Lethal Viral Sepsis/Septic Shock in Getah Virus Infected Ifnar1-/- Mice

1. Introduction

Getah virus (GETV) is a mosquito transmitted alphavirus found predominantly in South and East Asia [1,2]. GETV belongs to the Semliki Forest virus antigenic complex [3], which includes the Australasian Ross River virus (RRV) [4] and the globally distributed chikungunya virus (CHIKV) [5]. The geographic range of GETV is growing, in particular in China [2,6], with GETV contamination of a live porcine reproductive and respiratory syndrome virus vaccine potentially contributing [7]. Serosurveys suggest a range of wild and domestic ungulates are the primary targets of infection by GETV [8]. The virus has caused disease outbreaks, leading to significant economic losses, primarily in horses and pigs [1,9,10,11,12]. Although human infection with GETV is frequently claimed, the contention arises from a small number of primary references that are somewhat dated [13,14,15,16], with no recent studies emerging that use modern, and substantially more reliable, serodiagnostic methodologies [17]. To the best of our knowledge, a confirmed clinical GETV infection in humans has not been reported. Cross-reactivity with other serologically related alphaviruses can also provide false seropositive results for GETV. For instance, historical serosurveys did suggest GETV circulates in Australia; however, a series of analyses indicated that this is likely not the case, with serosurveys probably compromised by strong cross-reactivity of RRV immune sera with GETV [18].

GETV belongs to a group of Old World alphaviruses, such as RRV and CHIKV, where human symptomatic infections are usually associated with a spectrum of rheumatic disease manifestations, primarily fever, rash, myalgia and polyarthritis/polyarthralgia [4]. Symptomatic GETV infections in horses are also associated with fever, rash, hind limb oedema (mainly fetlock), as well as lymph node swelling lasting for 1–2 weeks [19]. GETV infection, like RRV [20], can affect racehorse performance, with a whole-virus formalin-inactivated combined vaccine for Japanese encephalitis virus (JEV) and GETV (JE/GETV vaccine) commercially available in Japan (Nisseiken Co., Ltd., Tokyo) for use in racehorses [3,21,22,23].

In contrast to GETV infection in horses, infection in neonatal and suckling piglets is often associated with mortality, preceded by a series of symptoms including fever, loss of appetite, diarrhoea, ataxia and cyanosis [11,24,25]. Increased virulence with 100% mortality [26,27] has recently been reported, and was associated primarily with genotype III viruses [6,25,28,29]. Fatal infections are characterized by a constellation of histopathological lesions across multiple organ systems. In lungs GETV infection caused widened alveolar septa, epithelial desquamation, alveolar-wall thickening, and intra-alveolar hemorrhage, with leukocyte infiltration [11,24,27,30]. In spleen, hemorrhagic splenitis was observed, characterized by reduced white-pulp volume, hemorrhage, and inflammatory-cell infiltration [11]. In liver, leukocyte infiltration, hepatocellular degeneration and hemorrhage was reported [11]. In brain, mild neuronal degeneration and vascular inflammatory-cell infiltration was seen [11]. In kidneys, histopathology included inflammatory infiltrates with erythrocyte extravasation, glomerular vasodilation, tubular epithelial swelling, and irregular tubular lumina [11]. For intestines, villous epithelial injury with inflammatory lesions was reported by several groups [11,27,30]. These findings indicate an acute, severe multi-organ inflammatory immunopathology.

The GETV viremia levels in adult pigs are relatively low, peaking at ≈4 log₁₀CCID₅₀/mL [31], similar to adult wild-type mice [3]. In newborn piglets the blood virus load would appear to be much higher, up to ≈7 log₁₀ viral copies per ml [30], and (as noted above) is often associated with mortality. Paralysis and mortality are well described features of alphavirus infections in neonatal mice, including GETV infections, with a permeable blood brain barrier often leading to brain infections and paralysis [32,33]. A key reason why neonates readily succumb to alphaviral infections is their immature/defective type I IFN responses [32,34,35,36,37]. The early and rapid innate type I IFN response represents a key mechanism that protects adult mice from excessive viremias and mortality [38,39,40,41,42,43].

A range of efforts are underway to develop new GETV vaccines [22,44,45], antivirals [46,47], and diagnostics [48,49]. We thus sought herein to characterize a mouse model of GETV infection that recapitulates the lethal infection seen in young piglets. The use of neonatal piglets and neonatal mice for evaluating new interventions is relatively expensive, ethically challenging and often inconvenient, as pregnant animals give birth to variable numbers of offspring, often with timing that is difficult to synchronise. We therefore characterized adult, type I IFN receptor 1 knock-out (Ifnar1-/-) mice as a potential model of lethal GETV infection, as these animals recapitulate the defective type I IFN responses often observed in neonates [32,34,35,36,37]. They also represent a widely available and widely used, well characterized, genetically modified mouse strain [42]. Cells in Ifnar1-/- mice are unable to respond to cytokine signalling mediated by IFNβ and a series of IFNαs [50], with these latter type I IFNs strongly induced by alphavirus infections in wild-type mice [43,51]. Infection of Ifnar1-/- mice (or other type I IFN response deficient mice) with most pathogenic alphaviruses, including GETV [52,53], generally results in high viremias, a surge in pro-inflammatory cytokines, and mortality within a few days [39,40,41,42,54]. To provide a detailed characterization of the cytokine responses, we describe herein a RNA-Seq analysis of the spleens of GETV infected Ifnar1-/- mice. Although RNA-Seq data has been reported for orthoflavivirus infected Ifnar1-/- mice [55,56], to the best of our knowledge, RNA-Seq data from alphavirus infected Ifnar1-/- mice has not previously been reported, and revealed surprisingly robust type I IFN responses despite the absence of a functional type I IFN receptor.

2. Materials and Methods

2.1. Ethics and Regulatory Statement

Mouse work was conducted in accordance with the “Australian code for the care and use of animals for scientific purposes” as defined by the National Health and Medical Research Council of Australia. Mouse work was approved by the QIMR Berghofer Animal Ethics Committee (P3746, A2108-612). Breeding and use of GMO mice were approved by the QIMRB Safety Committee under a Notifiable Low Risk Dealing (NLRD) Identifier: GTSC_075_2025: NLRD1.1(a). Research with live GETV was approved under the Queensland Biosecurity Act 2014, Scientific Research (prohibited matter)—Permit no. PRID000907. This legislation declares GETV to be an exotic virus [18], with specific containment and reporting procedures required to work with this virus in Queensland. All research was approved by the QIMR Berghofer Institutional Safety Committee.

2.2. The GETV Isolate, Culture and Titration

GETV_MM2021 (GenBank MN849355) was obtained from the Doherty Virus collection, located at QIMR Berghofer. GETV_MM2021 is the original 1955 prototype genotype I GETV isolate from Malaysia imported into Australia in the 1960s [18]. The virus was propagated in C6/36 cells and the titre determined using CCID₅₀ assays as described [22].

2.3. Mouse Infection and Monitoring

Ifnar1-/- mice on a C57BL/6J background [57] were originally provided by Dr P. Hertzog (Monash University, Melbourne, Australia) and were bred in-house at QIMRB [52,58]. These mice do not express a functional Ifnar1 protein and are thus unable to signal via the IFNα/β receptor. Details of mouse agistment conditions have been described previously [22], with sibling male mice housed together and male mice from different litters housed separately. Mice were injected subcutaneously on the top/side of each hind foot [41] with 40 µL per foot (see [3] for details), with either tissue culture supernatant containing infectious GETV or with heat inactivated (56 °C for 1 h) tissue culture supernatant containing inactivated GETV (Control). Mice were weighed and monitored daily [59] and were euthanized using CO₂ when ethically defined endpoints were reached. Weight loss reaching or approaching 20% was the main criterion triggering the ethical requirement for euthanasia.

2.4. Histology

Histology was undertaken as described [22,60] and was undertaken using a cohort of female mice. In brief, tissues were fixed in 10% formalin and embedded in paraffin (Sigma-Aldrich, Missouri, USA). Sections were stained with hematoxylin and eosin (H&E) (Sigma-Aldrich). Slides were scanned using Aperio AT Turbo (Aperio, Vista, CA, USA) and were viewed using Aperio ImageScope (v12.4.0.5043). Scanned slides were examined by a European board-certified veterinary pathologist using QuPath software v0.5.1 [61].

2.5. RNA-Seq and Bioinformatics

RNA-Seq was undertaken as described [62,63,64] using a cohort of male mice. In brief, spleens were cut open and placed in RNALater (Invitrogen) and RNA extracted using TRIzol (Invitrogen). The RNA Integrity Number for one sample was too low for further processing leaving GETV infected n = 5 and Controls n = 4. Extracted RNA cDNA libraries were generated using Illumina TruSeq Stranded mRNA library prep kit (polyA enrichment). Sequencing was undertaken using the Illumina Nextseq 2000 platform to generate 75 bp paired end reads. STAR aligner was used to align the RNA-Seq reads to a combined mouse (GRCm39, version M27) and GETV_MM2021 (GenBank MN849355) reference genome. A standard filter of 0.25 counts per million was applied. Differentially expressed genes (DEGs) were identified using EdgeR using a false discovery rate (q) cut-off of q < 0.01 and fold change > 2 (i.e., log₂ > 1). These latter filters were applied, as described previously for CHIKV infections [51,57], to reduce DEG numbers to approach the range “for best results” from Ingenuity Pathway Analysis (IPA) [65]. The latter was performed using IPA, v84978992 (QIAGEN) [64].

2.6. Statistics

Mouse data: the t-test (with Welch’s correction) was used if the difference in variances was <4 fold. The t test significance and variance were determined using Microsoft Excel. Otherwise, the non-parametric Kolmogorov-Smirnov exact test was performed using GraphPad Prism 10.

3. Results

3.1. GETV Virus Infection of Ifnar1-/- Mice, Weight Loss, Viremia and Haematocrit

GETV infection of adult male Ifnar1-/- mice resulted in rapid weight loss, which reached or approached the ethically defined end point for euthanasia of ≈20% weight loss within 4 days (Figure 1a). The viremias rose very rapidly in GETV infected mice reaching 8–9 log₁₀CCID₅₀/mL of serum on days 2–4 post infection (Figure 1b). Similar data were obtained for adult female Ifnar1-/- mice (Figure S1).

Increases in haematocrits have previously been reported in mice defective for type I IFN responses infected with CHIKV [40]. Increases in haematocrits are also well established for dengue virus infections in Ifnar1-/- mice [66] and in humans [67]. To the best of our knowledge, haematocrit levels have not been reported for lethal GETV infections in piglets. Haematocrit levels in GETV infected Ifnar1-/- mice were thus assessed when they had reached ethically defined endpoints. Haematocrit levels were quite variable, with no overall significant differences (n = 5 per group) emerging. However, an increased haematocrit was seen in 3/5 animals euthanized on day 3 or 4, with this selected subset (n = 3) significantly higher than Controls (n = 5) (Figure 1c). An increasing haematocrit is a sign of haemo-concentration that can be due to diarrhoea (see below), and can also be a feature of shock or pre-shock states [68], with cytokine storms (see below) able to induce vascular permeability [69,70,71]. These 3 mice also showed elevated levels of endogenous serum lactate dehydrogenase (Figure S2), a non-specific marker of general tissue damage and also an indicator of disease severity in sepsis patients [72].

3.2. GETV Virus Infection in Ifnar1-/- Mice Was Associated with Overt Diarrhoea

GETV infected mice that had reached the ethically defined endpoints on day 3 or 4 post infection were euthanized (Figure 1a,b), and at the same time mice that had been inoculated with heat inactivated GETV (Controls) were also euthanized. Intestines were dissected out and photographed (Figure 2). Consistent with the animal house observations of overt diarrhoea in the GETV infected mice, only a limited amount of faecal matter was visible in the intestines of these animals. In contrast, Control animals showed normal levels of faecal matter in the small intestine, cecum and colon (Figure 2).

Diarrhoea is a non-specific sign/symptom, but is well documented in GETV infected piglets [6,11,26,27,28,30]. Diarrhoea can also occur during acute infections with other alphaviruses [73], especially in infants and young children [4,74,75].

3.3. Splenomegaly in GETV Infected Ifnar1-/- Mice

Spleens from GETV infected Ifnar1-/- mice (euthanized when they reached ethically defined endpoints on days 3 or 4) and Control mice (euthanized at the same time points) were fixed, embedded in paraffin and longitudinal median sections stained using H&E. Stained sections were scanned (Figure 3a) and areas determined using Aperio ImageScope (Figure 3b). Overt and significant spleen enlargement (splenomegaly) was seen in GETV infected animals (Figure 3). Splenic enlargement was reported for piglets experimentally infected with GETV [11]. Reports of splenomegaly associated with infections by other Old World alphaviruses are, to the best of our knowledge, limited to mayaro virus infection of weanling Balb/c mice [76]. Splenomegaly is, however, a feature of dengue virus infections in children [77].

3.4. Histopathological Findings in GETV Infected Mice

Tissues from euthanized GETV infected and Control mice were fixed, embedded in paraffin, sections stained with H&E and examined by a European board-certified veterinary pathologist.

Lungs. The lungs of GETV infected Ifnar1-/- mice show (i) capillary damage with leakage (Figure 4a, white oval), (ii) leukostasis (Figure 4a, white arrowheads), (iii) fibrinous pleurisy (Figure 4a, black rectangle), and fibrin in alveoli (Figure 4a, black arrow). Although not specifically mentioned, these features do align with histopathology seen in GETV infected piglets [11,24,27,30]; (i) haemorrhage and leukocyte infiltration in alveolar cavities, aligns with damaged capillaries (Figure 4a, white oval) allowing blood into alveoli), (ii) widening septa and wall thickening, aligns with inflammatory cells (Figure 4a, white arrowheads) causing swelling and wall thickening, and (iii) epithelial desquamation (sloughing of damaged cells), can cause leakage of clotting factors (Figure 4a, black arrow).

Spleen. In the T cell areas and marginal zone of the white pulp of the spleen, areas of focal necrosis were observed (Figure 4b, white ovals), with a diffuse increase in the number of cells undergoing degeneration (perinuclear haloing) (Figure 4b, arrow). Similar features were seen in lymph nodes. Descriptions of histopathological lesions in spleens of GETV infected piglets are limited [27], although one study provided a diagnosis of haemorrhagic splenitis; reduced white-pulp volume, haemorrhage and inflammatory cell infiltration on day 4 post infection [11]. Cellular degeneration (necrosis and apoptosis) generally occurs before white pulp atrophy, haemorrhage and inflammatory cell infiltration, with the latter associated with the clearance of dead cells. So splenic manifestations would appear to be more advanced in piglets than in these mice, although they may reflect the same underlying disease processes.

Liver. The liver of GETV infected Ifnar1-/- mice displayed intravascular leukocytosis (Figure 4c, white ovals) and leukostasis (Figure 4c, white arrowheads); compare with blood vessels in Control liver (Figure 4c, Control). Focal infiltration of leukocytes in sinusoidal spaces (blood vessels) and inflammatory cell infiltration in central veins [11], was reported for GETV infected piglets and aligns well with the observations in mice. However, the hepatocellular degeneration and haemorrhage reported in piglets [11] was not a finding in mice.

Brain. Focal mild inflammatory cells were present in the meninges of GETV infected Ifnar1-/- mice (Figure 4d, white arrowheads). This aligns well with the mild vascular inflammatory cell infiltrate reported for GETV infected piglets [11]. However, the mild neuron degeneration seen in piglets [11] was not a significant observation in GETV-infectedmice.

No significant GETV infection-associated histopathological lesions were identified in kidneys, intestine or hind feet [40,78]. In summary, histopathological lesion in mice thus appear to be less severe overall than those seen in piglets. Nevertheless, histopathology from both species describe severe, acute, multi-organ inflammatory immunopathology.

3.5. RNA-Seq and Bioinformatic Analyses of Spleens from GETV Infected vs. Control Ifnar1-/- Mice

At euthanasia spleens were harvested from GETV infected mice and at the same time (day 3 or 4) spleens were also harvested from mice inoculated with heat-inactivated GETV (Controls). Spleens were analysed by RNA-Seq and the full data set is provided in Table S1. Viral read counts illustrated the presence of GETV RNA in the infected mice, with no viral read counts detected in the Controls (Figure 5a). The viral reads in infected mice may reflect virus in blood rather than infection of spleen cells as animals were euthanized at or close to peak viremia (Figure 1b). A Principal Component Analysis (PCA) plot shows a clear separation between infected and control mice (Figure 5b), and a volcano plot illustrates robust and significant changes in gene expression (Figure 5c).

Differentially expressed genes (DEGs) (GETV infected vs. Control; n = 2383, fold change > 2, q < 0.01) were analysed by Ingenuity Pathway Analysis (IPA), with Canonical pathways and Upstream Regulators (USRs) representing primary outputs from such analyses. The latter outputs provide much higher levels of cross-species concordance than comparisons of DEGs [79,80], and were thus the focus of our current reporting. Canonical and USR pathways are summarized in Figure 5d (full data set in Table S1). Innate sensing pathways feature prominently, as might be expected for an alphavirus infection [38,41,51,81]. The poly I:C, TLR3 and TLR7 annotations likely reflect detection of viral RNA. The TLR4 annotation may reflect detection of cell debris arising from virus-induced cytopathic effects (CPE) [82], although gene induction profiles for TLR4 and the other TLRs can overlap [83]. MyD88 is a key signal transduction adaptor for TLR signalling. The STING annotation may reflect detection of mitochondrial DNA from cells undergoing CPE [84].

Perhaps surprisingly for Ifnar-/- mice, robust type I IFN signatures were evident (Figure 5d), with upregulation of many IFNα mRNAs also seen (Figure S3). These type I IFN responses and signatures have also been reported in wild-type mice and humans and are believed to be central for controlling acute alphavirus infections [38,41,51,81,85,86,87]. Alphavirus infections in Ifnar-/- mice generally show high viremias and these mice generally succumb to acute infection [41,42]. This is also the case for GETV infection of Ifnar-/- mice (Figure 1a) [52], despite these robust type I IFN signatures (see Discussion).

3.6. mRNA Expression Signatures in Spleens of GETV Infected Ifnar1-/- Mice Indicate Viral Sepsis and/or Viral Septic Shock

Viral sepsis and septic shock are well described for dengue and COVID-19 and are also recognized for a number of other viral infections [88,89,90,91]. The top Canonical pathway and the top cytokine USRs (Figure 5d) suggest Ifnar1-/- mice infected with GETV succumb to viral sepsis and/or viral septic shock. TNF, IL-1 and IFNγ (Figure 5d, Cytokine storm) are considered the archetypal shock mediators [91,92,93,94], with IFNγ also recognized as a marker of hyper-inflammation [95]. Neutrophil degranulation (Figure 5d) is also a key feature of sepsis [96]. CHIKV infections can cause viral sepsis and septic shock in humans [97] and in Irf3/7-/- mice [40], although these are generally considered rare disease manifestations for Old World alphavirus infections [4,5]. A full literature search was thus undertaken, which illustrated that the majority of the cytokine/chemokines identified as top scoring USRs, have been reported to be elevated during sepsis and/or septic shock (Table 1). The bacterial sepsis and septic shock cytokine/chemokine literature was included as this is considerably larger and more expansive than the viral sepsis and septic shock literature, with the two (bacterial and viral) showing considerable overlap in upregulated cytokine/chemokine responses (Table 1).

In summary, RNA-Seq of spleens from euthanized GETV infected and Control Ifnar1-/- mice, and ensuing bioinformatic analyses, revealed a pattern of cytokine/chemokine signatures that illustrated a cytokine storm response (Figure 5d), which was strongly indicative of viral sepsis and viral septic shock (Table 1).

4. Discussion

The studies reported herein illustrate a series of parallels between GETV infected Ifnar1-/- mice and GETV infected piglets. As reported previously [52], the high viremias and lethality seen in GETV infected Ifnar1-/- mice (Figure 1a,c), recapitulates the high viremias (up to 7 logs per ml [30]) and lethality seen in GETV infected piglets. Splenomegaly (Figure 3) has also been reported for GETV infections in piglets [11]. The diarrhoea observed in GETV infected Ifnar1-/- mice (Figure 2) is also seen in GETV infected piglets [6,11,26,27,28,30], with diarrhoea a common non-specific symptom that can occur during inter alia sepsis and septic shock [68]. Overall the histopathological findings indicate that GETV infection-associated lesions are more severe in piglets than in Ifnar1-/- mice, which may reflect the evolutionary adaptation of GETV preferentially to infect ungulates and/or may arise from Australian ethical requirements to euthanize mice before they show unacceptable levels of overt disease. Other factors may also contribute to the lower disease severity observed in Ifnar1-/- mice. Detailed histopathology for several organs in GETV infected piglets relies on one reference [11], with all the reported histopathology (see Introduction) involving contemporary genotype III viruses, whereas the MM2021 strain used herein is genotype I. There are also inherent limitations in comparing histopathology in two distantly related mammalian species; mice (superorder Euarchontoglires, to which humans also belong) and pigs (superorder Laurasiatheria). Nevertheless, histopathological findings for the two species are both indicative of a severe, acute, systemic, multi-organ, inflammatory pathology [206,207,208]. The latter, and especially white pulp injury and focal lymphoid death in spleen (Figure 4b), are all consistent with severe sepsis and/or septic shock [209,210].

The RNA-Seq and bioinformatic analyses of cytokine and chemokine responses in GETV infected Ifnar1-/- mice show cytokine storm signatures that further support a diagnosis of sepsis/septic shock (Table 1). To the best of our knowledge, neither protein- or RNA-based cytokine/chemokine studies have been published for lethal GETV infections in piglets. Only RNA-Seq analyses of in vitro infections of a porcine cell line have been reported [211]. A diagnosis of viral sepsis and/or septic shock have also not (to the best of our knowledge) been formally enunciated or established for GETV infected piglets in the publicly available literature. Nevertheless, clinical observations associated with lethal GETV infections in piglets [29,30,212] are consistent with viral sepsis and/or septic shock. For instance, cyanosis is well described for piglets with lethal GETV infections [25,27,213], and cyanosis is a late stage sign of advanced sepsis or septic shock [68]. Unfortunately, cyanosis is difficult to observe in black mice, which were used herein; specifically Ifnar1-/- mice on a C57BL/6J background [57].

Conceivably, GETV infection leads to bacterial or endotoxin translocation from the gut into the circulation (with ensuing endotoxemia-mediated sepsis/shock). To the best of our knowledge, this has never been reported for alphavirus infections, although intestinal damage has been observed by histology for CHIKV infections [214]. No gut histopathology was observed in our study, with diarrhoea often associated with dysregulated ion and water transport in the gut, rather than histologically detectable mucosal injury [215]. Instead, sepsis-associated cytokines can mediate such dysregulated transport in gut epithelial cells [216,217,218], which can then lead to diarrhoea.

The type I IFN receptor is involved in the promotion of bacterial sepsis/septic shock in a number of settings [189,219,220,221]. A similar role for the type I IFN receptor in promoting viral sepsis/shock has not been reported, consistent with the key anti-viral (including anti-alphaviral [38]) role played by type I IFN responses. Infections in mice deficient in type I IFN responses often result in high viremias and acute lethality [42], with the data presented herein supporting previous reports [40,222] that this can lead to viral sepsis/septic shock.

A perhaps surprising observation is the robust type I IFN responses observed in GETV infected Ifnar1-/- mouse spleens (Figure 5d, Figure S3). Why do these responses fail to protect the Ifnar1-/- mice from the overwhelming lethal viremia? Cells infected with GETV would be able to generate type I IFN responses via cytoplasmic sensors [43], although these intracellular responses are generally insufficiently fast, robust and/or effective to achieve substantial curtailment of alphaviral replication already initiated in the infected cells [223,224]. Leukocytes can also generate type I IFN responses after TLR-mediated detection of cellular debris and viral RNA released from cells undergoing CPE [225,226]. The type I IFN receptor mediates a rapid feed-forward loop that rapidly amplifies secretion of IFNαs [227,228], a mechanism clearly absent in Ifnar1-/- mice. Nevertheless, consistent with the data presented in Figure S3, IFNαs can still be secreted in vivo in the absence of the type I IFN receptor [229,230,231]. However, the key issue in Ifnar1-/- mice is that circulating IFNβ and IFNαs, in the absence of the type I IFN receptor, are unable to protect uninfected cells from rapidly spreading viral infections via induction of anti-viral states [232,233]. Thus, even though robust type I IFN responses are seen in whole spleen tissues, GETV (and other viruses) can spread rapidly in Ifnar1-/- mice because IFNα/β, without the type I IFN receptor, is unable to induce anti-viral states in uninfected cells [39,234].

Understanding the aforementioned mechanisms may have relevance for the field of oncolytic virotherapy, with GETV being considered as a potential treatment for glioblastoma [235]. Viral infection and killing of cancer cells relies on antiviral resistance being induced in healthy cells (via infection-induced circulating IFNα/β), whereas antiviral resistance would need to be less effectively induced in cancer cells [236] to allow the spread of viral infection through the tumour. Thus, predicting effective GETV virotherapy might focus on tumours that are unable to generate an antiviral state after IFNα exposure. Type I IFN responses induced within GETV-infected glioblastoma cells are perhaps less informative for selecting patients with tumours that might be susceptible to virotherapy [209].

Perhaps perplexing is that genetically identical mice, all inoculated with essentially the same dose of GETV, can show such large mouse-to-mouse variation, for instance, with respect to viral loads (≈2 logs, Figure 1b, Figure 4a and Figure S1), haematocrit changes (Figure 1c) and serum lactate dehydrogenases levels (Figure S2). Age differences between adult mice (beyond 6–8 weeks) may (8 vs. 20 weeks of age [237]) or may not (6 to 12 weeks of age [60]) have a significant influence on alphavirus infection and/or disease in mice. However, infection outcomes for co-caged isogenic mice can actually differ because of social and behaviour factors. Both male and female mice establish social hierarchies [238], wherein subordinate status has been associated with higher basal cortisol [239], which can affect host resistance to viral infection [240,241,242]. In addition, mice held at standard animal house temperatures suffer from a degree of cold stress that can exacerbate viral infections [59,242,243,244]. Huddling can elevate body temperatures [245,246], with the centre of the huddle potentially providing a thermal advantage [246,247]. Social hierarchies are also established in piglets early post-partum, with a higher position improving suckling and weight gain [248]. Enhanced early piglet growth has been shown to positively influence innate immunity [249].

In summary, we characterize herein GETV infection in Ifnar1-/- mice and propose this represents a suitable model of GETV infection in piglets. However, the key finding of viral sepsis/septic shock has yet to be established for lethal GETV infections in piglets, although disease manifestations and histopathology in piglets are consistent with such a diagnosis.

5. Conclusions

GETV infections in Ifnar1-/- mice recapitulate many of the features seen for GETV infections of piglets. Infections with other Old World alphavirus are primarily associated with self-resolving inflammatory arthropathy [4], which is not recapitulated in Ifnar1-/- mouse models [250]. Even if sepsis/septic shock is confirmed in lethal GETV infection of piglets, treatment might be viewed as difficult and impractical. Nevertheless, supportive care and intravenous (or intra osseous) fluids may help, given its effective use for treating dengue shock in children [210]. Vaccination [3,22] of sows is also an option, with protective antibodies [3,23] transferred to the neonates via the colostrum.