Ablation of interleukin-19 improves motor function in a mouse model of amyotrophic lateral sclerosis

Neuroinflammation by activated microglia and astrocytes plays a critical role in progression of amyotrophic lateral sclerosis (ALS). Interleukin-19 (IL-19) is a negative-feedback regulator that limits pro-inflammatory responses of microglia in an autocrine and paracrine manner, but it remains unclear how IL-19 contributes to ALS pathogenesis. We investigated the role of IL-19 in ALS using transgenic mice carrying human superoxide dismutase 1 with the G93A mutation (SOD1G93A Tg mice). We generated IL-19–deficient SOD1G93A Tg (IL-19−/−/SOD1G93A Tg) mice by crossing SOD1G93A Tg mice with IL-19−/− mice, and then evaluated disease progression, motor function, survival rate, and pathological and biochemical alternations in the resultant mice. In addition, we assessed the effect of IL-19 on glial cells using primary microglia and astrocyte cultures from the embryonic brains of SOD1G93A Tg mice and IL-19−/−/SOD1G93A Tg mice. Expression of IL-19 in primary microglia and lumbar spinal cord was higher in SOD1G93A Tg mice than in wild-type mice. Unexpectedly, IL-19−/−/SOD1G93A Tg mice exhibited significant improvement of motor function. Ablation of IL-19 in SOD1G93A Tg mice increased expression of both neurotoxic and neuroprotective factors, including tumor necrosis factor-α (TNF-α), IL-1β, glial cell line–derived neurotrophic factor (GDNF), and transforming growth factor β1, in lumbar spinal cord. Primary microglia and astrocytes from IL-19−/−/SOD1G93A Tg mice expressed higher levels of TNF-α, resulting in release of GDNF from astrocytes. Inhibition of IL-19 signaling may alleviate ALS symptoms.


Introduction
Amyotrophic lateral sclerosis (ALS) is an adult-onset neurodegenerative disease characterized by motor neuron degeneration, which leads to progressive muscle weakness, amyotrophy, and death from respiratory paralysis within 3-5 years from onset [1]. ALS is thought to be caused by a multi-factorial mechanism, including glial neuroinflammation, oxidative stress, glutamate-mediated excitotoxicity, mitochondrial abnormalities, and impaired axonal transport [2]. In particular, recent evidence indicates that non-neuronal cells play a pivotal role in ALS pathogenesis [3]. Restricted expression of superoxide dismutase 1 (SOD1) mutation in neurons is not sufficient to induce ALS pathology [4], but wild-type neurons exhibit an ALS phenotype when their surroundings are altered by glial cells carrying the SOD1 mutation [5].
Interleukin-19 (IL-19) is a member of the IL-10 family of cytokines, which includes IL-20, IL-22, IL-26, IL-28A, IL-28B, and IL-29 [6]. Recently, IL-19 was grouped into the IL-20 subfamily, which includes IL-20, IL-22, and IL-24, because these cytokines share common receptor subunits (IL-20 receptor α/β heterodimer) and signaling pathways [7,8]. IL-19 is an anti-inflammatory cytokine, produced by activated microglia and macrophages, that acts as a negative-feedback regulator to limit pro-inflammatory responses by these cells in an autocrine and paracrine manner [9,10]. In the central nervous system (CNS), microglia and astrocytes are thought to be the main source of IL-19 production [9,11]. In regard to its anti-inflammatory role, IL-19 inhibits the symptoms of animal models of inflammatory bowel disease [10,12], suppresses hapten-dependent skin inflammation in contact hypersensitivity [13], decreases brain infarction area in a mouse model of cerebral ischemia [14], and improves locomotor function in a mouse model of spinal cord injury [15]. On the other hand, IL-19 also acts as a proinflammatory cytokine in T helper 2 cell (Th2)-mediated diseases such as asthma and rheumatoid arthritis. In fact, IL-19 stimulation induces the production of TNFα, IL-6, and reactive oxygen species in monocytes [16], TNF-α and IL-6 in synovial fibroblasts [17], and IL-1β, IL-6, IL-8, CCL5, and CXCL9 in lung epithelial cells [18]. Therefore, IL-19 exerts either anti-inflammatory or proinflammatory effects in accordance with immunological conditions. However, it remains uncertain whether IL-19 is involved in the pathomechanism of neurodegenerative diseases. In this study, we used a mouse model to investigate how IL-19 contributes to the pathogenesis of ALS.

Behavioral assessments
After the age of 8 weeks, mice were assessed weekly by body weight scaling and Rotarod test; assessments were performed in randomized order by an observer blinded to the genotype [20]. Time of disease onset was defined as the first day when the mice exhibited hindlimb motor deficits such as abnormal reflexes or tremor, which are early symptoms observed in SOD1 G93A Tg mice [20,21]. For the Rotarod test, the time that the mouse remained on a rotating cylinder (Ugo-Basile, Monvalle, Varese, Italy) at a constant speed of 15 rpm was measured. Each mouse was evaluated three times, and the longest latency to fall was recorded; 180 s was chosen as the arbitrary cut-off time.

Statistical analysis
Statistical significance was analyzed using Student's t-test or one-way analysis of variance (ANOVA) followed by post hoc Tukey's test. Survival time and onset time were analyzed by log-rank test. All statistical analyses were performed using GraphPad Prism version 8 (Graph Pad Software, La Jolla, CA, USA).

IL-19 is upregulated in primary microglia and lumbar spinal cord of SOD1 G93A Tg mice
First, we examined IL-19 expression levels in primary microglia and lumbar spinal cords of wild-type and SOD1 G93A Tg mice. qPCR analyses revealed that IL-19 expression in primary microglia was significantly higher in SOD1 G93A Tg mice than in wild-type mice (Fig. 1a). SOD1 G93A Tg mice also exhibited elevated expression of IL-19 in the lumbar spinal cords as the disease progressed (Fig. 1b). In addition, expression levels of IL-19 receptor (IL-20 receptor α/β heterodimer) were higher in lumbar spinal cords of SOD1 G93A Tg mice than in wild-type mice. (Fig. 1c, d).  (Fig. 2c), but it did significantly improve motor function at the late stage of the disease (Fig. 2d, b a c d Next, we examined the effect of IL-19 on glial cells in lumbar spinal cord at the late (20 weeks) and end (24 weeks) stages. Immunohistochemical analyses of microglia, astrocytes, and neurons did not reveal significant differences in the extent of gliosis or neuronal loss between SOD1 G93A Tg mice and IL-19 −/− /SOD1 G93A Tg mice at either disease stage (Fig. 3a-f ). However, ablation of IL-19 decreased positivity for CD68 (neurotoxic microglial marker; Fig. 4a, b, e) and increased positivity for arginase1 (neuroprotective microglial marker; Fig. 4c, d, f ) in lumbar spinal cord microglia of 20-week-old SOD1 G93A Tg mice. These data indicated that IL-19 modulates microglial phenotype in SOD1 G93A Tg mice.

Ablation of IL-19 upregulates both neurotoxic and neuroprotective factors in lumbar spinal cord of SOD1 G93A Tg mice
Next, we evaluated the expression levels of neurotoxic and neuroprotective factors in the lumbar spinal cords of SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg mice at the late stage (20 weeks) when motor function began to show a significant difference. qPCR analyses for pro-inflammatory factors revealed that expression levels of TNF-α and IL-1β were significantly upregulated in IL-19 −/− / SOD1 G93A Tg mice compared to SOD1 G93A Tg mice ( Fig. 5a-d). As for neurotrophic factors and anti-inflammatory factors, GDNF and TGF-β were significantly upregulated in IL-19 −/− /SOD1 G93A Tg mice compared to SOD1 G93A Tg mice (Fig. 5e-h). Interestingly, these data imply that IL-19 affects expression levels of both neurotoxic and neuroprotective factors in lumbar spinal cord of SOD1 G93A Tg mice.

Ablation of IL-19 increases TNF-α release from microglia and astrocytes, thereby promoting GDNF production in astrocytes, in SOD1 G93A Tg mice
Because IL-19 deficiency increased the expression of TNF-α and GDNF in lumbar spinal cord of SOD1 G93A Tg mice (Fig. 5d, f ), and in light of a previous observation that TNF-α induces GDNF release from astrocytes [25], we investigated the interaction between TNF-α and GDNF in microglia and astrocytes from ALS mice. First, we sought to determine whether IL-19 deficiency would alter the expression levels of TNF-α and GDNF in microglia and astrocytes from ALS mice. We used LPS to stimulate glial cell cultures because previous studies revealed that mutant SOD1 induces microglial activation via TLR4 and CD14, which serve as the receptor complex for LPS [26,27]. TNF-α expression was higher in 24-h LPS-stimulated primary microglia or astrocytes from IL-19 −/− /SOD1 G93A Tg mice than in those from SOD1 G93A Tg mice (Fig. 6a, c), whereas levels of GDNF were not affected (Fig. 6b, d).
Next, we investigated whether IL-19 deficiency alters the expression level of TNF-α-induced GDNF in microglia and astrocytes from ALS mice. For this purpose, we stimulated primary microglia and astrocytes from SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg mice with TNF-α (0-100 ng/ml) for 24 h. TNF-α stimulation dosedependently upregulated GDNF expression level in astrocytes (Fig. 7b), but not in microglia (Fig. 7a). Moreover, primary astrocytes from IL-19 −/− /SOD1 G93A Tg mice exhibited a significantly greater increase in GDNF expression level than those from SOD1 G93A Tg mice in response to TNF-α stimulation (Fig. 7b). These results suggested that astrocytes secrete GDNF upon stimulation with TNF-α released from microglia and astrocytes, which acts in an autocrine and paracrine manner.
In addition, we examined the levels of TNF-α and GDNF in the lumbar spinal cord over a time course. GDNF expression was significantly higher in IL-19 −/− / SOD1 G93A Tg mice after the late stage (20-24 weeks) than in SOD1 G93A Tg mice as the TNF-α upregulation (Fig. 7c, d). The chronological expression data in lumbar spinal cord are consistent with our in vitro observations of astrocytic release of GDNF in response to TNF-α stimulation.

Discussion
Previous studies showed that IL-19 acts in an autocrine and paracrine negative-feedback regulator to limit proinflammatory responses by microglia and macrophages [9,10], and that it can slow the progression of microglia/ macrophage-mediated CNS disorders such as stroke [14] and spinal cord injury [15]. In fact, we recently showed that IL-19 deficiency worsened microglia/ macrophage-mediated experimental autoimmune encephalitis, whereas IL-19 treatment abrogated this condition by inhibiting macrophage activation [28]. In this study, we observed upregulation of IL-19 expression in lumbar spinal cord of SOD1 G93A Tg mice in parallel with disease progression (Fig. 1). Therefore, we hypothesized that ablation of IL-19 would exacerbate disease progression in SOD1 G93A Tg mice.
Unexpectedly, however, we observed that ablation of IL-19 improved motor function at the late stage in SOD1 G93A Tg mice (Fig. 2). At this stage, immunohistochemical analyses detected that IL-19 deficiency resulted in conversion of microglia from a pro-inflammatory and neurotoxic phenotype to an anti-inflammatory and neuroprotective phenotype (Fig. 4). qPCR analyses revealed that expression levels of TNF-α and GDNF in lumbar spinal cord were higher in IL-19 −/− /SOD1 G93A Tg mice than in SOD1 G93A Tg mice (Fig. 5). Moreover, IL-19 deficiency enhanced TNF-α upregulation in microglia and astrocytes and augmented TNF-α-induced GDNF upregulation in astrocytes from SOD1 G93A Tg mice, which was corroborated by the chronological expression levels of GDNF and TNF-α in the lumbar spinal cord of these mice (Figs. 6 and 7). Due to the technical difficulties of isolating and maintaining glial cells from diseased adult mice, we used primary glial cell cultures from neonatal mouse brains to represent disease phenotypes in adult mice. In general, cell culture models recapitulated the disease progression of animal models on a compressed time scale. In ALS models, previous studies also demonstrated that neonatal primary cultures from mutant SOD1 transgenic mice underwent significant phenotypic changes corresponding to those in adult mice [29][30][31]. Therefore, we believe that our cell models at least partly recapitulated the disease pathology in ALS mice. Interestingly, previous studies documented that TNF-α promotes TNF-α release from microglia and astrocytes in an autocrine and paracrine manner,  [25,[32][33][34], and that TNF-α induces astrocytic GDNF production [25]. Other studies reported TNF-α elevation in the spinal cord of SOD1 G93A Tg mice prior to disease onset, and showed that ablation of TNFR1 but not TNFR2 enhanced motor neuron loss and accelerated disease progression in SOD1 G93A Tg mice [35,36]. Microglia-derived TNF-α plays a crucial role as a paracrine signal to regulate astrocytic neuroprotective functions [37], and microglia in SOD1 G93A Tg mice produce high levels of TNF-α [26]. Based on these findings, we hypothesized that the TNF-α-GDNF axis would exert neuroprotective effects in ALS pathogenesis (Fig. 8). Mutant SOD1-derived proinflammatory stimuli induce microglia and astrocytes to release TNF-α in an autocrine and paracrine manner. Secreted TNF-α induces astrocytic GDNF release, which exerts a neuroprotective effect. Ablation of IL-19 promotes neuroprotection by enhancing this TNF-α-GDNF neuroprotective cascade. Several lines of evidence indicate that activated microglia and astrocytes are pivotal players in ALS pathogenesis, and that glial activation exerts both neurotoxic and neuroprotective effects depending on its spatiotemporal distribution [29,38]. In particular, glial activation by cytokines often shows conflicting abilities in neurotoxicity and neuroprotection [39]. For example, overexpression of TGF-β1, an antiinflammatory cytokine, paradoxically accelerates the disease progression of SOD1 G93A Tg by astrocytic activation [40]. Furthermore, IL-10, a major antiinflammatory cytokine, exacerbates amyloid-β plaque burden and cognitive impairment, whereas IL-10 deficiency enhanced amyloid-β clearance in Alzheimer's disease model mice [41,42]. We also observed that the level of IL-1β in the lumbar spinal cord was higher in IL-19 −/− /SOD1 G93A Tg mice than in SOD1 G93A Tg mice. IL-1β is a major pro-inflammatory cytokine that enhances glial NF-κB activation and subsequent TNF-α production [43]. Therefore, it is possible that IL-1β elevation could induce and augment glial TNF-α production in IL-19 −/− /SOD1 G93A Tg mice.
IL-19 also exerts a potent anti-inflammatory effect on CNS-infiltrating monocytes and macrophages under Th1-prone condition [14,15], whereas it exerts a pro-inflammatory effect on tissue-infiltrating monocytes/macrophages and fibroblasts in Th2-prone conditions [16][17][18]44]. Previous studies also reported conflicting observations of IL-19 effects (pro-inflammatory vs. anti-inflammatory) even in similar mouse models of colitis [10,45], suggesting that the effects of IL-19 depend strongly on the condition and distribution of inflammation. In addition to the unexpected pro-inflammatory effect of IL-19 in ALS mice, we observed another small discrepancy in the trends of GDNF and TNF-α expression levels in lumbar spinal cord in the end stage (24 weeks) (Fig. 7c, d). Although differential expression levels of TNF-α were observed between SOD1 G93A Tg and IL-19 −/− / SOD1 G93A Tg mice in the late stage (20 weeks), rapidly increasing TNF-α expression in the end stage reached similar levels in both Tg mice (Fig. 7d). By contrast, GDNF expression was significantly higher in IL-19 −/− / SOD1 G93A Tg mice than in SOD1 G93A Tg mice from the late to the end stage (Fig. 7c). Nevertheless, these mice had similar lifespans. Strong upregulation of TNF-α in the end stage may have exacerbated neuroinflammation, disrupted CNS homeostasis, and overwhelmed the neuroprotective effect of GDNF. Further investigations are needed to elucidate these issues.
Our observations reveal a complex neuro-gliaimmune interaction under inflamed conditions in ALS mice. Spatiotemporal manipulation of glial activation might facilitate the development of novel therapeutic strategies for neurodegenerative diseases such as ALS. a b c d Fig. 6 Ablation of IL-19 upregulates TNF-α expression in primary microglia and astrocytes from SOD1 G93A Tg. a, b qPCR data for TNF-α and GDNF expression in LPS (1000 ng/ml)-stimulated primary microglia from SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg. c, d qPCR data for TNF-α and GDNF expression in LPS (1000 ng/ml)-stimulated primary astrocytes from SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg.

Conclusions
In this study, we demonstrated that IL-19 deficiency delayed disease onset and improved motor function at the late stage in the SOD1 G93A Tg mice. Ablation of IL-19 slowed the disease progression of ALS mice by enhancing glial neuroprotection through such glia-secreted factors as TNF-α and GDNF. Our results suggest that blockade of IL-19 signaling represents a potential therapeutic strategy for ameliorating neurodegenerative diseases including ALS. a b c d Fig. 7 Ablation of IL-19 enhanced TNF-α-induced upregulation of GDNF in primary astrocytes and lumbar spinal cord from SOD1 G93A Tg mice. a qPCR for GDNF mRNA in primary microglia stimulated with TNF-α (0-100 ng/ml). Values are means ± SD (n = 4). *p < 0.05. b qPCR for GDNF mRNA in primary astrocytes stimulated by TNF-α (0-100 ng/ml). Values are means ± SD (n = 4). *p < 0.05. c qPCR data for GDNF expression in lumbar spinal cord from SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg. Values are means ± SD (n = 6). *p < 0.05. d qPCR data for TNF-α expression in lumbar spinal cord from SOD1 G93A Tg and IL-19 −/− /SOD1 G93A Tg. Values are means ± SD (n = 6). *p < 0.05 when data of 12-20-week-old mice were analyzed. **p < 0.05 when data of 12-24-week-old mice were analyzed Fig. 8 Model of the roles of TNF-α and GDNF in glia-mediated neuroprotection in ALS. Stimuli such as mutant SOD1 induce microglia and astrocytes to release TNF-α. Secreted TNF-α induces GDNF release from astrocytes in an autocrine and paracrine manner, resulting in enhancement of neuroprotection. Blockade of IL-19 signaling may augment this TNF-α-induced GDNF neuroprotective cascade (red arrows)