Rabies virus

Rabies is the prototype neurotropic zoonotic encephalomyelitis and one of the oldest infectious diseases known to humanity. Once neurological symptoms begin it is effectively 100% fatal, yet it is fully preventable if timely wound care, vaccination and rabies immunoglobulin are given before the virus reaches the central nervous system. That asymmetry between near-universal fatality after symptom onset and near-universal preventability before it has shaped the clinical and public-health response for over a century. The global burden remains approximately 59,000 deaths annually, almost all from canine rabies in low-income settings.

Discovery and historical significance

Rabies has been recognised for millennia. The pre-Mosaic Eshnunna Code of Mesopotamia (c. 2300 BC) sets out fines for the owner of a rabid dog that kills a free person, and Hippocrates, Aristotle and Celsus describe the syndrome. Celsus coined the term hydrophobia and recognised that the saliva of a rabid animal was infectious. Through the seventeenth and eighteenth centuries, soldiers’ journals documented epidemics of “hydrophobia” in animals and people across Europe.

The modern history begins with Louis Pasteur, who developed an attenuated preparation of rabies virus by serial passage through rabbit spinal cord and on 6 July 1885 administered it to Joseph Meister, a nine-year-old severely bitten by a rabid dog. Meister survived. Pasteur’s achievement is regarded as opening the modern era of vaccinology, and it predated any understanding of viruses or the mature germ theory.

Negri described the eosinophilic cytoplasmic inclusion bodies that now bear his name in 1903, providing the first reliable histological marker of the disease. The virus itself was not visualised by electron microscopy until the 1960s. Goldwasser and Kissling introduced fluorescent-antibody staining in 1958, the basis for modern rapid diagnosis. Modern cell-culture-grown inactivated vaccines (human diploid cell vaccine, purified chick embryo cell vaccine, Verorab) replaced the nerve-tissue preparations of the Pasteur era from the 1960s onwards and have transformed both the safety and the efficacy of prophylaxis.

Classification, structure, and genome

Classification

Rabies virus belongs to the genus Lyssavirus (from the Greek lyssa, “frenzy”), family Rhabdoviridae, order Mononegavirales. The formal species name is Lyssavirus rabies; the common name “rabies virus” remains the everyday term in clinical and laboratory practice. Rhabdoviridae takes its name from the Greek rhabdos, “rod”, describing the characteristic bullet-shaped virion.

The genus contains seventeen recognised lyssavirus species. Classical rabies virus is the type species and causes over 99% of human rabies. The remaining lyssaviruses share the same essential biology, clinical syndrome and replication strategy, but differ enough at the antigenic level that vaccine cross-protection cannot be assumed across the whole genus.

Other lyssaviruses and phylogroup cross-protection. Six lyssaviruses other than RABV have caused documented human disease; many more circulate in bat populations without clinically apparent human spillover. The clinical syndrome is indistinguishable across the genus, but vaccine cross-protection is not. The genus is divided into three antigenic phylogroups, and standard rabies vaccines and rabies immunoglobulin reliably cross-protect within Phylogroup 1 only. Current ICTV taxonomy formally recognises two phylogroups, placing the most divergent lyssaviruses (the Phylogroup 3 grouping used here) outside both.

Phylogroup 1 (cross-protected by standard PEP and PrEP): classical rabies virus (RABV; global, dominant agent of human disease, principally dog-mediated); Duvenhage virus (DUVV; Southern African insectivorous bats; three documented human deaths, two from South Africa); Australian bat lyssavirus (ABLV; Australian flying foxes and microbats; three confirmed human fatalities); European bat lyssaviruses 1 and 2 (EBLV-1, EBLV-2; European bats; a handful of human fatalities across Europe); Irkut virus (IRKV; Russian Far East; one confirmed case); Aravan and Khujand viruses (Central Asian bats; no documented human cases).

Phylogroup 2 (not cross-protected by standard vaccines or RIG): Mokola virus (MOKV; sub-Saharan Africa including South Africa; two documented human cases); Lagos bat virus (LBV; African fruit bats; no documented human cases, but well-established in Southern African bats); Shimoni bat virus (East African bats).

Phylogroup 3 (poorly cross-protected): West Caucasian bat virus (WCBV; Caucasus region bats); Ikoma lyssavirus (single isolate, Tanzania); Matlo bat lyssavirus (MBLV; recently described from South African Miniopterus natalensis bats, divergent and provisionally Phylogroup 3).

The phylogroup distinction is the operational pivot. A bite from any recognised lyssavirus is managed as Category III by default, but the clinical reality is that standard PEP provides no proven protection against Mokola virus, Lagos bat virus or other Phylogroup 2/3 agents, a meaningful concern for bat exposures in sub-Saharan Africa where these viruses circulate.

Virion structure

The virion is bullet-shaped (rounded at one end, flat at the other), approximately 75 by 180 nm, and enveloped. A helical nucleocapsid coils within a host-derived lipid envelope studded with trimeric glycoprotein G spikes. G is the major neutralising antigen and the target of all current vaccines.

The matrix protein (M) lines the inside of the envelope and links it to the underlying ribonucleoprotein complex, also driving budding. Inside, the helical ribonucleoprotein contains the genome RNA coated by nucleoprotein (N), with the polymerase complex (L and its cofactor P) associated.

Genome organisation

The genome is a linear, non-segmented, negative-sense single-stranded RNA molecule of approximately 12 kilobases (around 11,932 nucleotides). It encodes only five proteins in a fixed 3’-N-P-M-G-L-5’ order:

• N (nucleoprotein): encapsidates the RNA into the helical nucleocapsid.
• P (phosphoprotein): polymerase cofactor; also blunts host interferon signalling.
• M (matrix): links the nucleocapsid to the envelope and drives budding.
• G (glycoprotein): surface trimer; mediates receptor binding and pH-dependent fusion; carries the neutralising epitopes.
• L (large polymerase): the RNA-dependent RNA polymerase.

A single 3’ promoter drives monocistronic transcription of all five mRNAs. Polymerase reads through with partial detachment between genes, producing a characteristic descending transcription gradient: N most abundant, L least. This concentration ratio matches the structural needs of the assembling virion.

Replication cycle

Rabies virus replicates entirely in the cytoplasm. The cycle is slow, non-cytopathic, and yields fewer progeny per cell than most other RNA viruses.

Entry. At the site of inoculation, virions in saliva first replicate locally in striated muscle. G binds neuronal receptors at the neuromuscular junction: principally the nicotinic acetylcholine receptor (nAChR), with neural cell adhesion molecule (NCAM) and p75 neurotrophin receptor (p75NTR) also implicated. Endocytosis is followed by pH-dependent fusion of the viral envelope with the endosomal membrane, releasing the nucleocapsid into the cytoplasm.

Transcription and replication. L (the RNA-dependent RNA polymerase), with P as its cofactor, transcribes five capped and polyadenylated monocistronic mRNAs from a single 3’ promoter. Stuttering polymerase release between gene junctions yields the characteristic descending gradient (N most abundant, L least). As N accumulates and reaches a threshold concentration, the polymerase switches from transcription to genome replication, producing full-length positive-sense antigenome and in turn new negative-sense genomes encapsidated by N.

Assembly and egress. M links the assembled nucleocapsid to membranes studded with G. Two egress patterns are critical:

• In neurons, the virus buds primarily through intracytoplasmic membranes, producing the eosinophilic cytoplasmic inclusions recognised histologically as Negri bodies.
• In salivary gland epithelial cells, the virus buds through the plasma membrane into saliva, the route by which the next host is infected via a bite.

This site-specific budding pattern is essential to the transmission cycle: the virus reaches the CNS, replicates there, spreads centrifugally to the salivary glands, and emerges via saliva to seed the next exposure.

Pathogenesis

Rabies is strictly neurotropic and produces clinical illness through neuronal dysfunction rather than destruction. The pathology is striking for its paucity of histological damage relative to the severity and uniform lethality of disease.

After inoculation in saliva, the virus undergoes a period of local replication near the bite site. It then binds neuronal receptors at the motor end-plate and enters peripheral nerve terminals. From there it is carried by retrograde fast axonal transport centripetally to the dorsal root ganglia, then to the spinal cord and brain. There is no viraemia. The virus is sequestered in muscle and within axons, and is not exposed to circulating antibody during the incubation period. This is the mechanistic reason that post-exposure prophylaxis works: vaccine-induced or passively administered antibody can reach the virus while it is still on its way to the CNS, but cannot follow it across the blood-brain barrier once it is there.

Within the CNS, the virus spreads along neuroanatomical pathways. The P protein interferes with host interferon signalling, blunting the innate response. Limbic system involvement produces the agitation, hyperexcitability and autonomic dysfunction of the furious form; neocortical and spinal-cord involvement produces the ascending paralysis of the paralytic form. Both progress through coma to death over days to weeks.

From the established CNS infection, the virus then spreads centrifugally along sensory and autonomic nerves to multiple peripheral tissues: heart, adrenal medulla, gastrointestinal tract, skin, and most critically the salivary glands, where high-titre replication and apical plasma-membrane budding release infectious virus into saliva. The skin involvement underlies the diagnostic value of the nuchal biopsy.

Epidemiology

The WHO estimates approximately 59,000 human deaths from rabies each year, almost certainly an underestimate given limited surveillance in endemic settings. Over 99% of cases are caused by domestic dog exposure. The burden is concentrated in Asia and Africa, where poverty, limited access to PEP, and large unvaccinated dog populations sustain transmission.

Wildlife reservoirs vary regionally:

• In the United States and Canada, dog rabies has been controlled; wildlife now dominates (bats in every state except Hawaii, raccoons along the eastern seaboard, skunks across the Midwest and prairie Canada, and grey and red foxes largely controlled by oral wildlife-vaccination programmes).
• In Europe, decades of oral red-fox vaccination have eliminated fox rabies from most of the continent; residual cases are bat-mediated.
• In Asia and Africa, domestic dogs remain the dominant reservoir.
• In Australia, classical RABV is absent; rabies-like disease is caused by Australian bat lyssavirus in flying foxes and microbats.

The bat lyssaviruses circulate globally and represent a structurally different problem from the canine cycle. They cannot be addressed by mass canine vaccination and require exposure-based prophylaxis. Sustained transmission within marine mammal populations had been considered exceptional until sustained seal-to-seal spread was documented in Cape fur seals off southern Africa from 2022.

Natural history

Once symptomatic, rabies is effectively 100% fatal. Documented survivors are exceedingly rare (fewer than fifteen in the medical literature) and the great majority had received at least some pre- or post-exposure vaccination before clinical onset. The 2004 Milwaukee case (Jeanna Giese, 15, bat-virus exposure with no prior vaccine) survived with significant neurological sequelae and remains the most-cited example of survival without prior immunity, but is likely attributable to a less neurovirulent bat-variant virus rather than to the treatment given.

Death follows symptom onset by 7 to 14 days through coma, autonomic collapse and respiratory or cardiac arrest. There is no chronic phase and no carrier state. The incubation period is typically 1 to 3 months, with a wide range from a few days to over a year and rare documented cases at six years or more. Incubation length depends on the distance from the inoculation site to the central nervous system (proximal bites are shortest), the inoculum size, and how richly innervated the bite site is. A facial bite has a notably shorter incubation than a peripheral limb bite.

Clinical presentations and complications

The prodrome lasts up to ten days and is non-specific: fever, chills, malaise, fatigue, insomnia, anorexia, headache and irritability. The earliest neurological symptom is often paraesthesia, pain or pruritus at or near the bite site, a localising and effectively pathognomonic sign, attributable to local sensory ganglion infection. The original wound may by then have healed.

Two clinical forms follow:

Furious (encephalitic) rabies (~80% of cases): episodes of hyperexcitability, agitation, aggression, hallucinations and confusion separated by lucid intervals; autonomic dysfunction (hypersalivation, sweating, piloerection, priapism, cardiac arrhythmias); hydrophobia (painful 5 to 15 second spasms of the diaphragm and inspiratory muscles triggered by swallowing, then by the sight, sound or even mention of water); and aerophobia, the same spasms triggered by an air draft. Hydrophobia is pathognomonic but more common with canine than bat exposure.

Paralytic (dumb) rabies (~20% of cases): ascending flaccid weakness starting in the bitten limb, sphincter involvement, sensory disturbance, then bulbar and respiratory weakness. Hydrophobia is rare in this form; survival from onset is somewhat longer.

Both forms progress to coma, multi-organ complications and death, typically within 7 to 14 days of symptom onset. The clinical differential includes psychiatric disorder, Guillain-Barré syndrome, herpes B virus encephalitis, tetanus, anti-NMDA-receptor encephalitis, post-vaccinal encephalomyelitis (from the nerve-tissue vaccines still used in some low-income settings), and cerebral malaria in endemic regions.

Diagnosis

Routine laboratory investigations are typically unremarkable in early disease. CT brain is usually normal; MRI may show non-specific signal changes in the brainstem, hippocampus and spinal cord, useful chiefly for excluding alternatives. CSF shows a mononuclear pleocytosis (under 100 cells/mL) with mildly raised protein.

No single ante-mortem test is sensitive at all stages. A panel of specimens is submitted and often repeated serially:

• Saliva (RT-PCR) is the most sensitive ante-mortem test. Submit every 3 to 6 hours to maximise yield.
• Nuchal skin biopsy: full-thickness, posterior neck at the hairline, including hair follicles. DFA or RT-PCR demonstrates virus in nerve fibres around the follicles.
• CSF: RT-PCR (lower sensitivity than saliva), and anti-rabies neutralising antibody in an unvaccinated patient (diagnostic if positive).
• Serum: neutralising antibody often appears late or not at all; a negative serum does not exclude.

A positive result on any of these confirms the diagnosis; a single negative does not exclude. Post-mortem remains the gold standard when ante-mortem confirmation is not achieved:

• Direct fluorescent antibody (DFA) test on brain tissue: typically brainstem and cerebellum. Sensitivity and specificity over 99%. The post-mortem test of choice and the test performed at the NICD reference laboratory.
• RT-PCR on the same tissue: complementary molecular confirmation.
• Histology for Negri bodies: supportive but less sensitive than DFA.

Management

Once symptomatic, rabies is managed with supportive and palliative care alone. There is no licensed antiviral and no proven specific therapy.

The Milwaukee protocol (drug-induced coma combined with ribavirin, amantadine and ketamine, popularised after the 2004 Wisconsin survivor) has subsequently failed in over thirty documented attempts. There is no proven viral clearance, no neuroprotective benefit at autopsy in treated patients, and ketamine is ineffective against rabies in cultured neurons. Authoritative reviews now conclude that the protocol has no role in current management.

Care focuses on symptomatic control of agitation and autonomic instability (sedation, often with benzodiazepines and opioids), hydration, prevention of secondary complications, and dignified end-of-life care. Family and clinical contacts should be assessed for infectious exposure (bites, mucosal contamination with saliva or CNS tissue, open-wound contamination) and offered PEP only where exposure is documented.

Prevention and public health

Prevention rests on three pillars: immediate wound care, active immunisation with rabies vaccine, and passive immunisation with rabies immunoglobulin (RIG). The combination, given correctly and in time, is essentially 100% effective.

Vaccination

Modern cell-culture-grown inactivated whole-virus vaccines have replaced the nerve-tissue preparations of the Pasteur era.

Examples of rabies vaccines:

• Chirorab®: a purified chick embryo cell vaccine used heavily in the South African public sector (formerly Rabipur®).
• Verorab®: a purified Vero cell vaccine, widely available in the South African private sector.
• RABAVERT®: a highly purified chick embryo cell vaccine used in North America and other regions.
• IMOVAX® Rabies: a human diploid cell vaccine used globally, including in Canada and the United States.

Pre-exposure prophylaxis (PrEP) is a 2-dose intramuscular schedule on days 0 and 7 in immunocompetent adults. Intradermal alternatives (2 sites per visit on days 0 and 7, 0.1 mL per site) are dose-sparing. Immunocompromised patients require a 3-dose schedule on days 0, 7 and 21 to 28. Boosters are titre-guided for ongoing high-risk exposure. PrEP is recommended for laboratory workers handling lyssaviruses, veterinarians, animal handlers, wildlife rehabilitators, spelunkers, and travellers to endemic regions where access to PEP may be delayed.

Post-exposure prophylaxis

PEP is the operational core of rabies prevention and combines wound care, vaccination, and RIG.

Wound care. The first intervention is also the most important. Wash all wounds with soap and running water for fifteen minutes, then flush thoroughly. Disinfect with chlorhexidine 0.05% or povidone-iodine 10%. Avoid suturing unless required for haemostasis; never infiltrate local anaesthetic into the wound, which risks forcing virus into deeper tissues. Update tetanus prophylaxis; consider antibiotic prophylaxis (amoxicillin- clavulanate) for a clearly contaminated bite.

WHO exposure categories.

Category	Exposure	Management
I	Touching or feeding animals, licks on intact skin	None (no skin breach); wash exposed area only
II	Nibbling of uncovered skin, minor scratches without bleeding	Wound care + vaccine alone, no RIG
III	Single or multiple transdermal bites or scratches, mucosal contact, licks on broken skin, any contact with a bat	Wound care + vaccine + RIG

The “any bat contact equals Category III” rule reflects the bat-bite paradox: bat teeth are tiny and bites may leave no visible wound. Many documented rabies fatalities have followed unrecognised bat exposures.

PEP schedule. 4-dose intramuscular schedule on days 0, 3, 7, and one day between 14 and 28. Same dose for paediatric and adult patients (one full vial per administration). Deltoid in adults and older children; anterolateral thigh in children under 2 years. Never gluteal (impaired immunogenicity).

Previously vaccinated patients with a fresh exposure require only a 2-dose schedule on days 0 and 3 and no RIG: endogenous immunity is rapidly boosted.

Rabies immunoglobulin (RIG) provides passive antibody covering the window before vaccine-induced immunity develops.

• HRIG (human rabies immunoglobulin): 20 IU/kg.
• ERIG (equine rabies immunoglobulin): 40 IU/kg.

The entire calculated dose is infiltrated into and around the wound(s); none is given at a distant intramuscular site. Dilute with normal saline if wound volume is insufficient. If wounds are healing or poorly defined, infiltrate as much as anatomically feasible.

Critical pitfalls:

• Never mix RIG with vaccine in the same syringe: they neutralise each other.
• Never inject RIG into the same anatomical site as the vaccine for the same reason.
• RIG is effective only within 7 days of the first vaccine dose: after that, endogenous antibody is already developing and RIG no longer confers benefit and may interfere with the patient’s own vaccine response.

10-day animal observation. A bite from a domestic dog, cat or ferret that can be safely observed is managed with the 10-day rule. These animals shed virus only within the 10 days preceding clinical signs. PEP is started immediately, and may be stopped at day 10 if the animal remains clinically well under veterinary observation. The rule does not extend to wild animals or stray dogs that cannot be reliably observed.

Surveillance and notification

Rabies is a globally notifiable condition under the WHO and the World Organisation for Animal Health (WOAH, formerly OIE). Reporting is on two axes: human cases through public-health notification systems, and animal cases through national veterinary services. WOAH receives quarterly returns on animal rabies and publishes the World Animal Health Information System (WAHIS). The WHO maintains a global database of human rabies deaths, recognising substantial under-reporting from endemic regions.

Outbreak response

Human rabies is managed at the individual rather than population level (PEP for the exposed individual, source-animal investigation and observation or euthanasia). Animal rabies outbreaks trigger quarantine, intensified surveillance, ring vaccination of domestic animals, and (where applicable) oral wildlife vaccination campaigns. Sustained rabies transmission in Cape fur seals off southern Africa since 2022 is the most recent example of a novel outbreak requiring a coordinated intersectoral response.

Elimination and eradication

The WHO and the Global Alliance for Rabies Control set the goal of zero human dog-mediated rabies deaths by 2030 (“Zero by 30”). The intervention package is:

• Mass dog vaccination at sustained coverage above 70%: a single-dose immunisation programme with sufficient coverage drives the basic reproduction number below one in dog populations.
• Universal access to PEP for bitten humans, ideally free at point of care.
• Public education on bite avoidance and PEP-seeking behaviour.
• One Health coordination between human health, veterinary, and wildlife sectors.

Sustained dog-vaccination programmes have successfully eliminated dog rabies in much of Latin America (Pan American Health Organization regional certification) and the western United States. Oral wildlife-vaccination programmes (recombinant vaccinia-rabies G protein vaccine baits) have eliminated red fox rabies across most of Western Europe and have controlled raccoon rabies along the eastern United States.

South African context

Rabies is endemic in South Africa. Approximately ten laboratory-confirmed human cases are reported per year, concentrated in KwaZulu-Natal, the Eastern Cape, Mpumalanga, the Free State and Limpopo.

Three established terrestrial cycles are recognised:

• A canid cycle maintained in domestic dogs (the principal source of human exposure) and black-backed jackals, dominant in the eastern provinces.
• A mongoose (herpestid) cycle maintained in the yellow mongoose (Cynictis penicillata) on the central plateau.
• A bat-eared fox cycle in the Western and Northern Cape, the western Free State and parts of the Eastern Cape and North West.

Five lyssavirus species (besides RABV) have been detected in South Africa: Duvenhage virus, Mokola virus, Lagos bat virus, Matlo bat lyssavirus, and Phala bat lyssavirus. Duvenhage virus has caused three confirmed African human deaths; Mokola virus is Phylogroup 2 and is not cross-protected by standard vaccine or RIG. Any bat contact in South Africa is managed as Category III regardless.

Notification operates on two axes: human cases through the NMC framework (reported to the Department of Health, typically via the NICD NMC App), and animal cases under the Animal Diseases Act, 1984 (Act No. 35 of 1984) through state Veterinary Services. The NICD is the national reference laboratory, and One-Health coordination with DALRRD veterinary services underpins surveillance.

Rabies in Cape fur seals (Arctocephalus pusillus). From 2024, sustained transmission of rabies has been documented in Cape fur seals along the Western and Northern Cape coastline, with retrospective laboratory evidence of the virus in the population since 2022 and a geographic range now extending to Namibia. The isolates appear to be canid-biotype, consistent with spillover from the coastal jackal cycle, but sustained seal-to-seal transmission is now established, representing the first documented rabies outbreak in marine mammals globally. No human cases have been reported to date, but pre-exposure prophylaxis is recommended for water-sports users (surfers, divers, open-water swimmers) and seal-handling personnel in affected areas, and beach advisories caution against approaching seals alive or dead.

South African operational framework. The 2021 NDoH National Guidelines for the Prevention of Rabies in Humans, South Africa is the primary source for wound care, vaccine and RIG schedules, exposure categorisation, special-population considerations (HIV, paediatric, pregnancy) and One-Health coordination. Free vaccine and RIG are supplied through the public sector for category II and III exposures.

References and recommended reading

• Jackson AC. Rhabdoviruses. In: Richman DD, Whitley RJ, Hayden FG (eds.), Clinical Virology, 4th edition, Chapter 41. Washington: ASM Press; 2017.
• Burrell CJ, Howard CR, Murphy FA. Rhabdoviruses. In: Fenner and White’s Medical Virology, 5th edition, Chapter 27. Academic Press / Elsevier; 2017.
• National Department of Health, South Africa. National Guidelines for the Prevention of Rabies in Humans, South Africa. September 2021.
• Wallace RM, Shlim DR. Rabies. In: CDC Yellow Book: Health Information for International Travel, 2026 edition. US Centers for Disease Control and Prevention; 2025.
• National Institute for Communicable Diseases. What you need to know about rabies in seals in South Africa (August 2025). NICD Centre for Emerging Zoonotic and Parasitic Diseases.