What persists after the virus is gone?
Symptoms persisting beyond four weeks were well documented clinically, including breathlessness, fatigue, altered mood, impaired memory and concentration, loss of smell and abnormal somatosensation, but their biological basis was unknown and there was no small animal system in which to study them. Two things were needed. The first was a comparator, because an inflammatory or histological change found a month after a severe respiratory infection means nothing until it is known whether another respiratory virus leaves the same mark. The second was a way to ask how a virus that barely enters a tissue, or does not enter it at all, changes what that tissue does.
The loss of smell posed that second problem in its sharpest form, since olfactory sensory neurons do not express the entry factors SARS-CoV-2 requires and are infected only rarely.
All three publications carry co-led, each pairing the laboratory's hamster infection platform and transcriptional framing with a group bringing different expertise.
Frere 2022, co-corresponding with Zachariou, made the comparator a methodological requirement. Hamsters were infected with SARS-CoV-2 or with 2009 pandemic H1N1 influenza at doses chosen to match peak titres, and profiled at peak infection, one week after clearance and at 31 days, across lung, heart, kidney and six nervous system regions. Acutely the two viruses looked similar. After clearance the pictures diverged, and the discipline of the design is that several findings which would have read as specific to SARS-CoV-2 turn out not to be, including the acute interferon response across tissues, peribronchiolar metaplasia, renal tubular atrophy and loss of olfactory receptor transcripts, all shared with influenza and differing in degree. What survived was narrower. At 31 days the olfactory bulb of influenza animals had returned to baseline while SARS-CoV-2 animals retained interferon signatures, elevated CXCL10 and CCL5 and microglial and myeloid activation, with no viral RNA detectable by quantitative PCR or in situ hybridization and no difference in apoptotic nuclei. The phenotype was present in both sexes. Animals buried fewer marbles at 26 days, and post-mortem olfactory tissue from donors recovered from documented COVID-19 showed correlated inflammatory programs.
Zazhytska 2022, led with the Lomvardas and Overdevest groups, supplied a mechanism for anosmia that runs entirely through cells the virus does not enter. In hamsters the virus infected sustentacular cells, which were transiently depleted and then restored, while neuronal representation stayed constant. Uninfected olfactory sensory neurons nonetheless mounted an antiviral response and then, with a delay, lost expression of olfactory receptor genes and of the signal transduction genes that make odour detection possible, a loss still present at ten days after clearance. In situ Hi-C showed that the long-range cis and trans contacts among receptor gene clusters, which normally converge into specialized interchromosomal compartments, were reduced from one day and remained reduced at ten days. Ultraviolet-inactivated serum from infected animals, given intranasally to naive animals for 12.5 hours, reproduced the loss of contacts with no viral genome transferred. Human autopsy tissue showed the same receptor downregulation and, in sorted neuronal nuclei, the same loss of contacts.
Serafini 2023, again co-corresponding with Zachariou, carried the question into the peripheral sensory nervous system. Viral nucleocapsid transcripts and Isg15 rose in cervical and thoracic dorsal root ganglia and spinal cord within one day and largely resolved by four to seven days, while plaque assay recovered infectious virus from lung and from no neural tissue and nucleocapsid protein was not detectable there. Influenza produced pronounced mechanical hypersensitivity at one day that resolved by four, whereas SARS-CoV-2 produced a slower, milder decline that returned at 28 days in both sexes. At 31 days the ganglia carried 1065 differentially expressed genes of neuropathic character. Upstream regulator analysis nominated ILF3, whose inhibitor YM155 reduced hypersensitivity in three mouse pain models including one driven purely by interferon beta.
The thread is a persistent inflammatory program decoupled from detectable virus, arrived at three times in three tissues. Frere 2022 establishes it in the olfactory bulb with the benchmark that makes the claim specific. Zazhytska 2022 provides the cellular route by which an infection confined to neighbouring cells reaches into a neuron, and does so at the level of nuclear organization rather than signalling. Serafini 2023 extends the pattern to dorsal root ganglia and adds a pharmacological endpoint, though that endpoint is displaced from the virus.
What none of these studies establishes should be stated plainly. Frere 2022 does not demonstrate causality between persistent brain inflammation and altered behaviour, offers two untested hypotheses for the persistence, residual defective genomes or debris and barrier damage admitting commensal organisms, and notes that absence of virus at 31 days is a limit of the assays used. Its behavioural assays were developed in mice and rats, group sizes were reduced by containment constraints, and the human comparison rests on two recovered donors against one control for bulb and two against three for epithelium, comparing a moderate hamster infection with human cases that ended in death. Zazhytska 2022 did not identify the circulating molecule or the receiving neuronal pathway, and did not establish that receptor downregulation causes anosmia, inferring that from knockout mouse phenotypes with no smell testing in either species. Its serum transfer shows sufficiency of something in serum rather than necessity of any identified factor, its nuclear memory proposal is explicitly a hypothesis, the human Hi-C rests on two control and four infected autopsies, and the hamster time course ends at ten days. Serafini 2023 states that the association between viral material and persistent hypersensitivity is correlative, its cell composition shift is bulk deconvolution rather than a cell count, and ILF3 inhibition was never tested in infected animals, so it is not shown to relieve SARS-CoV-2 associated hypersensitivity.
The premise comes from interferon as intervention, where Hoagland 2021 found antiviral transcription in olfactory bulb, brain and small intestine at viral loads orders of magnitude below the airway, and Carrau 2023 showed that this state is produced by circulating interferon from the lung. That finding is what makes a persistent response in a tissue the virus never productively infects an expected rather than a puzzling result. The observation in Eriksen 2021 that interferon-stimulated genes rise in uninfected cells across an infected ocular culture is the same non-cell-autonomous phenomenon in a different tissue. Read together with the imbalanced host response, the corpus supports a general statement that much of the damage attributed to this virus is done by the host response rather than by infection of the damaged cells, and that statement is synthesis across papers rather than a claim any one of them makes.
Intranasal SARS-CoV-2 infection of golden hamsters deposits viral RNA without infectious virus in dorsal root ganglia and spinal cord, producing a milder but longer-lasting mechanical hypersensitivity than influenza A virus and a neuropathic dorsal root ganglion transcriptome at 31 days, from which ILF3 emerges as an analgesic target validated in mouse pain models.
SARS-CoV-2 infection of the olfactory epithelium reorganizes the nuclear architecture of uninfected olfactory sensory neurons, dissipating the interchromosomal compartments that hold olfactory receptor genes and suppressing receptor and signal transduction transcription in both hamsters and human autopsy tissue.
Benchmarked against pandemic influenza in golden hamsters, SARS-CoV-2 uniquely sustains interferon signaling, chemokine production and myeloid activation in olfactory bulb and epithelium a month after clearance, alongside altered behavior and matching signatures in recovered human olfactory tissue.