In vivo Ebola virus infection leads to a strong innate response in circulating immune cells
Caballero et al. BMC Genomics
In vivo Ebola virus infection leads to a strong innate response in circulating immune cells
Ignacio S. Caballero 3
Anna N. Honko 1 2
Stephen K. Gire 5 6
Sarah M. Winnicki 5 6
Marta Melé 4 6
Chiara Gerhardinger 4 6
Aaron E. Lin 5 6
John L. Rinn 4 6
Pardis C. Sabeti 5 6
Lisa E. Hensley 1 2
John H. Connor 0
0 Department of Microbiology, Boston University School of Medicine , Boston, MA , USA
1 Integrated Research Facility, National Institute of Allergy and Infectious Diseases, National Institutes of Health , Fort Detrick, MD , USA
2 Virology Division, United States Army Medical Research Institute of Infectious Diseases , Fort Detrick, MD , USA
3 Bioinformatics Graduate Program, Boston University , Boston, MA , USA
4 Department of Stem Cell and Regenerative Biology, Harvard University , Cambridge, MA , USA
5 Broad Institute of MIT and Harvard , Cambridge, MA , USA
6 Center for Systems Biology, Department of Organismic and Evolutionary Biology, Harvard University , Cambridge, MA , USA
Background: Ebola virus is the causative agent of a severe syndrome in humans with a fatality rate that can approach 90 %. During infection, the host immune response is thought to become dysregulated, but the mechanisms through which this happens are not entirely understood. In this study, we analyze RNA sequencing data to determine the host response to Ebola virus infection in circulating immune cells. Results: Approximately half of the 100 genes with the strongest early increases in expression were interferonstimulated genes, such as ISG15, OAS1, IFIT2, HERC5, MX1 and DHX58. Other highly upregulated genes included cytokines CXCL11, CCL7, IL2RA, IL2R1, IL15RA, and CSF2RB, which have not been previously reported to change during Ebola virus infection. Comparing this response in two different models of exposure (intramuscular and aerosol) revealed a similar signature of infection. The strong innate response in the aerosol model was seen not only in circulating cells, but also in primary and secondary target tissues. Conversely, the innate immune response of vaccinated macaques was almost non-existent. This suggests that the innate response is a major aspect of the cellular response to Ebola virus infection in multiple tissues. Conclusions: Ebola virus causes a severe infection in humans that is associated with high mortality. The host immune response to virus infection is thought to be an important aspect leading to severe pathology, but the components of this overactive response are not well characterized. Here, we analyzed how circulating immune cells respond to the virus and found that there is a strong innate response dependent on active virus replication. This finding is in stark contrast to in vitro evidence showing a suppression of innate immune signaling, and it suggests that the strong innate response we observe in infected animals may be an important contributor to pathogenesis. Abbreviations: CPM, Counts per million; EBOV, Ebola virus; GEO, Gene expression omnibus; ISG, Interferon stimulated gene; LASV, Lassa virus; MARV, Marburg virus; PBMC, Peripheral blood mononuclear cells; PFU, Plaque-forming units; rhAPC, Recombinant human activated protein C; rNAPc2, Recombinant nematode anticoagulant protein c2; rVSV, Recombinant vesicular stomatitis virus with the ebola glycoprotein; VLP, Virus-like particle
Ebola virus; Transcriptional response; Transcriptomics; Interferon-stimulated genes
Ebola virus belongs to the family Filoviridae and is an
envelope, non-segmented, negative-stranded RNA
virus with filamentous virion morphology. Infection
with Ebola virus causes Ebola Virus Disease, a disease
associated with mortality rates between 25 and 90 %
]. The earliest clinical symptoms are non-specific
and flu-like, such as high fever, headache and myalgia
]. During the large outbreak in West Africa
(2014present), the symptoms also included diarrhea and
]. The late stage of disease is associated
with immune cell imbalances such as neutrophilia and
coagulation disorders like diffuse intravascular
It has been postulated that survival following Ebola
virus infection correlates with the ability of the host to
mount an early and robust interferon response. In
cultured human liver cells, the virus has been shown to
block specific aspects of the innate immune response,
preventing the expression of interferon-stimulated genes
]. Experiments in animal models, including mice [
] and guinea pigs [
], have provided evidence that
the suppression of interferon signaling is important for
Previous studies of primates exposed to wild-type
Ebola virus strains have suggested that there is a strong
interferon-like transcription signal in circulating
immune cells [
]. It has not yet been established
whether interferon signaling is seen only in circulating
immune cells or whether there is evidence for
interferon-induced mRNA expression in infected
tissues as well. It is also not well established whether this
response is dependent on active virus replication.
Several studies have reported that during infections
caused by respiratory and hemorrhagic fever viruses,
circulating immune cells undergo major gene
regulatory changes that result in the upregulation of many
innate immune system genes [
]. One of the most
interesting aspects about this surge in transcriptional
activity is that it constitutes an early measurable
indicator of infection.
In the present study, we studied the transcriptional
profile of circulating immune cells obtained from
several different studies. We investigated the most
robustly expressed genes that were observed in the
peripheral blood mononuclear cells (PBMCs) of
cynomolgus macaques following different routes of
infection (aerosol and intramuscular), as well as following
vaccination of the host. We further investigated
whether genes highly-upregulated in PBMCs also
showed changes in expression in tissues of infected
animals. Our results suggest that interferon-signaling is
an early and robust response to Ebola virus infection
throughout the body.
Macaque model of intramuscular exposure to Ebola virus
This RNA sequencing study used cynomolgus macaques
divided in two groups: vaccinated and Ebola-naïve [
The vaccinated group received an intramuscular
injection of recombinant Vesicular Stomatitis Virus with the
Ebola glycoprotein (rVSV/EBOV-GP) and a lethal dose
of EBOV strain Kikwit several days later. The
Ebolanaïve group was treated with a non-protective dose of
rVSV/MARV-GP and received the same lethal dose of
EBOV administered to the vaccinated group several days
later. Blood samples were taken at 0, 4 and 7 days
postinfection (see Table 1). PBMCs were isolated from the
18 blood samples, and RNA sequencing was performed
using the same methods specified in [
Macaque model of aerosol exposure to Ebola virus
This study used rhesus macaques exposed to EBOV via
aerosol (between 7.43×102 and 2.74×102 plaque-forming
units (PFU)) 12 PBMC samples were collected at 0, 3, 6
and 8 days post-infection (see Table 1 for details) [
19 spleen samples were obtained at 0, 1, 3, 4, 5, 6, 7 and
8 dpi. 12 adrenal gland samples were obtained at 1, 3, 4,
6, 7 and 8 dpi. 8 axillary lymph node samples were
obtained at 0, 1, 3, 4, 5 and 6 dpi. 11 brain samples were
obtained at 1, 3, 4, 6 and 7 dpi. 4 liver samples were
obtained at 0, 3, 5 and 8 dpi. 3 pancreas samples were
obtained at 0, 4 and 8 dpi. RNA was extracted from
PBMCs and the other tissues and sequenced [
For tissue homogenates, RNA was extracted from
TRIzol-inactivated samples using BCP:chloroform, and
total RNA was depleted of small RNAs (<200 nt) with a
mirVana miRNA Isolation Kit per manufacturer’s
instructions (Thermo Fisher, Waltham MA). Subsequently,
ribosomal RNA was depleted using complementary
DNA oligos and RNase H. For PBMC samples, mRNA
was enriched using the Dynabeads mRNA Purification
Kit (Thermo Fisher) per protocol. For all samples,
TruSeq v2 (Illumina, San Diego CA) libraries were
constructed as described previously [
using custom adaptors , and sequenced on a HiSeq
to obtain 2 × 100 bp paired-end reads. Reads were
demultiplexed with a 0 mismatch tolerance, and
crosssample contamination was back-calculated from
synthetic RNAs that were spiked in prior to library
]. The PBMC samples referenced above,
as well as 24 additional PBMC samples taken at 0, 1, 3,
4, 5, 6, 7 and 8 dpi (Table 1), were also quantified using
microarrays according to the methods specified in [
Mouse model of exposure to Ebola virus
This study used BALB/c mice divided into two groups
]. One group was inoculated intraperitoneally with
a 8 of these animals (and 12 of these samples) correspond to the 8 animals (and 12 samples) that were quantified using RNA sequencing)
wild-type EBOV. The other group was inoculated with
mouse-adapted EBOV. Spleen samples were collected
and homogenized at 0, 12, 24, 48 and 72 h
postinfection for both groups (see Table 1). RNA was
extracted from the 29 spleen samples and Agilent
microarray processing was performed as described in [
Macaque model of aerosol exposure to Lassa and Marburg
This study used cynomolgus macaques divided in two
]. One group was exposed to LASV Josiah
and the other to MARV Angola, both groups received
1000 PFU of virus via aerosol. RNA was extracted from
24 PBMC samples (see Table 1) and RNA sequencing
Fold change analysis of sequencing data
After the raw sequencing reads of each study were
generated, they were trimmed and the adapters were
removed using Trimmomatic [
]. The processed reads
were aligned to the Macaca mulatta genome (Ensembl
release 77) [
] using TopHat 2.0.11 [
] with default
parameters (segment length of 25, allowing up to 2
segment mismatches). Gene counts were obtained using
] to count reads that aligned uniquely to each
gene. Counts were normalized to compensate for
differences in library size using the trimmed mean of
Mvalues normalization method included in the edgeR
BioConductor package [
A gene was deemed to show statistically significant
changes in expression at a specific time after infection if
the moderated t-test between the infected and
preinfection samples resulted in a p-value lower than 0.05
(after Benjamini-Hochberg correction), with an absolute
fold change in expression greater or equal to 3, and an
average number of reads across all samples greater than
4 counts per million (CPM). When calculating relative
changes to the pre-infection samples, infected samples
were not subtracted from their individual uninfected
controls (since not every infected sample had a
preinfected control), but from an average of all the
preinfection samples. Statistical significance, however, was
calculated using the individual pre-infection samples,
not their average.
Fold change analysis of microarray data
Agilent two-color human gene expression microarrays
were processed using limma [
]. Fold changes were
obtained by calculating the log-ratio between the
intensities of the red channel (corresponding to experimental
samples) and the green channel (corresponding to
Human Universal Reference RNA [
]). These were later
background-corrected and normalized using the LOESS
algorithm. Agilent one-color mouse gene expression
microarrays were processed in a similar manner as the
two-color arrays, but using the reported expression value
for each probe instead of the log-ratio between the two
channels. Benjamini-Hochberg was used to correct the
false discovery rate of multiple testing.
To examine the response of the circulating immune
system of nonhuman primates infected with Ebola virus, we
used sequencing data from PBMCs of cynomolgus
macaques infected with Ebola virus via intramuscular
injection. These included pre-infection samples and samples
taken 4 and 7 days after infection, as described in [
] (unless specified otherwise, results and figures
discussed in this manuscript refer to this dataset). We were
first interested in connecting the transcriptomic data to
previously reported information about the host cytokine
response to Ebola virus infection and to examine
whether other cytokines showed evidence of being
upregulated during infection.
Cytokine responses during Ebola virus infection
Previous studies have shown strong increases in cytokine
protein levels in the serum of infected animals and
]. In our analysis, we observe that a subset
of these cytokines reported to be present in the serum also
show transcriptional upregulations in PBMCs by 4 days
post-infection (IL6, IL1B, IL1RN, CCL8, CXCL10) and by
7 days post-infection (CCL2, CCL3, CSF1, TNF, CXCL1,
FAS and CXCL8). Figure 1a shows the average
foldchange (log2) for each of these cytokines on the y-axis, as
well as the time after infection when the samples were
taken on the x-axis. These results suggest a biphasic
immune response to infection. IL1B is the only cytokine with
an unusual transcription pattern, peaking at day 4 and
returning to pre-infection levels by day 7.
Additionally, we identified several cytokines that were
not previously known to play a role during Ebola virus
infection. Among these, some became upregulated
during the early stage of infection (day 4) CCL7, CXCL11,
IL1R1 (cytokine receptor) and IL15RA, and others
became upregulated during the late stage of infection (day
7): IL2RA, CSF2RB and others (Fig. 1b).
The majority of cytokines during Ebola virus infection
in vaccinated animals did not reach a 3-fold change in
expression at 4 or 7 days post-infection. Three notable
exceptions were CCL7, CXCL10 and CXCL11, which
peaked (respectively) at 42.8, 35.5 and 6.6 times the level
of expression at 7 days post-infection compared to the
pre-infection levels. These genes mirrored the expression
patterns seen in the Ebola-naïve animals, but at a lower
magnitude (Fig. 2). For example, by 4 days post-infection
the expression level of CXCL10 had gone up 7.1-fold in
the Ebola-naïve group, but only 3.1-fold in the vaccinated
group. By 7 dpi, these values were 92.6-fold in the
Ebolanaïve group and 35.5-fold in the vaccinated group.
A strong and sustained early host transcriptional response after Ebola virus infection contains many genes associated with the innate immune response
We also analyzed the macaque intramuscular Ebola
dataset to identify genes that underwent strong,
Previously reported cytokines
Previously unreported cytokines
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Days post−infection Days post−infection
Fig. 2 Effects of vaccination in the expression of cytokines after Ebola virus infection. Each line represents the log2 fold-change (y-axis) of three
cytokines that become upregulated in vaccinated animals: CCL7, CXCL10, and CXCL11 at different times (x-axis) after Ebola virus infection. IL6 is
shown to illustrate the behavior of a cytokine that only becomes expressed in Ebola-naïve animals
significant increases in expression between pre-infection
and 4 days post-infection. The thresholds that we used
were 8 or higher fold-change with at least 4 read counts
per million, and an adjusted p-value of less than 0.05.
We found 125 genes that met these requirements, 25 of
which showed levels of expression that peaked at 4 days
post-infection and decreased by 7 dpi. The levels of
expression of the other 100 genes remained constant or
increased between days 4 and 7. Of these, 33 genes are
known to be responsive to innate immunity
transcription factors such as IRF3, IRF7 and STAT1 (also known
as interferon-stimulated genes, or ISGs) [
(Fig. 3a) and are herein referred to as canonical ISGs.
Figure 3a shows the log2-fold change of these
interferon-stimulated genes at 0, 4 and 7 days
postinfection, highlighting a few representative examples
(MX1, ISG15, OAS1, DHX58 (RIG-I), IFIT2, and
HERC5). The other 47 strong and sustained genes show
similar responses to the ISG group but they have not
been previously identified as downstream effectors of
the interferon response (Fig. 3b). This population
includes several neutrophil-associated genes (OLFM4,
CD177, SERPINB1, S100P, PTX3 and MMP8),
suggesting that there is an accumulation of neutrophils
in PBCMs during the course of infection.
The strong increase in canonical ISGs following Ebola
virus infection led us to investigate if this strong
interferon-like response was observed in other animal
models of Ebola virus infection. To do this, we used the
Ebola mouse model dataset to analyze the ISG response
in mice that had been infected with wild type EBOV
(which is non-pathogenic in mice) and mouse-adapted
EBOV (which is pathogenic in mice) [
we found that a strong ISG response was apparent 3 days
after infection in mice that were infected, irrespective of
pathogenesis (Fig. 4).
We validated these results using expression data from
two previous studies that infected rhesus macaques with
Ebola virus (intramuscular injection) and treated them
with an anticoagulant drug (recombinant human
activated protein C (rhAPC), and recombinant nematode
anticoagulant protein c2 (rNAPc2) [
]). We found
significant similarity for a majority of early response
genes even though the expression data was quantified
using microarrays. For example, in the rhAPC samples,
the expression of IFIT2 increased 4 log2-fold and 7.4
Other strong and sustained genes
log2-fold at 3 and 6 days post-infection, respectively; in
the rNAPc2 samples the increases were 4.6 log2-fold
and 6.9 log2-fold at 3 and 6 days post-infection,
respectively. Additionally, we were also able to confirm these
patterns in a previous study that quantified PBMC
samples from Ebola virus-infected macaques using
]. In this dataset, for example, IFIT2 expression
increases 3.4 log2-fold 3 dpi and remains sustained
throughout infection. The fact that this multiple
transcriptomic studies agree on this observation provides
additional evidence that the innate immune signaling is
upregulated after infection.
The sustained transcription of early responsive genes is not present in a non-productive infection
Given the repeated appearance of a type I interferon-like
signature following Ebola virus infection, we were
interested in determining whether the appearance of an early
innate immune response following a hemorrhagic fever
virus infection was a response to the injection of a
negative strand RNA virus. To do this, we used the
intramuscular Ebola dataset to compare the innate
immune response to Ebola virus in non-human primates
that had been previously vaccinated against Ebola virus
to those that had been immunized against a different
hemorrhagic fever virus and were therefore susceptible
to Ebola virus (Ebola-naïve group) [
]. Figure 5
shows the changes in expression of four ISGs that
illustrate the overall differences in response between the two
animal cohorts. In the Ebola-naïve group, for example,
SIGLEC1 undergoes a 5.72 log2-fold change 4 days
postinfection, while in the vaccinated group the change is
only 2.53 log2-fold change (Fig. 5a). By 7 days
postinfection the difference is even greater: 7.64 in the
Ebola-naïve group, and 3.36 in the vaccinated group. A
similar trend was seen for RSAD2, MX1, IFIT3, and the
majority of genes making up the early transcriptional
response: vaccinated animals showed significantly lower
levels of expression than Ebola-naïve animals. These
results argue that the strong interferon-like response is
the result of active virus dissemination, and is not a
non-specific response to the injection of viral material.
0.0 0.5 1.0 2.0 3.0 0.0 0.5 1.0 2.0 3.0 0.0 0.5 1.0 2.0 3.0 0.0 0.5 1.0 2.0 3.0 0.0 0.5 1.0 2.0 3.0
Fig. 4 Expression of interferon stimulated genes in spleens of Ebola virus infected mice. Each line represents the average fold-change in expression
(log2, y-axis) of each gene at different times (x-axis) after Ebola virus infection, for each of the conditions: uninfected control, non-pathogenic wild-type
Ebola virus infection, and pathogenic mouse-adapted Ebola virus infection
2 3 4 5
The early transcriptional response is common to multiple hemorrhagic fevers
The early host transcriptional response that we observed
in these studies appeared to be similar to strong innate
transcriptional responses observed in other hemorrhagic
]. To compare these responses, we used gene
expression data from cynomolgus macaques exposed via
aerosol to either Lassa virus [
] or Marburg virus
], and found that the innate response to both
infections is highly similar to that of Ebola virus infection.
This is illustrated in Fig. 6, which shows the changes in
expression of four canonical ISGs (MX1, ISG15, DHX58
and OAS1) during Ebola, Lassa, and Marburg virus
infection. All four genes are significantly upregulated at
the earliest infected timepoint, and they remain
sustained throughout the late stage of disease. For example,
MX1 goes up 4.7 log2-fold 3 days after Lassa infection,
4.1 log2-fold 3 days after Marburg infection, and 4.9
log2-fold 4 days after Ebola virus infection. By 6–7 days
post-infection, the log2-fold changes are 6.1, 6, and 5.4,
respectively. By 9–10 days post-infection, the expression
of MX1 seems to decrease mildly in Lassa and Marburg
infection, but it remains at very high levels compared to
the pre-infection baseline. During Lassa infection, ISG15
and OAS1 undergo similar expression changes to those
of MX1: a strong increase, followed by a slight decrease
in expression. For Marburg and Ebola virus infection,
the early levels of expression of these genes increase
more rapidly than during Lassa infection, and this trend
continues during the late stage of disease. These results
support the hypothesis that the immune system
responds to different viral infections via a common, early,
sustained and strong innate immune response.
Different routes of infection lead to a similar early transcriptional response
To determine if the route of infection could alter the
early transcriptional response to Ebola virus infection,
we looked at gene expression data from a study that
exposed macaques to Ebola virus via aerosol and we
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Day post−infection Day post−infection
Fig. 6 Changes in expression of interferon-stimulated genes during different hemorrhagic fever infections. Each line represents the average fold-change
in expression (log2, y-axis) of MX1, ISG15, DHX58 or OAS1 at different times (x-axis) after Ebola, Lassa or Marburg virus infection
compared it to the intramuscular injection data. Figure 7
shows the expression of the top 20 genes (Additional file
1) that have the strongest changes 4 days post-infection in
the intramuscular group (Fig. 7a), as well as the pattern of
expression of those same genes under aerosol exposure
(Fig. 7b). RNA sequencing was used to quantify the
expression in the intramuscular group, while microarrays
were used for the aerosol group. We find that both groups
show a strong increase in expression starting at 3–4 days
post-infection and increasing during the course of disease.
The average log2 fold-change in the intramuscular group
at 4 days post-infection is 6.05, and in the aerosol group at
3 days post-infection it is 2.17. Given that this difference
in magnitude takes place at different days, it is not
possible to determine if the aerosol model shows a delayed
response without obtaining additional samples at these
timepoints. By 6 days post-infection, the fold change in
expression for both models is similar.
The innate immune response takes place across most tissues
We next used the list of genes that were strongly
upregulated in PBMCs in the macaque intramuscular Ebola
dataset (125 genes, Additional file 1) and looked at their
expression in the macaque aerosol Ebola dataset. We
found that 88 % of the genes from the intramuscular list
showed strong upregulation in the aerosol model (111
genes, Additional file 2). We also looked at the expression
of these genes in the remaining tissues in the aerosol
model and found that the early immune response is
present across most tissues at varying magnitudes and
expression rates. Figure 8 shows the average expression rate
of these genes across each tissue type. PBMCs become
transcriptionally activated at 6 days post-infection, and
show the strongest activation compared to the remaining
tissues at 7 days post-infection. Liver starts becoming
transcriptionally activated 3 days post-infection and the
average expression level continues to increase by 5 and 8 days
post–infection. The gene expression response in the spleen
begins to increase at day 3 and also increases to similar
levels as in the liver. In the adrenal gland and the pancreas,
the expression level increases at 4 days post-infection and
remains activated until the end of the infection. The
axillary lymph node is the last to become activated, starting
around day 6. The brain shows the lowest levels of
activation, with only a slight increase that begins on days 4–6.
Intramuscular injection ISG15
Our analysis of the transcriptomic response to Ebola
virus infection highlights that there is a strong activation
of innate immune response genes at early times
postinfection, most of which are classified as
interferonstimulated genes. This finding is consistent with other
analyses of host responses to different viral infections
14, 24, 41
] and with an earlier report [
findings highlight a strong contrast between in vitro studies
of Ebola virus function. Previous studies have shown that
Ebola virus infection in vitro blocks the expression of
ISGs in liver cells [
]. This interferon-antagonism is
based on the ability of VP35 to block the activation of
IRF3, which inhibits the expression of interferon beta
and other ISGs. Mutation of VP35 in a manner that
allows IRF3 signaling has been studied in mice  and
guinea pigs [
] using a recombinant Ebola strain
containing a single-point mutation in VP35 (R322A). This
mutation reduced the virus’s ability to replicate and to
block interferon signaling in vitro; in vivo, this virus did
not cause pathogenesis in these animal models. These
studies are consistent with the hypothesis that Ebola
virus causes a systemic inhibition of the interferon
A caveat of these earlier studies is that they did not
measure the innate immune response in the animals
during Ebola virus infection to determine if there was an
observable change in interferon-induced signaling in the
animal model as well. Our analysis found that mice
infected with both pathogenic and non-pathogenic EBOV
show a strong expression of interferon-stimulated genes
in the spleen. The fact that strong interferon responses
Average Expression Across Tissues
Average PBMC Expression
occur in mice infected with non-pathogenic WT-EBOV
is perhaps not surprising, as the non-adapted virus
would be expected to induce a strong innate immune
response. That the same response was seen in pathogenic
MA-EBOV infected mice argues that a robust innate
immune response in the spleen is generated in response to
both pathogenic and non-pathogenic EBOV infection in
mice. It was reported that in the mouse model the
WTEBOV showed strongly attenuated and delayed growth,
while the MA-EBOV showed strong growth [
suggesting that the non-mouse adapted virus was
suppressed by innate signaling while the pathogenic virus
This observation that innate immune genes are
upregulated in the mouse-model of EVD is consistent with
earlier reports suggesting that primary cells exposed to
Ebola virus strongly activate interferon-like signaling
] and data showing that in NHP-infection
interferonlike signaling is upregulated [
]. Together, the
findings from two different model systems imply that innate
signaling is largely unimpeded during Ebola virus
infection and would be observed in human disease as well.
Either IRF3 is activated or interferon is expressed and
released into the circulation leading to an early, and
global innate response in vivo. Additional experiments are
needed to determine why we detect a significant increase
in the expression of interferon-stimulated genes at the
early stages of Ebola virus infection. One explanation
could be that infected cells stimulate uninfected
neighboring cells to produce interferon through a currently
unknown mechanism, and that this stimulates them to
express ISGs and to continue spreading the interferon
signal (Fig. 9a). Since we see viral titers and ISG
expression increasing throughout infection, further studies are
required to understand if the populations of cells that
are expressing ISGs are different from the ones that are
undergoing viral replication. Another hypothesis is that,
in vivo, some cells may retain their ability to translocate
IRF3 to the nucleus and to express interferon, perhaps
due to a missing VP35 (Fig. 9b). This is consistent with
earlier reports showing that there are early
immuneassociated transcriptional responses in primary target
cells exposed to either Ebola virions or virus-like
particles (VLPs) [
One of the hallmarks of Ebola virus infection is
dysregulated levels of circulating proinflammatory
cytokines. We observed the levels of expression of previously
reported cytokines such as IL6, and CXCL8 (IL8)
increase between 4 and 7 days post-infection. We also
found several cytokines, such as CCL7 and CXCL11
(ITAC), that showed significant changes in expression but
whose protein levels have not been previously observed
during Ebola pathogenesis. CXCL10 (IP-10) and CXCL11
were two of the cytokines that were expressed in animals
that succumb to infection. Interestingly, we saw a
significant increase in expression in vaccinated macaques
infected with Ebola virus. This suggests that they are part of
a conserved response in both the pathogenic and
nonpathogenic response to Ebola virus. Both cytokines are
induced by interferon alpha (they are ISGs), share a
Fig. 9 Model for the expression of interferon-stimulated genes during Ebola virus infection. a Ebola-infected cell (top) is not able to produce
interferon due to the VP35-inhibition of IRF3 translocation. We suggest that the infected cell, through an unknown mechanism, might be able to
induce neighboring cells (middle) to translocate IRF3 to the nucleus and start producing interferon. Once interferon is released by the neighboring
cells, it activates the receptors of additional cells (bottom) and leads to the transcription of ISGs. b An alternative model is that some cells (top)
can become infected with Ebola VLPs, which are not able to block IRF3 translocation, and therefore they can produce interferon, release it to
neighboring cells (middle and bottom) and they in turn start transcribing ISGs
common receptor (CXCR3), and are thought to be
involved in the recruitment of effector T cells and NK cells
]. They are also known to be induced during the acute
phase in other types of viral infections including dengue
], influenza [
], hepatitis B and C [
simplex  and HIV-1 [
]. Upregulation of CXCL10
has also been associated with hemorrhagic manifestations
in patients infected with Sudan virus [
observations suggest that these cytokines are part of a core innate
immune response that is triggered in all animals exposed
to Ebola virus, but that they are overexpressed in animals
that will succumb to disease.
By comparing how different routes of infection affect
the cellular circulating immune response in Ebola
virusinfected primates, we observed that infection via an
aerosol and via intramuscular injection resulted in
similar patterns of gene expression. Previous studies have
reported a delay in the aerosol model of exposure
compared to the intramuscular model [
], but given that we
do not have samples from both infections taken at
identical early time points, we were not able to confirm this
Infection with Ebola virus leads to early and robust
interferon-like responses that take place before the
appearance of virus in the blood. This response takes place
not only on circulating immune cells, but throughout
the majority of infected tissues. Our results extend
earlier observations of a strong innate immune response
and suggest the involvement of new cytokines in Ebola
virus infection. Further analysis of the cells responsible
for driving this response, and for producing the different
cytokine signals, will be important to understand the
ability of the virus to replicate virtually unchecked in
many tissues—even when these tissues show a strong
interferon-expression signal—, and to identify which
cells are undergoing an uncontrolled ISG response.
Additional file 1: 125 strongly upregulated genes in PBMCs in the
macaque intramuscular Ebola dataset. For each gene, we report the log
fold-changes at 4 and 7 days post-infection, the corresponding adjusted
p-values and its functional category. (XLSX 20 kb)
Additional file 2: 111 strongly upregulated genes in PBMCs in the
macaque aerosol Ebola dataset. For each gene, we report the log
fold-change at 6 days post-infection, the corresponding adjusted p-values
and its functional category. (XLSX 14 kb)
We thank Lee Wetzler for helpful discussions regarding data interpretation.
This work was supported by contracts W81XWH 100-02-0008, 11-02-0130,
and USAMRAA Biomarkers W81XWH-11-1-0141, and by a grant from the
National Science Foundation No. DGE 1144152 (AEL). This project has been
funded in part with Federal funds from the National Institute of Allergy and
Infectious Diseases, National Institutes of Health, Department of Health and
Human Services, under Grant Number U19AI110818 to the Broad Institute.
Marta Melé Messeguer is a Gilead Fellow of the Life Sciences Research
Availability of data and material
Macaque model of aerosol exposure to Ebola virus: The microarray expression
data is available in GEO under accession GSE68809.
Mouse model of exposure to Ebola virus: The microarray data is available in
the University of Washington Viromics server: bit.ly/Cilloniz_2011_microarray.
Macaque model of aerosol exposure to Lassa and Marburg virus: The sequencing
data is available in the Sequence Read Archive under accession numbers
PRJNA222891 (Lassa) and PRJNA222892 (Marburg).
Macaque model of intramuscular exposure to Ebola virus: The sequencing data
is available in Gene Expression Omnibus (GEO) under accession GSE64538.
Conceived and designed the experiments: IC, JC. Performed the experiments:
SG, SW, MM, CG, AL, JR. Analyzed the data: IC. Contributed reagents/materials:
AH, LH, PS. Wrote the paper: IC, JC. All authors read and approved the final
The authors declare that they have no competing interests.
Ethics approval and consent to participate
Ethics not required.
Submit your next manuscript to BioMed Central
and we will help you at every step:
1. Feldmann H , Geisbert TW . Ebola haemorrhagic fever . Lancet . 2011 ; 377 : 849 - 62 .
2. WHO Ebola Response Team . Ebola virus disease in West Africa - the first 9 months of the epidemic and forward projections . New Eng J Med . 2014 ; 371 : 1481 - 95 .
3. Wong G , Kobinger GP , Qiu X . Characterization of host immune responses in Ebola virus infections . Expert Rev Clin Immunol . 2014 ; 10 : 781 - 90 .
4. Bradfute SB , Swanson PE , Smith MA , Watanabe E , McDunn JE , Hotchkiss RS , Bavari S. Mechanisms and consequences of ebolavirus-induced lymphocyte apoptosis . J Immunol . 2010 ; 184 : 327 - 35 .
5. Reed DS , Hensley LE , Geisbert JB , Jahrling PB , Geisbert TW . Depletion of peripheral blood T lymphocytes and NK cells during the course of ebola hemorrhagic Fever in cynomolgus macaques . Viral Immunol . 2004 ; 17 : 390 - 400 .
6. Paessler S , Walker DH . Pathogenesis of the viral hemorrhagic fevers . Annu Rev Pathol . 2013 ; 8 : 411 - 40 .
7. Kash J , Mühlberger E , Carter V . Global suppression of the host antiviral response by Ebola-and Marburgviruses: increased antagonism of the type I interferon response is associated with enhanced virulence . J Virol . 2006 .
8. Hartman AL , Ling L , Nichol ST , Hibberd ML . Whole-genome expression profiling reveals that inhibition of host innate immune response pathways by Ebola virus can be reversed by a single amino acid change in the VP35 protein . J Virol . 2008 ; 82 : 5348 - 58 .
9. Cilloniz C , Ebihara H , Ni C , Neumann G , Korth MJ , Kelly SM , Kawaoka Y , Feldmann H , Katze MG . Functional genomics reveals the induction of inflammatory response and metalloproteinase gene expression during lethal Ebola virus infection . J Virol . 2011 ; 85 : 9060 - 8 .
10. Hartman AL , Bird BH , Towner JS , Antoniadou Z-A , Zaki SR , Nichol ST . Inhibition of IRF-3 activation by VP35 is critical for the high level of virulence of ebola virus . J Virol . 2008 ; 82 : 2699 - 704 .
11. Prins KC , Delpeut S , Leung DW , Reynard O , Volchkova VA , Reid SP , Ramanan P , Cárdenas WB , Amarasinghe GK , Volchkov VE , Basler CF . Mutations abrogating VP35 interaction with double-stranded RNA render Ebola virus avirulent in guinea pigs . J Virol . 2010 ; 84 : 3004 - 15 .
12. Rubins KH , Hensley LE , Wahl-Jensen V , Daddario DiCaprio KM , Young HA , Reed DS , Jahrling PB , Brown PO , Relman DA , Geisbert TW . The temporal program of peripheral blood gene expression in the response of nonhuman primates to Ebola hemorrhagic fever . Genome Biol . 2007 ; 8 : R174 .
13. Mejias A , Dimo B , Suarez NM , Garcia C , Suarez-Arrabal MC , Jartti T , Blankenship D , Jordan-Villegas A , Ardura MI , Xu Z , Banchereau J , Chaussabel D , Ramilo O . Whole blood gene expression profiles to assess pathogenesis and disease severity in infants with respiratory syncytial virus infection . PLoS Med . 2013 ; 10 : e1001549 .
14. Zaas AK , Chen M , Varkey J , Veldman T , Hero AO , Lucas J , Huang Y , Turner R , Gilbert A , Lambkin-Williams R , Øien NC , Nicholson B , Kingsmore S , Carin L , Woods CW , Ginsburg GS . Gene expression signatures diagnose influenza and other symptomatic respiratory viral infections in humans . Cell Host Microbe . 2009 ; 6 : 207 - 17 .
15. Zaas AK , Burke T , Chen M , McClain M , Nicholson B , Veldman T , Tsalik EL , Fowler V , Rivers EP , Otero R , Kingsmore SF , Voora D , Lucas J , Hero AO , Carin L , Woods CW , Ginsburg GS . A host-based RT-PCR gene expression signature to identify acute respiratory viral infection . Sci Transl Med . 2013 ; 5 : 203ra126 .
16. Caballero IS , Bonilla G , Yen JY , Connor JH . Diagnosing Lassa virus infection by tracking the antiviral response . BMC Bioinformatics . 2012 ; 13 Suppl 18 : A13 .
17. Marzi A , Engelmann F , Feldmann F , Haberthur K , Shupert WL , Brining D , Scott DP , Geisbert TW , Kawaoka Y , Katze MG , Feldmann H , Messaoudi I. Antibodies are necessary for rVSV/ZEBOV-GP-mediated protection against lethal Ebola virus challenge in nonhuman primates . Proc Natl Acad Sci U S A . 2013 ; 110 : 1893 - 8 .
18. Barrenas F , Green RR , Thomas MJ , Law GL , Proll SC , Engelmann F , Messaoudi I , Marzi A , Feldmann H , Katze MG . Next generation sequencing reveals a controlled immune response to Zaire Ebola virus challenge in cynomolgus macaques immunized with VSVΔG/EBOVgp . Clin Vaccine Immunol . 2015 ; 22 ( 3 ): 354 - 6 .
19. Twenhafel NA , Mattix ME , Johnson JC , Robinson CG , Pratt WD , Cashman KA , Wahl-Jensen V , Terry C , Olinger GG , Hensley LE , Honko AN . Pathology of experimental aerosol Zaire ebolavirus infection in rhesus macaques . Vet Pathol . 2013 ; 50 : 514 - 29 .
20. Adiconis X , Borges-Rivera D , Satija R , DeLuca DS , Busby MA , Berlin AM , Sivachenko A , Thompson DA , Wysoker A , Fennell T , Gnirke A , Pochet N , Regev A , Levin JZ . Comparative analysis of RNA sequencing methods for degraded or low-input samples . Nat Methods . 2013 ; 10 : 623 - 9 .
21. Matranga CB , Andersen KG , Winnicki S , Busby M , Gladden AD , Tewhey R , Stremlau M , Berlin A , Gire SK , England E , Moses LM , Mikkelsen TS , Odia I , Ehiane PE , Folarin O , Goba A , Kahn SH , Grant DS , Honko A , Hensley L , Happi C , Garry RF , Malboeuf CM , Birren BW , Gnirke A , Levin JZ , Sabeti PC . Enhanced methods for unbiased deep sequencing of Lassa and Ebola RNA viruses from clinical and biological samples . Genome Biol . 2014 ; 15 : 519 .
22. Gire SK , Goba A , Andersen KG , Sealfon RSG , Park DJ , Kanneh L , Jalloh S , Momoh M , Fullah M , Dudas G , Wohl S , Moses LM , Yozwiak NL , Winnicki S , Matranga CB , Malboeuf CM , Qu J , Gladden AD , Schaffner SF , Yang X , Jiang P-P , Nekoui M , Colubri A , Coomber MR , Fonnie M , Moigboi A , Gbakie M , Kamara FK , Tucker V , Konuwa E , et al. Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak . Science. 2014 ; 345 ( 6202 ): 1369 - 72 .
23. Geisbert TW , Hensley LE , Jahrling PB , Larsen T , Geisbert JB , Paragas J , Young HA , Fredeking TM , Rote WE , Vlasuk GP . Treatment of Ebola virus infection with a recombinant inhibitor of factor VIIa/tissue factor: a study in rhesus monkeys . Lancet . 2003 ; 362 : 1953 - 8 .
24. Caballero IS , Yen JY , Hensley LE , Honko AN , Goff AJ , Connor JH . Lassa and Marburg viruses elicit distinct host transcriptional responses early after infection . BMC Genomics . 2014 ; 15 : 960 .
25. Bolger AM , Lohse M , Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data . Bioinformatics . 2014 ; 30 : 2114 - 20 .
26. Cunningham F , Amode MR , Barrell D , Beal K , Billis K , Brent S , Carvalho-Silva D , Clapham P , Coates G , Fitzgerald S , Gil L , Giron CG , Gordon L , Hourlier T , Hunt SE , Janacek SH , Johnson N , Juettemann T , Kahari AK , Keenan S , Martin FJ , Maurel T , McLaren W , Murphy DN , Nag R , Overduin B , Parker A , Patricio M , Perry E , Pignatelli M , et al. Ensembl 2015 . Nucleic Acids Res . 2014 ; 43 : D662 - 9 .
27. Trapnell C , Pachter L , Salzberg SL . TopHat: discovering splice junctions with RNA-Seq . Bioinformatics . 2009 ; 25 : 1105 - 11 .
28. Anders S. HTSeq: analysing high-throughput sequencing data with Python . URL http://www-huber.embl.de/users/anders/HTSeq/doc/overview.html 2010 .
29. Robinson MD , McCarthy DJ , Smyth GK . edgeR: a Bioconductor package for differential expression analysis of digital gene expression data . Bioinformatics . 2010 ; 26 : 139 - 40 .
30. Smyth GK . Linear models and empirical bayes methods for assessing differential expression in microarray experiments . Stat Appl Genet Mol Biol . 2004 ; 3 : Article3 .
31. Novoradovskaya N , Whitfield ML , Basehore LS , Novoradovsky A , Pesich R , Usary J , Karaca M , Wong WK , Aprelikova O , Fero M , Perou CM , Botstein D , Braman J . Universal reference RNA as a standard for microarray experiments . BMC Genomics . 2004 ; 5 : 20 .
32. Wauquier N , Becquart P , Padilla C , Baize S , Leroy EM . Human fatal zaire ebola virus infection is associated with an aberrant innate immunity and with massive lymphocyte apoptosis . PLoS Negl Trop Dis . 2010 ; 4 .
33. Ebihara H , Rockx B , Marzi A , Feldmann F , Haddock E , Brining D , LaCasse RA , Gardner D , Feldmann H . Host response dynamics following lethal infection of rhesus macaques with Zaire ebolavirus . J Infect Dis . 2011 ; 204 Suppl Suppl 3 : S991 - 9 .
34. McElroy A , Erickson B . Ebola hemorrhagic fever: novel biomarker correlates of clinical outcome . J Infect Dis . 2014 ; 210 ( 4 ): 558 - 66 .
35. Schoggins JW , Wilson SJ , Panis M , Murphy MY , Jones CT , Bieniasz P , Rice CM . A diverse range of gene products are effectors of the type I interferon antiviral response . Nature . 2011 ; 472 : 481 - 5 .
36. Uchiumi F , Miyazaki S , Tanuma S. The possible functions of duplicated ets (GGAA) motifs located near transcription start sites of various human genes . Cell Mol Life Sci . 2011 ; 68 : 2039 - 51 .
37. Hensley LE , Stevens EL , Yan SB , Geisbert JB , Macias WL , Larsen T , DaddarioDiCaprio KM , Cassell GH , Jahrling PB , Geisbert TW . Recombinant human activated protein C for the postexposure treatment of Ebola hemorrhagic fever . J Infect Dis . 2007 ; 196 Suppl Suppl 2 : S390 - 9 .
38. Yen JY , Garamszegi S , Geisbert JB , Rubins KH , Geisbert TW , Honko A , Xia Y , Connor JH , Hensley LE . Therapeutics of Ebola hemorrhagic fever: wholegenome transcriptional analysis of successful disease mitigation . J Infect Dis . 2011 ; 204 Suppl Suppl 3 : S1043 - 52 .
39. Malhotra S , Yen JY , Honko AN , Garamszegi S , Caballero IS , Johnson JC , Mucker EM , Trefry JC , Hensley LE , Connor JH . Transcriptional profiling of the circulating immune response to Lassa virus in an aerosol model of exposure . PLoS Negl Trop Dis . 2013 ; 7 : e2171 .
40. Connor JH , Yen J , Caballero IS , Garamszegi S , Malhotra S. Transcriptional profiling of the immune response to Marburg virus infection . J Virol . 2015 ; 89 : 9865 - 74 .
41. Fink J , Gu F , Ling L , Tolfvenstam T , Olfat F , Chin KC , Aw P , George J , Kuznetsov VA , Schreiber M , Vasudevan SG , Hibberd ML . Host gene expression profiling of dengue virus infection in cell lines and patients . PLoS Negl Trop Dis . 2007 ; 1 .
42. Messaoudi I , Amarasinghe GK , Basler CF . Filovirus pathogenesis and immune evasion: insights from Ebola virus and Marburg virus . Nat Rev Microbiol . 2015 ; 13 : 663 - 76 .
43. Hensley LE , Young HA , Jahrling PB , Geisbert TW . Proinflammatory response during Ebola virus infection of primate models: possible involvement of the tumor necrosis factor receptor superfamily . Immunol Lett . 2002 ; 80 : 169 - 79 .
44. Wahl-Jensen V , Kurz S , Feldmann F , Buehler LK , Kindrachuk J , DeFilippis V , da Silva Correia J , Früh K , Kuhn JH , Burton DR , Feldmann H . Ebola virion attachment and entry into human macrophages profoundly effects early cellular gene expression . PLoS Negl Trop Dis . 2011 ; 5 : e1359 .
45. Ayithan N , Bradfute SB , Anthony SM , Stuthman KS , Bavari S , Bray M , Ozato K. Virus-like particles activate type I interferon pathways to facilitate postexposure protection against Ebola virus infection . PLoS One . 2015 ; 10 : e0118345 .
46. Groom JR , Luster AD . CXCR3 ligands: redundant, collaborative and antagonistic functions . Immunol Cell Biol . 2011 ; 89 : 207 - 15 .
47. Hsieh M-F , Lai S-L , Chen J-P , Sung J-M , Lin Y-L , Wu-Hsieh BA , Gerard C , Luster A , Liao F . Both CXCR3 and CXCL10/IFN-inducible protein 10 are required for resistance to primary infection by dengue virus . J Immunol . 2006 ; 177 : 1855 - 63 .
48. Fadel SA , Bromley SK , Medoff BD , Luster AD . CXCR3-deficiency protects influenza-infected CCR5-deficient mice from mortality . Eur J Immunol . 2008 ; 38 : 3376 - 87 .
49. Dufour JH , Dziejman M , Liu MT , Leung JH , Lane TE , Luster AD . IFN-gammainducible protein 10 (IP-10; CXCL10 ) -deficient mice reveal a role for IP-10 in effector T cell generation and trafficking . J Immunol . 2002 ; 168 : 3195 - 204 .
50. Zeremski M , Petrovic LM , Chiriboga L , Brown QB , Yee HT , Kinkhabwala M , Jacobson IM , Dimova R , Markatou M , Talal AH . Intrahepatic levels of CXCR3- associated chemokines correlate with liver inflammation and fibrosis in chronic hepatitis C. Hepatology . 2008 ; 48 : 1440 - 50 .
51. Nakanishi Y , Lu B , Gerard C , Iwasaki A. CD8(+) T lymphocyte mobilization to virus-infected tissue requires CD4(+) T-cell help . Nature . 2009 ; 462 : 510 - 3 .
52. Brainard DM , Tager AM , Misdraji J , Frahm N , Lichterfeld M , Draenert R , Brander C , Walker BD , Luster AD . Decreased CXCR3+ CD8 T cells in advanced human immunodeficiency virus infection suggest that a homing defect contributes to cytotoxic T-lymphocyte dysfunction . J Virol . 2007 ; 81 : 8439 - 50 .
53. Geisbert TW , Daddario-Dicaprio KM , Geisbert JB , Reed DS , Feldmann F , Grolla A , Ströher U , Fritz EA , Hensley LE , Jones SM , Feldmann H . Vesicular stomatitis virus-based vaccines protect nonhuman primates against aerosol challenge with Ebola and Marburg viruses . Vaccine . 2008 ; 26 : 6894 - 900 .