Acute viral infection accelerates neurodegeneration in a mouse model of ALS

Abstract

While several viral infections have been associated with amyotrophic lateral sclerosis (ALS), the mechanism(s) through which they promote disease remains elusive. Here we investigate the impact of common, acute viral infections on ALS disease onset and progression in the SOD1^G93A mouse model. A single sublethal infection prior to onset of ALS clinical signs is associated with markedly accelerated ALS disease progression characterized by rapid loss of hindlimb function. Prior infection results in gliosis in the lumbar spine and upregulation of transcriptional pathways involved in inflammatory responses, metabolic dysregulation, and muscular dysfunction. Therapeutic suppression of gliosis with an anti-inflammatory small molecule, or administration of a direct-acting antiviral, is associated with significantly improved ALS clinical signs, akin to what is observed in uninfected animals. Our study provides causal and mechanistic evidence that the immune response elicited by acute viral infections may be an important etiological factor that alters ALS disease trajectory.

Common, acute, non-neurotropic respiratory virus infections accelerate ALS disease progression well after the virus has been cleared by augmenting inflammation of glial cells in the spinal cord and motor neuron loss. Anti-inflammatory and antiviral treatment delay disease progression, pointing to infection as a modifiable ALS risk factor.

Introduction

Amyotrophic Lateral Sclerosis (ALS) is an incurable motor neuron disease which is characterized by progressive deterioration of upper and lower motor neurons leading to muscle wasting and paralysis^1,2. The disease is remarkably heterogeneous in age-of-onset and in presentation of symptoms, even for individuals harboring the same ALS-associated mutations^3–6. A combination of genetic and environmental factors is therefore believed to play critical roles in disease development and progression⁶.

Viruses have been characterized as environmental risk factors for ALS and other neurodegenerative diseases⁷. However, several of these associations have been erroneous, and study designs have led to an enrichment bias favoring neurotropic viruses, and viruses that cause chronic/latent infections⁸. Poliovirus was the first virus associated with ALS due to its neurotropism and ability to cause post-poliomyelitis, which phenotypically resembles ALS⁷. Subsequent studies have demonstrated that the enterovirus-associated 3C protease is able to cleave TAR DNA binding protein 43 (TDP-43), a pathophysiological feature observed in 97% of ALS cases^9,10. Additionally, a sublethal intracerebroventricular infection with the coxsackie virus B3, a type of enterovirus, accelerated ALS disease progression in SOD1^G85R mice¹¹. There has also been compelling evidence linking endogenous retroviruses such as Human Endogenous Retrovirus-K (HERV-K) to ALS, as transgenic mice constitutively expressing the env protein of HERV-K develop symptoms of progressive motor neuron disease^12,13. Such observations have supported on-going clinical trials evaluating antiretroviral therapy in ALS patients^14–16.

Although common, acute non-neurotropic viral infections may also be associated with ALS and other neurodegenerative diseases, they have been widely understudied since ubiquitous infections, especially those that occur significantly prior to the onset of disease (i.e. hit-and-run), are difficult to associate using conventional epidemiological approaches. This is due, in part, to the fact that these viruses are likely cleared long before ALS symptom onset begins and seropositivity is high amongst the general population. Elegant recent work using data from biobank repositories across multiple countries showed an increased risk of developing neurodegenerative diseases such as ALS from acute viral infections, with increased hazard ratios persisting even 15 years following infection¹⁷. However, much remains to be understood regarding the mechanism through which viral infections impact ALS.

Influenza A viruses (IAV) are a group of single-stranded RNA viruses which predominantly result in acute, self-limiting respiratory infections¹⁸. Infection is associated with systemic elevation of pro-inflammatory soluble mediators which can potentiate neuroinflammation. Indeed, non-neurotropic IAV strains have been shown to directly induce long-lasting neuroinflammation with accompanying decreases in dendritic spine density to hippocampal neurons—a typical hallmark of neurodegeneration¹⁹. The COVID-19 pandemic, caused by SARS-CoV-2, has also led to lasting neurological manifestations during long COVID, with studies showing long-term neurological sequelae in approximately 33% of recovered COVID-19 patients, even 6 months post-infection^20,21. In addition, SARS-CoV-2 RNA and antigens have been found in the central nervous system (CNS) of COVID-19 patients²². Strikingly, SARS-CoV-2 infection has already been associated with increased susceptibility to Parkinson’s disease as well as other neurodegenerative diseases^23,24. In addition to inducing pathological immune responses, viral infections have been shown to perturb many of the same pathophysiological pathways underlying motor neuron death in ALS⁸.

A common mechanism through which diverse viral infections may exacerbate ALS is through the induction of aberrant host immune responses that can lead to neuroinflammation²⁵. Inflammation is a hallmark characteristic of both viral infections and ALS, with elevated peripheral inflammation and neuroinflammation positively correlating with disease severity²⁶. ALS patients often experience heightened levels of gliosis and pro-inflammatory cytokines in the peripheral blood and cerebral spinal fluid (CSF)²⁷. Given this close interplay between inflammation and ALS, common acute viral infections may exacerbate ALS disease progression through further stimulating the immune system and elevating pre-existing CNS and systemic inflammatory mediators. However, many studies suggesting an association between ALS and prior infection have been correlative and lack mechanistic insights as to how viral infections potentiate disease. Furthermore, it remains unclear whether common acute respiratory infections impact ALS disease progression, and the mechanism(s) through which this might occur.

Here, we demonstrate that a single acute infection with either IAV or SARS-CoV-2 during the pre-clinical stages of ALS disease significantly accelerates ALS progression in SOD1^G93A mice. Interestingly, intramuscular administration of inactivated IAV resulted in similar disease acceleration, a phenomenon not observed following intranasal delivery. Mechanistically, this acceleration of ALS progression was accompanied by activation of microglia and astrocytes in the lumbar spinal cord of SOD1^G93A mice following infection, and transcriptomic enrichment of pathways relating to inflammation, immunity, and muscular dysfunction. Importantly, our findings further reveal that the inhibition of microglial activation during infection, or treatment with antivirals, helps mitigate ALS acceleration in these mice. This study thus provides direct causal and mechanistic evidence of acute respiratory infections exacerbating ALS progression in a mouse model of disease.

Results

Acute IAV infection prior to ALS onset accelerates disease progression in SOD1^G93A mice

Despite over 30 years of reported associations between viral infections and ALS, no prior studies have systematically explored the role of acute, non-neurotropic viral infections on the progression of ALS. To address this critical gap, we used influenza A virus (IAV) as a model given its ubiquity. Indeed, most individuals have experienced their first influenza infection by five years of age and are re-infected regularly throughout the course of their life^28–30. To determine whether IAV infection during the pre-clinical stages of ALS disease accelerates the onset and/or progression of ALS, wildtype BL6SJL (WT) and B6SJL-TgN(SOD1*G93A)1Gur/J (SOD1^G93A) mice were infected intranasally (i.n.) with a sublethal dose (300 plaque forming units (PFU)) of mouse-adapted influenza A virus (strain A/Puerto Rico/8/1934 H1N1, (PR8)) at day 60 of age (Fig. 1A). This age was chosen based on the well-characterized kinetics of disease onset in this model as reported in the literature (typically, early onset is detectable at day 90)³¹. Indeed, there were no statistical differences observed between WT and SOD1^G93A mice in rotarod performance at day 60 of age (mean 33.13 and 28.58 s), indicating no observable phenotypic signs of ALS onset in SOD1^G93A mice (Fig. 1B). Following infection, no infectious virus was detected in brain homogenates of WT or SOD1^G93A mice at 4- or 14-days post-infection (dpi) (Supplementary Fig. 1A, B) and IAV infection was fully resolved in the lungs by 14 dpi (Supplementary Fig. 1C). This is consistent with previous literature demonstrating that infection with this strain of IAV is acute and non-neurotropic³². After recovering from infection, mice were tested on the rotarod apparatus to measure motor function and general coordination and were compared to uninfected controls. Mice were tested weekly, and the latency for falling off the rotating rod was recorded. Both WT and SOD1^G93A mice demonstrated initial improvements on the rotarod test as they become experienced with the rotarod test^33,34. Both uninfected and IAV-infected WT mice performed similarly on the rotarod test, never reaching 50% of their initial fall time (Fig. 1C). However, we observed that previously infected SOD1^G93A mice performed significantly worse on the rotarod test when compared to uninfected SOD1^G93A mice (8.153 weeks and 10.37 weeks, respectively, p = 0.0241), with a decline in performance occurring earlier, indicating accelerated disease onset (Fig. 1C). In addition, these mice reached endpoint due to ALS clinical signs, as defined by complete hind-limb paralysis and loss of the righting reflex, significantly faster than uninfected SOD1^G93A mice (Fig. 1D). Given that sex is an important biological variable in ALS disease progression^35–37, with higher incidence and predominantly spinal onset reported in men, and sex-dependent differences in disease onset and survival documented in SOD1^G93A mice, we additionally assessed the contribution of sex to ALS disease progression following IAV infection. By stratifying the data presented in Fig. 1D by sex, we observed that infected male SOD1^G93A mice reached endpoint due to ALS clinical signs significantly faster than females (Supplementary Fig. 2). Taken together, these data demonstrate that a single, acute, sublethal IAV infection during the pre-clinical stages of ALS significantly accelerates disease progression.

Fig. 1. Acute IAV infection prior to ALS onset accelerates disease progression in SOD1^G93A mice.

A Schematic of the experimental timeline for IAV infection of mice. B Raw rotarod test fall times of wildtype (WT) (n = 7) and SOD1^G93A (n = 8) mice at day 60 of age. C Mean weekly rotarod fall times for uninfected WT (n = 13), IAV infected WT (n = 5), SOD1^G93A uninfected (n = 11), and SOD1^G93A IAV infected (n = 14), starting at 2 weeks post-infection expressed as a percentage of week 1 fall time (week 1 = 100%). D ALS-related endpoint (survival) of uninfected WT (n = 11) and SOD1^G93A (n = 17) and IAV infected WT (n = 11) and SOD1^G93A (n = 14) mice. E Weight loss following IAV infection of WT (n = 8) and SOD1^G93A (n = 12) mice as compared to uninfected WT (n = 9) and SOD1^G93A (n = 9) mice. F Survival of uninfected WT (n = 9), uninfected SOD1^G93A (n = 9), IAV infected WT (n = 8) and IAV infected SOD1^G93A (n = 12) mice. G Lung viral titers 4 days post-IAV infection of WT (n = 3) and SOD1^G93A (n = 4) mice. H Viral growth curves in WT and SOD1^G93A mouse embryonic fibroblasts (MEFs) supernatants infected with IAV (MOI = 3) (n = 3). Means and standard error of the mean (SEM) are shown. Statistics were obtained by a two-tailed Student’s T-test (B, G), one-way ANOVA with Tukey post-hoc test (E, H), one-tailed Extra Sum of Squares F-test (C), and two-tailed Mantel-Cox test (D, F).

WT and SOD1^G93A mice experience similar acute disease characteristics following viral infection

Given that a single acute IAV infection was associated with marked acceleration in ALS disease progression, we next sought to determine whether this was related to differences in infection kinetics and/or severity in SOD1^G93A mice compared to WT controls. To this end, mice were infected i.n. with a sublethal dose of IAV at day 60 of age and were subsequently monitored for influenza-associated morbidity and mortality. There were no significant differences when comparing weight loss of IAV-infected WT and SOD1^G93A mice during active IAV infection (Fig. 1E). Likewise, no significant differences in survival following acute infection were observed (Fig. 1F). Consistent with the lack of observed differences in infection kinetics and severity, there were no significant differences in lung viral titres when comparing SOD1^G93A and WT mice as assessed by plaque forming unit (PFU) assay during peak viral load (4 dpi) (Fig. 1G).

Next, we sought to determine whether IAV replication kinetics were directly affected by the WT and SOD1^G93A genotypes in vitro. Using susceptible culture-based systems allowed us to evaluate viral replication and dynamic host responses to IAV infection. To this end, we isolated mouse embryonic fibroblasts (MEFs) from WT and SOD1^G93A mice and infected them ex vivo with IAV at a multiplicity of infection (MOI) of 3. There were no significant differences in IAV titres in the supernatant of infected cells over the course of 48 h (Fig. 1H).

To determine whether variability in SOD1 copy number between WT and SOD1^G93A mice could confound our viral replication findings, we performed an orthogonal in vitro experiment using HEK-293T cells transfected with plasmids encoding either WT SOD1 or SOD1^G93A. Successful overexpression was confirmed at the transcript level by RT-qPCR, with plasmid-derived SOD1 mRNA upregulated at comparable levels in both transfection conditions (Supplementary Fig. 3A), and at the protein level by western blot (Supplementary Fig. 3B). Transfected cells were subsequently infected with IAV at an MOI of 3, consistent with our MEF experiments, and viral replication was assessed up to 48 h post-infection by plaque assay. No differences in viral titres were detected between SOD1 WT- and SOD1^G93A-transfected cells at any timepoint examined (Supplementary Fig. 3C), indicating that SOD1 overexpression does not alter IAV replication kinetics. These findings are consistent with our in vivo data and reinforce the observation the WT and SOD1^G93A mice respond to acute IAV infection in the same manner.

Taken together, these data demonstrate that WT and SOD1^G93A mice respond similarly to acute IAV infection and thus, that differences in acute infection characteristics are unlikely to explain the observed acceleration in ALS disease progression.

WT and SOD1^G93A mice mount similar antiviral- and cytokine/chemokine responses following IAV infection

Peripheral- and neuro-inflammation are key hallmarks of ALS and related neurodegenerative disorders, correlating with disease severity in both patients and pre-clinical models, including SOD1^G93A mice^26,27. Additionally, there is growing recognition that the cGAS-STING pathway, which is involved in the host antiviral response, is important for disease progression^38–40. Given that acute viral infections induce pro-inflammatory and antiviral responses, we next sought to determine whether differences exist in such responses following IAV infection in WT and SOD1^G93A mice. To this end, WT and SOD1^G93A MEFs were stimulated with an MOI of 10 of UV-inactivated IAV (Fig. 2A). Viral inactivation was performed to eliminate antiviral gene antagonism by virus-encoded proteins, which would confound measurement of the host antiviral response⁴¹. RNA was isolated from cellular lysates collected at 4-, 8-, and 24-hours post-infection for RT-qPCR to determine relative gene expression values for bona fide antiviral genes, including interferon-induced transmembrane protein 3 (IFITM3), RIG-I (DDX58), and interferon-induced protein with tetratricopeptide repeats 1 (IFIT1). Relative expression levels were determined using the 2^−ΔΔCt method, and values were normalized to untreated, uninfected WT MEFs (Fig. 2B). As expected, all antiviral genes were significantly upregulated following exposure to inactivated virus in a time-dependent manner, with no significant differences in antiviral gene expression levels between the two genotypes (Fig. 2B).

Fig. 2. WT and SOD1^G93A mice mount similar antiviral- and pro-inflammatory responses following IAV infection.

A Schematic of experimental timeline for assessing antiviral gene and cytokine/chemokine expression in MEFs. B RT-qPCR of antiviral gene transcript levels in MEF cellular lysates following UV-inactivated IAV stimulation (MOI = 10) (n = 3). Values are normalized to unstimulated WT MEFs C Pro-inflammatory cytokine expression levels in the supernatant of MEFs following stimulation with UV-inactivated IAV (MOI = 10 equivalent) (n = 5). Values are normalized to unstimulated (US) WT MEFs. D Experimental timeline for assessing cytokine/chemokine expression in the lungs and brain. Pro-inflammatory cytokine expression levels in E the lungs (n = 3) and F the brain (n = 3) of uninfected WT and SOD1^G93A mice and IAV-infected WT and SOD1^G93A mice at 4- and 30- dpi. Values are normalized to uninfected WT mice. Means and standard error of the mean (SEM) are shown. Statistics were obtained by two-way ANOVA with Tukey post-hoc test (B).

To assess pro-inflammatory cytokines/chemokines induced by IAV infection, a multi-plex pro-inflammatory cytokine/chemokine array was performed on the supernatants collected following UV-inactivated IAV stimulation of WT and SOD1^G93A MEFs (Fig. 2A). Once again, we observed no significant differences in the magnitude of cytokine/chemokine examined at each timepoint when comparing IAV-stimulated WT and SOD1^G93A MEFs (Fig. 2C and Supplementary Fig. 3D).

To determine whether these similarities were recapitulated in vivo, WT and SOD1^G93A mice were infected i.n. with a sublethal dose of IAV at day 60 of age and lung and brain homogenates were collected at 4- and 30- dpi to assess for pro-inflammatory cytokines/chemokines induced by IAV infection (Fig. 2D). As expected, multiple pro-inflammatory analytes were significantly elevated in the lungs (Fig. 2E, left panel; and Supplementary Fig. 4A) and brain (Fig. 2F, left panel; and Supplementary Fig. 5A) following IAV infection at 4 dpi in both mouse strains, with no significant differences between WT and SOD1^G93A mice post-infection. Strikingly, several of these analytes remained elevated at 30 dpi in the lung (Fig. 2E, right panel; and Supplementary Fig. 4B) and brain (Fig. 2F, right panel; and Supplementary Fig. 5B), well after resolution of IAV infection (Supplementary Fig. 1C). In concordance with the observations in MEFs, there were no significant differences in pro-inflammatory cytokine/chemokine relative expression levels between IAV-infected WT and SOD1^G93A mice. Taken together, these data suggest that WT and SOD1^G93A mice mount similar antiviral and inflammatory responses to IAV infection; results that are in alignment with the observed similarities in weight loss and viral replication between the two genotypes.

Virus-induced systemic and neuroinflammation is associated with accelerated ALS disease progression

Having established that IAV infection accelerates ALS disease progression in SOD1^G93A mice, we next went on to determine the contribution of the immune response stimulated by the virus on disease acceleration. By inactivating the virus prior to inoculation, the contribution of viral replication to observed phenotypes is eliminated, and any differences can be attributed primarily to the host response resulting from viral detection. To this end, we administered 50 μg of formalin-inactivated IAV intramuscularly (i.m.) within the quadricep muscle and compared these to mice receiving i.m. vehicle control (equimolar formalin diluted in PBS) (Fig. 3A). Mice were then subjected to the rotarod test and monitored until endpoint was reached due to progression of ALS clinical signs. SOD1^G93A mice administered formalin-inactivated IAV i.m. performed significantly worse on the rotarod test, reaching 50% of their initial fall time significantly faster than mice administered i.m. vehicle control (8.457 weeks and 10.04 weeks, respectively, p = 0.0355) (Fig. 3B). Additionally, i.m. administration significantly shortened lifespan of SOD1^G93A mice compared to those administered vehicle control (Fig. 3C). Since IAV infection results in considerable systemic inflammation, and the hindlimbs where inactivated IAV was injected is the site where ALS disease manifests in the SOD1^G93A mouse model, we set out to determine whether local inflammation in tissues other than the quadriceps also accelerated disease. To this end, we administered inactivated IAV or vehicle control i.n. (Fig. 3A). However, in contrast to inactivated IAV given i.m., when an equal dose of formalin-inactivated IAV was administered i.n., there were no significant differences in rotarod performance between SOD1^G93A administered formalin-inactivated IAV i.n. compared to SOD1^G93A vehicle control i.n. (7.055 and 7.394 weeks, respectively, p = 0.4273) (Fig. 3D) and there were no differences in survival due to ALS clinical signs for mice administered inactivated IAV i.n. compared to mice administered vehicle control (Fig. 3E).

Fig. 3. Virus-induced systemic and neuroinflammation is associated with accelerated ALS disease progression.

A Schematic of experimental timeline. B Mean weekly rotarod fall times for mock-injected WT (n = 4) and SOD1^G93A (n = 4) and i.m. inactivated-IAV treated (50 μg) WT (n = 11) and SOD1^G93A (n = 5) mice, starting 2 weeks post-administration expressed as a percentage of week 1 fall time (week 1 = 100%). C ALS-related endpoint of WT (n = 11) and SOD1^G93A (n = 5) mice administered inactivated-IAV i.m. compared to WT (n = 4) and SOD1^G93A (n = 4) mice given vehicle control i.m. D Mean weekly rotarod fall times for mock-inoculated WT (n = 6) and SOD1^G93A (n = 5) and i.n. inactivated-IAV treated (50 μg) WT (n = 8) and SOD1^G93A (n = 8) mice, starting 2 weeks post-administration expressed as a percentage of week 1 fall time (week 1 = 100%). E ALS-related endpoint of WT (n = 8) and SOD1^G93A (n = 8) mice administered inactivated-IAV i.n.compared to WT (n = 6) and SOD1^G93A (n = 5) mice given vehicle control i.n. Means and standard error of the mean (SEM) are shown. Statistics were obtained by a one-tailed Extra Sum of Squares F-test (B, D), and two-tailed Mantel–Cox test (C, E).

To measure the magnitude of systemic and neuroinflammation induced by administering inactivated IAV relative to a bona fide IAV infection, we administered 50 μg inactivated IAV either i.n. or i.m. and compared these mice to those given 300 PFU i.n. of live IAV. At 4 days post-administration, serum and brain homogenates were prepared, and a multi-plex pro-inflammatory cytokine/chemokine array was performed (Supplementary Fig. 6A). In serum (Supplementary Fig. 6B/C) and brain (Supplementary Fig. 6D/E), the highest magnitude of pro-inflammatory analytes was observed in mice infected with 300 PFU IAV. However, there were no statistically significant differences between WT and SOD1^G93A mice.

The more pronounced inflammatory responses induced by live virus infection and i.m. administration of inactivated virus were therefore associated with disease progression relative to i.n. administration of inactivated IAV, which induced minimal systemic and neuroinflammation and did not accelerate disease.

Collectively, these data suggest infection-induced immunity appears to be an important factor in eliciting ALS disease acceleration. This is especially evident when inflammation is systemic or localized within disease susceptible tissues, such as the hind-limbs of SOD1^G93A mice.

IAV infection is associated with elevated gliosis, ALS-related transcripts, and motor neuron loss in the lumbar spine

Neuroinflammation and pathological non-cell autonomous processes such as aberrant glial cell activation are critical drivers of motor neuron degeneration in patients and pre-clinical models⁴². In the SOD1^G93A model, selective ablation of this mutant SOD1 in microglia and astrocytes leads to slower disease progression, suggesting a key role for these cells in ALS pathophysiology^43,44. Furthermore, it has been well established that neuroinflammation can prime microglia to a predominantly pathogenic cytotoxic/pro-inflammatory M1 phenotype⁴⁵. Given these findings, and that acute infection with a non-neurotropic strain of IAV was associated with sustained increased levels of pro-inflammatory cytokines/chemokines in the brain (Fig. 2F), we next investigated whether IAV infection induced inflammation and/or gliosis in the lumbar spine of SOD1^G93A mice.

We first characterized soluble inflammatory mediators in the lumbar spine following IAV infection in WT and SOD1^G93A mice. Lumbar spine homogenates were collected at 4 dpi (acute infection) and 90 dpi (late-stage ALS disease onset) and assessed using a multiplex pro-inflammatory cytokine/chemokine array (Supplementary Fig. 7). During acute infection (4 dpi), IAV infection was associated with minimal changes in pro-inflammatory mediators within the lumbar spine. Only IP-10 showed a statistically significant increase in infected SOD1^G93A mice compared to uninfected SOD1^G93A controls (Supplementary Fig. 7A, B). No other significant changes were observed following infection in either WT or SOD1^G93A mice, and no differences in pro-inflammatory mediators were detected between WT and SOD1^G93A mice with or without IAV infection.

We next assessed pro-inflammatory mediators in the lumbar spine at 60 dpi, well after resolution of acute IAV infection. In line with previously published studies demonstrating that pro-inflammatory mediators increase in the lumbar spinal cord of SOD1^G93A mice as a function of disease progression^46,47, we observed statistically significant increases in pro-inflammatory cytokines in the lumbar spines of uninfected SOD1^G93A mice in comparison to WT controls (Supplementary Fig. 7C, D). Importantly, there were no significant differences in cytokine levels between uninfected and IAV-infected SOD1^G93A mice, indicating that the elevated inflammatory mediators reflect ongoing ALS pathology rather than infection-induced neuroinflammation.

To quantify gliosis in the lumbar spine, WT and SOD1^G93A mice were infected i.n. with a sublethal dose of IAV at day 60 of age, and immunohistochemistry was performed at day 90 of age (30 dpi) (Fig. 4A). This timepoint was chosen since sustained peripheral and central inflammation following sublethal IAV infection was observed (Fig. 2E, F), and day 90 typically corresponds to onset of ALS clinical signs in this model. Spinal sections were isolated and fixed sections were probed with anti-GFAP and anti-Iba1 antibodies to measure levels of astrocyte and microglia activation, respectively (Fig. 4B/C). In agreement with previously published studies⁴⁸, we observed elevated gliosis as measured by GFAP+ astrocyte cells (Fig. 4B and Supplementary Fig. 8A), and by Iba+ microglial cells (Fig. 4C and Supplementary Fig. 8B) in lumbar spines from uninfected SOD1^G93A mice in comparison to uninfected WT controls. Furthermore, we observed a significant increase in gliosis of previously infected SOD1^G93A mice relative to uninfected SOD1^G93A mice, in line with the observed accelerated disease phenotype reported in Fig. 1 (Fig. 4C and Supplementary Fig. 8B). Importantly, there were no significant increases in gliosis in WT mice following IAV infection (Fig. 4B/C and Supplementary Fig. 8A/B). These data suggest that a single, acute infection 30 days prior to typical onset of ALS clinical signs was associated with significantly elevated gliosis in the lumbar spine of SOD1^G93A mice.

Fig. 4. Acute IAV infection is associated with elevated gliosis and ALS-related transcripts in the lumbar spine.

A Schematic for assessing glial cell activation and transcriptomics in the lumbar spine. B Quantification of GFAP+ cells for each lumbar spine at day 90 of age (30 dpi) of WT (n = 3) and SOD1^G93A (n = 4) uninfected, WT (n = 3) and SOD1^G93A (n = 4) IAV infected C Quantification of Iba1+ cells for each lumbar spine at day 90 of age (30 dpi) of WT (n = 3) and SOD1^G93A (n = 6) uninfected, WT (n = 3) SOD1^G93A (n = 5) IAV infected D Gene set enrichment analysis of pathways significantly enriched in IAV infected SOD1^G93A mice (n = 3) compared to uninfected SOD1^G93A mice (n = 3) at day 90 of age (30 dpi). Enrichment map (top left), select enriched pathways with the number of genes associated with each pathway listed beside the bars (bottom left), individual enrichment score plots (top right). E Quantification of the number of ChAT+ cells in the ventral horn per lumbar spinal section at day 90 of age (30 dpi) of SOD1^G93A uninfected (n = 4) and SOD1^G93A IAV infected (n = 4). F Quantification of the number of ChAT+ cells in the ventral horn per lumbar spinal section at day 120 of age (60 dpi) SOD1^G93A uninfected (n = 4) and SOD1^G93A IAV infected (n = 3). (right). Means and standard error of the mean (SEM) are shown. Statistics were obtained by a one-way ANOVA with Tukey post-hoc test (B, C) and by a two-tailed Student’s T-test (E, F).

We next isolated RNA from the lumbar spine of both previously IAV infected and uninfected SOD1^G93A mice at 90 days of age (30 dpi) and carried out RNA sequencing to identify differentially expressed genes. Gene set enrichment analysis (GSEA) revealed significant upregulation in pathways associated with innate and adaptive immunity, lipid metabolism, apoptotic pathways, and muscle dysfunction in previously IAV infected SOD1^G93A mice compared to uninfected SOD1^G93A mice (Fig. 4D). Specifically, pathways related to inflammatory and immune responses were prominently elevated, indicating a robust immune activation in the lumbar spine well after IAV infection was resolved. Taken with the results above, these data suggest that prior acute viral infection potentiates gliosis in the lumbar spine and induction of transcriptional programs relating to inflammation, immunity, and lipid and muscular dysfunction.

To determine whether the observed gliosis in SOD1^G93A mice was associated with motor neuron loss, lumbar spines were isolated at day 90 of age (30 dpi) and day 120 of age (60 dpi), and motor neurons were quantified by ChAT+ cell counts as per convention^49–51. Aberrant glial activation is well documented to precede motor neuron loss in ALS, and SOD1^G93A mice typically begin to develop signs of disease around day 90 of age, suggesting that this earlier time point may precede substantial motor neuron degeneration. Consistent with this, fewer ChAT+ motor neurons were observed in the lumbar spinal cord of SOD1^G93A mice at day 90 of age (30 dpi), although these differences were not statistically significant (Fig. 4E and Supplementary Fig. 8C). However, by day 120 of age (60 dpi), significant motor neuron loss was observed (Fig. 4F and Supplementary Fig. 8D).

Collectively, these data suggest that a single acute respiratory viral infection is sufficient to potentiate gliosis, drive neuroinflammatory transcriptional programs, and accelerate motor neuron loss in the lumbar spine of SOD1^G93A mice.

Inhibition of microglial activation during acute IAV infection protects SOD1^G93A mice from accelerated disease progression

We have thus far demonstrated that a single acute infection event prior to onset of ALS clinical signs accelerates ALS disease progression in SOD1^G93A mice and resulted in elevated inflammatory signatures, gliosis, and motor neuron loss in the lumbar spine (Fig. 4 and Supplementary Figs. 7 and 8). Consequently, we next sought to determine whether inhibiting this activation during acute infection could prevent ALS disease acceleration. Minocycline is known to have anti-inflammatory properties and inhibit aberrant microglial activation⁵². Importantly, minocycline has been shown to improve ALS outcomes in transgenic SOD1 mice by inhibiting microglial activation⁵³. To test whether inhibiting microglia activation during acute infection prevents IAV-induced ALS acceleration, SOD1^G93A mice were administered 50 mg/kg minocycline via intraperitoneal (i.p.) injection daily^54–56. Treatment was initiated one day prior to infection (day 59 of age) and continued until 3 weeks post-infection (day 81 of age) (Fig. 5A). Both minocycline treated and untreated mice infected with IAV experienced similar kinetics of weight loss following infection, with weight loss peaking around 8 dpi (Fig. 5B), indicating that treatment did not alter acute disease kinetics in comparison to controls. At 2 weeks post-infection, mice were assessed on the rotarod. SOD1^G93A mice administered minocycline performed significantly better on the rotarod test with a significant delay in minocycline treated mice reaching 50% of their initial fall time compared to untreated mice infected with IAV (5.156 weeks and 6.803 weeks, respectively, p = 0.0017) (Fig. 5C). However, there were no significant differences in survival due to ALS clinical sign progression between the two groups (Fig. 5D). The shape of the Kaplan-Meyer curve alongside the rotarod data suggest that the benefits of minocycline may diminish upon cessation of treatment. We next assessed levels of microglial activation in the lumbar spine at the time of treatment cessation (day 81 of age—3 weeks post-infection). Although we observed a trend towards fewer Iba1+ cells in minocycline treated SOD1^G93A mice infected with IAV compared to those infected receiving vehicle control, these differences were not statistically significant (Fig. 5E and Supplementary Fig. 9). Altogether, these results indicate that inhibiting microglial activation during acute infection helps to prevent infection-induced disease acceleration.

Fig. 5. Inhibition of microglial activation during acute IAV infection protects SOD1^G93A mice from accelerated disease progression.

A Schematic of experimental timeline. SOD1^G93A mice were administered 50 mg/kg minocycline daily at day 59 of age (one day prior to IAV infection) until day 81 of age (3 weeks post-infection). B Weight loss of minocycline treated (n = 12) and control (n = 8) SOD1^G93A mice following IAV infection. C Mean weekly fall times for minocycline treated and IAV infected SOD1^G93A (n = 11) and IAV infected SOD1^G93A vehicle control (n = 7) mice, starting 2 weeks post-infection expressed as a percentage of week 1 fall time (week 1 = 100%). D ALS-related endpoint survival of IAV infected minocycline treated mice (n = 11) compared to IAV infected vehicle control (n = 7). E Iba1 quantification in the lumbar spine at day 81 of age (administration cessation) of IAV infected SOD1^G93A mice given minocycline (n = 4) or vehicle control (n = 4). Means and standard error of the mean (SEM) are shown. Statistics were obtained by one-way ANOVA with Tukey post-hoc test (B), one-tailed Extra Sum of Squares F-test (C), two-tailed Mantel–Cox test (D), and by a two-tailed Student’s T-test (E).

Oseltamivir treatment during IAV infection protects SOD1^G93A mice from accelerated disease progression

We have shown that prior IAV infection potentiates gliosis in lumbar spine of SOD1^G93A mice, and that administering minocycline reduces IAV-induced glial activation, resulting in subsequent motor function improvements. We next examined the protective role of administering a direct-acting antiviral on infection-induced ALS disease acceleration. Oseltamivir is a direct-acting antiviral that prevents influenza illness by inhibiting the activity of neuraminidase on influenza viruses⁵⁷. To this end, SOD1^G93A mice were infected i.n. with a sublethal dose of IAV at day 60 of age and were subsequently administered 10 mg/kg oseltamivir phosphate via oral gavage twice daily. Oseltamivir administration was initiated 1 dpi and continued for 7 consecutive days (Fig. 6A). As expected, oseltamivir administration attenuated disease, reducing weight loss following IAV infection at peak weight loss (8 dpi) compared to mice administered vehicle control (Fig. 6B). Additionally, IAV-infected mice administered oseltamivir had significantly better motor performance on the rotarod test compared to infected mice that were administered vehicle control (4.790 weeks and 5.980 weeks, respectively, p = 0.0333) (Fig. 6D). However, similar to minocycline treatment, this did not result in a significant lengthening of survival (Fig. 6C). Collectively, these data demonstrate the therapeutic benefit of antiviral treatment in preventing accelerated ALS motor decline resulting from IAV infection.

Fig. 6. Oseltamivir treatment during IAV infection protects SOD1^G93A mice from accelerated disease progression.

A Schematic of experimental timeline. SOD1^G93A mice were infected with 300 PFU IAV and treated with oseltamivir (OSLT) via oral gavage twice daily at a dose of 10 mg/kg starting 1 day following infection. B Weight loss of IAV infected OSLT treated (n = 7), IAV infected vehicle treated (n = 9) SOD1^G93A mice. C ALS-related endpoint survival of IAV infected OSLT treated (n = 7) and IAV infected vehicle control treated (n = 8) SOD1^G93A mice. D Mean weekly fall times for IAV infected OSLT treated (n = 7) and IAV infected vehicle control treated (n = 8) mice, starting 2 weeks post-infection expressed as a percentage of week 1 fall time (week 1 = 100%). Means and standard error of the mean (SEM) are shown. Statistics were obtained by two-tailed Student’s T-test, one-tailed Extra Sum of Squares F-test and two-tailed Mantel–Cox test.

SARS-CoV-2 infection prior to ALS onset accelerates disease progression in SOD1^G93A mice

Given the widespread global impact of COVID-19, we next sought to determine whether similar ALS disease acceleration was observed in SOD1^G93A mice acutely infected with SARS-CoV-2. To this end, SOD1^G93A mice were infected i.n. with a sublethal dose (10⁵ PFU) of a mouse-adapted SARS-CoV-2 (strain MA10) (Fig. 7A). We observed that SOD1^G93A mice infected with SARS-CoV-2 had significantly worse motor performance assessed through the hanging wire test (p = 0.0019) (Fig. 7B). However, in contrast to acute IAV infection, acute SARS-CoV-2 infection did not accelerate time to endpoint due to ALS clinical signs faster than uninfected controls (Fig. 7C). Next to assess SARS-CoV-2 infection disease kinetics in this model, SOD1^G93A mice were infected i.n. with a sublethal dose of SARS-CoV-2. We observed no appreciable weight loss throughout the course of infection (Fig. 7D), and in concordance with IAV infection, WT and SOD1^G93A mice experienced similar viral burden within the lungs (Fig. 7E). Taken together, these results suggest that a sublethal SARS-CoV-2 infection during the pre-symptomatic stages of ALS significantly accelerates motor decline.

Fig. 7. SARS-CoV-2 infection prior to ALS onset accelerates disease progression in SOD1^G93A mice.

A Schematic of the experimental timeline for SARS-CoV-2 infection of mice. B Hanging wire evaluation (area under the curve (AUC)) of uninfected SOD1^G93A (n = 11) and SARS-CoV-2 infected SOD1^G93A (n = 12) mice. C) Survival of uninfected SOD1^G93A (n = 11) and SARS-CoV-2 infected SOD1^G93A (n = 14) mice due to ALS-related endpoint. D Weight loss following SARS-CoV-2 infection of SOD1^G93A uninfected (n = 5) and SARS-CoV-2 infected SOD1^G93A (n = 6). E Viral titres in the lungs at 2 dpi (n = 4). AUC analysis depicts standard deviation. Statistics were obtained by a two-tailed Student’s T-test and two-tailed Mantel–Cox test.

Discussion

ALS remains an incurable disease, and current treatments have relatively modest therapeutic benefits. Tremendous efforts have been made to develop new treatments; however, many clinical trials have eventually failed to slow disease progression as measured by the ALSFRS-R^58,59. The immense heterogeneity of ALS is an important factor contributing to this lack of success, a product of genetic and environmental factors⁶⁰. Viral infections have long been associated with ALS and other neurodegenerative diseases⁶¹. However, most prior studies have been based on observational studies that are correlative in nature⁸. Additionally, these studies have largely focused on chronic and neurotropic infections, as these are more easily detected in post-mortem and serological analyses of ALS patients¹⁷. Indeed, infections that occur and are resolved throughout the life of an individual can have profound long-term impacts (e.g., long COVID). Importantly, the mechanism underlying viral-mediated potentiation of ALS disease progression has been widely understudied. Our study demonstrates that a single, acute infection with common pathogens, such as IAV and SARS-CoV-2, during the pre-symptomatic stages of disease accelerates the progression of ALS in a murine model (Fig. 8). Our work also suggests that potentiation of aberrant inflammatory immune processes in the CNS may be a unifying mechanism to explain why so many viruses with diverse biologies have been implicated in ALS.

Fig. 8. Acute viral infection accelerates neurodegeneration in a mouse model of ALS.

WT and SOD1^G93A mice respond similarly to acute viral infection during the presymptomatic stages of ALS disease (left panel). Acute viral infection in SOD1^G93A mice is associated with elevated gliosis and ALS-related transcripts in the lumbar spine during ALS disease onset (middle panel). Acute viral infection in SOD1^G93A mice is associated with accelerated ALS disease progression (right panel). Created in BioRender. Ahmed, I. (2026) https://BioRender.com/309t75s.

Here, we used the SOD1^G93A model to explore the mechanisms through which viral infections impact ALS disease. This is one of the most extensively characterized mouse models of ALS with tractable disease features that closely resemble classical ALS^62,63. Additionally, timing of disease onset and rate of progression are predictable and have been widely documented, allowing us to investigate whether infection during pre-symptomatic stages has an influence on disease progression⁶³. A mouse-adapted strain of IAV subtype H1N1 (PR8) and mouse-adapted SARS-CoV-2 (strain MA10) were used as model pathogens due to their ubiquity. These particular isolates are not neurotropic, yet infection induces signatures of neuroinflammation through peripheral perturbations^19,64,65.

We found that a sublethal i.n. infection with IAV at day 60 of age resulted in poorer motor function and significant acceleration of disease relative to uninfected SOD1^G93A mice. Disease acceleration was also observed in SOD1^G93A mice infected with SARS-CoV-2. This acceleration occurred despite clearance of the virus, as no detectable virus was found in the lungs or brains of infected WT and SOD1^G93A mice following infection resolution at 14 dpi¹⁹. Furthermore, we observed no differences in the dynamics of acute infection experienced by WT and SOD1^G93A mice following IAV and SARS-CoV-2, with both genotypes demonstrated similar weight loss trajectories and viral burden in the lungs following infection. This suggests that individuals at risk of developing ALS may not be inherently more susceptible to infection prior to onset of clinical disease.

Neuroinflammation is a well-established feature of ALS and other related neurodegenerative diseases, like Parkinson’s Disease and Alzheimer’s Disease. Indeed, as we and others have shown, many genes associated with ALS have important roles in regulating innate and adaptive immune responses to infection^66–68. ALS disease progression is correlated with elevated levels of soluble pro-inflammatory mediators in the peripheral blood and CSF, and there is evidence of substantial cross-talk between the periphery and CNS^26,27. Interestingly, we observed elevated levels of pro-inflammatory cytokines/chemokines in the lungs and brains of infected WT and SOD1^G93A mice at 4 dpi that remained elevated 30 dpi, well after the initial infection had been resolved. This suggests long-term neuroinflammation can be caused by acute infection originating from peripheral responses. To distinguish the relative contribution of infection-induced immunity on ALS acceleration relative to that of virus replication itself, inactivated IAV was administered through i.m. or i.n. inoculation. We observed that i.m. administration in the quadriceps resulted in accelerated ALS disease progression of SOD1^G93A mice relative to those administered vehicle control. However, this was not observed when inactivated IAV was administered i.n. These findings suggest that peripheral immune activation alone, independent of active viral replication, may be sufficient to influence ALS disease progression. This raises broader questions about the potential impact of diverse inflammatory stimuli on disease course in susceptible individuals and is in part supported by rodent studies showing that single or chronic administration of lipopolysaccharide (LPS) into disease-susceptible tissues (the brain through intracerebroventricular (i.c.v.) administration) is associated with elevated microgliosis and astrogliosis^69,70. It is important to note that administration of inactivated virus, as was performed in this study, is significantly different than administration of vaccines in humans in terms of dose, composition, and presence of inflammatory stimuli. Individuals living with ALS and their caregivers are recommended to receive several vaccines, particularly those that prevent respiratory infection, due to the impacts of ALS on lung function. There is currently no convincing evidence that vaccination triggers or accelerates ALS progression. However, several studies have reported associations of vaccine-preventable infections with ALS risk¹⁷.

To gain further mechanistic insight, immunohistochemical analysis of the lumbar spine provided evidence of elevated astrocyte and microglial activation levels in previously infected SOD1^G93A mice compared to control mice at day 90 of age (30 dpi). Given the reported importance of glial cells in ALS pathophysiology, the observation that acute infection results in sustained gliosis, long after the infection is cleared, provides a compelling mechanism through which these infections potentiate ALS. The immunohistochemical findings were further corroborated through bulk RNA sequencing of the lumbar spinal cord of previously IAV infected and control SOD1^G93A mice. GSEA revealed significantly enriched transcripts for pathways relating to inflammation, immunity, lipid and muscle dysregulation, and apoptosis in IAV infected SOD1^G93A mice. As previously mentioned, soluble pro-inflammatory mediators and the induction of innate and adaptive immunity are positively correlated with ALS progression⁷¹. However, interestingly, enrichment of lipid dysregulation and apoptotic pathways suggests additional changes in bioenergetics and intrinsic cellular processes in the lumbar spine following IAV infection. Indeed, dysregulated lipid profiles and apoptotic signaling have been observed in ALS patients^72–74. Notably, gene sets related to muscle dysfunction and weakness were upregulated in IAV infected SOD1^G93A mice, indicating exacerbated neuromuscular impairment characteristic of this ALS model. Muscle weakness and atrophy are primary symptoms of ALS, and their exacerbation following viral infection suggests a potential interaction between acute peripheral IAV infection and neuromuscular health⁷⁵.

Given our findings that acute viral infection accelerates disease, presumably via potentiation of pathogenic immune responses in the CNS, we sought to determine the therapeutic potential of two separate approaches: chemically inhibiting microglial activation, and treating acute viral infection using antivirals. We found that minocycline and OSLT treatment improved performance of SOD1^G93A mice on the rotarod. This suggests that acute treatment is efficacious but must likely be sustained following infection in order to extend survival (since we observed the resumption of accelerated disease upon treatment cessation). Indeed, prior studies showing efficacy of antivirals and immunomodulatory drugs in pre-clinical models have maintained treatment for the duration of the disease^11,76–78.

The data presented here demonstrate that a single sublethal infection with common, acute viruses exacerbates ALS disease. These results also underscore inherent difficulties in conducting these types of studies in humans. In our model, were no observable differences in susceptibility to, or severity of, acute viral infection when comparing WT and SOD1^G93A mice. Furthermore, acceleration of disease was not evident until several weeks after the acute infections had been cleared. These types of hit-and-run mechanisms are inherently challenging to study in epidemiological contexts – especially for ubiquitous pathogens wherein differences in seroprevalence would not be expected. Elegant recent work using data from European biobanks, highlighted extensive associations between infections and many neurodegenerative diseases, including ALS, even 15 years following an exposure¹⁷. Here, we build on those studies and provide mechanistic insight that may explain how and why pathogens with inherently different biologies all contribute to neurodegenerative disease – namely, through their shared propensity to elicit inflammatory immune responses. To our knowledge, this is one of the first studies investigating whether a direct causal relationship exists with acute viral infections (particularly non-neurotropic infections), and ALS progression. This should set the stage for additional studies in human cohorts at risk for developing ALS.

Limitations

For practical reasons, our studies were limited to the SOD1^G93A mouse model, which is one of the few ALS models that recapitulates key features of neuromuscular disease with predictable and tractable kinetics. In the future, it would be interesting to validate that our findings extend to other ALS models with distinct genetic backgrounds (most of which do not show a neuromuscular phenotype, but may demonstrate features consistent for frontotemporal dementia, for example). Likewise, it would be interesting to determine whether infection could serve as a trigger for ALS in a model wherein animals do not spontaneously develop disease. However, a suitable inducible model of ALS has yet to be developed, to our knowledge.

Methods

Study design

The objective of our study was to establish whether an acute viral infection accelerates the progression of ALS in SOD1^G93A mice. We infected WT and SOD1^G93A mice with a sublethal dose of IAV i.n. (300 PFU) during the presymptomatic stages of ALS. Following viral clearance, we evaluated mice on the rotarod test and assessed endpoint due to ALS clinical sign progression, defined as complete hind-limb paralysis and loss of the righting reflex. To assess primary responses to IAV infection, weight loss, survival from acute IAV infection (defined as 20% loss of initial weight), and viral load in the lungs at 4 dpi were evaluated. Viral load was quantified by plaque assays on MDCK cells. Additionally, we evaluated antiviral gene expression in WT and SOD1^G93A MEFs and cytokine/chemokine induction in MEFS, and the lungs and brain of WT and SOD1^G93A mice at 4 and 30 dpi. We next determined whether i.m. and i.n. administration of inactivated-IAV accelerates ALS progression. Gliosis in the lumbar spine and RNA sequencing of lumbar spine homogenates were assessed at day 90 of age (30 dpi). We next sought to determine whether inhibiting microglial activation through minocycline administration mitigates IAV-induced ALS acceleration. To this end, minocycline was administered daily at a dose of 50 mg/kg i.p. beginning one day prior to infection and given up to 3 weeks post-infection. Furthermore, we assessed the therapeutic potential of the direct-acting antiviral OSLT in mitigating disease acceleration. Here, we administered 10 mg/kg OSLT twice daily via oral gavage beginning one day following infection and administered for 7 consecutive days. Finally, we evaluated whether sublethal infection with SARS-CoV-2 resulted in ALS acceleration in SOD1^G93A mice.

Animal colony and ethics statement

All mouse experiments were approved by the McMaster University Animal Research Ethics Board. SOD1 Tg mice expressing the G93A mutation, raised on the B6SJL-TgN(SOD1*G93A)1Gur/J background (SOD1^G93A) (Strain #:002726), and wildtype BL6SJLF1/J (WT) (Strain #:100012) mice were purchased from the Jackson Laboratory. Mice were maintained on a 12 h light/dark cycle and low-fat dietary chow, and water was provided ad libitum. Breeding pairs consisted of male SOD1^G93A mice and female WT mice, and mice were bred up to the second generation. Litters were weaned at 21 days of age and housed at a maximum of five mice per cage. All experimental mice were age-matched. All experiments were conducted with littermate controls and groups included both male and female animals.

Viruses and cells

Influenza A/Puerto Rico/8/1934/H1N1 (PR8) was propagated in 10 day-old specific-pathogen-free embryonated chicken eggs (Canadian Food Inspection Agency)⁷⁹. Viral stock concentrations were quantified by performing plaque assay on Madin-Darby Canine Kidney (MDCK) cells (ATCC, VA, USA). SARS-CoV-2 MA10 was kindly provided by Dr. Ralph Baric⁸⁰. MDCKs and VeroE6 cells (ATCC, VA, USA) and HEK-293T cells were maintained in Dulbecco’s minimal essential media (DMEM) (Gibco) supplemented with 1 X GlutaMAX^TM supplement (Gibco), 10% heat inactivated fetal bovine serum (FBS) (Gibco), 100 U/mL penicillin and 100 µg/mL streptomycin (Gibco), and 1% of 1 M HEPES (ThermoFisher) at 37 °C and 5% CO₂. Mouse embryonic fibroblasts (MEFs) were isolated from embryos 13.5 days post-conception as previously described⁸¹. MEFs were maintained in DMEM supplemented with 1 X GlutaMAX^TM supplement, 15% FBS, and 100 U/mL penicillin and 100 µg/mL streptomycin and incubated at 37 °C and 5% CO₂.

IAV and SARS-CoV-2 infection

WT and SOD1^G93A mice were anesthetized using isoflurane and infected with IAV i.n. in a 40 µL inoculum of PBS. Mice were infected with 300 plaque forming units (PFU) of IAV, equating to a 0.1 median lethal dose (LD₅₀) infection. SARS-CoV-2 infections were performed using a mouse-adapted strain of SARS-CoV-2 MA10. Mice were anesthetized using isoflurane and infected with SARS-CoV-2 MA10 i.n. at a dose of 10⁵ PFU. Weight loss, lethargy, and general body condition were monitored following infection for up to 2 weeks post-infection. Endpoint due to IAV- and SARS-CoV-2-associated mortality was defined as mice reaching 80% of their pre-infection weight.

Cytokine and chemokine analysis

Evaluation of cytokine and chemokine induction in the lungs, brain, and lumbar spinal cord was performed through a mouse 32-plex cytokine/chemokine array (MD32) (Eve Technologies, Alberta, Canada).

IAV Inactivation

100 X inactivation solution (0.925% formaldehyde in PBS) (Anachemia) was combined with 4.2 mg/mL of purified IAV suspension to a final concentration of 1 X inactivation solution. Samples were incubated at 4 °C on constant rotation for 72 h, followed by storage at −80 °C.

Inactivated IAV administration

WT and SOD1^G93A mice were administered 50 µg of formaldehyde-inactivated IAV in a 100 µL inoculum via i.m. injection into the quadricep muscle, and i.n. administration was performed with a 40 µL inoculum of 50 µg of formaldehyde-inactivated IAV.

Rotarod test

Motor performance and general coordination were measured using the rotarod test (Panlab, Harvard Apparatus, Holliston, MA, USA). The rotarod was set to acceleration mode (4 rpm to 15 rpm in 60 s, or 0.2 rpm/s) for a maximum rotation speed of 15 rpm. WT and SOD1^G93A mice were tested for a maximum of 180 s and the latency for falling off the apparatus was recorded. Mice were assessed once weekly until they were unable to support themselves on the apparatus, with the average of six trials being used. As with other motor tasks, mice exhibit a learning curve on the rotarod, characterized by a progressive increase in latency to fall across initial trials as they acquire the motor coordination necessary to perform the task.

Wire hang test

SOD1^G93A mice were placed on a metal wire mesh grid that was inverted 180 degrees. Latency to fall was recorded weekly across three independent trials, with each trial being performed at a maximum duration of three minutes. Average fall time was recorded for each mouse, and testing continued until animals were no longer able to support themselves on the grid.

ALS clinical sign assessment and endpoint monitoring

At day 120 of age, mice were monitored daily until reaching endpoint due to ALS clinical signs. Mice were assessed based on a 4-point neurological scoring system. Score 0: full extension of hind-limbs when suspended by the tail, Score 1: collapse or partial collapse of the hind-limbs and/or trembling of the hind-limbs when suspended by the tail, Score 2: Toe curling during locomotion, Score 3: paralysis of one or both of the hind-limbs, Score 4: unable to right themselves within 30 s of being placed on either side. Endpoint was defined as mice reaching a neurological score of 4.

IAV plaque assay

MDCK cells were seeded in a 6-well plate at a density of 6.5 × 10⁵ cells per well and incubated for 48 h at 37 °C in 5% CO₂. Following incubation, samples were serially diluted in 1 X Minimum Essential Media (MEM), 1 X GlutaMAX^TM supplement, 1.5% sodium bicarbonate, 200 mM HEPES, 1 X MEM Vitamin Solution, 1 X MEM Amino Acid Solution, 2 mg/mL penicillin/streptomycin, 7% BSA. Cells were infected for 1 h at 37 °C in 5% CO₂. Cells were subsequently overlayed with 1% (m/v) agar bacteriological combined with one part 2 X MEM, 0.01% DEAE-Dextran, 1 µg/mL N-p-Tosyl-L-phenylalanine chloromethyl ketone (TPCK)-treated trypsin, and plates were incubated at 37 °C in 5% CO₂ for 48 h. Cells were fixed with 3.7% formaldehyde diluted in PBS for 30 min at RT and stained with crystal violet.

IAV growth curves

WT and SOD1^G93A MEFs were infected with IAV at a multiplicity of infection (MOI) of 3. Supernatant was collected at 0, 2, 4, 8, 12, 24, and 48 h post-infection, and viral titres were assessed using plaque assays on MDCK cells. For in vivo studies, WT and SOD1^G93A mice were infected with 300 PFU of IAV i.n. and lungs were collected in serum-free DMEM and homogenized using the Bullet Blender Gold High-Throughput Bead Mill Homogenizer (Next Advance). Viral titres from lung homogenates were determined via plaque assay on MDCK cells.

Gene expression quantification and cytokine analysis

WT and SOD1^G93A MEFs were plated and stimulated with UV-inactivated IAV at an MOI of 10 equivalent, and supernatant was collected at 0, 4, 8, and 24 h post-stimulation. For gene expression quantification, RNA from cell lysates was isolated using the RNeasy kit (Qiagen), and 1 μg of complimentary DNA was synthesized using the Maxima First Strand cDNA Synthesis kit, according to manufacturers instructions (Thermo Fisher Scientific). Gene expression levels were assessed using SensiFAST SYBR reaction kit (Bioline Reagents Ltd.) for quantitative real-time PCR (qRT-PCR). Primers used (Thermo Fisher Scientific): IFITM3 Forward Primer (5′ – GACAGCCCCCAAACTACGAA – 3′), Mouse IFITM3 Reverse Primer (5′ – ATTGAACAGGGACCAGACCAC – 3′), DDX58 Forward Primer (5′ – ATGTGCCCCTACTGGTTGTG – 3′), DDX58 Reverse Primer (5′ – CCCCAGAAATGCTCGCAATG – 3′), IFIT1 Forward Primer (5′ – CTTTACAGCAACCATGGGAGAGA – 3′), IFIT1 Reverse Primer (5′ – TGATGTCAAGGAACTGGACCTG – 3′), 18 s Forward Primer (5′ – GTAACCCGTTGAACCCCATT – 3′), 18 s Reverse Primer (5′ – CCATCCAATCGGTAGTAGCG – 3′), SOD1 Forward Primer (5′ – AAGTAATGGACCAGTGAAGG – 3′), HA-tag Forward Primer (5′ – CATACGATGTTCCAGATTAC – 3′), SOD1 Reverse Primer (5′ - GCCTGCTGTATTATCTCCAA – 3′). For cytokine analysis, supernatants from stimulated MEFs were collected at 0, 4, 8, and 24 h post-stimulation. For in vivo studies, mice were infected with 300 PFU IAV i.n. and lung and brain homogenates were harvested in PBS containing Pierce Protease Inhibitor (Thermo Fisher Scientific) at 4- and 30- days post-infection. Cytokine levels were assessed using a 32-plex mouse cytokine/chemokine array (EVE Technologies Corp., Calgary, AB, Canada).

Fluorescent immunohistochemistry

Spinal cord sections were isolated using hydraulic extrusion as previously described, with slight modifications⁸². Briefly, mice were anesthetized and transcardially perfused with PBS. The spinal cord was removed and cut immediately distal to the 13th rib to isolate the lumbar spine. The lumbar spinal cord was hydraulically removed from the vertebrae using 5 mL of PBS and stored in 3.7% PFA for 72 h. Subsequently, samples were sliced 5 µm thick and mounted onto paraffin. Sections were deparaffinized with xylene and subsequently dehydrated using decreasing ethanol concentrations of 100%, 95%, 95%, 75%, and 75%, for 5 min at each step. Slices were then rehydrated in H₂O for 5 min. Antigen retrieval was performed with 1 X citrate buffer (Abcam) for 20 min at 95 °C. Thereafter, slices were washed 3 times with PBS-T (0.1% Tween 20, Sigma-Aldrich). Slices were permeabilized using 0.2% Triton^TM X-100 diluted in PBS-T for 20 min at RT, and then washed 3 times with PBS-T. Following permeabilization, slices were blocked with 10% normal goat serum (Abcam) and 1% BSA diluted in PBS for 1 hour at RT. Spinal cord slices were incubated with primary antibodies overnight at 4 °C. Primary antibodies included: Iba1 [1:300] (Abcam), GFAP [1:400] (Abcam), ChAT [1:100] (Abcam). Slices were then washed 3 times with PBS-T and probed with compatible AlexaFluor-conjugated secondary antibodies at a 1:1000 dilution (Invitrogen). Slices were washed 3 times with PBST-T and were subsequently stained with Hoechst 33342 nuclear counterstain (ThermoFisher) for 5 min at RT, then washed 3 times with PBST-T, followed by being mounted with EverBrite mounting medium (Biotium Inc.).

Images were acquired using a Zeiss Imager M2 microscope. Nucleated cells were quantified based on GFP expression by a trained, blinded researcher using ImageJ software. Quantification was performed on technical duplicates per animal, which were averaged to generate a single data point per animal.

Minocycline and oseltamivir treatment

Minocycline hydrochloride (Sigma-Aldrich) was dissolved in PBS and administered intraperitoneally (i.p.) once daily to SOD1^G93A at a dose of 50 mg/kg beginning one day prior to infection (day 59 of age) for a duration of three weeks following infection (day 81 of age). Mice were monitored for weight loss, assessed on the rotarod, and monitored until reaching endpoint due to ALS clinical signs. For therapeutic oseltamivir treatment, mice were infected at day 60 of age and treated twice daily with 10 mg/kg oseltamivir phosphate (Toronto Research Chemicals, O701000) via oral gavage for 7 consecutive days starting at 1 dpi. Mice were monitored for weight loss, assessed on the rotarod, and monitored until reaching endpoint due to ALS clinical signs.

RNA sequencing

Lumbar spinal cord was extracted and flash frozen in liquid nitrogen. Once frozen, samples within a cryotube were disrupted using a pestle and homogenized by pipetting with 600 µL of RLT buffer (Qiagen). Homogenized samples were then transferred to a Qiashredder spin column (Qiagen), and the flowthrough was subsequently spun in a gDNA eliminator column (Qiagen). Equal volume of 70% ethanol was added, and RNA isolation was carried out as per the RNeasy instruction manual (Qiagen). Sample libraries were prepared using the NEBNext Ultra^TM II Directional RNA Library Prep Kit for Illumina (New England Biolabs), and cDNA libraries were sequenced using an Illumina HiSeq machine (2 × 50 bp sequence reads). The reads were trimmed using TrimGalore and then aligned with GRCm39 reference using STAR⁸³. Next, the reads were counted by using HTSeq count⁸⁴. Genes showing low levels of expression were removed using EdgeR package in R, resulting in 12,503 genes⁸⁵. Counts for these remaining genes were normalized with TMM normalization method and then transformed using voom transformation^86,87. The limma package in R was used to identify genes differentially expressed between the infected and the uninfected groups⁸⁸. For differential expression analysis, corrected p-values < 0.05 were considered to be significant. Gene Set Enrichment Analysis (GSEA) was performed to compare SOD1^G93A IAV infected to their uninfected counterparts, respectively, using C2 canonical pathways and C5 GO collections⁸⁹. Enrichment with corrected p-value < 0.1 was considered to be significant. GSEA results were shown using Enrichment Maps in Cytoscape environment, and barplots created in R⁹⁰. Additional results were visualized in PCA plot using rgl package in R.

Transfection

HEK293T cells were transfected using the Lipofectamine 3000 Reagent (Thermo Fisher Scientific) for 24 h with 2.5 µg of SOD1 wildtype (SOD1^WT), SOD1^G93A, and Empty Vector (EV) PCI-Neo plasmids as per the manufacturer’s instructions.

Western blotting

Protein was isolated using RIPA buffer (Thermo Fisher Scientific) and quantified using the Pierce BCA Protein Assay Kit (ThermoFisher Scientific). 25 µg of protein was loaded onto a 4–12% Bis-Tris gel (ThermoFisher Scientific) and ran at 100 V for 110 min at RT with 1X Bolt MOPS SDS running buffer (Invitrogen). Protein was transferred onto a nitrocellulose membrane using 1X Bolt transfer buffer (ThermoFisher Scientific), containing 20% methanol, at 100 V for 60 min at 4 °C. The membrane was blocked in 5% skim milk in Tris-buffered saline (TBS) for 1 h and then incubated with mouse anti-HA Tag (1:1000; ThermoFisher Scientific) and mouse anti-GAPDH (1:2000; ProteinTech) antibodies, diluted in 5% milk in TBS at 4 °C overnight. Membranes were then washed 5× with TBS with 0.05% Tween (TBS) and subsequently incubated with goat anti-mouse IRDye 800CW IgG secondary antibody (LI-COR Biosciences) diluted 1:3000 in 5% skim milk in TBS for 1 h at RT. Following incubation, the membranes were washed 5× with TBS. Membranes were imaged using the Odyssey CLx (LI-COR Biosciences).

Statistics and reproducibility

All data and figures were generated, and statistical analyses were performed using GraphPad Prism (GraphPad Software v.10., La Jolla, CA, USA). Statistical tests that were performed are presented in the figure legends. No data were excluded in this study. Animals were randomised based on their genotypes to experimental groups. Blinded analysis was performed, where applicable.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Conceptualization, A.M., J.M., B.C. and M.S.M.; methodology, A.M., J.M., I.A., B.C. and M.S.M.; formal analysis, A.M., J.M., I.A., B.C., A.D.G., S.A. and M.S.M.; investigation, A.M., J.M., I.A., B.C., D.C., M.R.D, S.A., J.C.A., A.T.C, V.S., Y.T., H.L., K.W., A.Z., H.D.S., M.L., Y.K., K.M. and M.S.M.; writing, A.M., S.A., I.A., J.M. and M.S.M.; supervision, M.S.M.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

M.S.M. was supported, in part, by funding from ALS Canada, the Natural Sciences and Engineering Research Council of Canada (NSERC), the M.G. DeGroote Institute for Infectious Disease Research, and by a Canada Research Chair in Viral Pandemics. A.M. was supported by an Ontario Graduate Scholarship and an ALS Canada-Brain Canada Trainee Award.

Data availability

The data generated in this study are provided in the Source Data file. The RNASeq data used in this study are available in the GEO database under accession code GSE342036. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77353-y.