R-Ras1 and R-Ras2 Expression in Anatomical Regions and Cell Types of the Central Nervous System

Since the optic nerve is one of the most myelinated tracts in the central nervous system (CNS), many myelin diseases affect the visual system. In this sense, our laboratory has recently reported that the GTPases R-Ras1 and R-Ras2 are essential for oligodendrocyte survival and maturation. Hypomyelination produced by the absence of one or both proteins triggers axonal degeneration and loss of visual and motor function. However, little is known about R-Ras specificity and other possible roles that they could play in the CNS. In this work, we describe how a lack of R-Ras1 and/or R-Ras2 could not be compensated by increased expression of the closely related R-Ras3 or classical Ras. We further studied R-Ras1 and R-Ras2 expression within different CNS anatomical regions, finding that both were more abundant in less-myelinated regions, suggesting their expression in non-oligodendroglial cells. Finally, using confocal immunostaining colocalization, we report for the first time that R-Ras2 is specifically expressed in neurons. Neither microglia nor astrocytes expressed R-Ras1 or R-Ras2. These results open a new avenue for the study of neuronal R-Ras2’s contribution to the process of myelination.


Introduction
The neural structures and circuits that receive, transform, ship, and integrate the light signals that allow human beings to see are deeply complex; information must be received by the retinal ganglion cells (RGCs), which then project their axons through the optic nerve (ON), decide whether to cross the optic chiasm, and arrive at the lateral geniculate nucleus. From there, new fibers emerge to connect with the primary visual cortex, where information is correctly integrated to form images. In some diseases-such as Leber's hereditary optic neuropathy, or some subtypes of Charcot-Marie-Tooth disease-neurons are predominantly affected [1,2]. However, neurons are not the only cell type involved in the delivery of visual information.
Myelin, produced in the central nervous system (CNS) by oligodendrocytes (OLs), is critically required in the vertebrate nervous system in order to enable fast and efficient synaptic transmission [3]. In this sense, many demyelinating diseases of the CNS disturb the visual system. Neuromyelitis optica (Devic's disease) is an inflammatory demyelinating disease that preferentially affects the ON and spinal cord (SC) [4]. Initially described as a variation of multiple sclerosis (MS), it is now classified as a distinct disease due to the presence of autoantibodies against aquaporin-4 water channels or myelin oligodendrocyte
(1.02 ± 0.22). No significant differences were observed between the R-Ras1 −/− (1.07 ± 0.15) and control mice. Lastly, for R-Ras3 we observed no significant changes in mRNA expression levels in R-Ras1 −/− (1.00 ± 0.09), R-Ras2 −/− (0.94 ± 0.05), or R-Ras1 −/− ;R-Ras2 −/− mice (0.99 ± 0.10) relative to controls (1.01 ± 0.15) ( Figure 1A). These results show that, despite the high degree of homology shared among R-Ras family members, loss of R-Ras1 and/or R-Ras2 cannot be compensated by one another or by R-Ras3, suggesting that they have specific non-redundant functions within the CNS.  levels normalized to GAPDH demonstrated no significant differences between the mutant and control mice. Bar graph represents the mean ± SD of the change as a percentage, relative to the control measurements. A two-tailed Student's t-test was used for statistical analysis. GAPDH: glyceraldehyde 3-phosphate dehydrogenase; ON: optic nerve; SD: standard deviation. n = 3.

R-Ras1 and R-Ras2 Are Expressed at Different Levels throughout the CNS
Little is known about R-Ras1 and R-Ras2 expression patterns in the CNS. For this reason, we first performed relative RT-qPCR experiments on several CNS anatomical regions of control adult mice, including the corpus callosum (CC), cerebral cortex (CX), hippocampus (HP), thalamus-hypothalamus (THT), cerebellum (CB), and SC. We found that both R-Ras1 and R-Ras2 are expressed throughout several tissues of the CNS at similar levels. For R-Ras1, the fold change values relative to the CB (1.28 ± 0.22) were 1.28 ± 0.12 in the CC, 1.01 ± 0.79 in the CX, 1.09 ± 0.16 in the HP, 0.89 ± 0.26 in the THT, and 1.37 ± 0.15 in the SC (Figure 2A). On the other hand, the fold change values for R-Ras2 relative to the CB (1.18 ± 0.17) were 1.21 ± 0.0.81 in the CC, 0.88 ± 0.13 in the CX, 1.31 ± 0.32 in the HP, 1.18 ± 0.24 in the THT, and 1.59 ± 1.09 in the SC ( Figure 2B). No significant differences were found in the relative expression of R-Ras1 or R-Ras2 across these tissues.   Next, we wanted to investigate whether the expression of R-Ras1 and R-Ras2 was different within the same tissue. To this end, we performed absolute RT-qPCR experiments in the CC, CX, HP, THT, CB, and SC of control adult mice. The results revealed that, although the levels of expression of R-Ras1 and R-Ras2 were similar in most tissues, R-Ras1 was expressed more than R-Ras2 in the CX and CB. Specifically, R-Ras1 expression values (in copy numbers) were 499 ± 47.3 in the CC, 458 ± 65.3 in the CX, 636 ± 140 in the HP, 450 ± 118 in the THT, 587 ± 66 in the CB, and 639 ± 249 in the SC. In comparison with R-Ras1, R-Ras2 expression values (in copy numbers) were 343 ± 253 in the CC, 278 ± 9.89 in the CX (p < 0.01), 539 ± 75.2 in the HP, 421 ± 56.2 in the THT, 423 ± 73.1 in the CB (p < 0.05), and 431 ± 164 in the SC ( Figure 2C). Taken together, these data show that, in addition to the ON, CC, and SC [30], R-Ras1 and R-Ras2 are also expressed in the CX, HP, THT, and CB. Moreover, expression of R-Ras1 tends to be higher in tissues such as the CX and CB, possibly due to the described expression of R-Ras1 in neurons [32][33][34][35].
To corroborate the expression data obtained from the RT-qPCR assays, we performed WB experiments using the ON, CC, CX, HP, THT, CB, and SC of adult control mice. Given the high homology between R-Ras1 and R-Ras2, most antibodies recognize both proteins. Nevertheless, as previously described by Sanz-Rodriguez et al. [30], both proteins can be specifically identified by their molecular weight: R-Ras1 appears as a band at 25 kDa [35], while R-Ras2 shows a band at 21 kDa [36].  Figure 2C,D).
Given that R-Ras2 −/− mice have a reporter sequence for LacZ that encodes β-galactosidase, we performed X-Gal staining on brain coronal sections of R-Ras2 −/− adult mice ( Figure 2E). This assay led us to recognize regions where R-Ras2 was highly expressed. Specifically, we found high levels of R-Ras2 in the CX and HP. In the CX, greater staining was observed in the upper layers, while in the HP reactivity was higher in the CA1 and dentate gyrus. Surprisingly, R-Ras2 expression was higher in less myelinated tissues than in highly myelinated tissues such as the ON or SC. In addition, we studied the presence of R-Ras2 in other anatomical regions of the visual pathway by performing X-Gal staining on whole retinas and longitudinal sections of the ON (Figure 2G,I) of control and R-Ras2 −/− adult mice. The results showed expression of R-Ras2 in the ON and punctate staining on the retinas of R-Ras2 −/− mice, compatible with the RGCs. These results confirm the expression of R-Ras2 in cell types other than OLs.

R-Ras1 and R-Ras2 Are Expressed by Neurons in the CNS, but Not in Astrocytes or Microglia
To finally determine which CNS cell types express R-Ras1 and R-Ras2, we performed double immunostaining against R-Ras1 and R-Ras2 combined with antibodies against microglia, astrocytes, or neurons in coronal sections of areas stained by the X-Gal reaction from R-Ras1 −/− and R-Ras2 −/− adult mice.
First, we used an antibody against the specific microglial membrane protein ionized calcium-binding adapter molecule 1 (Iba1) to identify co-expression with R-Ras1 and/or R-Ras2 in the dentate gyrus of the HP (ML: 0.5 mm, AP: −2 mm from bregma, DV: 2.2 mm) and the upper layers of the somatosensory CX (ML: 3 mm, AP: −2 mm from bregma, DV: 1.5 mm). Results showed an absence of colocalization between both markers. Quantification revealed no colocalization between R-Ras1 and Iba1 (r = −0.042) or R-Ras2 and Iba1 (r = −0.013), confirming that neither R-Ras1 nor R-Ras2 are expressed in microglial cells ( Figure 3A). The same procedure was followed to identify astrocytes, using glial fibrillary acidic protein (GFAP) as a specific marker. Again, no colocalization was found between GFAP and R-Ras1 (r = −0.018) or R-Ras2 (r = −0.021) ( Figure 3B). Lastly, to detect expression of R-Ras1 and/or R-Ras2 in neurons, we used the marker NeuN. We observed that both markers were located within the same cell, consistent with higher Pearson's correlation values for R-Ras1 and NeuN (r = 0.375), and for R-Ras2 and NeuN (r = 0.43). The results showed that both R-Ras1 and R-Ras2 are expressed in neurons ( Figure 3C) of the CX and the dentate gyrus of the HP, which would explain the observed differences in R-Ras expression between the CNS anatomical regions, as well as their degree of myelination. Together, our results show for the first time that R-Ras2 is expressed in neurons.
sion of R-Ras1 and/or R-Ras2 in neurons, we used the marker NeuN. We observed that both markers were located within the same cell, consistent with higher Pearson's correlation values for R-Ras1 and NeuN (r = 0.375), and for R-Ras2 and NeuN (r = 0.43). The results showed that both R-Ras1 and R-Ras2 are expressed in neurons ( Figure 3C) of the CX and the dentate gyrus of the HP, which would explain the observed differences in R-Ras expression between the CNS anatomical regions, as well as their degree of myelination. Together, our results show for the first time that R-Ras2 is expressed in neurons.

Discussion
The Ras family was the first small GTPase group to be discovered, more than 30 years ago, and is particularly involved in development, proliferation, differentiation, and survival [21,24]. Within the R-Ras subfamily, R-Ras3 was long considered to be the most important R-Ras member in the CNS [27,29], while R-Ras1 and R-Ras2 were thought to play more secondary roles. However, recent studies from our laboratory demonstrated in vivo that R-Ras1 and R-Ras2 are expressed in OLs, where they play essential roles in cell survival and differentiation [30]. A loss of R-Ras1 or R-Ras2 resulted in reduced numbers of OLs, with a more dramatic decrease observed in the absence of both GTPases. Furthermore, the mutant mice had reduced numbers of mature OLs, with a concomitant increase in immature OL populations. Therefore, mice lacking R-Ras1 and/or R-Ras2 showed varying degrees of hypomyelination in CNS tracts related to the visual pathway, such as the ON and CC. Analysis of the ON by electron microscopy revealed thinner myelin sheaths in the absence of R-Ras1, and 30% fewer myelinated axons due to the lack of R-Ras2, while an absence of both GTPases resulted in a dramatic hypomyelination, where up to 80% of axons were unmyelinated. As myelin thickness and node and internode length condition the efficacy of neural transmission, we further studied the functional consequences of the absence of R-Ras1 and R-Ras2 on the visual pathway. Consistent with studies where a lack of myelin slowed down impulse transmission [37][38][39], we found reduced conduction velocity in the optic tracts of the single-and double-mutant mice. The findings presented here show the expression of these GTPases in the retina, although their low expression would not explain the dramatic alterations observed in conduction velocity. For this reason, we believe that this effect is mediated by the functional role of R-Ras1 and R-Ras2 in OLs, rather than that in neurons. In future studies, we will analyze the roles of these GTPases in neurons in greater depth.
Our current data demonstrate that R-Ras1 and R-Ras2 have functions that are distinct from those of other R-Ras members and the classical Ras. RT-qPCR and WB studies showed that the absence of R-Ras1 and/or R-Ras2 does not change expression levels of R-Ras3 or the most closely related Ras members. Moreover, we analyzed the expression of R-Ras1 and R-Ras2 in order to explore their potential importance in other CNS anatomical regions, showing them both to be widely expressed throughout all of the brain structures tested. In addition to the previously described expression in the ON, CC, and SC, we found that R-Ras1 and R-Ras2 are also expressed at similar levels in the CX, HP, THT, and CB. We further performed absolute RT-qPCR experiments to see whether there were significant differences in the expression of R-Ras1 and R-Ras2 within the same tissues. We found that the number of copies of R-Ras1 was significantly higher than that of R-Ras2 in the CX and CB, probably due to the fact that R-Ras1 is expressed in neurons other than OLs [32][33][34][35]40]. Although we did not find differences in the relative RT-qPCR assays, WB experiments showed that R-Ras1 and R-Ras2 protein levels vary between tissues. Even though we had expected higher levels of R-Ras1 and R-Ras2 in highly myelinated tissues such as the ON and SC, the results revealed that the ON and SC were actually the tissues with the least R-Ras1 and R-Ras2 expression. It has been described that R-Ras1 is expressed in cortical and hippocampal neurons other than OLs [32][33][34][35]40], which may justify the higher levels of R-Ras1 seen in less myelinated tissues such as the CX or HP. However, higher levels of R-Ras2 were also found in the CX, HP, and retina, suggesting that they may be widely expressed in other cell types that have never been described before. Interestingly, the pattern of expression was higher in the CX and dentate gyrus of the HP, again suggesting that R-Ras2 may be expressed in other cellular types. Finally, confocal microscopic imaging allowed us to disregard expression of R-Ras1 and R-Ras2 in astrocytes and microglia, though we found for the first time that R-Ras2 is expressed in neurons.
Given the close relationship between neurons and OLs in the myelination process, and knowing that R-Ras1 and R-Ras2 are expressed in both cell types, we cannot disregard that an absence of neuronal R-Ras1 or R-Ras2 contributes to the visual deficits observed in mutant mice. Future work is essential in order to segregate the functions that these GTPases may play in each cell type during myelination, pointing us toward a new field studying the role of R-Ras2 in neurons.

Animals
C57BL6 mice were housed in specific pathogen-free conditions in a humidity-and temperature-controlled room with a 12 h light/dark cycle, receiving water and food ad libitum. All experiments were performed in male and female mice, and all animal procedures were approved by the corresponding institutional ethical committee (Centro de Biología Molecular Severo Ochoa, CBMSO, Madrid, Spain) and were performed in accordance with Spanish and European guidelines. All efforts were made to minimize animal suffering.
R-Ras1 −/− mice were generated at genOway (Lyon, France) using the targeting construction BAL1-HR with a neomycin resistance cassette flanked by FRT sequences, inserted in the intron 1 and LoxP sites flanking exons 2 and 6. The construction was electroporated into embryonic stem cells derived from murine 129Sv/Pas and selected by G418 antibiotic. Southern blotting was used to verify correct homologous recombination. Heterozygous mice were crossed, and offspring littermates were genotyped by PCR to ensure that all animals used were homozygous for the knockout (Table 1). PCR was performed on an Applied Biosystems™ 2700 Thermal Cycler (Waltham, MA, USA) as follows: 94 • C for 5 min; 30 cycles of 94 • C for 30 s, 60 • C for 30 s, and 72 • C for 30 s; and a final step of 72 • C for 7 min. Samples were then run on a 2.5% agarose gel in Tris/Borate/EDTA (TBE) buffer at 100 V for 30 min, and visualized in a transilluminator. R-Ras2 −/− mice were generated at Lexicon Pharmaceuticals (The Woodlands, TX, USA), and were derived from embryonic stem cell clone OST361011 with insertion of retroviral VICTR37 in the middle of intron 1 of R-Ras2. Heterozygous mice were crossed and offspring littermates were genotyped by PCR to ensure that all animals used were homozygous for the knockout (Table 1) [36]. The same PCR conditions were used as for the R-Ras1 −/− mice. R-Ras1 −/− and R-Ras2 −/− mice were kindly provided by Professor B. Alarcón (CBMSO). R-Ras1 −/− ;R-Ras2 −/− mice were generated by backcrossing individual lines of R-Ras1 −/− and R-Ras2 −/− mice. Offspring littermates were genotyped by PCR to ensure that all animals used were homozygous for both knockout genes. All experiments were performed using wild-type mice (R-Ras1 +/+ ;R-Ras2 +/+ ) as controls. Animals were maintained in a C57BL6J background.

Reverse Transcription Quantitative PCR (RT-qPCR)
The different CNS anatomical regions of interest were harvested from adult (P90) control, R-Ras1 −/− , R-Ras2 −/− , and R-Ras1 −/− ;R-Ras2 −/− mice after cervical dislocation and decapitation (n = 3 per genotype). To preserve RNA, freshly extracted tissue was immediately frozen in TRIzol™ (Thermo Fisher Scientific, Waltham, MA, USA, catalog #15596026) at −80 • C. RNA extraction and purification were performed using an RNeasy Mini Kit (Qiagen, Germantown, MD, USA, catalog #74106), following the manufacturer's protocol. RNA concentration and integrity were measured using a NanoDrop (Thermo Scientific NanoDrop One) and a Bioanalyzer (Bioanalyzer Agilent 2100, Santa Clara, CA, USA), respectively. Most of the samples showed 260/280 and 260/230 ratios around 2, and RNA integrity values (RIN) of the samples varied between 7.4 and 9.5. Complementary DNA (cDNA) was generated from 500 ng of total ON RNA using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA, catalog #1708891) in a 20 µL final reaction volume. SsoFast EvaGreen Supermix (Bio-Rad, Hercules, CA, USA, catalog #1725204) reagent was used for qPCR expression profiling, carried out on a CFX384 Real-Time System C1000 Thermal Cycler (Bio-Rad, Hercules, CA, USA) in hard-shell 384-well PCR plates (White Well Clear Shell; Bio-Rad, Hercules, CA, USA, catalog #HSP3805). A total of 5 ng of cDNA per sample was used in a 10 µL final volume reaction. Technical triplicates were performed in order to correct for pipetting errors in plate loading. No-template control (NTC) reactions were carried out using all reagents except the sample to disregard potential contamination. Valid-Prime assay was performed to control the presence of genomic DNA background signals during qPCR expression profiling [41]. Relative quantification was performed using the 2 ∆∆Cq method [42], taking GAPDH, ActB, Hprt1, Tbp, Arbp, and GusB as reference genes for sample normalization. The gene expression levels in control groups were normalized to 1. For absolute quantification experiments, standard curves were performed with a qPCR over an eight-point 1 4 dilution curve made from 1.56 × 10 5 copies/µL of R-Ras1 or R-Ras2. Samples were normalized to the expression levels of GusB and Arbp. The RT-qPCR primers used in this study can be found in Table 2.

X-Gal Reaction
X-Gal staining was performed on fresh, non-perfused brains from R-Ras2 −/− mice following widely established protocols [43]. After cervical dislocation, brains were extracted, washed in 1× PBS, and mounted in a solution of 10% sucrose and −4% agarose to increase tissue robustness. Then, they were sectioned on a Leica VT-1200S sliding-blade vibratome to produce 300-400 µm thick sections that were stained with X-Gal reactive at 37 • C for 4 h or overnight. During this reaction, X-Gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) was hydrolyzed by the β-galactosidase present in the cassette of R-Ras2 −/− mice to form galactose and 5-bromo-4-chloro-3-hydroxyindole-an insoluble blue compound. Tissues were later fixed in 4% paraformaldehyde (PFA) and photographed on a Leica MZ6 magnifying glass.

Immunohistochemistry
Animals were anesthetized (ketamine 100 mg/kg and xylazine 10 mg/kg intraperitoneally) and perfused transcardially with 0.1 M PBS (pH 7.4) followed by 4% PFA in PBS. Perfused tissues were removed and postfixed in 4% PFA at 4 • C overnight. Then, they were cryoprotected in 30% sucrose in PBS and embedded and frozen in a 7.5% gelatin in 15% sucrose solution. Then, they were sectioned on a cryostat to produce 20 µm cryosections on SuperFrost Plus microscope slides (Thermo Fisher Scientific, Waltham, CA, USA, catalog # 22-037-246). Sections were blocked for 1 h at room temperature with 10% fetal bovine serum in PBS containing 0.5% Triton-X 100 (blocking solution), and then incubated overnight at 4 • C with the pertinent primary antibodies (Table 4). After 3 washes, fluorescent-tagged secondary antibodies (Table 4) were applied for 1 h at room temperature, and sections were counterstained with DAPI (Sigma-Aldrich, St. Louis, MO, USA, catalog# 32670) and mounted in Aqua-Poly/Mount mounting medium (PolySciences, Warrington, PA, USA, catalog# 18606).

Confocal Microscopy
Fluorescence images were obtained using a confocal multispectral Leica TCS SP5 system (Leica Microsystems, Wetzlar, Germany) controlled by Las AF v 2.7 software (Leica). Image acquisition was performed sequentially using a 40×/1.4 NA oil immersion objective and appropriate fluorochrome excitation lines (405 nm, 488 nm, and 560 nm for DAPI, Alexa-488, and Alexa-555, respectively).

Statistical Analysis
Quantitative data are shown as the mean ± SD. The experimental groups were compared using a two-tailed Student's t-test. Statistical numeric data are provided in the figure legends. (*): p < 0.05; (**): p < 0.01; (***): p < 0.001. An alpha level of 0.05 was considered significant. Statistical analysis was performed using GraphPad Prism 8 statistical software.