Treatment with human antibodies partially protected ferrets from EBOV

Two groups of ten ferrets each were inoculated with a target dose of 1000 TCID₅₀ EBOV via either the IN or IM route, after which four animals in each group were treated with 30 mg/kg each of two human mAbs on days 2 and 5 post-infection (“early” treatment) or days 3 and 6 (“late” treatment) (S1A and S1B Fig and S1 Table). As controls, two ferrets from each inoculation group were treated with phosphate buffered saline (PBS) on 2 and 5 days post-infection (DPI). All four control animals developed severe disease and succumbed or were euthanized on day 5 (IM) and 6 (IN) post-infection, as expected (Fig 1A). Three of the IM-inoculated and treated animals also developed severe disease, with one early-treated animal requiring euthanasia on 6 DPI and two late-treated animals requiring euthanasia on 7 and 8 DPI (Fig 1A). None of the 13 animals remaining at 8 DPI had displayed any significant signs of disease up to that point, and all were gaining or maintaining their body weight (Fig 1B). On 12 DPI, one animal (ID# 653M) in the IM-inoculated and late-treated group began to show signs of disease that progressed rapidly, with the animal meeting euthanasia criteria later the same day. Between 15 and 18 DPI, five more animals succumbed to disease that appeared suddenly and progressed rapidly: two animals that were IM-inoculated and treated early, one animal that was IM-inoculated and treated late, and two animals that were IN-inoculated and treated late. A total of seven animals survived until the end of the study on 29 DPI, including all four animals that were IN-inoculated and treated early. None of the survivors exhibited any pronounced signs of disease at any point after virus inoculation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Human mAb provides limited protection from Ebola virus disease in ferrets.

Ferrets were inoculated with EBOV variant Makona-C07 and treated with a human mAb cocktail in two independent experiments. In both the first (A, B) and second (C, D) experiment, ferrets were inoculated with a target dose of 1000 TCID₅₀ EBOV via the intranasal (IN; n = 10/experiment) or intramuscular (IM; n = 10/experiment) routes. In the first experiment, 8 ferrets from each inoculation group were treated with 30 mg/kg each antibody delivered via the intraperitoneal (IP) route “early”, on 2 and 5 days post-infection (DPI), or “late”, on 3 and 6 DPI. Two ferrets from each inoculation group served as control animals and were treated with PBS only on 2 and 5 DPI. Animals were monitored for 29 days for survival (A) and weight change (B). In the second experiment, 8 ferrets from each inoculation group were treated with 30 mg/kg each antibody delivered via the IP route “very early”, on 2 and 4 DPI, or “mid”, on 3 and 5 DPI. One ferret from the IN group served as a control animal and was treated with PBS only on 2 and 4 DPI. Animals were monitored for 36 days for survival (C) and weight change (D). On all graphs, the vertical area shaded grey, from 12 to 18 DPI, represents the window in which atypical disease was observed.

https://doi.org/10.1371/journal.ppat.1013916.g001

We repeated the study and decreased the time between antibody doses from two days to one. Thus, animals were treated with 30 mg/kg of each human mAb on days 2 and 4 post-infection (“very early” treatment) or days 3 and 5 (“mid” treatment) (S1C and S1D Fig and S1 Table). A single control animal, which was IN-inoculated and treated with PBS on 2 and 4 DPI, succumbed to disease on 5 DPI (Fig 1C). Four mid-treated animals—three that were IM-inoculated and one that was IN-inoculated—developed severe disease and were euthanized on 5, 6, and 8 DPI (Fig 1C). Of the 12 remaining animals on 8 DPI, all but one appeared clinically normal, with no pronounced signs of disease observed up to that point and weights increasing (Fig 1D). One animal (752M) from the IN-inoculated, mid-treated group showed signs of disease beginning around 4 DPI and remained ill until meeting euthanasia criteria on 11 DPI (Fig 1C). As in the first experiment, four animals developed a rapid-onset disease late post-infection and were euthanized between 14 and 18 DPI: one IN- and two IM-inoculated animals treated very early and one IN-inoculated animal that received the mid treatment. A total of seven animals survived to study end, which was extended to 36 DPI for this study. Six of these animals showed no significant signs of disease throughout the experiment, while one animal (977F) in the IM-inoculated, mid-treated group became acutely ill between 3 and 6 DPI, but recovered completely by 7 DPI.

Of all the animals evaluated at different inoculation routes and doses, fewer than half survived (Tables 1 and S1). However, we did note a significant difference in survival between animals inoculated via the IN route and animals inoculated via the IM route: 10 of the 16 (62.5%) IN-inoculated animals survived, whereas only 4 of the 16 (25%) IM-inoculated animals survived (S2A Fig). Indeed, of the IN-inoculated animals treated early or very early, only one succumbed, suggesting that IN inoculation followed by early treatment may be the most effective way to demonstrate antibody efficacy in the ferret model. When the data were stratified by sex, no significant difference in survival was observed (S2B Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Outcomes in mAb-treated ferrets.

https://doi.org/10.1371/journal.ppat.1013916.t001

Among all 18 antibody-treated, non-surviving animals, only 7 (~40%) succumbed to disease on day 8 or sooner, around the same time that the 5 untreated control animals died and consistent with previous results using the ferret model [21]. One animal (752M) died on day 11, but considering that it first showed signs of disease on 4 DPI that persisted until euthanasia, this animal likely experienced a protracted form of acute disease. The remaining 10 of the 18 (~60%) treated, non-surviving animals succumbed to disease between 12 and 18 DPI, with 7 animals succumbing as late as 16–18 DPI, despite showing no or very mild signs of disease early after inoculation. All 10 of these animals exhibited up to 15% weight loss over the day or two preceding euthanasia (Fig 1B and 1D), and a few of them exhibited elevated temperatures, although none exhibited conspicuous signs of disease until the day before—or even the day of—euthanasia. Disease in all of these animals appeared abruptly and progressed rapidly, with all animals exhibiting dramatically decreased activity. Several animals also exhibited signs consistent with neurological problems, including limited or no movement, rolling movement, lack of coordination, or the inability to right themselves. Based on clinical signs alone, it was clear that these animals did not succumb to disease that is typical for EBOV in ferrets. For clarity and concision throughout the manuscript, we refer to the disease experienced by all animals that died on 12 DPI or later as atypical EVD, and we refer to these ferrets as “atypical animals”. We use the term acute EVD to refer to the disease experienced by all animals that died on 11 DPI or earlier, and we refer to these ferrets as “acute animals”. There was no statistically significant association between disease outcome (i.e., acute, atypical, or survivor) and either inoculation route or sex (S2C and S2D Fig).

Viral RNA levels rebounded with atypical EVD

In almost all animals, viral RNA levels in the blood peaked within 2–6 DPI before decreasing substantially following mAb treatment (Fig 2A and S2 Table). Interestingly, we noted that, prior to mAb treatment (on 2 and 3 DPI), viral RNA levels were significantly lower in animals inoculated via the IN route, in line with the lower overall mortality in IN-inoculated ferrets (S2E Fig). No differences were observed over the same time period between male and female animals (S2F Fig). By 10 DPI, viral RNA was reduced to very low or undetectable levels in most remaining animals, and, by the end of the studies (29 or 36 DPI), viral RNA was no longer detectable in the blood of any of the survivors. Of the animals that experienced atypical EVD, viral RNA levels disappeared in eight of them by 14 DPI. The two animals (653M, 043F) that succumbed to atypical EVD earlier than the rest—on 12 and 14 DPI—showed drops in their viral RNA levels following mAb treatment, but the RNA was never cleared. After 14 DPI, however, viral RNA reappeared in the blood of all eight remaining atypical animals, coincident with their clinical deterioration. Levels of infectious virus (TCID₅₀/ml) largely paralleled the levels of viral RNA, albeit with a lower level of sensitivity (S2 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Levels of viral RNA and infectious virus differ among acute, atypical, and surviving ferrets.

Viral RNA loads (genome equivalents [GEQ] per millilitre) in the blood were determined over time for the acute, atypical, and surviving ferrets (A). The vertical area shaded grey, from 12 to 18 DPI, represents the window in which atypical disease was observed. Viral RNA loads (in GEQ/ml or GEQ/g) at the terminal time points are depicted separately for the blood (B), liver (D), spleen (F), and brain (H) samples, with means and standard deviations indicated. Mean levels were compared using a one-way ANOVA with Tukey’s multiple comparison test. Infectious virus loads at the terminal time points are depicted in median tissue culture infectious dose (TCID₅₀) per ml or g, with means and standard deviations indicated, for the blood (C), liver (E), spleen (G), and brain (I) samples. Mean levels were compared using a t test. ns, not significant; *, p ≤ 0.05; ***, p ≤ 0.001; ****, p ≤ 0.0001.

https://doi.org/10.1371/journal.ppat.1013916.g002

To compare viral RNA levels among the three different disease groups, we grouped all data from the terminal time points of animals that experienced acute disease, animals that experienced atypical disease, and animals that survived. This analysis revealed that, although viral RNA levels in the atypical animals increased just prior to their death, the mean viral RNA load in the blood of these animals was still significantly lower than what was observed in the acute animals (Fig 2B). A similar pattern was observed with respect to infectious virus found in the blood (Fig 2C). The majority of acute animals had 6–8 log₁₀ TCID₅₀/ml of virus at the terminal time point, although infectious virus was not recovered from two animals. Conversely, infectious virus was only recovered from a single blood sample from an atypical animal (698M), with all other samples negative. Thus, animals were dying of atypical EVD with significantly lower levels of virus in their blood than those dying of acute EVD, suggesting that other factors may be influencing disease course.

Animals dying of atypical EVD had high levels of virus in the brain

We next examined the systemic spread of the virus using a limited set of liver, spleen, and brain samples taken from most ferrets at the time of euthanasia (Fig 2D–2I). Similar to what was observed in the blood, viral RNA levels were significantly higher in the liver and spleen from the acute animals than they were from the atypical animals, suggesting that infection of these organs was much more active in the animals that died of acute disease (Fig 2D and 2F). This conclusion was supported by the isolation of infectious virus from 5 of 6 liver and spleen samples from acute animals, compared to only 1 of 4 liver and spleen samples from atypical animals (Fig 2E and 2G), although the differences in mean levels of infectious virus were not statistically significant. Lower levels of viral RNA were detected in most spleen samples assessed from the surviving animals, but no infectious virus was detected in the 6 samples that were available for virus isolation, suggesting that the presence of viral RNA was not the result of a concurrent productive infection. Strikingly, the levels of viral RNA in brain samples from the acute and atypical animals were essentially equal and significantly higher than what was observed in the surviving animals, where most animals had no detectable viral RNA in the brain (Fig 2H). These data were corroborated by the isolation of infectious virus from 5 of 6 brain samples from the acute animals and all four brain samples from the atypical animals (Fig 2I), and they suggest that both acute and atypical animals experienced high levels of virus replication in the brain.

Virus genomes were sequenced from the brain samples of six acute animals and nine atypical animals, as well as from the liver samples of nine acute animals and ten atypical animals (S3 Fig). A few synonymous and non-synonymous mutations were identified in some of the samples, but almost all mutations were present at very low frequencies and none were present only in acute or atypical animals. A non-synonymous mutation in VP24 (E180V) was identified in many of the brain and liver samples, mostly from the acute animals; however, with the exception of three animals (one acute and two atypical), the frequency of this mutation was very low. Two GP mutations (T544I, D552N) and one VP30 mutation (t9267c) were present in the inoculum virus at relatively high frequences. The frequencies of these mutations increased slightly in the atypical animals compared to the acute animals, but the relevance of this increase is unclear.

Severe meningoencephalitis, ventriculitis, and choroid plexitis in animals with atypical EVD

To further characterize the brain infection in animals with atypical EVD, we performed RNA in situ hybridization to detect EBOV genomic RNA in brain tissues. In ferrets that died of acute EVD, EBOV RNA was limited in the endothelial vessels of the brain (Fig 3A, 3B and 3E), which are primary targets of virus infection. In contrast, in animals that succumbed to atypical EVD, EBOV RNA and protein were highly abundant in the infiltrates of the meninges, choroid plexuses, ventricles, and their adjacent neuropils (Fig 3C and 3D). Viral RNA was primarily detected in the inflammatory cells rather than the endothelial vessels of the brains in these animals (Fig 3E). Histological analysis of brain tissues from the acute and atypical animals demonstrated that no detectable histopathology was observed in the brains of acute animals, but meningoencephalitis, ventriculitis, and choroid plexitis characterized by infiltrates of lymphocytes, plasma cells, and macrophages in the meninges, ventricles, and their adjacent cerebral perivascular spaces and in the choroid plexuses were detected in the atypical animals (Fig 3F–3L).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Brain histopathology in animals that died of atypical EVD.

Ferret brain tissue sections stained by RNA in situ hybridization (red; A–C) or immunohistochemistry (brown; D) from an acute ferret (A, B) and an atypical ferret (C, D). EBOV genomic RNA (red) was only detected in endothelial vessels in the brain (A, B) of the acute ferret. Both EBOV genomic RNA (red; C) and antigen (brown; D) were detected in the inflammatory cells of meninges and ventricular system of the atypical animal. Quantitative analysis of EBOV genomic RNA distribution was performed by scoring the severity of RNA ISH signal in the meninges and ventricular systems in the acute and atypical ferrets (E). Representative brain tissue sections stained with hematoxylin and eosin (H&E) from an acute ferret (F, G) and an atypical ferret (H–K), revealing meningoencephalitis (H, I), ventriculitis (J), and choroid plexitis (K) in the brains of atypical ferrets (H–K) in comparison to unnoticeable changes in the brains of acute animals (F, G). Quantitative analysis of histopathological lesions detected in the brain regions of atypical and acute ferrets, including the ventricular system, meninges, and cerebral cortex (L). Nuclei were counterstained blue with hematoxylin (A-B, F-K). ISH (E) and histopathology (L) severity scores are indicated for each analyzed animal (represented by a single dot), with means and standard deviations indicated. Mean levels were compared using unpaired t-tests (Holm-Sidak). ns, not significant; *, p ≤ 0.05.

https://doi.org/10.1371/journal.ppat.1013916.g003

Inflammation of the brain was further substantiated by transcriptomic and proteomic analyses, which revealed variable expression and abundance of genes and proteins associated with a range of biological processes (Fig 4). Of the fifty named genes with the most variable expression, many were associated with immune system processes and inflammatory responses, defence responses, and programmed cell death, as well as various cellular signalling, differentiation, and transport processes; however, none was differentially expressed only in animals that succumbed (Fig 4A). Among the fifty most variably expressed genes overall, a cluster of eleven unnamed genes with no known functions was found to be significantly upregulated only in animals that experienced atypical disease (S4 Fig). While one of these genes had no known homologues, the other ten genes had homologues in other species (50–90% similarity) and were all identified as immunoglobulin-like domain-containing proteins.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Transcriptomic and proteomic analyses of ferret brain tissue.

Brain tissue samples from a subset of ferrets (all from the second challenge experiment) were subjected to transcriptomic and proteomic analyses. A heatmap from the transcriptomic analysis shows the differential expression of the fifty named genes with the most variable expression, with log₂ fold change centered for each gene (A). A heatmap from the proteomic analysis shows the differential abundance of the fifty proteins with the highest variance among all animals, with log₂ fold change centered for each protein (B).

https://doi.org/10.1371/journal.ppat.1013916.g004

Proteomic analysis complemented the transcriptomic findings and identified many of the same biological processes (Fig 4B). The fifty most variably abundant proteins fell into two main clusters: those higher in survivors compared to non-survivors and those higher in non-survivors compared to survivors. Proteins more abundant in survivors were associated with processes related to recovery from acute illness, such as anatomical structure development, signaling, cell differentiation, adhesion and motility, and cellular transport. Conversely, proteins found more abundantly in acute or atypical animals were associated with processes involved in the immune and nervous systems, defence responses, and programmed cell death, as well as cellular signaling, differentiation, and transport processes. Notably, haptoglobin (LOC101692672), which binds free hemoglobin resulting from the lysis of red blood cells, was found in much higher abundance in the brains of acute and atypical animals than survivors, suggesting intracerebral hemorrhage and/or a break down in the blood-brain barrier [29]. Moreover, several proteins associated with nervous system processes, namely ITPR1, S100B, CALB1, and CDH23, were more abundant in the brains of atypical animals, especially 698F, in concordance with the apparent neurological nature of disease that these animals experienced.

Blood chemistry in ferrets that died of atypical disease was mostly normal

Dramatic changes in certain blood chemistry parameters are hallmarks of EVD and generally indicate organ damage caused by infection. In the ferrets that died of atypical EVD, however, the majority of these parameters were normal (Figs 5 and S5). Alanine aminotransferase (ALT), a marker of liver function, was increased in all acute animals, as well as several atypical and surviving animals, during the week following inoculation (Fig 5A). These levels decreased in atypical and surviving animals around 9 DPI and remained within their normal range for the rest of the study. Thus, at the terminal time points, the mean level of ALT present in the atypical animals was significantly lower than that in the acute animals and indistinguishable from the ALT level in the survivors (Fig 5B). Two other markers of liver function—alkaline phosphatase (ALP) and total bilirubin (TBIL)—showed a similar trend. ALP and TBIL levels were elevated in most acute animals shortly after inoculation but remained low in both the atypical and surviving animals (Fig 5C and 5E). At the terminal time points, mean levels of ALP and TBIL were significantly lower in the atypical animals compared to the acute animals, once again resembling what was observed in the surviving animals at the end of the study (Fig 5D and 5F). The same trends were seen for amylase, blood urea nitrogen (BUN), and K⁺ (S5A–S5F Fig). Albumin levels decreased immediately following inoculation in most animals, but most dramatically in those that died of acute disease (Fig 5G). Levels rebounded in both the atypical and surviving animals such that, at terminal time points, the mean level of albumin in the atypical animals was significantly higher than the mean level in the acute animals (Fig 5H). Likewise, Ca²⁺ levels in these animals followed the same general pattern (S5G and S5H Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Blood biochemistry parameters for acute, atypical, and surviving ferrets.

Levels of alanine aminotransferase (ALT) (A, B), alkaline phosphatase (ALP) (C,D), total bilirubin (E, F), albumin (G, H), globulin (I, J), glucose (K, L), and total protein (M, N) were quantified in all blood samples from all animals. Levels of each of these analytes are depicted over time (A, C, E, G, I, K, M). The vertical area shaded grey, from 12 to 18 DPI, represents the window in which atypical disease was observed, while the horizontal area shaded grey represents the normal range for each parameter. The levels of each analyte at the terminal time points are also depicted separately (B, D, F, H, J, L, N), with means and standard deviations indicated. Mean levels were compared using a one-way ANOVA with Tukey’s multiple comparison test. ns, not significant; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.

https://doi.org/10.1371/journal.ppat.1013916.g005

Not all blood chemistry parameters in the atypical animals resembled the survivors. While globulin levels increased in many animals immediately following inoculation, they remained high in the acute animals, which succumbed around the same time that globulin levels peaked (Fig 5I). Levels subsequently decreased in atypical and surviving animals, only to increase again in the atypical animals at the time of death. Thus, at the terminal time points, the levels of globulin in the acute and atypical animals were equivalent and significantly higher than what was observed in the surviving animals (Fig 5J). A similar trend, albeit less well pronounced, was observed for Na⁺ levels (S5I and S5J Fig). Conversely, the changes in glucose and total protein levels in the atypical animals differed from both the acute and surviving animals. Glucose levels increased significantly at the time that animals succumbed to atypical disease, while they remained relatively normal in both acute and surviving animals at their terminal time points (Fig 5K and 5L). Total protein levels increased gradually for all animals after inoculation, peaking with the atypical animals at their time of death and then decreasing again in the surviving animals (Fig 5M and 5N). These data further suggest that the atypical disease observed in some ferrets was distinct from acute EVD.

Hematology parameters differed in animals that died of atypical disease

Changes in the cellular composition of the blood of the atypical ferrets also differed from the changes observed in both the acute animals and survivors (Fig 6). A drop in lymphocyte numbers was observed in all animals around 3 DPI, which is typical for EVD (Fig 6A). While most of the acute animals died shortly thereafter with low numbers of lymphocytes in their blood, numbers rebounded in both the atypical and surviving animals, with some of these animals showing dramatic increases in their lymphocyte counts around 7 DPI. Lymphocyte numbers eventually returned to normal in the surviving animals, but they dropped once again in the atypical animals at the time of death. Thus, at the terminal time points, both the acute and atypical animals had similar mean lymphocyte counts that were significantly lower than what was observed in the surviving animals (Fig 6B). Conversely, neutrophil numbers trended upwards for all animals following infection, with several atypical animals showing high peak numbers between 6 and 10 DPI (Fig 6C). Numbers decreased in most surviving and atypical animals before increasing again in many atypical animals or eventually returning to normal levels in the surviving animals. As a result, at the terminal time points, atypical animals showed mean neutrophil numbers that were higher than both acute and surviving animals, although the difference between the atypical and acute levels was not statistically significant (Fig 6D). Similarly, the mean number of monocytes in the atypical animals at the terminal time points was significantly higher than that of both the acute and surviving animals (Fig 6E and 6F). Prior to the terminal time points, there was an early peak in monocyte numbers in most animals around days 2 and 3, followed by a second, more dramatic peak in some of the atypical animals around days 7–9 (Fig 6E). Monocyte numbers remained relatively high or increased around the time that the atypical animals succumbed (Fig 6F). Total white blood cell counts reflected the sum of the changes in lymphocyte, neutrophil, and monocyte counts (Fig 6G). A general decrease in numbers was evident within the first 3–6 days after inoculation, perhaps driven by concomitant drop in lymphocytes. A peak around 7 DPI in many atypical and surviving animals aligned with similar peaks in lymphocyte and neutrophil numbers; this peak was followed by a return to normal or near normal white blood cell numbers for most remaining animals. Accordingly, no differences in the mean numbers of white blood cells were observed among acute, atypical, and surviving animals at their terminal time points (Fig 6H). Platelet counts varied considerably throughout the study, although the numbers in the acute animals trended downwards at the terminal time points (Fig 6I). Platelet numbers in the atypical animals also trended downwards at or near the time of death, but the mean numbers were still significantly higher than what was observed in the acute animals (Fig 6J). The surviving animals also showed variability in their platelet counts, but they returned to the normal range at study end, when the mean level was significantly higher than both the acute and atypical animals at their endpoints (Fig 6J).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Blood cell counts for acute, atypical, and surviving ferrets.

Lymphocytes (A, B), neutrophils (C, D), monocytes (E, F), white blood cells (G, H), and platelets (I, J) were enumerated in all blood samples from all animals. Cell counts are depicted over time (A, C, E, G, I). The vertical area shaded grey, from 12 to 18 DPI, represents the window in which atypical disease was observed, while the horizontal area shaded grey represents the normal range for each cell type. Cell counts at the terminal time points are also depicted separately (B, D, F, H, J), with means and standard deviations indicated. Mean levels were compared using a one-way ANOVA with Tukey’s multiple comparison test. ns, not significant; *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.

https://doi.org/10.1371/journal.ppat.1013916.g006

Differential cytokine responses were induced during acute and atypical disease

The levels of twelve different cytokines increased in nearly all animals over the first 9–10 days following virus inoculation, with the most dramatic increases observed in the animals that died of acute EVD (Figs 7 and S6). In almost all cases, cytokine levels peaked in acute animals at their terminal time point, between 5 and 8 DPI, with IP-10, MCP-1, MIP-1β, IL-6, IL-12p40, and IL-8 reaching particularly high levels (Fig 7). The cytokine expression patterns in the atypical and surviving animals were generally similar to each other, with peak levels reached around 7–9 DPI, after which they typically declined. Interestingly, however, increases in some cytokines—namely IP-10, MCP-1, MIP-1β, IL-6, and to a lesser extent, IL-4—coincided with the presentation of atypical EVD (Fig 7). With the exceptions of MIP-1β and IL-4, the mean levels of these cytokines in the atypical animals at their terminal time points were significantly higher than what was observed in the surviving animals, although they were also significantly less than what was observed in the acute animals. These data once again demonstrate a stark difference in the disease phenotype between animals that succumbed to acute EVD and those that succumbed to atypical EVD.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. The cytokine response in acute, atypical, and surviving ferrets.

Levels of IP-10 (A, B), MCP-1 (C, D), MIP-1β (E, F), IL-6 (G, H), IL-12p40 (I, J), IL-8 (K, L), and IL-4 (M, N) were quantified in all serum samples from all animals via Luminex assay. Levels of each cytokine are depicted over time (A, C, E, G, I, K, M). The vertical area shaded grey, from 12 to 18 DPI, represents the window in which atypical disease was observed. The levels of each cytokine at the terminal time points are also depicted separately (B, D, F, H J, L, N), with means and standard deviations indicated. Mean levels were compared using a one-way ANOVA with Tukey’s multiple comparison test. ns, not significant; *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.

https://doi.org/10.1371/journal.ppat.1013916.g007

Both atypical and surviving animals mounted a humoral immune response

To understand the humoral immune response mounted by infected animals, we quantified the amount of EBOV GP-specific IgG present in the plasma of atypical and surviving animals, beginning with samples collected on 9 DPI (Fig 8A). All animals exhibited a detectable IgG response against EBOV GP, with endpoint titers ranging from 1:800 to as high as 1:25,600. In general, IgG levels increased over time, as expected based on the continued maturation of the humoral immune response. Although the geometric mean IgG endpoint titer for the survivors was significantly higher than that of the atypical animals at the terminal time points (Fig 8B), this can likely be attributed to the difference in time between when the terminal samples were collected (12–18 DPI for the atypical animals and 29 or 36 DPI for the surviving animals). Indeed, comparing the mean titer of the atypical animals at their terminal timepoints with the mean titer of the survivors at 14 DPI revealed no significant difference (Fig 8B). Thus, both atypical and surviving animals appeared to mount a humoral immune response of equivalent magnitude, making it unlikely that this response was a factor in the development of atypical EVD.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Antibody quantification in acute, atypical, and surviving ferrets.

Levels of EBOV GP-specific IgG were determined via ELISA (A, B) in all serum samples from the atypical and surviving animals, starting with samples collected on 9 DPI. Levels of antibody are depicted over time (A), with each line in the graph off-set slightly relative to the x-axis to ensure that data from individual animals could be discriminated. EBOV GP-specific IgG levels at the terminal time points, as well as at 14 DPI for the surviving animals, are also depicted separately (B), with means and standard deviations indicated. The red dotted line indicates the lower limit of quantification for this assay. Total levels of human IgG were determined via ELISA (C, D) in all serum samples from all animals. Levels of antibody are depicted over time (C). Human IgG levels at the terminal time points are also depicted separately (D), with means and standard deviations indicated. The vertical area shaded grey, from 12 DPI to 18 DPI, in (A) and (C) represents the window in which atypical disease was observed. Mean levels in (B) and (C) were compared using a one-way ANOVA with Tukey’s multiple comparison test. ns, not significant; *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001.

https://doi.org/10.1371/journal.ppat.1013916.g008

Levels of circulating human mAbs differed among acute, atypical, and surviving animals

To better understand the dynamics of human mAb concentration following administration, we used an ELISA specific for human IgG to quantify the total amount of human antibody present in all ferret plasma samples (Fig 8C). Human antibody levels peaked between 3 and 7 DPI, following administration of the antibodies between 2 and 6 DPI. By 14 DPI, human antibody levels were universally very low, and they effectively disappeared by the terminal time points in the atypical and surviving animals.

Comparing the peak levels of human antibody in the acute, atypical, and surviving animals revealed significant differences (Fig 8D). The animals that died of acute EVD exhibited very low to undetectable levels of human antibody in most samples, with a mean peak level that was significantly less than what was observed in the survivors. The mean peak level of human antibody in the animals that died of atypical disease was higher than that observed in the acute animals, which died earlier, but significantly lower than that observed in the surviving animals. These data suggest that there were differing total levels of human mAb circulating in each ferret, and that the peak levels may correlate with disease outcome. However, caution is required in interpreting these data since the concentration of human antibody detected in each serum sample is likely affected by a number of different factors, including antibody uptake following intraperitoneal injection and antibody half-life, the latter of which may be further influenced by the binding of these antibodies to the virus and host responses against the foreign human antibodies.

To determine the half-lives of the human mAbs in ferrets, we conducted a pharmacokinetic experiment in a group of uninfected animals. Ferrets were injected with 30 mg/kg of one of two human IgG1 antibodies, either 1C3 or 1C11, and serum samples were collected regularly up to 504 hours post-treatment. Levels of each antibody were then quantified using the same human IgG ELISA as described above, and the half-life of each antibody in each animal was calculated (S7 Fig). The half-lives of these antibodies proved to be strikingly short, with 1C3 ranging from 29.76 hours to 78.41 hours (mean 50.2 hours) and 1C11 ranging from 33.51 hours to 65.64 hours (mean 54.85 hours). It seems likely that such short half-lives might have contributed to the reduced efficacy of these mAbs in the ferret model.

Ethics statement

All animal work involving infectious filoviruses was conducted in the containment level 4 facility at the Canadian Science Center for Human and Animal Health (CSCHAH) in Winnipeg, Canada. All animal experiment protocols were reviewed and approved by the Animal Care Committee at the CSCHAH, in accordance with guidelines from the Canadian Council on Animal Care (CCAC). All staff working on animal experiments completed education and training programs according to the standard protocols appropriate for this level of biosafety.

Viruses and cells

EBOV variant Makona C07 (H.sapiens-wt/GIN/2014/Makona-Gueckedou-C07; GenBank accession no. KJ660347.2) was used as the challenge virus in this study. The virus stock was mycoplasma negative, sequence confirmed, and stored at -80^oC. Vero E6 cells were obtained from the American Type Culture Collection and maintained at 37^oC and 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM; HyClone) supplemented with 5% heat-inactivated fetal bovine serum (HyClone), 2 mM L-glutamine (HyClone), and 50 U/ml penicillin and 50 ug/ml streptomycin (HyClone).

Ferret challenge studies

Two independent ferret challenge experiments were performed (S1 Fig) with domestic ferrets (Mustela putorius furo) purchased from Marshall BioResources (New York, USA), where they had been implanted with a temperature and ID transponder subcutaneously over the dorsal aspect of the caudal region. Upon arrival at CSCHAH, animals were acclimated for at least 7 days prior to the commencement of the experiment. Throughout the study period, animals were given food and water ad libitum, provided with environmental enrichment, and monitored at least twice daily.

For the first challenge experiment, twenty ferrets (10 male and 10 female; ~ 8 weeks old) were randomly divided into one of two groups and inoculated with a target dose of 1000 TCID₅₀ EBOV via the intramuscular (IM) or intranasal (IN) route (S1 Table). Both inoculations were delivered in a total volume of 500 ul, split evenly across two sites in the rear quadriceps (IM) or between each nostril (IN). Following virus inoculation, eight animals (four from each group; equal males and females) were treated via intraperitoneal (IP) injections with 30 mg/kg each of 1C3 and 1C11 diluted in PBS on 2 and 5 DPI; another eight animals (four from each group; equal males and females) were treated with the same dose of mAb on 3 and 6 DPI. Four animals (two from each group; equal males and females) were treated with PBS only. For the second challenge experiment, seventeen ferrets (7 male and 10 female; ~ 9 weeks old) were randomly divided into one of two groups and once again inoculated with EBOV via the IM or IN route, as described above (S1 Table). Following virus inoculation, eight animals (four from each group; equal males and females) were treated via IP injections with 30 mg/kg each of 1C3 and 1C11 on 2 and 4 DPI, as above; another eight animals (four from each group; 3 males and 5 females) were treated with the same dose of mAb on 3 and 5 DPI. One animal (female) that had been inoculated via the IN route was treated with PBS only. Antibody doses were calculated separately for male and female animals, based on the average weight of all animals of each sex. Thus, for the first experiment, all females received 13.2 mg of each mAb, while all males received 15.9 mg of each mAb. For the second experiment all females received 15.4 mg of each mAb and all males received 18.7 mg. Animals were monitored daily for clinical signs of disease including changes in body weight, temperature (via transponder), physical activity, and food/water intake. Animals that reached the clinical scoring criteria for humane endpoint were euthanized. On predetermined days (-2, 2 or 3, 5 or 6, 9, 14, 21, and 29 DPI for the first experiment and -4, 2 or 3, 4 or 5, 7, 10, 14, 21, 28, and 36 DPI for the second) or when an animal reached the humane endpoint, animals were anesthetized and given a physical examination, including weight and rectal temperature measurements, by qualified study personnel. Whole blood (EDTA) and plasma (lithium heparin) samples were also collected during these examinations. Terminal blood samples were obtained from animals that reached the humane endpoint, after which the animals were euthanized and tissues (liver, spleen, brain) were collected. Note that, from the first experiment, liver and spleen samples were harvested for RNA extraction from all treated animals (but not the control animals), while brain samples were harvested from all animals that were euthanized after 12 DPI. No tissue samples were saved from this study for the isolation of infectious virus. From the second experiment, liver, spleen, and brain samples were harvested from all animals for both RNA extraction and virus isolation.

Antibody pharmacokinetics study in ferrets

Eight ferrets (four male and four female; ~ 9 weeks old) were purchased and acclimated as described above. The animals were randomly divided into one of two groups and treated via IP injection with 30 mg/kg of 1C3 or 30 mg/kg of 1C11. Antibody doses were calculated based on the individual weight of each animal; the males therefore received 18-20.4 mg of mAb, while the females received 14.7-15.9 mg of mAb. Serum samples were obtained from each animal one day prior to mAb treatment and at regular intervals after mAb treatment: 0 (i.e., immediately after treatment), 1, 6, 24, 48, 72, 96, 120, 144, 168, 336, and 504 hours post-treatment.

Viral RNA extraction and quantification

Viral RNA was extracted from EDTA blood samples using the QIAamp Viral RNA Mini Kit (Qiagen) following manufacturer’s instructions. Tissue samples were stored in RNAlater (ThermoFisher Scientific), after which RNA was extracted using the RNeasy Plus Mini Kit (Qiagen) according to manufacturer’s instructions. EBOV RNA levels were quantified via reverse transcription quantitative PCR (RT-qPCR) using the LightCycler 480 RNA Hydrolysis Probes kit (Roche) on a QuantStudio 3 Real-Time PCR System (Applied Biosystems) using the following EBOV-specific primers and probes: forward primer, 5’-CAGCCAGCAATTTCTTCCAT-3’; reverse primer, 5’-TTTCGGTTGCTGTTTCTGTG-3’; probe 1, 5′-6-FAM-ATCATTGGC/ZEN/GTACTGGAGGAGCAG-3IABkFQ-3′; and probe 2, 5′-6-FAM-TCATTGGCG/ZEN/TACTGGAGGAGCAGG-3IABkFQ-3’ (Integrated DNA Technologies). The thermocycling conditions were as follows: 63^oC for 3 min, 95^oC for 30 s, followed by 45 cycles of 95^oC for 15 s and 60^oC for 30 s. The cycle threshold (Ct) values were converted to genome equivalents per milliliter (GEQ/mL) or gram (GEQ/g) using standard curves generated from a plasmid encoding the EBOV polymerase (L) gene.

Quantification of infectious virus

Whole blood and tissues samples that were positive for EBOV RNA via RT-qPCR were assessed for the presence of infectious virus. Tissue samples were homogenized in DMEM using 5-mm stainless steel beads in a Bead Ruptor Elite Tissue Homogenizer (Omni International, Inc.) at a frequency of 4 m/s. Homogenates were clarified by centrifugation at 1000 x g for 10 minutes, after which they were tenfold serially diluted from 10^-1 to 10^-8 in DMEM containing 2% FBS. Whole blood samples were tenfold serially diluted from 10^-2 to 10^-9. A volume of 200 μl of each dilution was then added to a confluent monolayer of Vero E6 cells in 96-well plates, in triplicate. The plates were incubated at 37^oC with 5% CO₂ and assessed on 13 DPI for cytopathic effect. The median tissue culture infectious dose (TCID₅₀) result was calculated for each sample using the Reed-Muench method [47].

Genome sequencing

EBOV whole genome sequencing was performed as described previously [48]. Briefly, extracted viral RNA samples (8 μl) were treated with ezDNAse (Invitrogen) according to the manufacturer’s instructions. RNA extracts were combined with 50 ng/μl random hexamer primers and 0.2 mM dNTPs and heated to 65^oC for 5 min, after which the primers were annealed by cooling on ice for 1 min. The Superscript III First-Strand Synthesis System (Invitrogen) was used to carry out cDNA synthesis from the hexamer-annealed RNA, per manufacturer’s instructions. cDNA was synthesized at 55°C for 90 min followed by inactivation of enzymes at 85°C for 5 min. Second strand synthesis was performed at 16°C for 16 hours in a total of 40 μl reaction, which included 10X Second Strand Synthesis Buffer, 10 U DNA Polymerase I, 0.35 U E. coli RNase H, 1.25 U E. coli DNA Ligase, and 0.375 mM dNTP mixture (all reagents from New England Biolabs). Resulting dsDNA was used as template for the TruSeq DNA Nano LP kit (Illumina) following the manufacturer’s protocol.

A total of 316 120-nt custom biotinylated lockdown DNA probes covering the entire EBOV genome in a double tiling approach were designed based on GenBank accession no. KJ660347.2 and purchased from Integrated DNA Technologies (Coralville, IA, USA). For EBOV-specific library enrichment, equal volumes of up to eight final libraries were pooled with the lock-down probes, per the xGen hybridization capture of DNA libraries protocol Version 2 (Integrated DNA Technologies). Biotinylated probes hybridized to EBOV libraries were captured with Dynabeads M270-streptavidin magnetic beads (Invitrogen) and washed extensively to remove unbound nucleic acids, following the manufacturer’s protocol. Subsequently, the EBOV libraries were subjected to twelve cycles of PCR amplification using universal Illumina primers that recognize the ligated adapter sequence of the libraries and enzymes from the Illumina TruSeq DNA Nano LP kit. PCR fragments were purified with Agencourt AMPure XP beads (Beckman Coulter) and eluted with Resuspension Buffer provided in the Illumina TruSeq DNA Nano LP kit. The DNA concentration of the library was determined using the dsDNA HS Assay Kit (Invitrogen) on a Qubit 4.0 Fluorometer (Invitrogen). The enriched libraries were sequenced without further size selection on a MiSeq instrument (Illumina) using the MiSeq Reagent Kit v3 (600-cycle) (Illumina). To balance GC content and low genomic diversity of the EBOV libraries, pools of up to five libraries were spiked into bacterial genomic MiSeq runs at 5%. For those brain samples from which reads were recovered, the average read depths ranged from 54 to 4,176, the median read depth ranged from 40 to 2,416, and the interquartile range varied from 32 to 3,444. For those liver samples from which reads were recovered, the average read depth ranged from 216 to 13,102, the median read depth ranged from 86 to 8,857, and the interquartile range varied from 216 to 10,498.

Raw sequence files were mapped using the nf-ViralMutations pipeline (https://github.com/phac-nml/nf-ViralMutations) v1.0.1 using KJ660347.2 as the reference sequence. Single nucleotide polymorphism (SNP) calling was manually verified and curated. The curated SNPs were imported into R v4.3.3 [49] and plotted using the tidyverse package v2.0.0 [50]. Visuals were annotated in Inkscape v1.2.2. The raw sequencing data are available on the Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra): PRJNA1454895. The R scripts used to generate the figures are available on GitHub: https://github.com/jaudetnml/Atypical_EBOV_ferrets.

Transcriptomics

RNA extracted from the brains of eight ferrets from the second challenge experiment were subjected to RNA sequencing. RNA quality was assessed on a Bioanalyzer using the Agilent RNA 6000 Pico Kit (Agilent Technologies), and quantified using the RNA High Sensitivity Assay Kit (Invitrogen) on a Qubit Flex fluorometer (Invitrogen). Samples were evaluated for DNA contamination using the dsDNA HS Assay Kit (Invitrogen). An aliquot of diluted ERCC ExFold RNA Spike-Mix 1 (Invitrogen) was added to the RNA, per manufacturer’s guidelines, and the RNA was DNase-treated using ezDNase enzyme (Invitrogen). RNA-seq libraries were prepared from 100, 500, or 1000 ng of the DNase-treated RNA using the TruSeq Stranded mRNA Library Prep Kit (Illumina) following a slightly modified manufacturer’s protocol, where Superscript II was replaced with Superscript III reverse transcriptase to generate cDNA. The purified cDNA was then adenylated at the 3’ end and ligated to IDT for Illumina TruSeq RNA Unique Dual index adapters (Illumina). DNA fragments containing adapters on both ends were enriched via PCR and purified using AMPure XP beads (Beckman Coulter Genomics). Libraries were quantified on a Qubit Flex fluorometer using the dsDNA BR Assay Kit (Invitrogen), and DNA size distribution was assessed on a Bioanalyzer using the Agilent High Sensitivity DNA Kit (Agilent Technologies). Libraries were sequenced on a NextSeq 2000 (Illumina) using the NextSeq 2000 P3 Reagent Kit (100-cycle) (Illumina) for 1 x 100 bp reads. The reads were aligned to the ENSEMBL ferret genome using HiSAT.

Data analysis and visualization was performed using R v4.4.1 [49]. Briefly, aligned data was normalized based on the ERCC spike-in control material using the BRGenomics v1.17.1 package [51] and then normalized further by regularized log transformation within the DESeq2 v1.46.0 package [52]. The log₂ fold change was centered for each gene, and the fifty genes with the highest variance were used to generate heatmaps using pheatmap v1.0.12 [53]. For genes with no names, ENSEMBL identifiers were queried against the UniProt database (https://www.uniprot.org/) to determine if there were homologous genes found in other species. As some genes did not have annotated names, a heatmap displaying the top fifty named genes with the highest variance was also generated. Gene names were queried against the GOslim database using GO Term Mapper (https://go.princeton.edu/cgi-bin/GOTermMapper) to summarize the shared ontologies of the genes identified. Additional R packages used included dplyr v1.1.4 [54], rtracklayer v1.66.0 [55], and tidyverse v2.0.0 [50]. The raw transcriptomic data are available on the Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra): PRJNA1454531. The R scripts used to generate the figures are available on GitHub: https://github.com/jaudetnml/Atypical_EBOV_ferrets.

Proteomics

Brain tissue homogenates were combined with an equal volume of 4X SDS Buffer (0.25 M Tris-HCl, pH 7.5, 4% SDS, 35% glycerol, 0.5% (w/v) bromophenol blue) and boiled for 10 minutes at 100^oC prior to protein extraction using a modified SPEED method [56]. The lysed brain homogenate (50 µL) was first acidified with 100 µL of trifluoroacetic acid (TFA) and after 15 minutes of shaking, was neutralized with 1,350 µL of a 20 mL solution of 2 M Tris base containing a cOmplete EDTA-free Protease Inhibitor Cocktail (PIC) tablet (Roche). Reduction and alkylation was performed separately with 10 mM dithiothreitol (DTT) and 55 mM iodoacetamide (IAA). Following an overnight acetone precipitation, samples were quantified by bicinchoninic acid (BCA) assay (Thermo Scientific).

Each sample (100 µg) was brought to 5% Sodium Dodecyl Sulfate (MP Biochemicals, Solon OH) in a total volume of 100 µl. The samples were processed using S-Trap mini spin columns (Protifi, Fairport NY USA) according to the manufacturer’s protocol. The samples were digested overnight with trypsin (Pierce, Rockford IL). The following morning, the peptides were concentrated to near-dryness (1–5 µl) using a vacuum centrifuge and labeled with the TMT 18-plex Isobaric Label Reagent Set (Thermo Fisher Scientific, Waltham, USA) per the manufacturer’s protocol; the TMT-labelled peptides were then mixed one-to-one. TMT sample mixtures were resuspended in 40 µL of buffer A (20 mM ammonium formate) and fractionated by high-pH C18 reversed-phased liquid chromatography using an Agilent 1260 Infinity liquid chromatography system (Agilent Technologies, Santa Clara, CA) equipped with a XBridge C18 3.5-µm, 2.1 x 100-mm column (Waters, Wexford Ireland) using a linear gradient of 20–56% buffer B (20 mM ammonium formate, 90% acetonitrile) over 52.5 minutes. The twelve TMT fractions were concentrated to near-dryness using vacuum centrifugation, and resuspended in 80 µL of nano buffer A (2% acetonitrile, 0.1% formic acid).

Each fractionated sample was analyzed by a liquid chromatography-tandem mass spectrometry using a nano-flow Easy-nLC 1200 connected in-line to an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) with the ion source set to 2.1 kV. The samples were loaded on a 50-cm Easy-Spray c18 column (Thermo Fisher Scientific, ES903) with buffer A (2% acetonitrile, 0.1% formic acid) and run using a 120-minute gradient of 3% to 100% buffer B (80% acetonitrile, 0.1% formic acid). The following settings were used for data acquisition. The precursor scans were acquired in positive mode in the orbitrap with range of 375–1500 m/z at 120,000 resolution. Fragmentation scans were done in the IonTrap using CID fixed at a 35% collision energy with a 1.2 m/z isolation window, fragmenting as many targets as possible in a 2 second window. Each of the targets were also fragmented again using MS3 with HCD set to 65% normalized collision energy, 5 dependent scans, and analyzed in the orbitrap with a resolution set to 50,000.

Data was searched against the UniProt database for Mustela putorius (taxon ID: 9669) using Sequest within the Proteome Discoverer v3.1 software (Thermo Scientific). Subsequent data analysis and visualization was performed using R v4.4.1 [49]. Briefly, data was normalized using variance stabilizing transformation within the DEP v1.28.0 package [57]. Similarly to the transcriptomic data, the log₂ fold change was then centered for each protein and the fifty proteins with the highest variance were used to generate heatmaps as above. Protein identifiers were queried using GO Term Mapper, as above, to summarize the shared ontologies of the proteins identified. Additional R packages used not listed above included readxl v1.4.3 [58] and stringr v1.5.1 [59]. The raw proteomics data and R scripts used to make the figures are available on GitHub: https://github.com/jaudetnml/Atypical_EBOV_ferrets.

Histopathology, immunohistochemistry (IHC), and in situ hybridization (ISH)

Ferret brain samples were fixed in 10% neutral buffered formalin. Tissues were processed in a HistoCore Pearl Tissue Processor (Leica) followed by paraffin embedding with the Arcadia H/C embedding station (Leica). Sections were cut on a Leica model RM2255 microtome at 5 μm, air dried, left to dry overnight at 39^oC, deparaffinized, and stained with hematoxylin and eosin (H&E) using standard procedures.

To detect EBOV genomic RNA in formalin-fixed, paraffin-embedded tissues, RNA in situ hybridization (ISH) was performed using the RNAscope 2.5 HD RED kit (Advanced Cell Diagnostics, Newark, CA) according to the manufacturer’s instructions and USAMRIID SOP PT-04–10. Briefly, an ISH probe targeting the genomic fragment of EBOV nucleoprotein (NP) gene was designed and synthesized by Advanced Cell Diagnostics (Cat # 448581, Advanced Cell Diagnostics). Tissue sections were deparaffinized with xylene, underwent a series of ethanol washes and peroxidase blocking, and were then heated in kit-provided antigen retrieval buffer and then digested by kit-provided proteinase. Sections were exposed to ISH target probe and incubated at 40°C in a hybridization oven for 2 h. After rinsing, ISH signal was amplified using kit-provided pre-amplifier and amplifier conjugated to alkaline phosphatase and incubated with a Fast Red substrate solution for 10 minutes at room temperature. Sections were then stained with hematoxylin, air-dried, and mounted.

Histopathology and in situ hybridization severity scores were determined based on the following criteria: Negative (score 0), no cells in the section are affected; minimal (score 1), 10% or less of the cells in the section are affected; mild (score 2), 11 – 25% of the cells in the section are affected; moderate (score 3), 26 – 50% of the cells in the section are affected; marked (score 4), 51 – 79% of the cells in the section are affected; severe (score 5), 80% or more of the cells in the section are affected.

EBOV immunohistochemistry (IHC) was performed using the Dako Envision system (Dako Agilent Pathology Solutions, Carpinteria, CA, USA). After deparaffinization, rehydration, and methanol-hydrogen peroxide blocking, slides were stained using a mixture of USAMRIID mouse monoclonal anti-EBOV GP1,2 (M-DA01-A5-B11) antibody at a dilution 1:8,000 and USAMRIID mouse monoclonal anti-VP40 antibody (B-MD04-BD07-AE11) at a dilution 1:8,000, followed by a horseradish peroxidase-conjugated secondary anti-mouse polymer (Cat # K4003, Dako Agilent Pathology Solutions). All slides were exposed to brown chromogenic substrate, 3,3’-diaminobenzidine (DAB; K3468, Dako Agilent Pathology Solutions), counterstained with hematoxylin, dehydrated, and coverslipped. All images were collected using an Olympus BX46 microscope.

Blood biochemistry and hematology

Complete blood cell counts in EDTA blood were performed using the VetScan HM5 hematology analyzer (Abaxis Veterinary Diagnostics) where lymphocytes, neutrophils, monocytes, white blood cells, and platelets were enumerated. Blood biochemistry was analyzed in lithium heparin blood use the VetScan VS2 chemistry analyzer (Abaxis Veterinary Diagnostics), and the following analytes were quantified: alanine aminotransferase (ALT), alkaline phosphatase (ALP), total bilirubin, albumin, globulin, glucose, total protein, amylase, blood urea nitrogen (BUN), potassium (K⁺), calcium (Ca²⁺), sodium (Na⁺), inorganic phosphate, and creatinine. For each parameter, the normal range was depicted as the mean ± standard deviation from all pre-infection samples.

Quantification of cytokines

Cytokines in plasma samples were quantified with the Ferret Cytokine Panel 1A kit (Cat # F103-Ka, Ampersand Biosciences) following the manufacturer’s instructions. Each sample was diluted 1:3 in sample diluent buffer and assayed in duplicate on a Luminex MAGPIX instrument. Data analysis was performed with the xPONENT software v4.2 (Luminex). The following cytokine levels were assessed: interferon gamma-induced protein 10 (IP-10; CXCL10), monocyte chemoattractant protein-1 (MCP-1; CCL2), macrophage inflammatory protein-1β (MIP-1β; CCL4), interferon alpha (IFN-α), tumour necrosis factor alpha (TNF-α), interleukin 2 (IL-2), IL-4, IL-6, IL-8, IL-17, IL-12p40, and IL-12p70.

ELISAs

EBOV GP-specific IgG levels in serum samples were quantified by indirect enzyme-linked immunosorbent assays (ELISA), as previously described [60]. Briefly, half-area high-binding 96-well assay plates (Corning) were coated at 4^oC overnight with 30 µl of 1 µg/ml of transmembrane domain-deleted EBOV GP protein (IBT Bioscience) prepared in carbonate buffer (pH 9.5). The plates were then blocked with 100 ul of 5% skim milk (BD Biosciences) prepared in PBS at 37^oC for 1 h. Two-fold serial dilutions (starting at 1:200) of serum samples prepared in 2% skim milk were added to the plates for incubation at 37^oC for 1 h, followed by washing with PBS containing 0.1% Tween-20 (PBST). The plates were incubated at 37^oC for 1 h with an HRP-conjugated goat anti-ferret IgG secondary antibody (Novus) diluted 1:70,000 in 2% skim milk, and then washed again with PBST. Following addition of TMB solution (Life Technologies), chemiluminescence signal as measured at 650 nm using the Synergy HTX plate reader (Biotek) after 30 min. Endpoint dilution titers were calculated by determining the highest dilution that gave an average OD 650 reading greater than or equal to the cut-off OD value, which was set as the mean OD value for pre-infection serum samples plus three times the standard deviation.

The levels of human IgG (i.e., 1C3 and 1C11) were quantified in serum samples from the two ferret challenge experiments as well as the antibody pharmacokinetics experiment using the Human IgG ELISA Kit (Cat # ab195215, Abcam), according to manufacturer’s instructions. A standard curve was generated in GraphPad Prism and used to calculate the concentration of total human IgG in each sample. For the pharmacokinetics experiment, the half-life of each antibody in each animal was calculated using a one-phase decay curve incorporating data from the 96–504 hour time points. Because human antibody was not detected at any time point in one of the animals treated with 1C11, the half-life for this mAb was calculated from only three animals.

Statistical analyses

GraphPad Prism (version 10) was used to perform all statistical tests (as indicated in the figure legends) and generate all graphs, with the exception of the transcriptomic and proteomic analyses, which used R v4.4.1 [49]. Statistically significant differences are indicated with asterisks, where a P value less than or equal to 0.05 is marked with one asterisk (*), less than or equal to 0.01 marked with two asterisks (**), less than or equal to 0.001 marked with three asterisks (***), and less than or equal to 0.0001 marked with four asterisks (****).