Since its discovery in 1976, and especially since Richard Preston published The Hot Zone in 1994, Ebola has been an evocative word in the American consciousness. A 1995 outbreak in Zaire (now the Democratic Republic of the Congo), reported on extensively by The New York Times, cemented the word “Ebola” into the American lexicon. The result: Ebola virus, named after the Ebola River in the Democratic Republic of the Congo, has both frightened and fascinated for decades. Many Americans know the rough outlines: that it is an African virus, that it has killed 50–90 percent of those infected in most outbreaks, that it emerges suddenly from the forest before disappearing again. Its emergence and explosion in West Africa in December 2013 has elevated it from a fascinating curiosity to the front pages and to the front of minds. Along with its high-profile appearance on the global stage have come misunderstanding and fear. These stem from a lack of knowledge—not from lack of interest, but rather from the relative inaccessibility of scientific information. Among these gaps in the public knowledge are what Ebola virus actually is, where it comes from (as best we know), how it causes disease, and how it spreads. We begin with the simplest question of the four: What is Ebola virus?
Viruses represent a vast taxon apart from the recognized kingdoms of life. They are vanishingly small, yet infect every known living organism. Viral taxonomy is a sort of poor man’s approximation of taxonomy among the three kingdoms of life. Within the viral universe, Ebola virus fits into a small, poorly understood family of viruses, the Filoviridae. The filoviruses fall into three genera within Filoviridae: Ebolavirus, Marburgvirus, and Cuevavirus. Of these, Cuevavirus is the simplest; its sole constituent virus, Lloviu virus, is known only by a genome sequence recovered from dead European bats.1
Marburg virus, discovered in 1967, was the first filovirus to be identified and has since caused small, sporadic outbreaks throughout Africa. The genus Marburgvirus consists of one viral species, Marburg marburgvirus, which consists of two member viruses, Marburg virus and Ravn virus.2 These two viruses, though slightly different, cause the same clinical disease in infected individuals, as do the ebolaviruses, to which we now turn.
The genus Ebolavirus contains five viral species, four of which are known to cause human disease and three of which have caused outbreaks. Zaire ebolavirus comprises Ebola virus, the particular virus that has caused the West Africa Ebola epidemic and which has been responsible for the majority of prior human outbreaks as well. The other viruses that cause the same human disease as Ebola virus are Sudan virus (Sudan ebolavirus species), Bundibugyo virus (Bundibugyo ebolavirus species), and Tai Forest virus (Tai forest ebolavirus species); however, these three viruses have thus far been less fatal than Zaire ebolavirus.3 Reston virus, the fifth Ebolavirus species, belongs to the species Reston ebolavirus and was the subject of The Hot Zone, which described a Reston virus outbreak within a Virginia primate holding facility. Reston virus is thought to be nonpathogenic in humans and also the only ebolavirus found outside of Africa. All available epidemiological evidence suggests that it is of Asian origin.
One of the troubling aspects of Ebola virus’ emergence has been its capacity to cause sporadic outbreaks and disappear back into the forest. After the first recognized outbreak in 1976, the virus vanished for nearly twenty years before reappearing in 1995 in the city of Kikwit, Zaire. The scientists and epidemiologists who led the international response in 1976 prioritized identifying the reservoir host of Ebola virus; they hoped that understanding where the virus came from might aid prevention or prediction of future outbreaks. Despite testing hundreds of animals from dozens of species, they came up empty. Similar work following the 1995 outbreak failed to identify the virus or past evidence of infection in any of the 3,066 animals collected.4
In 2005 a research group led by Eric Leroy of the Centre International de Recherches Médicales de Franceville in Gabon finally made a critical discovery. They collected over 1,000 birds, bats, and small terrestrial mammals in the vicinity of an ongoing Ebola outbreak among gorillas and chimpanzees. Among these, individual bats of three species—Hypsignathus monstrosus, Epomops franqueti and Myonycteris torquata—contained antibodies against Ebola virus, indicating past infection. Other bats of the same three species contained fragments of the Ebola virus genome in their livers and spleens. However, the amount of viral RNA in the samples was very low, and attempts to isolate live Ebola virus from these samples were unsuccessful.5 Later studies that tested other bat species for previous infection with Ebola virus identified additional possible reservoirs, including Rousettus aegyptiacus, a reservoir of Marburg virus.6,7 The discovery that multiple species of bat might serve as reservoirs greatly complicates the task of predicting and preventing spillovers into the human population.
Although we now know that various bat species are the probable reservoir, there are still important unknowns about the behavior of Ebola virus in its natural host. Some of these questions have implications for the pattern of Ebola virus outbreaks in humans. We don’t know, for example, why outbreaks are so rare, or why they occur when they do. It may be that Ebola virus prevalence is very low in the reservoir population but surges intermittently. It is also possible that Ebola virus loads in the natural host are so low that viral shedding by bats is typically minimal, yet viral load may also fluctuate and result in increased likelihood of spillover into humans at certain times. The answer may be as simple as that contact between bats and humans is relatively rare, but becomes more frequent under certain environmental conditions.
Additionally, many Ebola virus outbreaks in humans have occurred concurrently with outbreaks in great apes and some have been associated with human exposure to infected ape carcasses. The dynamics of ape-bat interactions are largely unknown, though their consumption of similar types of fruit provides an obvious opportunity for Ebola virus spillover into primates, and the existence of an intermediate host increases the complexity of preventing spillover into humans.
Everything a virus does to promote its own replication and transmission to a new host is driven by its structure and the functions of the proteins encoded by its genome. Ebola virus consists of a protein-coated genome, matrix proteins that support the structure of the virion, or single virus particle, and a lipid envelope that is co-opted from the host-cell membrane when new virions exit the host cell. Together, these factors give Ebola virus and other filoviruses a distinctive thread-like appearance that is strikingly different than that of other known viruses.
Unlike animals, plants, and bacteria, which all use DNA to store genetic information, viruses may use DNA or RNA. DNA viruses include herpesviruses and poxviruses while RNA viruses include influenza viruses, flaviviruses like West Nile virus, and the filoviruses. The RNA viruses are further classified by the organization of their genome, and by how many steps are required to make proteins encoded by their RNA.
The simplest distinction is between positive and negative-sense viral RNA genomes. Positive-sense RNA genomes, such as that of West Nile virus, are essentially viral versions of our own protein-coding messenger RNAs (mRNA); both can be directly translated into proteins by host-cell ribosomes. Often, the proteins of viruses with positive-sense RNA genomes are translated as one or two large polyproteins that are then cleaved into individual proteins.
Negative-sense RNA genomes differ significantly in that they cannot be directly translated into protein. Rather, they must be transcribed into full-length positive-sense RNA anti-genomes to begin genome replication and smaller viral mRNAs as the first step in protein synthesis. Either of these can then be translated into protein by the host protein synthesis machinery. Some RNA genomes exist in multiple pieces, such as the negative-sense segmented RNA genome of influenza virus. Most, however—including filovirus genomes—are a single strand of RNA.
Ebola virus has a 19 kilobase-long negative-sense RNA genome containing seven genes, which encode eight proteins. The nucleoprotein (NP), VP30, and VP35 proteins comprise the protein component of the nucleocapsid complex, the assembly of proteins that tightly coats the viral genomic RNA.8 The major component of this complex is NP, which forms a helical structure that surrounds the viral genome. The L protein is the viral RNA-dependent RNA-polymerase, the enzyme required for transcribing positive-sense copies of the viral genome and transcribing viral mRNAs.
The glycoprotein (GP) is the primary viral surface protein; it is sugarcoated and mediates attachment to the cellular receptor and subsequent entry into the host cell. After GP is made, three different GP molecules associate with each other to form a trimer. This trimer is then coated in sugars, or glycosolated, and is incorporated into the viral envelope, and will ultimately play a critical role in Ebola virus entry into a host cell.
In addition to GP, which is anchored to the surface of the Ebola virion and sticks outward, Ebola virus also produces a soluble version of GP (sGP) that is secreted from infected cells. Though the other Ebola virus proteins are similar structurally and functionally to those produced by Marburg virus, synthesis of sGP is unique to Ebola virus. Soluble GP likely serves a role in immune evasion, specifically in rendering the host antibody response ineffective. Though its role was poorly understood for years, recent evidence suggests sGP induces the host to produce antibodies that will bind itself in addition to full-length GP.9 As a result, sGP effectively soaks up antibodies that might otherwise bind full-length GP and neutralize the virus. Indeed, this role is seemingly so important that Ebola virus actually produces more sGP than full-length GP.
VP40 is the matrix protein that provides structure to virions and lies just under the lipid envelope that the virus steals from its host as it exits the cell.10 VP24 is a poorly understood minor matrix protein that plays a role in assembling the Ebola nucleocapsid complex and also is active in disrupting the host response to infection.11
Through the activity of these different proteins, Ebola virus is able to infect and replicate to tremendous levels inside a host, prevent an effective immune response, and—in the case of primates—cause a devastating disease. All of this requires an initial, complex step mediated by GP: entry into a host-cell.
Viruses as a rule lack the ability to replicate on their own; thus, they must co-opt host cell machinery to replicate and produce new virus particles. For any virus, and Ebola virus is no exception, the first step to replication is entry into a susceptible host cell. Virus entry into a host cell is a multistep process involving both viral and host proteins. Most enveloped viruses enter cells via receptor-mediated endocytosis, which begins after the viral glycoprotein attaches in a highly specific fashion to a receptor molecule on the surface of the host cell. After attachment, the receptor and virus are brought into the cell inside an endosome, which is an acidic vesicle containing enzymes required for the glycoprotein to facilitate exit from the endosome. The virion will effectively disassemble in the endosome, allowing its protein-coated genome to enter the cytoplasm of the host cell, where its replication cycle can begin.
Ebola virus, which is unusually large, uses a somewhat unusual pathway to enter host cells, and identifying its receptor was a challenge not overcome until 2011.12 Rather than entering cells by surface receptor-mediated endocytosis like most enveloped viruses, Ebola virus and other filoviruses enter cells via a process called macropinocytosis, which does not require a specific cell-surface receptor (unlike, say, influenza).13 Rather, macropinocytosis occurs when cell-surface molecules like DC-SIGN or TIM-1,14 which have been identified as Ebola virus entry factors, recognize a large molecule, or in this case, a virion, in the space outside the cell. In the case of Ebola virus, these cell-surface molecules likely recognize host cell-derived molecules on the viral envelope called phosphatidylserines. Ultimately Ebola virus ends up in the same place, an acidified endosome, as other viruses; it just takes a slightly different route in getting there.
Although DC-SIGN and TIM-1 are known Ebola virus entry factors, they do not appear to be strictly required for entry.15 Rather, a cellular cholesterol transport protein, NPC-1, serves as the actual entry receptor inside the endosome.16 After binding NPC-1, the Ebola virus GP trimer is cleaved by specialized enzymes, called proteases, in the endosome causing GP to change its shape, or undergo a conformational change. This conformational change causes GP to insert into the endosomal membrane, causing a fusion of the viral envelope and the endosomal membrane. This fusion facilitates delivery of the Ebola virus genome, wrapped tightly in a protein coat, into the host cell cytoplasm, where the rest of its life cycle proceeds.
After uncoating of the genome in the host-cell cytoplasm, Ebola virus initiates two distinct processes to produce new viral particles. It must both replicate its genome and manipulate the host cell into synthesizing the eight viral proteins. Replication of the genome occurs in two steps, involving both viral and host-cell proteins.
The principal viral protein involved is the L protein, the RNA-dependent RNA-polymerase (RdRp) that each Ebola virion carries into a host cell. Ebola and other RNA viruses must encode an RdRp in their own genomes because host-cell polymerases will only use DNA templates to make RNA, whereas the virus provides an RNA template. The viral polymerase, L in the case of Ebola virus, makes a positive-sense copy of the negative-sense RNA genome. This copy, or anti-genome, serves as a template for L to make a new negative-sense genome (in the same way a negative image can be used to create a copy of a photograph). Multiple host proteins participate in this process as well, likely in the role of stabilizing the RNA-polymerase complex and making sure the viral genome is in a configuration that allows it to be transcribed by L.
Viral protein synthesis, or translation, also requires significant involvement of viral and host proteins. Specifically, viruses lack any ability to synthesize their own proteins, and so both nucleic acid and protein components of the host protein synthesis machinery are essential for the viral life cycle. The first step in viral protein synthesis is L transcribing each of the seven Ebola virus genes into positive-sense viral mRNA, which can be recognized by the host protein translation machinery. Sometimes a less than full length mRNA is made from the GP gene, and synthesis of sGP is the result. Following transcription, the viral mRNAs are modified at each end so the host translation machinery can recognize them.
After a new genome has been produced and the viral proteins synthesized, the RNA and proteins need to assemble into new virions that can leave the cell that produced them and find new targets. The first step in this process is the formation of the nucleocapsid, the RNA-protein complex that comprises the structural core of the virion. NP, by now the most abundant protein in the entire cell, is the key player in this complex. It tightly wraps newly transcribed Ebola virus genomes and is the scaffold for formation of the nucleocapsid that also includes VP35, VP40, VP24, and L. The nucleocapsid has an elongated, helical structure that provides the distinctive shape of an Ebola virion. The nucleocapsid is transported to the cell membrane, where GP awaits. The virion then buds from the cell, stealing some of the cell membrane with GP incorporated in it as it leaves.
Overall, viral replication is so efficient and robust that viral RNA may ultimately become the most abundant RNA in a host cell and viral proteins the most abundant proteins. Over time, a single Ebola virus particle will turn a cell into a virus-producing factory that is able to pump out millions of new virions before it dies. Ultimately, an infected cell may be recognized and killed by host immune cells, or may die due to the relentless production of new virions, each of which steals a piece of the cell membrane on its way out. Exactly what happens to an infected cell is irrelevant to the virus as long as it lives long enough to propagate the infection. The same is true on a larger scale regarding transmission of Ebola virus, a topic that has become particularly fraught over the course of the ongoing epidemic.
Two great misconceptions predominate the public discourse on Ebola. The first is that patients bleed to death; in fact, the percentage of cases with hemorrhaging is much lower than observed for other symptoms. By a large margin the most common symptoms of EVD are fever, malaise, diarrhea, abdominal pain, and vomiting, which are observed in 55–90 percent of patients.17 Hemorrhagic symptoms are observed less frequently and are not necessarily correlated with a fatal outcome. In a large 2014 Sierra Leone study, only 1 of 106 patients exhibited signs of hemorrhage,18 while 19 out of 37 patients experienced some bleeding in a contemporaneous study in Guinea.19 During the 1995 Kikwit outbreak, up to 50 percent of patients exhibited some hemorrhagic symptoms.20,21 While these data suggest that different variants even within an outbreak may be more or less likely to cause hemorrhage, most hemorrhagic symptoms were minor, such as bleeding gums, bleeding from injection sites, and some blood in vomit or diarrhea rather than massive hemorrhage. Hemorrhage has never been evaluated as a leading cause of death due to Ebola virus disease. Rather, death is primarily due to complications of catastrophic dehydration as a result of fluid loss from vomiting and diarrhea that can exceed 5–10 liters/day, resulting in life-threatening electrolyte deficiencies.
The second misconception, by far the more dangerous, is that Ebola can spread through the air like a cold or influenza. The science on how Ebola virus is transmitted is actually quite clear. Both experimental research in macaques and careful epidemiological work during and after human outbreaks have failed to demonstrate any credible evidence of airborne spread. One widely cited 2012 article showed that Ebola virus might be able to spread through the air between pigs and macaques,22 but a follow-up study showed that this could not occur between macaques—a much more relevant model for a human outbreak.23
The epidemiological data from human outbreaks clearly demonstrates that direct bodily contact or contact with the fluids of a person actually suffering from EVD puts one at the highest risk of transmission.24 Contact with blood, vomit, and diarrhea of EVD patients carries a high risk of transmission, as may contact with sweat and saliva late in infection.25 Finally, the current West African epidemic has featured the first documented sexual transmission of Ebola virus in a woman with no risk factors other than having had sex with a survivor. This man’s semen tested positive for virus nearly identical to that isolated from the woman.26
Historically, people with limited contact with EVD patients, or who only were in contact with infected individuals before the onset of disease, have not become sick themselves. This is almost certainly because virus levels in the blood of infected individuals before symptom onset are very low. Early in infection, the virus is largely confined to the liver and the spleen, where it replicates vigorously and causes tremendous damage. Virus only becomes reliably detectable in the blood two to three days after infection, which is when the risk of transmission begins to rise.27 Until then, even people who may ultimately progress to severe EVD and die are extremely unlikely to transmit the virus.
The 2013–2016 West African Ebola epidemic has opened the world’s eyes to what had—until recently—been a frightening but little seen virus. Nevertheless—however terrifying, however (seemingly) unique and horrible the disease—Ebola is bound by the same rules and made of the same kinds of molecules as any other virus. Through a basic understanding of the science underlying Ebola virus we might beat back the fear that has coursed through West Africa and around the world, move toward better, science-based public health policy, and build a framework for responding more rapidly, comprehensively, and effectively to the next epidemic.