NPY-IR neurons in the gerbil visual cortex were labeled by parvalbumin. Similarly, NPY-IR neurons lacked parvalbumin in the cat [49,88] and mouse visual cortex [40]. In contrast to the findings of the present study, NPY-IR neurons in the cat visual cortex are not co-labeled with calbindin-D28K [49]. Some NPY-IR neurons in the mouse visual cortex have been reported to contain calretinin [40]. The present results show that all parvalbumin (100%), a majority of calbindin-D28K (96.25%), and over two-thirds of calretinin (69.19%) neurons in the gerbil visual cortex are distinct subpopulations that differ from those that co-express NPY. These results and the finding that approximately 10% of NPY-IR neurons did not co-localize GABA and somatostatin suggest diversification of neuronal cell types in the gerbil visual cortex. Thirteen distinct GABAergic neurons have been found in the mouse visual cortex based on the expression of CBPs and neuropeptides [40]. Moreover, Masland [89] suggested that there could be 1000 different cell types in the cortex. Markram et al. [90] suggested that "the diversity of interneurons might be required to achieve a balance between inhibition and excitation in the neocortex". Genome-wide analysis of gene expression studies will help to clarify the diversity and reveal subtype-specific roles of these neurons [91].
The precise effects of NPY on the excitation and inhibition of cortical neurons are not fully understood. NPY mediates inhibitory synaptic transmission by releasing GABA onto cortical pyramidal neurons [92,93]. Our observation of NPY-IR fibers surrounding GABAergic cells suggests that some NPY-IR neurons may form synaptic contacts with GABAergic inhibitory neurons. In the present study, we found that at least 10% of NPY-IR neurons were not GABAergic cells. The close apposition of the processes of non-GABAergic NPY-IR neurons and GluR2 or CaMKII may suggest that these NPY-IR neurons can affect excitatory synaptic activity. Previous studies have shown that NPY can affect inhibitory as well as excitatory synaptic activity [92]. Although previous studies have shown that the immunopuncta observed through the confocal microscope represent synaptic sites [94,95], true synapses can only be identified with certainty using electron microscopy; this raises the question of whether these are true synapses. Therefore, electron microscopic studies will be necessary to confirm the synaptic features observed in the present study.
NPY performs a wide range of significant modulatory functions in the brain [7,8,11,12]. NPY also appears to play important roles in the visual cortex. For instance, NPY-IR fibers and varicosities near blood vessels in the visual cortices of humans and cats play a role in blood flow regulation [29,30,35,96]. In the macaque monkey, NPY-IR neurons are primarily located outside of cytochrome oxidase patches, indicating that NPY-IR neurons may play a role in pattern perception and binocular processing [32,33]. However, in humans, no relationship with cytochrome oxidase patches has been found [29]. NPY is related to neurological diseases and exerts neuroprotective functions [97,98]. Therefore, the decreased number of NPY-positive cells in the adult mouse visual cortex was found to be related to animal model of autism [99]. However, the particular role of NPY in the visual cortex is still poorly understood and remains to be elucidated.

Conclusions
Our study demonstrated that NPY-IR neurons were mostly present in layers IV and V, and the highest density was found in layer V in the gerbil visual cortex, with no labeled cells in layer I. A large number of NPY-IR neurons were round/oval cells, although other types of cells were also present. Unlike many other mammalian visual cortices where NPY neurons are almost exclusively GABAergic, at least 10% of NPY-IR neurons were not GABAergic or somatostatinergic interneurons in the gerbil visual cortex. Many of the NPY-IR neurons were distinct subpopulations of interneurons, which are independent from calbindin-D28K, calretinin, or parvalbumin-containing interneurons. Some non-GABAergic NPY-IR neurons may form excitatory synaptic contacts. Our findings should contribute to a better understanding of the rodent visual system and may provide fundamental insights for further physiological studies.