Non-respiratory sites of influenza-associated disease: mechanisms and experimental systems for continued study

Abstract

The productive replication of human influenza viruses is almost exclusively restricted to cells in the respiratory tract. However, a key aspect of the host response to viral infection is the production of inflammatory cytokines and chemokines that are not similarly tissue restricted. As such, circulating inflammatory mediators, as well as the resulting activated immune cells, can induce damage throughout the body, particularly in individuals with underlying conditions. As a result, more holistic experimental approaches are required to fully understand the pathogenesis and scope of influenza virus-induced disease. This review summarizes what is known about some of the most well-appreciated non-respiratory tract sites of influenza virus-induced disease, including: neurological, cardiovascular, gastrointestinal, muscular, and fetal developmental phenotypes. In the context of this discussion, we describe the in vivo experimental systems currently being used to study non-respiratory symptoms. Finally, we highlight important future questions and potential models that can be used for a more complete understanding of influenza virus-induced disease.

Introduction

Each year, seasonal influenza epidemics are correlated with hospitalizations for severe respiratory illnesses like pneumonia and acute respiratory distress syndrome (ARDS) [1], as well as heart attacks and strokes [2], miscarriages [3] and febrile seizures [4–6]. In an effort to prevent these well-recognized influenza-associated negative health outcomes, at-risk groups such as pregnant individuals or people with cardiovascular disease are encouraged to receive annual influenza vaccines [7,8]. However, with national vaccination rates remaining around 50% in the United States [9] and influenza-associated hospitalizations numbering in the hundreds of thousands each year [10], it is critical to understand the effects of the systemic inflammatory state that accompanies respiratory influenza virus infections.

Human influenza viral infections begin with the inhalation of viral particles that enter and replicate in the upper respiratory tract (URT) epithelium. For seasonal human influenza viral strains, entry requires the viral surface glycoprotein hemagglutinin (HA) to bind to sialic acids present on respiratory epithelial cells along with cleavage of HA by host cell proteases to facilitate membrane fusion [11]. These trypsin-like proteases are mainly expressed in airway tissues, restricting influenza viral tissue tropism to the respiratory tract [12]. Within days of exposure, the infected person will typically begin to experience respiratory symptoms such as a sore throat, sneezing, cough and congestion [13]. During most infections, these symptoms and the underlying viral replication are resolved within a week of symptom onset. However, initial influenza URT infections can progress to bronchitis or pneumonia, placing patients at higher risk for further, more serious complications. Even in cases of severe disease with seasonal strains of human influenza viruses, infectious viral particles and viral RNA are primarily detected in the respiratory tract, and productive infection is rare in non-respiratory tissues [14].

While influenza virus replication is typically restricted to human airways, some of the hallmarks of “influenza-like” illness are systemic symptoms, such as fever, chills, body aches and fatigue. These common non-respiratory manifestations of disease are recognized to be a result of circulating cytokines produced as a part of the immune response [15]. In patients with underlying conditions, or in cases where an uncontrolled inflammatory response leads to a “cytokine storm”, systemic inflammatory signaling results in significant damage to, and disease associated with, organs outside the site of infection [16,17]. Studies observing patients with confirmed influenza cases or during experimental influenza virus challenge studies have revealed an array of circulating cytokines that are consistently detected during human influenza virus infections, including interferons (IFNs), IL-6, and TNF-α [18–20]. Studies in experimental models have helped demonstrate how these individual cytokines interact with immune cells and uninfected tissues to enact an inflammatory antiviral state throughout the body [15]. Nonetheless, a complete picture of how a respiratory-restricted influenza infection leads to specific negative health outcomes in an uninfected organ requires a model system with two key characteristics: 1) The model system must support influenza viral infection in the respiratory tract; 2) The model system must recapitulate the non-respiratory influenza disease symptoms.

A wide range of animal models for the study of influenza virus disease have been used [21–23]. Despite several differences with human disease [23,24], mice have been disproportionately used to study the mechanisms of respiratory disease pathogenesis. During a typical influenza viral infection in mice, the virus primarily infects the lungs (rather than the URT), leading to pneumonia and observable evidence of infection including weight loss due to anorexia, hunching and huddling indicating discomfort, and hypothermia (rather than hyperthermia) [25,26]. However, like humans, the virus is usually cleared within 10 days of infection. A key advantage of mice is their utility for replicating the non-respiratory aspects of influenza-associated disease, particularly when studying underlying conditions. Mouse models for many neurological, cardiovascular, gastrointestinal, and muscular diseases and conditions exist [27–30]. Further, there is a growing research in modeling conditions that confer high-risk during influenza virus infections (i.e. comorbidities), such as obesity, pregnancy, aging, and chronic diseases, in mice [31]. Thus, although murine influenza infections are not entirely reflective of the human experience, studies using mice are the basis for much of what we know regarding the mechanisms of pathology and immunology related to influenza disease.

This review aims to cover previous research into the mechanisms behind well-described, non-respiratory disease associated with influenza virus infections, including neurological, cardiovascular, gastrointestinal, muscular, and fetal developmental phenotypes (Figure 1). These studies will be used to illustrate what is known about how influenza virus-induced inflammation intersects with other organ systems and underlying conditions to produce negative health outcomes. In the final section of the review, we discuss important future research directions and emerging experimental systems that may be leveraged to further illuminate the effectors of non-respiratory influenza-associated disease and lead to a more complete understanding of the scope of the disease.

Figure 1.

Effects of influenza viral disease outside sites of viral replication and effectors identified using animal models. Influenza virus is a respiratory infection that rarely disseminates throughout the body, while its effects can include neurological, cardiovascular, gastrointestinal, muscular, and fetal disease. Animal model studies, primarily using mice, have identified candidate inflammatory cytokines and organ-specific factors associated with damage at non-respiratory sites during influenza virus infection. Created with BioRender.com.

Neurological Effects and Effectors

Influenza A viruses (IAVs) are thought to induce most of the total influenza disease and are broadly separated into neurotropic or non-neurotropic strains. “Neurotropic” IAVs in humans, often highly pathogenic avian influenza viruses, though rare, have the ability to enter and replicate within the central nervous system (CNS) [32]. In contrast, the more common “non-neurotropic” IAVs, whether pandemic or seasonal strains, generally fail to invade the nervous system while successfully replicating throughout the respiratory tract. Despite their accepted inability to replicate and spread within the nervous system, non-neurotropic IAVs have been implicated in neurological disease.

Observations of neurological disease associated with IAV infection were first broadly recognized following the 1918 pandemic [33]. The 1918 H1N1 strain has also been controversially implicated in the contemporaneous epidemic of the neurologic disease von Economo’s encephalitis lethargica [34,35]. More recently, an increase in neurological symptoms, ranging from headaches to comas, during severe IAV disease was observed during the 2009 H1N1 pandemic [36]. However, even seasonal IAVs can result in significant neurological effects during serious infections, particularly among pediatric populations [37]. Interestingly, despite the severity of the neurological disease associated with IAVs, these so-called “non-neurotropic” viruses are only rarely detected within the central nervous systems of symptomatic patients [38], although this could be attributed to limited sample collection and time of sample collection (often post-mortem). It is also important to note that transient or extremely low levels of viral replication below the limit of detection could occur within the CNS, therefore contributing to neurological disease manifestations while virus remains “undetected” within these tissues. Whether “non-neurotropic” influenza viruses contact the CNS or not, it is well-accepted that the host immune response, whether fever or inflammation in the form of circulating cytokines or immune cells, contributes to influenza-associated neurological disease.

Influenza-associated febrile seizures and encephalopathy

Today, the most commonly observed severe neurological complications of influenza, primarily reported in young children, are febrile seizures and encephalopathy [37]. Nearly one-fifth of children hospitalized with influenza experience febrile seizures [39], which occur in young children with fevers above 38°C (100.4°F) due the vulnerability of their incompletely developed central nervous systems [40]. Influenza viral infections in mice do not result in hyperthermia, so febrile seizures alone have primarily been studied using other methods (notably the “hair dryer model” and “heated chamber model”) to raise body temperatures and induce seizures in young rodents [41]; however, these methods of study do not take into account the influence of inflammation on the development of febrile seizures. In contrast to mice, ferrets do experience increased body temperature in response to influenza viral infections [42,43], although influenza-associated seizures are predominately observed with neuroinvasive H5N1 viruses [44,45].

Influenza-associated encephalopathy is typified by altered consciousness and frequently leads to death or lasting neurological sequalae [37]. Influenza-associated encephalopathy (IAE), and its most severe form acute necrotizing encephalopathy (ANE), occurs at higher rates in East Asian populations compared to children of other ancestries, indicating a genetic link [17]. The identification of genetic predisposition for these conditions has led to clinical studies demonstrating the importance of the carnitine palmitoyl transferase II [46], ADORA2A [47], and RANBP2 [48] genes in IAE and ANE. Mouse models of IAE require lipopolysaccharide (LPS) administration in addition to influenza viral infection to replicate the hypercytokinemia (TNF-α, IL-6) and blood-brain barrier (BBB) permeability found in human patients [49,50]. In 2019, Imakita et al. used a murine model of IAE to identify a role for Caveolin-1 in regulating BBB permeability in the context of encephalopathy (Table 1) [51].

Table 1.

Key Animal Models of Neurological Disease with Respiratory Influenza Virus Infection

Disease Model	Animal Model	Influenza Strain (and treatment)	Observed Effects	Proposed Effectors or Mechanisms	Ref
Influenza-associated encephalopathy (IAE)	Neonatal ICR mice	H3N2 A/Aichi/2/68 followed by LPS injection	Brain edema Compromised BBB	Cytokines in plasma (TNF-α, MCP-1, IL-6, IL-10)	[49]
	BALB/c mice	H1N1 A/Puerto Rico/8/34 followed by i.v. LPS	Brain edema Compromised BBB	Increased brain cytokine expression levels (Tnfa, Il6, Il1b, Ifna4, Ifnb1, Ifng) Decreased brain Caveolin-1 levels	[51]
Experimental autoimmune encephalitis (EAE) model for multiple sclerosis (MS)	C57BL/6J mice	H1N1 A/Puerto Rico/8/34 followed by injection of MOG35–55 peptide	Exacerbated EAE Increased demyelination	Increased IFN-γ and TNF-α from mononuclear cells in CNS post-flu Increased CCL5 expression and CCR5+ T cells in lung and spinal cord during EAE	[53]
	2D2 mice C57BL/6J mice	H1N1 A/Puerto Rico/8/34	In 2D2: Tail weakness and unilateral paralysis Demyelination	In BL6: Gene upregulation associated with glial activation, IFN signaling, and MAPK signaling in CNS Sera IL-6 and IFN-γ Brain CXCL5 and immune cell recruitment	[55]
Respiratory influenza infection	BALB/c mice	H1N1 A/Puerto Rice/8/34	Cognitive dysfunction Hippocampal neuron architectural changes Microglial activity	Increased cytokines (IL-1β, IL-6, TNF-α, IFN-α) Decreased neurotrophic factors (BDNF, NGF) Decreased immunomodulatory factors (CD200, CX3CL1)	[64]
	C57BL/6J mice	H1N1 (pandemic) A/California/04/2009	Microglial activity	Reduced neurotrophic factor expression (bdnf, gdnf) Increased immune modulatory ligand (ccl4)	[66]
	C57BL/6J mice	H3N2 mouse-adapted A/Hong Kong/1/68	Cognitive dysfunction Spine loss in the hippocampus Microglial activity Compromised BBB	Increased cytokines (TNF-α) Increased inflammatory and immune gene expression Decreased neurotrophic gene expression (Bdnf, Ntf3)	[67]
	Ferrets	H1N1(pandemic) A/Netherlands/602/2009	No observed nervous symptoms	Detected virus in cerebrum and cerebellum Upregulated IL-8 in cerebrum	[43] [73]

Exacerbation of neuroimmune conditions with influenza viral infection

Influenza virus infection is also recognized to exacerbate or initiate certain neuroimmune conditions, such as Guillen-Barré syndrome, multiple sclerosis, Parkinson’s disease, and Alzheimer’s disease. These findings are connected to a larger hypothesis that inflammation associated with repeated, acute viral diseases could in turn lead to chronic, neurodegenerative diseases [52]. Most in vivo research investigating the connection between influenza viral infections and exacerbated neuroimmune conditions have relied on mouse models of experimental autoimmune encephalitis (EAE), which is meant to replicate human multiple sclerosis (MS) flare-ups by targeting T cells to myelin oligodendrocyte glycoprotein (MOG), a component of the myelin sheath (Table 1). In 2017, Chen et al. first infected C57BL/6 mice with an H1N1 virus then allowed the mice to recover (50 days) before inducing EAE, finding mice previously exposed to influenza virus were unable to recover from EAE due to increased T cell infiltration and cytokine secretion (TNF-α, IFN-γ) into the CNS as a result of a potentially autoreactive T cells residing in the lung following infection [53]. In 2017, Blackmore et al. used the autoimmune 2D2 mouse model, transgenic mice that express T cells specific to MOG [54], to investigate the potential for influenza virus infection to induce a “flare-up” of multiple sclerosis [55]. Using this model, they found the non-neurotropic H1N1 virus in mouse lungs, but not the cerebellum or spinal cord; they did observe evidence of EAE, demyelination and inflammation in the CNS tissues and an upregulation of genes associated with glial activation and indicate a role for elevated chemokines in the CNS including CXCL5 and IFN-γ. In contrast, in 2017, Glenn et al. immunized mice against MOG [56] to induce autoimmune disease before infecting with influenza virus [57]. Using this model for MS and influenza virus co-morbidity, Glenn et al. did not find influenza virus infection caused an EAE flare-up, and instead the mortality associated with this autoimmune-infection model led to uncontrolled viral replication in the lungs due to myeloid-derived suppressor cell (MDSC) recruitment to the lungs limiting essential immune responses.

Although these three models of the interaction between MS and influenza virus utilize different experimental set ups and present somewhat conflicting results, each emphasize the increased mortality and risk respiratory viral infections pose to individuals with neuroimmune conditions. Further, and also in 2017, Sadasivan et al. tested the hypothesis that influenza infection could exacerbate neuroinflammatory damage in the context of Parkinson’s disease through systemic inflammation by infecting mice with pandemic H1N1 followed by (30 days) administration of the parkinsonian toxin (MPTP), finding a synergistic loss of dopaminergic neurons compared to mice either infected or treated with MPTP alone [58]. Similarly, in 2021, Hosseini et al. found a transgenic mouse model of Alzheimer’s disease (APP/PS1) showed more pronounced measures of cognitive impairment and amyloid plaques follow non-neurotropic H3N2 infection [59].

Microglial involvement in neurological effects of influenza

Mechanistically, animal model studies of neurological phenotypes during non-neurotropic IAV infections have nearly universally identified a role for microglial cell activation (Table 1) [60]. Microglia are macrophage-like cells in the central nervous system that mediate inflammatory responses within the brain [61]. During viral disease, microglia may be activated by pathogen-associated molecular patterns (PAMPs), damage-associated molecular patterns (DAMPs), and cytokines leading to morphological and transcriptional changes that facilitate cytokine release and other functional changes like antigen presentation [62]. In particular, the long term effects of viral infection on microglia have been implicated in brain aging and disease [63].

In 2012, Jurgens et al. infected BALB/C mice with a non-neurotropic H1N1 IAV and despite a lack of virus in the CNS, observed behavioral changes as measured using the Morris water maze, increased inflammatory cytokine expression in the hippocampus, and indicators of microglial activation [64]. The same group found these negative neurological effects as a result of IAV infection could be mediated by environmental enrichment [65]. In 2015, Sadasivan et al. infected C57BL/6J mice with a pandemic (but non-neurotropic) H1N1 virus and made similar observations about the activation of microglia, in addition to finding increases in expression of inflammatory cytokines and decreased neurotropic factors [66]. In 2018, Hosseini, et al. followed the long-term effects of influenza infection on neurological phenotypes by comparing the effects of a neurotropic H7N7 virus and a non-neurotropic H3N2 virus. They reported more extreme phenotypes with the neuroinvasive virus, but again observed similar changes in the hippocampus gene expression and microglial activation status with the non-neurotropic H3N2 virus. Further, these phenotypes fully resolved within 120 days post infection [67]. In 2021, Bantle et al. exposed young mice to manganese, a neurotoxin, and found with H1N1 infection in adulthood the exposed mice showed increased microglial activation, demonstrating how environmental factors can interact with viral infection to enhance neurological damage [68]. These studies together show that general observations about the neurological effects of non-neurotropic influenza viral infections are observable in mouse models at the level of behavioral changes, transcriptional changes within the brain, and microglial activation.

Ferrets as a model of influenza-associated neurological disease

The prominence of CNS involvement means the gold standard animal model for influenza, ferrets, have also been employed to understand how this respiratory virus impacts the brain. Influenza-induced neurological disease in ferrets has been frequently studied in the context of neuroinvasive viruses like highly pathogenic avian influenza H5N1, which readily invades and replicates within the CNS in both humans and animals [69]. In ferrets, these neuroinvasive and often systemic [70,71], H5N1 infections result in observable neurological phenotypes, while less pathogenic H1N1, pandemic H1N1 and H3N2 infections failed to induce neurological disease [43,72]. However, a follow-up study by Short et al. in 2017 measuring the induction of proinflammatory genes in the CNS of pandemic H1N1 and H5N1 infected ferrets found significant IL-8 expression in the olfactory bulb and cerebrum [73]. Further, in 2018 Sun et al. performed swine-origin H3N2 infections in ferrets and revealed inconsistent and low levels of virus within the brain and olfactory bulb, indicating these viruses were likely not neurotropic but were able to make contact with the CNS [74]. These findings show ferrets are another promising model of influenza-associated neurological effects.

Despite the general inability of non-neurotropic influenza viruses to enter the CNS, evidence from both mouse and ferret models of infection indicate the CNS is still negatively impacted by the systemic inflammation that accompanies respiratory infections. Although influenza infection alone induced behavioral changes indicative of cognitive dysfunction and neuroinflammation at the level of microglial activation, combining viral infection with additional conditions to replicate various neurological diseases resulted in extensive inflammatory damage to the CNS. Further, mouse models of influenza vaccination have also shown vaccination itself is often protective against neurological disease [75]. These findings along with epidemiological studies demonstrate the potential benefits of influenza vaccination for individuals with or at risk for ANE [76], MS [77], and other neurological conditions.

Cardiovascular Effects and Effectors

The close physical proximity and shared role in gas exchange of the respiratory and cardiovascular systems has made influenza virus infection of cardiovascular tissues an obvious area of investigation. Indeed, avian influenzas have been reported to be cardiotropic in birds but the detection of influenza virus in mammalian cardiac tissues is relatively rare [78] and mainly observed in the context of H5N1 highly pathogenic avian influenza infections [79,80]. Recently, it was found that IFITM3 is a key restriction factor preventing replication of influenza virus in the heart using an IFITM3 KO mouse model of infection, perhaps explaining why the virus only rarely accesses those tissues [81]. However, the hearts of mice fed a high fat diet had weaker antiviral responses and allowed higher viral replication in cardiac tissues during H1N1 infection compared to mice fed a low fat diet [82], identifying a potential risk factor for direct infection of the heart by influenza virus. Irrespective of the role of active cardiac infection by influenza virus, strong correlations suggest this respiratory infection leads to cardiac involvement in as many as 10% of influenza virus infections [83]. Acute influenza-associated cardiac events are especially prevalent in patients with pre-existing cardiovascular disease and include heart attacks and failure [84]. Due to strong support for the role of systemic inflammation in cardiovascular disease following respiratory viral infection [85], a number of studies have modeled these effects within mice and other animal models absent direct infection of cardiac tissues.

Cholesterol, atherosclerosis, and influenza infection

Cardiovascular disease is complex, and in animal models, influenza virus infections have been shown to exacerbate many aspects of the disease from cholesterol to plaques to coagulation (Table 2). High density lipoprotein (HDL) is well-recognized to reduce risk of coronary artery disease with roles in cholesterol transport and antioxidant and anti-inflammatory properties [86]. In 2001, Van Lenten et al. found that during H1N1 virus infection of C57BL/6 mice, while there was no evidence of viremia, HDL lost its anti-inflammatory properties and had a reduced ability to prevent low density lipoprotein (LDL) oxidation [87]; oxidized LDL is recognized to have atherogenic properties [88]. Following this work, Van Lenten et al. hypothesized treatment with a peptide mimetic of the main protein in HDL ApoA-I (D-4F) would combat the loss of HDL anti-inflammatory activity during influenza virus infection [89], particularly in LDL receptor deficient mice which develop hypercholesteremia and atherosclerotic plaques [90]. Indeed, the group found that although influenza infection led to arterial macrophage trafficking, which is associated with plaque instability, D-4F treatment was protective and reduced IL-6 lung and plasma levels.

Table 2.

Key Animal Models of Cardiovascular Disease with Respiratory Influenza Virus Infection

Disease Model	Animal Model	Influenza Strain (and treatment)	Observed Effects	Proposed Effectors or Mechanisms	Ref
Respiratory influenza infection	C57BL/6J mice	H1N1 A/WSN/1933	HDL had reduced anti-inflammatory properties	HDL lost PON and PAF-AH activity Increased plasma IL-6 and SAA	[87]
	C57BL/6N mice MLKL KO mice	H1N1(pandemic) A/California/04/2009 H1N1 A/Puerto Rico/8/34	Changes in cardiac proteome and phosphoproteome	IAV particles in myocardium Necroptosis deficiency ameliorated cardiac damage	[98]
Atherosclerosis	LDLR-null mice	H1N1 A/WSN/1933 followed by D-4F injection	Treatment prevented LDL increase Treatment prevented macrophage recruitment to arterial arch	Treatment reduced plasma IL-6 Treatment increased HDL PON activity	[89]
	ApoE−/− mice	H3N2 A/Hong Kong/1/68	Increased atherosclerotic plaque lesions	Increased MMP-13 expression from infiltrating macrophages	[93]
Atherothrombosis	C57BL/6J mice TMpro/pro mice PAI-1−/− mice	H1N1 A/Puerto Rico/8/34	Increased intravascular thrombi	Increased thrombin Reduced levels of protein C	[94]
	Ferrets	H3N2 A/Netherlands/177/2008 H1N1 (pandemic) A/Netherlands/602/2009 H5N1 A/Indonesia/5/2005	Procoagulant state	Increased prothrombin time and activated partial thromboplastin time Increased D-dimer levels	[95]

Other mouse models of atherosclerosis in combination with influenza viral infection have made similar conclusions, that respiratory infection resulted in systemic inflammation able to reduce atherosclerotic plaque stability, which in turn increases the risk of life-threatening thrombosis and ischemia [91]. In 2003, Naghavi et al. used the apoE−/− mouse model of atherosclerosis during influenza viral infection and found atherosclerotic plaques were subject to increased inflammation and fibrosis [92]. In 2020, Lee et al. found that in apoE−/− mice influenza infection increased MMP-13 expression from macrophages in the aorta, destabilizing atherosclerotic plaques [93].

Coagulation, atherothrombosis, and influenza infection

Research has also identified a role for influenza infection in modulating coagulation. In 2006, Keller et al. investigated the effects of influenza virus infection on thrombosis and, using a variety of mouse models, concluded influenza virus infection induced a prothrombotic state through increases in thrombin [94]. In 2014, Goeijenbier et al. used a ferret model of influenza virus infection and found similar pro-fibrotic and pro-coagulation phenotypes as observed in mouse models [95]. This association of severe influenza infection with pathogenic coagulation [96] is potentially relevant due to the prevalence of coagulation disorders in COVID-19 patients [97].

Influenza viral particles at or in cardiac tissue

Finally, influenza virus itself may play a more direct role, independent of productive cardiotropic infection, in cardiac disease. In 2021, Lin et al. infected mice with a pandemic H1N1 virus and found that although influenza viral particles were present in the heart after lung viral clearance, these particles did not appear to undergo productive replication but were able to alter the cardiac proteome [98]. They found eliminating the necroptosis pathway using MLKL KO mice reduced the harmful effects of influenza virus particles in cardiac tissues. Similarly, pandemic H1N1 influenza virus has been detected in the hearts of infected ferrets, although this did not appear to result in proinflammatory cytokine expression in cardiac tissues [43,73].

Thus, it has long been recognized that a systemic inflammatory state likely contributes to cardiovascular disease. Seasonal correlations of cardiac events with influenza viral infections build a strong epidemiological basis for this idea. Further, animal models of infection have shown that the immune response during influenza infections does create an environment conducive to atherosclerotic plaque rupture and thrombus formation. Furthermore, in 2011 Bermúdez-Fajardo & Oviedo-Orta demonstrated that vaccinating the apoE−/− mouse model of artherosclerosis against influenza virus promoted plaque stability [99]. Underlying conditions place patients at an even greater risk for cardiac complications of influenza infection. In 2021, Sinclair et al. employed a mouse model of type I diabetes and found influenza infection resulted in significant damage to the heart, without observing the same effect in healthy infected mice [100].With increasing evidence that respiratory viral infections play a role in exacerbating cardiovascular pathology, annual influenza vaccination is strongly recommended for individuals with cardiovascular and related disease [101].

Gastrointestinal and Metabolic Effects and Effectors

Forms of gastrointestinal “distress” during human influenza viral infections are relatively common [102]. Gastrointestinal symptoms of influenza virus infection like vomiting and diarrhea are not generally severe, although influenza-associated appendicitis, hemorrhagic gastritis and liver damage have been observed [17,102]. While influenza virus infections in birds are primarily gastrointestinal in nature [103], influenza viral RNA is only sometimes detected in the gastrointestinal tract of mammals and active viral replication in these tissues is controversial [43,102,104–106]. In order to understand how these alterations to the gastrointestinal tract and liver can occur without appreciable direct viral infection, a number of studies have illustrated that virally induced inflammation can alter the intestinal microbiome and overall metabolism in an infected person (Table 3) [107].

Table 3.

Key Animal Models of Gastrointestinal Disease with Respiratory Influenza Virus Infection

Disease Model	Animal Model	Influenza Strain (and treatment)	Observed Effects	Proposed Effectors or Mechanisms	Ref
Salmonella-induced colitis	C57BL/6J mice Ifnar1−/− mice	H1N1 A/Puerto Rico/8/34 followed by S. Typhimurium infection	Decreased obligate anaerobic bacteria Increased Proteobacteria including E. coli Secondary Salmonella intestinal colonization and systemic dissemination	Lung produced type I IFNs: Inhibit gut antimicrobial responses (Ifng, S100A9, Lcn2) Inhibit gut inflammatory responses (Il6, Cxcl2)	[109]
	C57BL/6J mice	H5N1 A/Viet Nam/1203/2004 followed by S. Typhimurium infection	Reduction in intestinal microbiota mass and diversity Decreased Bacteroidetes Small intestine mucosal tissue damage Increased susceptibility to S. typhimurium infection	IFN-α independent mechanism	[110]
	C57BL/6J mice	H3N2 A/Scotland/20/1974 followed by S. Typhimurium infection	Compromised intestinal barrier Inflammation in the liver Increased susceptibility to S. typhimurium infection	Decreased production of short-chain fatty acids	[111]
Respiratory influenza infection	C57BL/6J mice IFN-γ−/− mice Tcrd−/− mice IL-17A−/− mice	H1N1 A/Puerto Rico/8/34	Increased Enterobacteriaceae including E.coli Decreased Lactobacillus/Lactococcus Shortened colon Mild diarrhea	Lung-derived T cells recruited to the small intestine produced IFN-γ Microbiota changes increase IL-17A production, causing intestinal injury	[112]
	BALB/c mice	H1N1 (pandemic) A/Eng/195/2009	Altered gut microbiota diversity Decreased Firmicutes	Increased airway and colon Muc5ac levels	[114]
	BALB/c mice specific pathogen free	H1N1 (pandemic) A/Eng/195/2009	Reduced food consumption	CD8+ T cells induce inappetence resulting in changes to microbiome consumption	[115]

Pathogenic enteric bacterial infections following respiratory influenza viral infections

It is well recognized that viral respiratory infections can lead to secondary respiratory bacterial infections; however, observations that influenza infections alter the gut microbiome have led some researchers to investigate the potential for “secondary” enteric bacterial infections [108]. In 2016, Deriu et al. infected WT and Ifnar KO mice with PR8 and observed increases in Proteobacteria (including E. coli) in the infected WT, but not Ifnar KO mice [109], indicating this effect is dependent on type I IFN. The group also observed increases in Enterobacteriaceae, and subsequently investigated whether influenza infection could impact the course of a mouse model of acute colitis associated with Salmonella infection. Infection with S. Typhimurium in previously virally infected mice led to increased bacterial colonization and further systemic dissemination. In 2018, Yildiz et al. also found influenza viral infection altered gut microbial communities, although treatment with IFN-α was not able to replicate the effect despite interferon stimulated gene expression in the intestine [110], indicating type I IFN is not sufficient to explain changes in the microbiome during viral infection. However, they did find influenza viral infection increased susceptibility to subsequent S. Typhimurium infection.

In 2021, Sencio et al. identified a role for short-chain fatty acids (SCFAs), a metabolic byproduct of gut microbiota, in protecting against S. Typhimurium secondary infections following respiratory influenza infection in mice [111]. Despite no detection of influenza viral RNA in the intestine, they observed an inflammatory gut transcriptional profile in addition to reduced SCFA levels during respiratory infection; accordingly, supplementation with SCFAs during influenza infection partially mitigated subsequent colonization by S. Typhimurium. This research indicates respiratory viral infections, like influenza, could leave patients more vulnerable to pathogenic intestinal bacteria.

Respiratory influenza infections and alterations to the gut microbiome

More general effects of respiratory influenza viral infection on the gut microbiome have also been observed. In 2014, Wang et al. infected mice with PR8 and observed intestinal changes and diarrhea independent of viral detection within the intestinal tract [112]. The group concluded these damaging effects were due to increases in E. coli within the intestine that led to Th17 cell increases. They supported a model where lung-derived T cells migrated to the intestine, secreted IFN-γ to disrupt the microbiome, subsequently inducing IL-15 secretion that led to Th17 cell polarization. In 2017, Bartley et al. explored the effects aging could have on the microbiome in the context of influenza infection [113]. Again, this group found increases in Proteobacteria during influenza infection no matter the age group. In 2018, Groves et al. also observed changes in the murine microbiome during respiratory virus infection, notably a decrease in Firmicutes, implicating increased Muc5ac levels in both the lung and gut [114].

However, in 2020, Groves et al. concluded that the main reason for the observed effects of respiratory viral infections (IAV and respiratory syncytial virus - RSV) on the microbiome were due to inappetence (reduced food consumption) of the mice during infection, with increases in TNF and CD8+ T cells during RSV infection leading to the infected mouse’s reduced food intake [115]. In further support, Sencio et al. found mice with restricted food intake mimicked the weight loss observed during an influenza viral infection and had similar gut bacterial profiles to influenza-infected mice [116]. These recent findings concerning the effect of dramatic weight loss in mice during influenza infections put to question whether mouse models of infection are appropriate models for how respiratory viral infections impact the human gut, although patient samples have similarly revealed increases in Escherichia and decreases in Firmicutes during influenza infections [117–119]. In further support, the shift in fecal microbiome observed during influenza-like illness in patients, including increased abundance of Escherichia, has been proposed as a biomarker for respiratory disease [120].

While the significance of virus-induced microbiome alterations remains incompletely understood, Zhang et al. showed changes in the murine gut microbiome following influenza infection can ultimately be protective [121]. Specifically, during H7N9 infection they observed increases in Bifidobacterium animalis in mice that survived, and when naive mice were then treated with either the fecal microbiota of survivor mice or Bifidobacterium animalis they became more resistant to viral infections. Thus, understanding the effects of virally-altered microbiome compositions during subsequent disease processes is an important area of future study.

Alterations to metabolism during influenza viral infections

As evidenced by the general impact of weight loss on the gut microbiome, the systemic effects of respiratory influenza viral infection, whether cytokine circulation or other metabolic changes, impact organs outside their source. In general, influenza infection is known to impact glucose and lipid metabolism across tissues, with the induction of interferons indicated as the main mediators in this metabolic shift [122]. In 2020, Ohno et al. found that mice infected with influenza virus had reduced tricarboxylic acid cycle substrate levels, dysregulated insulin signaling, and impaired fatty acid metabolism, implicating that the liver, and its role in metabolism, is impacted during infection [123]. IL-6 and IFNs induced during influenza infection have been shown to impact gene expression in the liver [124], and changes in the liver’s ability to respond to oxidative stress during influenza infection, independent of virus at the liver, have long been recognized [125]. Interestingly, Sencio et al. identified disruption of the intestinal barrier during respiratory influenza infection as a potential source of liver damage, due to release of bacteria, that could explain some of these large shifts in metabolism during infection [111].

It is well appreciated that our microbiomes are often reflective of our overall health [126], so it is no surprise that respiratory viral infections can throw off the balance of bacteria within the gut and leave individuals vulnerable to subsequent disease. Even further, recognition of the lung-gut axis, the concept that these microbiome-supporting organ systems are especially interconnected, has led to increasing research on the role of gut microbiota in defense against respiratory infections [127,128]. For example, the reduction in gut microbiome SCFA production during influenza viral infection was found to leave mice more vulnerable to secondary pulmonary S. pneumoniae infection [116]. Further, mice with a gut microbiome derived from wild mice, compared to laboratory mice, were protected from severe disease during respiratory influenza infection [129]. Interestingly, the studies in this review describing the impacts of viral infection on gut bacteria did not utilize any models of underlying gastrointestinal disease to understand the impact of pre-existing gut dysbiosis during respiratory viral infection. With respect to co-morbidities, it is highly recommended that patients with irritable bowel disease (IBD) are vaccinated against influenza because they are at increased risk of contracting and being hospitalized for influenza virus infections [130] in addition to the frequent use of immunomodulators to mediate IBD. There is a general understanding that an influenza infection could lead to a “flare-up” of IBD symptoms [131,132] though this connection appears mostly uninvestigated at the epidemiological or experimental level. Further work is needed to determine the exact interplay between influenza infections and gastrointestinal disease, both independent of severe inappetence and dependent on pre-existing conditions.

Muscular Effects and Effectors

One of the most recognized symptoms of “influenza-like” disease is mild muscle aches, or myalgia. Compared to some other organ systems where the role of replicating virus in disease is debated, there is limited evidence of influenza viruses actively infecting myocytes. Instead, investigations into the mechanism behind muscle complications during influenza virus infections have predominantly focused on indirect effects of the immune response. Influenza viral infections are generally associated with mild effects on the muscles; however, more detrimental effects on the musculoskeletal system have also been reported. Breakdown of muscle, or rhabdomyolysis, has been reported with influenza virus infection in more than 20 cases [17], while pediatric cases of influenza-associated-acute myositis present as calf muscle pain resulting in difficulty walking that resolves within a few days [133,134]. For elderly patients experiencing severe influenza virus complications, such as pneumonia and ARDS, there are also concerns about long term effects of muscle weakness following viral infection and hospitalization [135].

Primarily, mouse models reporting the effects of respiratory influenza associated myopathy have identified circulating IL-6 as a potential effector through increasing muscular atrogin-1 expression (Table 4). In 2019, Radigan et al. infected mice with H1N1 virus and proposed a model where increased IL-6 during infection induced FoxO3 expression, through STAT3, to further induce muscle atrophy F-box protein (atrogin-1) resulting in muscle protein degradation [136]. In 2016, Bartley et al. also found a role for IL-6 in differentiating effects of H1N1 infection in young and aged mice, focusing on impacts on mobility and muscles. The group found reductions in voluntary activity and alterations in gait during infection in both groups, with the aged mouse losses lingering longer. Similarly, increased inflammatory (Il6, Tnf) and muscle-breakdown associated gene expression (atrogin-1) was detected in both groups during infection, while in aged mice expression took longer to return to base levels and recovery of muscle growth associated genes was delayed compared to younger mice [137]. These findings appear independent of active infection within mouse muscular tissues. Treatment with an antibody that prevents IL-6 signaling mitigated the effect of influenza virus on muscle phenotypes in mice. In general agreement, in 2020 Runyan et al. compared the recovery of muscle function following influenza infection in young and old mice, and also observed increases in sera IL-6 and muscular atrogin-1 [138]. The group found activity levels recovered in young mice but not old, accompanied by an expansion of muscle progenitors in young but not old mice. During infection, they observed expansions in skeletal muscle macrophages in young but not old mice, further concluding the loss of phagocytic functions within macrophages was a driver of muscle recovery losses in infected, aged mice. These reports demonstrate how acute systemic inflammation such as during an influenza viral infection can lead to damage outside the infection site resulting in long-term reductions in mobility, particularly in elderly subjects.

Table 4.

Key Animal Models of Muscular Disease with Respiratory Influenza Virus Infection

Disease Model	Animal Model	Influenza Strain (and treatment)	Observed Effects	Proposed Effectors or Mechanisms	Ref
Respiratory influenza infection	C57BL/6J mice Atrogin-1−/− mice	H1N1 A/WSN/1933	Decreased muscle weight Decreased forelimb strength	Increased muscular atrogin-1 expression Increased sera IL-6 levels IL-6 induces FoxO3a increasing atrogin-1 expression	[136]
Aging	C57BL/6J mice Young: 2.5–4 months Aged: 19–22 months	H1N1 A/Puerto Rico/8/34	Decreased mobility, more pronounced in aged mice Altered gait kinetics, more pronounced in aged mice	IL6 expression in skeletal muscle of mice at all ages TNF and CXCL10 expression in skeletal muscle of aged mice Increased expression of muscle degradation genes (FOXO1, ATROGIN1, MuRF1) Decreased expression of muscle growth genes (IGF1, PAX7, MYOD1, MYOG)	[137]
	C57BL/6J mice Young: 9–11 weeks Aged: 18 months	H1N1 A/WSN/1933	Decreased muscle weight and cross-sectional areas Aged mice fail to recover muscle function Muscle regeneration is impaired in aged mice	Increased sera IL-6 levels Increased muscular atrogin-1 expression Tissue-resident CX3CR1+ skeletal muscle macrophages expand in young mice to induce regeneration Skeletal muscle macrophages in aged mice have persistent MHCII expression	[138]
	C57BL/6J mice Young: 2.5–4 months Aged: 19–22 months	PR8 NP vaccination followed by H1N1 A/Puerto Rico/8/34	Vaccination prevented reductions in voluntary activity Vaccination prevented reductions in grip strength Vaccination prevented losses in IIB fiber cross-sectional areas	Vaccination prevented persistent sera cytokine expression (IL-6, CXCL10) in aged mice Vaccination prevents muscular MuRF-1 expression	[140]

Interestingly, in addition to the immune effects on muscles, influenza virus may impact physical attributes of muscles during respiratory infection. In 2020, Straight et al. tested whether respiratory viral infections in mice would impact the physical properties of skeletal muscles at the cellular and molecular level, finding myosin fibers showed reduced cross sectional areas and increased myofilament stiffness, demonstrating that influenza infection causes changes within muscle properties, potentially through modulation of calcium sensitivity [139].

Although muscle aches are a highly experienced influenza symptom, most infected individuals likely expect any weakness or tenderness to fade quickly along with their fever or headaches. However, these mouse studies demonstrate the potentially lingering effects respiratory infections, and their accompanying systemic inflammation, can have on muscle weakening and regeneration, particularly in older hosts. Individuals over the age of 65 are already recognized to be an at-risk group in terms of severe influenza infections and are strongly recommended to receive annual influenza vaccinations; further, in 2020 Keilich et al. demonstrated that prior influenza vaccination prevented many of the negative muscle-associated phenotypes observed in aged mice [140]. These findings demonstrate an additional reason to highly recommend influenza vaccination in elderly populations.

Fetal Effects and Effectors

During pregnancy, infections with TORCH (Toxoplasmosis, Other – syphilis, varicella-zoster, parvovirus B19, Rubella, Cyomegalovirus, and Herpes) pathogens are well-recognized to threaten fetal health through direct infection. In contrast, influenza viruses are generally thought to remain in maternal respiratory tissues during infection of pregnant women [141], although in some cases highly pathogenic avian influenza H5N1 virus has been isolated from human fetal tissue [142]. However, despite the general lack of viral replication in fetal tissues, influenza viral infections during pregnancy still place the developing fetus at risk through activation of the maternal immune response [143]. The effects of maternal influenza disease on fetal development are well documented, including increased incidence of miscarriage, low birth weight, and preterm birth [144–146]; additionally, there is an increased risk of offspring developing schizophrenia or other neurological conditions [147–149]. Animal model studies have quantified how the state of pregnancy impacts maternal immunity, generally finding increased maternal severe infections and morbidity due to altered hormonal and immune responses (reviewed by van Riel et al. [150]). Only some of these experimental animal studies, however, have focused on investigating the impacts of maternal influenza infection on fetal health (Table 5) [151,152].

Table 5.

Key Animal Models of Fetal Disease with Respiratory Influenza Virus Infection

Disease Model	Animal Model	Influenza Strain (and treatment)	Observed Effects	Proposed Effectors or Mechanisms	Ref
Pregnancy and Fetal Health	BALB/c mice E12	H1N1 A/Brisbane/59/07	Increased pre-term labor Impaired fetal growth Increased fetal mortality Decreased pup body weight Disrupted placental architecture	Decreased maternal serum progesterone Reduced maternal serum cytokine expression (IL-1β, IL-6, RANTES) Upregulated placental labor inducing hormones (PGF2α) Increased placental MMP expression Reduced placental expression of immune-responsive hormone regulators (COX-2, PIBF)	[156]
	C57BL/6J mice E12	H3N2 A/Hong Kong/1/68	Placental growth retardation Maternal major artery dysfunction	Increased placental IFN-β expression Placental and fetal brain hypoxia (Hif1a) Increased placental angiogenesis factor expression (VEGF-A, P1GF) Increased placental MMP expression IAV disseminates to artery, resulting in pro-inflammatory vascular storm	[153]
	C57BL/6J mice E10	G15 (GPER1 inhibitor) treatment with H1N1 A/Puerto Rico/8/34	GPER1 inhibition: Decreased pup weight Decreased pup numbers and viability Increased fetal reabsorption Placental endothelial degradation	GPER1 inhibition: Increased placental IFN stimulated gene expression	[157]
Pregnancy and Neurological Disease in Offspring	BALB/c mice C57BL/6J mice E9.5	H1N1 A/NWS/33CHINI	In offspring: Reduced exploratory behavior Reduced social behavior Reduced prepulse inhibition Loss of Purkinje cells in cerebellum	Non-viral immune responses to poly(I:C) resulted in similar effects so they are a result of maternal immune system rather than viral infection specifically	[163] [165]
	Rhesus monkeys Week 17 (24-week pregnancy)	H3N2 A/Sydney/5/97	In offspring: Reduced time in contact with mother Reduced cortical gray matter Enlarged lateral ventricles	No evidence of direct viral infection in neonates Offspring lateral ventricle size correlated with maternal antibody response	[166]

Impaired fetal development with maternal influenza infection

Several recent studies have begun to describe potential mechanisms behind impaired fetal development during maternal influenza infection. In 2020, Liong et al. infected pregnant mice with influenza virus on E12 and found signs of hypoxia and altered angiogenesis responses in developing fetal brains and placentas, attributed to a maternal inflammatory “vascular storm” typified by proinflammatory leukocytes in the maternal blood vessels that were observed only in the context of pregnancy and infection [153]. They also observed reduced pup and placental weights with maternal influenza infection. Interestingly, the fetal effects of maternal influenza infection and its impact on offspring immune responses have also been reported. In 2021, Jacobsen et al. showed the offspring of pregnant mice infected on E5.5 with pandemic H1N1 were more vulnerable to early life bacterial and viral infections [154]. This vulnerability was attributed to reduced alveolar macrophage functions.

Hormones such as estrogens, which are upregulated during pregnancy, have been implicated as immunomodulatory (and specifically anti-inflammatory) compounds that potentially delay viral clearance during a maternal infection in a mouse model [155]. In 2017, Littauer et al. infected pregnant mice on E12 to closely observe impacts on maternal and fetal health and correlating hormonal and immune cytokines responses [156]. The group found infected mothers produced more stillborn pups, had dysregulated levels of pregnancy hormones (lower progesterone, higher PGF2α), showed damaged placental architecture, and experienced preterm labor. Specifically, activated matrix metalloproteinases (MMPs) were indicated as mediators of placental degradation. Our group has provided evidence consistent with the immunomodulatory hormone model, showing that estrogen-regulated GPER1 is a key suppressor of fetal type I interferon responses during maternal influenza infection [157]. In experiments where pregnant mice were infected with influenza (E10) and treated with a GPER1 inhibitor, enabling harmful IFN signaling in placental tissues, pups had lower birth weights and higher occurrences of reabsorption compared to only infected or only drug-treated dams. Treatment with an IFNAR antibody to prevent interferon signaling was able to rescue the combined effects of infection and the GPER1 inhibitor. These infection-pregnancy models emphasize the roles hormones, cytokines, and maternal changes in vasculature have in regulating infection resolution and fetal and offspring health.

Neuropsychiatric disease in offspring with maternal influenza infection

In addition to overall fetal health and neonatal viability, maternal infections, including influenza infections, are thought to have an impact on the development of autism and schizophrenia [158]. In fact, influenza infections in pregnant mice are used as a maternal immune activation (MIA) model in order to study the development of neurological conditions in offspring [159,160]. Like with more general observations of impaired fetal development with maternal influenza infections, circulating cytokines, rather than maternal-to-fetal transmission, are implicated in inducing damage to the fetal brain in clinical [161] and animal model studies [162]. In 2003, Shi et al. found that infecting pregnant mice with influenza virus on E9.5 led to offspring with behavioral phenotypes associated with schizophrenia and autism [163]. The group determined this effect was due to maternal immune responses rather than direct fetal infection by treating pregnant mice with viral mimetic poly(I:C), enacting a maternal cytokine response independent of infection, and continuing to observe these same behavioral phenotypes in offspring. Shi et al. further showed influenza viral RNA could not be detected in the placenta or fetal brain during maternal infection in mice [164]. These findings led to a follow-up in 2009 by Shi et al. where they identified a loss of Purkinje cells in the cerebellum, similar to the pathology of autism, in the offspring of influenza infected mice [165]. In 2010, Short et al. infected pregnant rhesus monkeys with H3N2 A/Sydney/5/97 and found that although offspring did not exhibit many behavioral changes, there were significant reductions in cortex gray matter and parietal cortex white matter with prenatal exposure to influenza virus [166].

Much of the work using mouse models to define the impact of maternal influenza infections on fetal neurological condition development has been performed by Fatemi et al. [162], identifying changes in fetal brain gene [167,168] and protein [169] expression, brain structural alterations [170], and reduced neurotransmitter levels [171], and further indicating placental abnormalities [172] as a potential mediator of effects on the developing CNS and eventual offspring. MIA studies using non-infectious immunogens, such as poly(I:C), have implicated maternal cytokines IL-6 and IL-17 in particular as effectors of neurological disease in offspring; however, further studies are required to determine the disease threshold that can result in schizophrenia and autism [173]. For example, in 2021, Antonson et al. found that infecting pregnant mice with a moderately pathogenic H3N2 virus resulted in maternal and placental inflammatory responses and placental damage, but failed to induce inflammation in the fetal brain or elevate circulating IL-17 as expected from non-infectious MIA models [174]. This finding demonstrates a need for use of infectious agents, like influenza, in future MIA model studies to determine the levels of disease expected to impact the outcomes of maternal infections.

As a result of the high-risk influenza viral infections pose to maternal and fetal health, influenza vaccination is highly recommended during pregnancy. The protective effect of influenza vaccination has been directly demonstrated in animal pregnancy models. For example, in 2018 Wu et al. found influenza vaccination early in pregnancy prevented autism-like phenotypes in a mouse model [175]. Further, influenza vaccination has been repeatedly shown to be associated with positive postnatal outcomes for offspring in epidemiological studies [176]. Additionally, prenatal vaccination also allows maternal antibodies against influenza to be transferred to the fetus in utero and subsequently in breast milk, thus providing protection to the immunologically vulnerable neonate [177]. These studies highlighting the impact of influenza vaccination on fetal and neonatal health again underscore the importance of understanding and preventing indirect, non-respiratory, manifestations of viral disease.

Conclusions and Future Directions

Although influenza virus tropism is predominantly restricted to cells in the respiratory tract, it is common for symptoms to occur in many organ systems independent of active replication at these sites. This review has outlined many studies that, primarily using mouse models, attempted to identify the mechanism behind these phenotypes in the cardiovascular, nervous, muscular, gastrointestinal, and fetal systems. The systems explored here are not the only non-respiratory ones impacted during influenza infection; however, they are the best examples with multiple animal studies to interrogate the mechanisms behind the influenza-associated diseases. For example, the nephrotic, hepatic, endocrine and ocular systems [17,178–180] have also been observed to be negatively impacted during respiratory influenza infections. Animal models observing the effects of respiratory influenza viral infections on these organ systems could further reveal the impacts of systemic inflammation throughout the body.

Many of the studies described above used wild-type mouse models, while other groups leveraged genetically modified mouse models for specific underlying conditions, such as LDLR−/− and apoE−/− models for atherosclerosis. In the future, it is likely other specialized models (genetic or other approaches for inducing physiological perturbations) will be important to understand how infection with influenza viruses affects non-respiratory sites. For example, murine models of irritable bowel disease [181] could be useful in defining whether influenza viral infections are indeed potential triggers for flare-ups. A newly developed mouse model for Alzheimer’s disease could also be utilized to help determine the role of viral disease in the development of neurodegenerative diseases [182]. Additionally, because obesity is recognized as a risk factor that leads to more severe influenza disease; evaluation of extrapulmonary effects of respiratory viral infection obese animal models is an important area of future research [183].

Outside of mice, ferrets (Mustela furo) are another widely used animal model of influenza that better replicate some aspects of human pathogenesis compared to the mouse model [42]. Ferrets develop fevers in response to influenza viral infections [42], perhaps providing an opportunity to study extrapulmonary influenza disease associated with raised temperatures. Virus-associated seizures, which are known to occur during CNS viral infections in ferrets [45,184,185], are one such example of a hyperthermia induced phenotype difficult to model in mice. In particular, ferrets are an appropriate model for the impact of age on the response to influenza infection [186], an issue highlighted in this review by pediatric patients’ vulnerability to neurological effects of influenza and older populations greater risk for cardiac and muscular manifestations of disease. In further support of future ferret research, tools and reagents for studies in ferrets are increasingly available [187].

Additionally, other animal models less commonly used for influenza research could be particularly applicable for different systems of interest. Guinea pigs (Cavia porcellus) are another animal model with some specific attributes that could be especially useful for studying the extrapulmonary effects of influenza virus infection. Although guinea pigs become readily infected with and transmit many influenza virus strains without requiring adaption, guinea pigs do not exhibit as severe respiratory disease phenotypes as observed with ferrets or even mice [23]. This model system may be particularly impactful in better replicating milder courses of disease in humans and the corresponding effects of influenza virus outside sites of active replication. For example, guinea pigs do not tend to lose weight during influenza infection, as observed during ferret and mouse infections [23], which could be advantageous for gastrointestinal studies to remove weight loss as a variable which frequently confounds interpretations. Additionally, guinea pigs more closely model human pregnancy based on similarities in placental structure, remodeling of the maternal spiral arteries, prolonged gestation relative to mice, and advanced maturity of the offspring [188]. Pigs (Sus domesticus) also represent an animal model with some unique advantages for study of non-respiratory impacts of influenza viral infection. Pigs are a natural host of influenza and are similar to humans in many ways, from their size and anatomy to their behaviors around sleep and eating [189]. Further, pigs have already been used to study a number of underlying conditions associated with exacerbated influenza viral disease, such as cardiovascular diseases and cystic fibrosis [190]. The similarities in anatomy (guinea pigs – placenta, pigs – size and structure) and course of disease (mild) make these animal models particularly suited to studying the impact of respiratory-restricted influenza infection on non-respiratory manifestations of disease.

Regardless of the experimental model, it is clear that many non-respiratory consequences of influenza viruses, although rare, have devastating, long term, or lethal consequences, particularly for high-risk individuals [191]. Understanding the mechanisms behind these disease effects also enables the development of therapeutic strategies to prevent specific non-respiratory tract complications influenza virus infection. As of now the only approach to prevent these aspects of disease is vaccination, which is required annually, or antiviral treatment, which is only effective early after infection. This review highlighted some common phenomena associated with negative extra-respiratory effects, such as systemic IL-6 and IFNs. Additionally, we also highlighted some system specific findings, like the ability of GPER1 to protect the developing fetus from the harmful effects of IFN exposure and that D-4F treatment can prevent the loss of the anti-inflammatory properties of HDL during influenza infection [89,157]. For individuals with underlying conditions, greater knowledge of the effectors of extrapulmonary disease could lead to more targeted therapeutic strategies to prevent negative outcomes in those at high risk.

Finally, although this review focused on the mechanisms behind extrapulmonary disease associated with respiratory influenza virus infections, other respiratory viral pathogens are also appreciated to cause significant systemic disease. These have been dramatically demonstrated recently during the SARS-CoV-2 [192] pandemic, but it is also recognized to be true for endemic pathogens such as respiratory syncytial virus (RSV) [38,193–195]. Every virus has different tropisms and a differential ability to disseminate, so mechanisms underlying disease systems outside of the primary sites of viral replication are likely different. Nonetheless, the continued study of non-respiratory effects of influenza viruses and other related infections will be important not only to more fully define the true scope of the disease, but also potentially to identify new intervention strategies capable of reducing the burden of these viruses on human health.

Abbreviations

ANE

acute necrotizing encephalopathy

ARDS

acute respiratory distress syndrome

BBB

blood-brain barrier

Cal09

H1N1 A/California/04/2009

CNS

central nervous system

D-4F

apolipoprotein (apo)A-I mimetic peptide

DAMP

damage-associated molecular pattern

EAE

experimental autoimmune encephalitis

HA

hemagglutinin

HDL

high density lipoprotein

IAE

influenza-associated encephalopathy

IAV

influenza A virus

IBD

irritable bowel disease

IFN

interferon

LDL

low density lipoprotein

LPS

lipopolysaccharide

MDSC

myeloid-derived suppressor cell

MIA

maternal immune activation

MMP

metalloproteinase

MOG

myelin oligodendrocyte glycoprotein

MPTP

parkinsonian toxin; 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine

MS

multiple sclerosis

PAMP

pathogen-associated molecular pattern

PR8

H1N1 A/Puerto Rico/8/1934

RSV

respiratory syncytial virus

SCFA

short-chain fatty acid

TORCH

Toxoplasmosis, Other – syphilis, varicella-zoster, parvovirus B19, Rubella, Cyomegalovirus, and Herpes

URT

upper respiratory tract

WSN

H1N1 A/WSN/1933