The Selective Loss of Purkinje Cells Induces Specific Peripheral Immune Alterations.

fncel-15-773696 November 27, 2021 Time: 12:41 # 1 ORIGINAL RESEARCH published: 30 November 2021 doi: 10.3389/fncel.2021.773696 Edited by: Ulises Gomez-Pinedo, Instituto de Investigación Sanitaria del Hospital Clínico San Carlos, Spain Reviewed by: Teresa Guillamon-Vivancos, Instituto de Neurociencias de Alicante, Spain Ana Laura Márquez-Aguirre, CONACYT Centro de Investigación y Asistencia en Tecnología y Diseño del Estado de Jalisco (CIATEJ), Mexico Regina Rodrigo, Principe Felipe Research Center (CIPF), Spain *Correspondence: Eduardo Weruaga [email protected] David Díaz [email protected] †These authors have contributed equally to this work Specialty section: This article was submitted to Cellular Neuropathology, a section of the journal Frontiers in Cellular Neuroscience Received: 10 September 2021 Accepted: 27 October 2021 Published: 30 November 2021 Citation: del Pilar C, Lebrón-Galán R, Pérez-Martín E, Pérez-Revuelta L, Ávila-Zarza CA, Alonso JR, Clemente D, Weruaga E and Díaz D (2021) The Selective Loss of Purkinje Cells Induces Specific Peripheral Immune Alterations. Front. Cell. Neurosci. 15:773696. doi: 10.3389/fncel.2021.773696 The Selective Loss of Purkinje Cells Induces Specific Peripheral Immune Alterations Carlos del Pilar1,2, Rafael Lebrón-Galán3,4, Ester Pérez-Martín1,2, Laura Pérez-Revuelta1,2, Carmelo Antonio Ávila-Zarza2,5, José Ramón Alonso1,2,6, Diego Clemente3,4, Eduardo Weruaga1,2*†and David Díaz1,2*† 1INCyL, Institute for Neuroscience of Castile and Leon, Universidad de Salamanca, Salamanca, Spain, 2IBSAL, Institute of Biomedical Research of Salamanca, Salamanca, Spain, 3Grupo de Neuroinmuno-Reparación, Hospital Nacional de Parapléjicos, Toledo, Spain, 4SESCAM (Servicio de Salud de Castile-La-Mancha), Castilla–La Mancha, Spain, 5Applied Statistics Group, Department of Statistics, Universidad de Salamanca, Salamanca, Spain, 6Instituto de Alta Investigación, Universidad de Tarapacá, Arica, Chile The progression of neurodegenerative diseases is reciprocally associated with impairments in peripheral immune responses. We investigated different contexts of selective neurodegeneration to identify specific alterations of peripheral immune cells and, at the same time, discover potential biomarkers associated to this pathological condition. Consequently, a model of human cerebellar degeneration and ataxia -the Purkinje Cell Degeneration (PCD) mousehas been employed, as it allows the study of different processes of selective neuronal death in the same animal, i.e., Purkinje cells in the cerebellum and mitral cells in the olfactory bulb. Infiltrated leukocytes were studied in both brain areas and compared with those from other standardized neuroinflammatory models obtained by administering either gamma radiation or lipopolysaccharide. Moreover, both myeloid and lymphoid splenic populations were analyzed by flow cytometry, focusing on markers of functional maturity and antigen presentation. The severity and type of neural damage and inflammation affected immune cell infiltration. Leukocytes were more numerous in the cerebellum of PCD mice, being located predominantly within those cerebellar layers mostly affected by neurodegeneration, in a completely different manner than the typical models of induced neuroinflammation. Furthermore, the milder degeneration of the olfactory bulb did not foster leukocyte attraction. Concerning the splenic analysis, in PCD mice we found: (1) a decreased percentage of several myeloid cell subsets, and (2) a reduced mean fluorescence intensity in those myeloid markers related to both antigen presentation and functional maturity. In conclusion, the selective degeneration of Purkinje cells triggers a specific effect on peripheral immune cells, fostering both attraction and functional changes. This fact endorses the employment of peripheral immune cell populations as concrete biomarkers for monitoring different neuronal death processes. Keywords: selective neurodegeneration, neuroinflammation, brain infiltration, biomarkers, peripheral immune alterations Abbreviations: AD, Alzheimer’s disease; ALS, amyotrophic lateral sclerosis; CCP1, cytosolic carboxypeptidase 1; DAPI, 4’,6diamidino-2-phenylindole; GL, granular layer; LPS, lipopolysaccharide; MFI, mean fluorescence intensity; ML, molecular layer; OB, olfactory bulb; PB, phosphate buffer; PBS, phosphate buffered saline; PCD, Purkinje Cell Degeneration; PCL, Purkinje cell layer; PD, Parkinson’s disease; WM, white matter; WT, wild-type. Frontiers in Cellular Neuroscience | www.frontiersin.org 1November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 2 del Pilar et al. Immune Alterations by Selective Neurodegeneration INTRODUCTION Neurodegenerative diseases are generally accompanied by local inflammatory reactions that have been associated with an altered immune cell infiltration into the damaged nervous system (Zlokovic, 2011;Ransohoff and Brown, 2012). This phenomenon is critical for microglial regulation, a pivotal player in neuroinflammation known to be involved in the progression of the neurodegenerative process (Boyko et al., 2017;McManus and Heneka, 2017). Interestingly, alterations have not only been observed in the central nervous system, but also at the peripheral level, multiple sclerosis being a good example in which a clear participation of immune cells has been reported (Fakhoury, 2016). Moreover, changes in the peripheral immune system have also been described in Alzheimer’s disease (AD; Ciaramella et al., 2016;Gupta et al., 2018;Le Page et al., 2018), Huntington’s disease (Björkqvist et al., 2008), Parkinson’s disease (PD; Scherzer et al., 2007;Fuzzati-Armentero et al., 2019) and amyotrophic lateral sclerosis (ALS; Liu and Wang, 2017;Zhao et al., 2017). These alterations include changes in the distribution and activation of lymphocytes and macrophages (Lucin and Wyss-Coray, 2009), the latter usually presenting a pro-inflammatory phenotype in humans suffering from neurodegenerative disorders (González and Pacheco, 2014). Furthermore, increased levels of proinflammatory cytokines have been detected in both blood and cerebrospinal fluid of patients with AD or PD (Boyko et al., 2017;Chen et al., 2018). Therefore, a better comprehension of peripheral immune responses is becoming essential to understand neurodegenerative diseases to predict, halt or delay their progression. All the studies mentioned above have addressed neurodegenerative diseases in which distinct neural types perish and, ultimately, variable brain regions result affected. In this study, we wanted to ascertain how peripheral immune cells are influenced by selective neurodegeneration, i.e., by the death of a specific neuronal population. Our aim was to elucidate the behavior of such immune cells in this pathological condition and uncover potential biomarkers for predicting or monitoring its progression. Here we used the Purkinje Cell Degeneration (PCD) mouse, an excellent model of human cerebellar degeneration and ataxia (Shashi et al., 2018;Karakaya et al., 2019;Sheffer et al., 2019). Humans affected by this impairment (as well as PCD mice) hold a mutation in the CCP1 gene (also known as AGTPBP1 or NNA1) that leads to the postnatal loss of the Purkinje cells in the cerebellum (Mullen et al., 1976;Wang and Morgan, 2007), both species displaying very similar pathological alterations (Shashi et al., 2018;Karakaya et al., 2019;Sheffer et al., 2019). Interestingly, PCD mice also experience the loss of mitral cells in the olfactory bulb (OB). Both neurodegenerative events occur at different and well-defined periods of time (Wang and Morgan, 2007). More precisely, Purkinje cell loss starts at postnatal day 18 (P18) and progresses quickly in such a way that at P25 about 50% of Purkinje cells in the vermis have degenerated, at P30 these neurons mainly survive in the nodulus (lobule X) and ventral side of the uvula (lobule IX), and at P40 only a few Purkinje cells persist, most of which are in lobule X (Mullen et al., 1976;Wang and Morgan, 2007). Moreover, mitral cell degeneration in the OB begins around P60 and then advances more slowly for an additional 2 months (Greer and Shepherd, 1982;Valero et al., 2006). Therefore, the PCD mouse is a suitable model for studying selective neurodegeneration, allowing the assessment of two different and well-characterized scenarios in the same animal and with the advantage of not having to cause neuronal death using invasive techniques that would introduce additional variability. In particular, we explored possible alterations at the central level by analyzing both leukocyte infiltration and distribution, as well as at the peripheral level by evaluating the phenotype of splenic leukocytes in the PCD mutant mouse in comparison with wild-type (WT) animals. Additionally, in order to compare different neuroinflammatory scenarios, we used lipopolysaccharide (LPS) and gamma radiation to generate two standard models of neuroinflammation in which an enhanced peripheral recruitment into the brain parenchyma has been described (Cazareth et al., 2014;Morganti et al., 2014;Moravan et al., 2016). Our findings highlight the specific effects on peripheral immune cells in the central nervous system caused by the degeneration of Purkinje cells, as well as the intriguing alterations in the myeloid compartment at the peripheral level that could be employed as early biomarkers of this selective neurodegeneration. MATERIALS AND METHODS Animals Both WT and PCD mice from the C57BL/DBA hybrid strain were used. These mice were obtained by mating C57BL/6J and DBA/2J strains, both originally purchased from the Jackson Laboratory (Bar Harbor, ME, United States). Animals were separated into groups depending on their genotype and age at the time of analysis: P15, P20, P25, P30, P40, and P70 (n= 5 per age and genotype). These ages were chosen considering the neurodegenerative processes that occur in PCD mice (Mullen et al., 1976;Wang and Morgan, 2007). Additionally, we used another three groups of treated WT mice at P25: one received LPS, and the remaining two groups were gamma-irradiated (n= 5 per group). Mice were housed at the animal facility of the University of Salamanca at constant temperature and relative humidity, with a 12/12 h photoperiod, and were fed ad libitum with water and special rodent chow (Rodent toxicology diet, B&K Universal G.J., S.L., Barcelona, Spain). Animals were housed, handled, and sacrificed following the guidelines established by European (Directive 2010/63/UE, Recommendation 2007/526/CE) and Spanish (Law 32/2007, RD 53/2013) legislation. All experiments were approved by the Bioethics Committee of the University of Salamanca (reference numbers: #00291 and #00344). Genotyping As PCD mice are not suitable for breeding (Wang and Morgan, 2007), the colony was kept by mating heterozygous animals which are indistinguishable from their WT littermates. As a Frontiers in Cellular Neuroscience | www.frontiersin.org 2November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 3 del Pilar et al. Immune Alterations by Selective Neurodegeneration consequence, the offspring were genotyped by PCR as previously described (Díaz et al., 2012). Lipopolysaccharide Administration and Gamma Irradiation LPS-treated animals were administered with LPS at P24, while irradiated animals were exposed to radiation at P18 or P24. LPS (strain O26:B6, Sigma-Aldrich, St. Louis, MO, United States) was freshly dissolved in saline solution and injected intraperitoneally with a dose of 2 mg/kg body weight, as used in other studies (Cazareth et al., 2014;Fu et al., 2014). In addition, the irradiated animals received a single dose of 3 Gy whole-body gamma irradiation that was supplied using a 137Cs source for mice (Gammacell 1000 Elite, 243 cGy/min, 0.662 MeV, MDS Nordion, Ottawa, Canada). Tissue Preparation The mice were deeply anesthetized and subsequently perfused intracardially with 0.9% w/v NaCl for 1 min, followed by Somogyi’s fixative solution containing 4% w/v paraformaldehyde and 15% v/v saturated picric acid in 0.1 M phosphate buffer, pH 7.4 (PB) for 15 min. Brains were dissected out and immersed in the same fixative for 2 h at room temperature. Then, they were rinsed with PB and cryoprotected with 30% w/v sucrose in PB overnight at 4◦C. Afterward, cerebella and OBs were cut in 40-µm sagittal or coronal sections, respectively, by employing a freezing-sliding microtome (Jung SM 2000, Leica Microsystems, Wetzlar, Germany). The sections collected were rinsed with PB to remove fixative and sucrose residues and immunostained to visualize the leukocytes. Immunofluorescent Labeling Free-floating sections were washed with phosphate buffered saline, pH 7.4 (PBS; 3 ×10 min) and incubated for 72 h at 4◦C under continuous rotary shaking in a medium containing 0.2% v/v Triton X-100, 5% v/v normal donkey serum, and the primary antibodies: rat anti-CD45 (1:1,000; MCA1388; BioRad Laboratories, Hercules, CA, United States), rabbit antiIba1 (1:1,000, 019-19741, Wako Pure Chemical Industries, Ltd., Osaka, Japan) or rabbit anti-CD3 (1:200; ab5690, Abcam, Cambridge, United Kingdom) in PBS. These antibodies were used to stain leukocytes (CD45), macrophages/microglia (Iba1) and T lymphocytes (CD3). Only CD45high cells displaying rounded morphology were considered as infiltrated leukocytes (Perego et al., 2016). The CD45 labeling of microglial cells was imperceptible in most cases as previously described (Cuadros et al., 2006) and, when present, it was clearly different in morphology (branched) and intensity (low) compared to the positive staining of leukocytes. Sections were washed with PBS (3 ×10 min) and then incubated in a second medium for 1 h and 30 min at room temperature under continuous rotary shaking. This second medium contained an appropriate secondary antibody conjugated to Cy2 or Cy3 (1:500; Jackson ImmunoResearch Laboratories, Cambridge, United Kingdom) in PBS. Ten minutes prior to the end of the incubation, DAPI (40,6-diamidino-2-phenylindole; Sigma-Aldrich) at 1:10,000 was added to the medium to obtain a nuclear counterstain. Finally, the sections were rinsed with PBS, mounted on gelatincoated slides, and covered using a freshly prepared antifading mounting medium. Appropriate negative controls without the primary antibodies were performed and no staining was observed in any case. To better ascertain the peripheral/blood nature of the counted cells and rule out that they were not microglia, we performed an additional experiment consisting of a triple immunohistochemistry against Iba1, CD45 and TMEM119, the latter being expressed on microglia-derived cells but not on recruited blood-derived macrophages (Bennett et al., 2016). Freefloating sections were washed with phosphate buffered saline, pH 7.4 (PBS; 3 ×10 min) and incubated at 4◦C for 24 h under continuous rotary shaking in a medium containing 0.5% v/v Triton X-100, 10% v/v normal donkey serum, 0.3 M glycine, and the primary antibodies: rat anti-CD45 (1:1,000; MCA1388; Bio-Rad Laboratories), goat anti-Iba1 (1:1,000; ab5076, Abcam) or rabbit anti-TMEM119 (1:300; ab209064, Abcam) in PBS. Sections were washed with PBS (3 ×10 min) and then incubated in a second medium for 90 min at room temperature under continuous rotary shaking. This second medium contained an appropriate secondary antibody conjugated to Cy2, Cy3, or Cy5 (1:500; Jackson ImmunoResearch Laboratories) in PBS, to mark CD45, TMEM119 or Iba1, respectively. The rest of the protocol was the same as described above. Microscopy Visualization and Cell Counting Sections were observed under either an epifluorescence microscope Olympus Provis AX70 equipped with an Olympus DP70 digital camera (12.5 MP, Olympus, Tokyo, Japan) or a confocal microscope STELLARIS 8 (Leica Microsystems). Epifluorescence images were taken with a spatial resolution of 6,4 px/µm. Confocal images were taken with a spatial resolution of 1.75 px/µm (20X objective), 3.5 px/µm (40X objective) or 14 px/µm (4X magnification with the 40X objective). Digital images were processed using Adobe Photoshop CC 2015 (Adobe Inc., San Jose, CA, United States) to slightly adjust contrast, brightness and color balance. Leukocytes were manually counted during examination of all sections. For the cerebellar histological analyses, three sections of vermis per animal were chosen, where all cerebellar lobules were clearly seen, and the degeneration occurs earlier (Wang and Morgan, 2007). Regarding the OB, central coronal sections of a one-in-six series (8–10 sections per animal) were analyzed for comparisons, where both mitral cell and glomerular layers were clearly observed, excluding the accessory olfactory bulb (Weruaga et al., 1999). All counts were performed by the same person (CdP) and following the same criteria. Analysis of Splenic Subsets by Flow Cytometry For the exploration of peripheral immune cells, both myeloid and lymphoid populations from four WT and four PCD spleens per Frontiers in Cellular Neuroscience | www.frontiersin.org 3November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 4 del Pilar et al. Immune Alterations by Selective Neurodegeneration age were analyzed by flow cytometry. As soon as the mice were euthanized, their spleens were placed in HibernateTM medium (Thermo Fisher Scientific, Waltham, MA, United States) until subsequent processing. A single-cell suspension was obtained from spleens by passing them through a 40 µm nylon cell strainer (BD Biosciences, San Jose, CA, United States). Then, the red blood cells were lysed in ACK lysis buffer (0.83% w/v NH4Cl, 0.1% w/v KHCO3, 1 mM EDTA in distilled H2O, pH 7.4; Panreac Química, Barcelona, Spain) and 106splenocytes were resuspended in 50 µL of staining buffer containing 25 mM HEPES, 2% v/v penicillin/streptomycin, and 10% v/v fetal bovine serum (Cultek, Madrid, Spain) in sterile PBS. Fc receptors were blocked with anti-CD16/CD32 antibodies (10 µg/mL; BD Biosciences) in staining buffer for 10 min at 4◦C. Thus, all nonspecific bonds between Fc receptors present on immune cells and the subsequent antibodies were interrupted. After blocking, cells were labeled for 30 min at 4◦C in darkness using 50 µL of staining buffer containing the corresponding antibodies (see below). Splenocytes were washed twice with staining buffer, recovered by centrifugation at 1,500 rpm for 5 min at room temperature, resuspended in PBS and, finally, analyzed using a FACSCantoTM II cytometer (BD Biosciences). For the lymphoid subset analysis, we used the following anti-mouse antibodies: FITC-conjugated CD8 (5 µg/mL), PE-conjugated CD4 (2 µg/mL), Pacific Blue-conjugated CD3ε(4 µg/mL; BD Biosciences), PE-Cy5.5-conjugated CD25 (4 µg/mL) and APC-conjugated CD69 (4 µg/mL; Thermo Fisher Scientific). For the myeloid subset analysis, the following anti-mouse antibodies were used: FITCconjugated Ly-6C (10 µg/mL), PE-conjugated Ly-6G (4 µg/mL), PerCP-Cy5.5-conjugated CD11b (4 µg/mL; BD Biosciences), PE-Cy7-conjugated MHC-II (4 µg/mL), APC-conjugated CD11c (4 µg/mL) and eFluor 450-conjugated F4/80 (4 µg/mL; Thermo Fisher Scientific). The data were analyzed with FACSDiva 6.1 (BD Biosciences) and FlowJo 7.6.4 software (TreeStar Inc., Ashland, OR, United States) from the Flow Cytometry service of the Hospital Nacional de Parapléjicos. In all cases, the gating strategy involved the exclusion of dead cells, debris, and doublets. Statistical Analysis Once normality had been checked using the KolmogorovSmirnov test, the two-way ANOVA test was employed when the WT and PCD groups at different ages were compared. Once we had confirmed there was no interaction between age and genotype, we performed the Student’s t-test in order to evaluate possible differences between PCD and WT mice within each age, or one-way ANOVA and Bonferroni post hoc tests to evaluate differences between ages. In addition, the one-way ANOVA test was used to compare several groups at the same age. In this case, Dunnett’s post hoc test was performed to compare all groups with their WT or PCD mouse counterparts. The minimum significance level was set at p<0.05, except for detecting possible interactions between factors in the two-way ANOVA test, where significance was considered at p<0.1. All analyses were performed using the IBM SPSS Statistical 25 software (IBM, Armonk, NY, United States). RESULTS Number of Leukocytes in Wild-Type and Purkinje Cell Degeneration Mouse Olfactory Bulb Firstly, we investigated the effect of selective mitral cell loss on leukocyte infiltration. For this, we analyzed the number of leukocytes in OB sections from WT and PCD mice at P25, when the degeneration of mitral cells has not started, and P70, when the death of these cells is in progress (Valero et al., 2006). The peripheral nature of the leukocytes counted was verified by ascertaining that they were negatively immunostained for TMEM119, which was only expressed by microglial cells (Supplementary Figure 1). Leukocytes (CD45-positive and TMEM119-negative cells) were found in all bulb layers of both WT and PCD mice (Figures 1A–F). The location of the different layers of the OB is depicted in Figure 1G. No statistically significant differences were detected between WT and PCD mice at P25 (p= 0.154; Figure 1H) or at P70 (p= 0.747; Figure 1I). Therefore, it seems that the moderate affectation and, consequently, the mild neurodegenerative microenvironment in the OB of PCD mice does not trigger leukocyte infiltration. Amount and Distribution of Leukocytes in Wild-Type and Purkinje Cell Degeneration Mouse Cerebellum Afterward, we explored whether the severity of the inflammatory/degenerative process is important for leukocyte infiltration. To this end, we analyzed the number of leukocytes in the cerebellum, where a stronger neurodegenerative and inflammatory microenvironment exits in PCD animals (Baltanás et al., 2013). In this case, we employed WT and PCD mice at P15, P20, P25, P30 and P40, taking into account the temporal progression of the degeneration afflicting PCD mice. As in the OB, leukocytes were found in all cerebellar layers of both WT and PCD mice (Figures 2A–F). However, in this case, the quantitative analysis determined a statistically significant increase in the number of leukocytes in PCD mice exclusively at P25 (p= 0.017) and P30 (p= 0.027), when compared with WT ones (Figure 2G). Both ages correspond to advanced neurodegenerative stages in PCD mice. This is indicative that the degree of the neurodegenerative process in each brain area is related to the extent of cell recruitment. In addition, a striking peak in the number of leukocytes appeared at P20 in both WT and PCD mice (Figure 2G), which was statistically higher in relation to the rest of the ages of WT mice (p= 0.038 for P20 vs. P15, p= 0.002 for P20 vs. P25, p= 0.004 for P20 vs. P30, and p<0.001 for P20 vs. P40). This is probably connected with developmental refinements occurring in the cerebellum at this age. Given that the number of leukocytes in the cerebellum was not as limited as in the OB, we considered it appropriate to assess their distribution within the different layers of the cerebellar cortex and the underlying white matter. For this evaluation, Frontiers in Cellular Neuroscience | www.frontiersin.org 4November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 5 del Pilar et al. Immune Alterations by Selective Neurodegeneration FIGURE 1 | Leukocyte infiltration into the OB of WT and PCD mice. (A–F) Immunolabeling for CD45 (green) showing leukocytes in the different layers of the OB of both WT and PCD mice; nuclei are stained with DAPI (blue). (G) Typical coronal section of the mouse OB obtained with confocal microscopy; nuclei are stained with DAPI (gray). (H,I) Charts showing the quantification of leukocytes per OB section at P25, when the death of mitral cells has not started in PCD mice, and at P70, when the mitral cell loss is in progress. No differences were detected between WT and PCD mice in any case. Data are expressed as dot plots, where each dot represents one individual animal and horizontal lines represent the mean of each group. GL, glomerular layer; EPL, external plexiform layer; MCL, mitral cell layer; IPL, internal plexiform layer; GCL, granule cell layer; ONL, olfactory nerve layer. ns, not significant (p>0.05). Scale bar: 25 µm for (A–F); 500 µm for (G). we studied the molecular and Purkinje cell layers together (ML +PCL), where Purkinje cell dendritic arbors and somas are located, and, separately, the granular layer and the white matter (GL +WM). The percentage of leukocytes was notably higher in the ML +PCL in PCD mice regarding WT mice at P20 (p= 0.003), P25 (p= 0.002) and P30 (p= 0.019), i.e., throughout the cerebellar degenerative process (Figure 2H). The results concerning the percentage of leukocytes in GL +WM Frontiers in Cellular Neuroscience | www.frontiersin.org 5November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 6 del Pilar et al. Immune Alterations by Selective Neurodegeneration FIGURE 2 | Leukocyte infiltration and distribution in the cerebellum of WT and PCD mice. (A–F) Immunolabeling for CD45 (green) showing leukocytes in the different layers of the cerebellar cortex of both WT and PCD mice; nuclei are stained with DAPI (blue). (G) Chart showing the quantification of leukocytes per section of cerebellar vermis. (H,I) Graphs showing the proportional distribution of leukocytes in molecular and Purkinje cell layers (H) or granular layer plus the underlying white matter (I); these results are complementary. Note that differences in both the quantification and distribution were detected when the degeneration of Purkinje cells is in progress (from P20 to P30). Data are expressed as dot plots, where each dot represents one individual animal and horizontal lines represent the mean of each group. *p<0.05; **p<0.01. GL, granular layer; PCL, Purkinje cell layer; ML, molecular layer. Scale bar: 25 µm. were the opposite (Figure 2I). All these data suggest that the loss of Purkinje cells modifies the infiltration pattern of peripheral immune cells. Amount, Distribution, and Characterization of Leukocytes in the Mouse Cerebellum Under Several Neuroinflammatory Scenarios Next, we wanted to compare the effect of selective cerebellar neurodegeneration on the recruitment and distribution of leukocytes with other types of brain inflammation. Thus, we compared the following experimental paradigms: lack of inflammation (WT), inflammation associated with selective neurodegeneration (PCD), inflammation induced by LPS and inflammation induced by gamma radiation. The objective of this comparison was to ascertain whether the previously observed effects are specifically due to the degeneration of Purkinje cells. All groups were studied at P25, when the greatest difference in leukocyte infiltration and distribution between WT and PCD mice was appreciated. There was an evident decrease in the number of leukocytes in the cerebellum of irradiated animals with respect to the WT mice, regardless of the time elapsed from irradiation to tissue analysis, either 1 day (p= 0.002) or 7 days (p= 0.017) after treatment (Figure 3A). By contrast, leukocyte infiltration increased dramatically in LPS-treated mice as compared to the WT animals (p<0.001; Figure 3A). These results point to moderate whole-body irradiation or LPS as potential agents for hampering or facilitating leukocyte infiltration into the cerebellum, respectively. As noted above, the leukocytes in PCD mice clearly tended to localize in those layers of the cerebellar cortex mostly affected by PCD in relation to WT mice (Figures 3B,C). Interestingly, this phenomenon was not observed in LPS-treated or irradiated mice (Figures 3B,C). In fact, LPS-mediated inflammation even Frontiers in Cellular Neuroscience | www.frontiersin.org 6November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 7 del Pilar et al. Immune Alterations by Selective Neurodegeneration FIGURE 3 | Leukocyte characterization in the cerebellum of different animal models of neuroinflammation. (A) Quantification of the number of leukocytes per section of cerebellar vermis. In this case a logarithmic scale was used to achieve a better visualization of the data, since the number of leukocytes in LPS-treated mice was much higher than in the other groups. (B,C) Distribution of leukocytes in ML + PCL (B) and GL + WM (C); these results are complementary. Note the remarkable differential distribution of leukocytes in PCD mice. (D–F) Immunolabeling for CD3 (red) and CD45 (green) of a T cell in the molecular layer of a PCD mouse. (G) Percentage of T cells in the cerebellum. (H–J) Immunolabeling for Iba1 (red) and CD45 (green) of a monocyte in the molecular layer of a LPS-treated mouse. Arrow: microglial cells, which were negatively marked with CD45. (K) Percentage of monocytes in the cerebellum; note that radiation-induced or LPS-induced inflammation triggered monocyte infiltration, which was extremely low in both WT and PCD mice. (L,M) Temporal analysis of the percentage of T cells (L) and monocytes (M) in the cerebellum of WT and PCD mice; the crossing between temporal lines (dotted lines joining the mean values of each age) in both charts is indicative of the interaction between the factors analyzed. Note the fluctuations over time in PCD animals, while WT mice showed a relatively constant percentage of both T cells and monocytes throughout the whole period studied. Data are expressed as dot plots, where each dot represents one individual animal and horizontal lines represent the mean of each group. *p<0.05 and **p<0.01, regarding WT mice; ##p<0.01, regarding PCD mice. Scale bars: 25 µm. induced an opposite, albeit slighter, effect on the leukocyte disposition, toward the innermost layers of the cerebellum (p= 0.032). If we compare PCD mice with respect to the others, the differential effect on leukocyte distribution is even more evident (p= 0.002 for WT mice, p<0.001 for irradiated and LPS-treated mice; Figures 3B,C). Representative images of the amount and distribution of leukocytes in each experimental group can be appreciated in Supplementary Figure 2. Altogether, Frontiers in Cellular Neuroscience | www.frontiersin.org 7November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 8 del Pilar et al. Immune Alterations by Selective Neurodegeneration these findings verify that the selective degeneration of Purkinje cells triggers a specific attractive effect on peripheral leukocytes. Additionally, we wanted to characterize the leukocytes in all of these experimental groups. In particular, we studied the percentage of T cells (CD3+/CD45+cells; Figures 3D–F) and monocytes (Iba1+/CD45+cells; Figures 3H–J) in these scenarios, since they constitute the main peripheral immune populations that enter the brain parenchyma under neuronal damage (González and Pacheco, 2014;Solleiro-Villavicencio and Rivas-Arancibia, 2018). No statistical changes could be observed in the percentage of T cells among the experimental groups (one-way ANOVA p-value = 0.476; Figure 3G). Regarding the monocyte population, there was a significant increase in the percentage of monocytes in both irradiated groups in comparison with WT animals (p= 0.025 and p= 0.032 for 1 day and 7 days post-irradiation, respectively), which was even higher in LPStreated mice (p<0.001; Figure 3K). In contrast, no differences between PCD and WT animals were detected (p= 0.465; Figure 3K). These findings suggest that PCD does not alter the ratio of infiltrated T cells or monocytes in the cerebellar parenchyma as other less specific neuroinflammatory models do. Although no differences in the percentage of both T cells and monocytes between the cerebellum of PCD and WT mice were identified at P25, we wanted to complete this study by analyzing possible temporal changes in both genotypes throughout the degenerative process of mutant animals. In both cases, interaction between age and genotype was detected (two-way ANOVA: p= 0.090 for T cells, and p= 0.003 for monocytes), hence no reliable further statistical analyses can be performed. Accordingly, these results are described alluding to their graphic representation (Figures 3L,M). In WT animals, the percentage of T cells and monocytes remained constant over time (Figures 3L,M). On the contrary, in PCD mice the percentage of T cells presented some fluctuations at P15 and P30 (Figure 3L). Regarding the percentage of monocytes, remarkable high cell levels in PCD mice were identified between P15 and P20 which rapidly reverted to WT levels (Figure 3M). This could well mean that initial stages of the neurodegenerative process induce a prompt and transient recruitment of monocytes into the mutant cerebellum. Altogether, these results reflect certain alterations in two of the main leukocyte types that can be found in the central nervous system, before and during the degeneration of Purkinje cells in PCD mice. Phenotypic Analysis of Peripheral Immune Cells Finally, since the degeneration of Purkinje neurons exerted a special influence on the recruitment of immune cells, we decided to explore whether the phenotype of these cells is also altered during the cerebellar degeneration. To do so, the percentage of several splenic leukocyte populations was evaluated by flow cytometry, as well as the mean fluorescence intensity (MFI) of diverse myeloid and lymphoid markers. The analyzed cells were obtained from both WT and PCD spleens at P15, P20, P30 and P40. To facilitate comprehension only the significant results will be presented, which correspond to the myeloid compartment (CD11b+cells). Accordingly, all the percentages were provided with respect to CD11b+cells. On one hand, we found a lower percentage of several myeloid cell subsets. For instance, the percentage of inflammatory monocytes, known as Ly-6Chigh (CD11b+Ly-6G−/low Ly6Chigh), was significantly reduced in PCD mice at P15 (p= 0.015; Figures 4A,B), before PCD had begun. Interestingly, there were no differences at the other ages studied (Figure 4B). Furthermore, the percentage of dendritic cells (CD11b+CD11c+) and macrophages (CD11b+F4/80+) showed variations between the two animal groups. In the case of dendritic cells, the percentage was significantly reduced at P15 (p= 0.034) and P30 (p= 0.011; Figures 4C,D), while the percentage of macrophages presented a generalized reduction from P15 to P30, although it was only statistically significant at P20 (p= 0.033; Figures 4E,F). On the other hand, in PCD animals, several alterations in the MFI of markers related to antigen presentation and functional maturity of myeloid cells were detected. In this sense, antigen-presenting cells (CD11b+MHC-II+) displayed a significantly reduced surface expression of MHC-II in PCD mice from P15 to P30 (p= 0.009 at P15, p= 0.048 at P20, p= 0.045 at P30; Figures 5A,B). Besides, the MFI of both F4/80 and MHC-II was down-regulated in macrophages (Figures 5C–F), at P20 (p= 0.033) and P30 (p= 0.042) for F4/80, and at P30 for MHC-II (p= 0.006). Moreover, a significant decrease of the MFI of both CD11c and MHC-II was detected in dendritic cells at P20 (p= 0.040) and P30 (p= 0.032), respectively (Figures 5G–J). Altogether, these findings reflect alterations in the peripheral myeloid compartment during the PCD that may also appear prior to the beginning of the neurodegenerative process. Concerning the lymphoid lineage, we did not find statistically significant differences in any of the parameters studied (data not shown): the percentage of CD3+, CD4+,and CD8+cells with respect to all splenocytes, the percentage of CD4+and CD8+cells respecting the CD3+population, the percentage of CD25+and CD69+(T-cell activation markers) cells respecting the CD3+, CD4+, or CD8+subsets, the MFI of CD25 respecting the CD3+CD25+, CD4+CD25+, or CD8+CD25+subsets, and the MFI of CD69 respecting the CD3+CD69+, CD4+CD69+, or CD8+CD69+subsets. DISCUSSION The purpose of this study was to investigate how selective neurodegeneration and its associated inflammation influence peripheral immune cells. To this end, we analyzed the presence and distribution of leukocytes in the central nervous system of PCD and WT mice, as well as their phenotype at the peripheral level in the spleen. The PCD mouse was the chosen model of neurodegeneration, since it experiences the selective death of specific neuronal populations in different regions and at welldefined moments (Mullen et al., 1976;Wang and Morgan, 2007). Another point favoring the use of PCD mice, is that analogous mutations have recently been discovered in humans showing quite similar symptoms (Shashi et al., 2018;Karakaya et al., 2019; Frontiers in Cellular Neuroscience | www.frontiersin.org 8November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 9 del Pilar et al. Immune Alterations by Selective Neurodegeneration FIGURE 4 | Flow cytometry analysis of the percentage of different myeloid subsets in the spleen of WT and PCD mice. (A,C,E) Representative dot plots showing the Ly-6Chigh cell (A), dendritic cell (C) and macrophage (E) content within the whole myeloid population after having excluded dead cells, debris and doublets. (B,D,F) Charts showing the percentage of the Ly-6Chigh cells (B), dendritic cells (D) and macrophages (F) within the myeloid population. Note that PCD mice presented an overall reduction in the percentage of several myeloid populations. Data are expressed as mean ±standard error of the mean. *p<0.05. Sheffer et al., 2019). Moreover, we employed two models of general brain damage, one induced by a physical agent (gamma radiation) and the other by a biological agent (LPS), with the aim of comparing the inflammation of PCD mice (due to a selective neuronal death) with other standardized neuroinflammatory scenarios. The selective neurodegeneration afflicting the PCD mouse cerebellum induced a specific attractive effect on leukocytes Frontiers in Cellular Neuroscience | www.frontiersin.org 9November 2021 | Volume 15 | Article 773696 fncel-15-773696 November 27, 2021 Time: 12:41 # 16 del Pilar et al. Immune Alterations by Selective Neurodegeneration Ye, B., Li, C., Yang, Z., Wang, Y., Hao, J., Wang, L., et al. (2014). Cytosolic carboxypeptidase CCP6 is required for megakaryopoiesis by modulating Mad2 polyglutamylation. J. Exp. Med. 211, 2439–2454. doi: 10.1084/jem.20141123 Ye, B., Liu, B., Hao, L., Zhu, X., Yang, L., Wang, S., et al. (2018). Klf4 glutamylation is required for cell reprogramming and early embryonic development in mice. Nat. Commun. 9:1261. doi: 10.1038/s41467-018-03008-2 Zhao, W., Beers, D. R., Hooten, K. G., Sieglaff, D. H., Zhang, A., KalyanaSundaram, S., et al. (2017). Characterization of gene expression phenotype in amyotrophic lateral sclerosis monocytes. JAMA Neurol. 74, 677–685. doi: 10.1001/jamaneurol.2017.0357 Zlokovic, B. V. (2011). Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat. Rev. Neurosci. 12, 723–738. doi: 10.1038/nrn3114 Zondler, L., Müller, K., Khalaji, S., Bliederhäuser, C., Ruf, W. P., Grozdanov, V., et al. (2016). Peripheral monocytes are functionally altered and invade the CNS in ALS patients. Acta Neuropathol. 132, 391–411. doi: 10.1007/s00401-0161548-y Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Copyright © 2021 del Pilar, Lebrón-Galán, Pérez-Martín, Pérez-Revuelta, ÁvilaZarza, Alonso, Clemente, Weruaga and Díaz. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Frontiers in Cellular Neuroscience | www.frontiersin.org 16 November 2021 | Volume 15 | Article 773696