Dear Editor,
Mpox/monkeypox virus (MPXV) belongs to the Orthopoxvirus genus of the Poxviridae family.
It has a large, linear, double-stranded DNA genome. Since the first identification
in 1970, MPXV is generally confined to tropical regions within African countries.
Alarmingly, multi-national mpox outbreaks occurred in 2022 across a large number of
non-endemic countries particularly in Europe and America
1,2
. Skin lesions are the most common symptom, but the infection can also cause systemic
manifestations such as diarrhea, neurological and respiratory complications
1
.
A recent large-scale clinical study examining 528 mpox cases diagnosed between April
and June 2022 across 16 countries reported that 70 severely affected patients required
hospitalization. Among these hospitalized cases, two patients developed acute kidney
injury, but no further investigation was documented to explain the underlying pathophysiology
1
. In addition, MPXV is frequently detected in urine samples from infected patients
3,4
. These findings triggered our hypothesis that MPXV may directly infect the kidney.
Unfortunately, kidney biopsies from infected patients are not available for investigating
this. Human pluripotent stem cell-derived kidney organoids reconstitute the key cell
types and nephron-like structures of the kidney
5
. Its resemblance with the kidney and human origin makes kidney organoids an excellent
model for studying renal disease, pathogen infection, and drug development
6
. In this study, we evaluated the susceptibility of human kidney organoids to MPXV
infection, mapped virus-host interactions and tested the response to antiviral treatment.
We generated human kidney organoids from induced pluripotent stem cells as described
earlier, which display tubular and glomerular structures and contain endothelial and
stromal cells
7,8
. To investigate the susceptibility for MPXV infection, we inoculated kidney organoids
with cell culture-propagated MPXV virus particles originating from a patient isolate
of the 2022 outbreak (Supplementary methods). Viral DNA levels were quantified by
qRT-PCR and expressed as virus copy numbers (Supplementary Fig. S1). The viral kinetics
from 1, 48, and 96 h to 7 days post-inoculation showed an over 4 log10 increase of
intracellular viral DNA (Fig. 1a). This corresponded to an increase of infectious
titers from 2.8 to 6.8 log10 plaque forming units (PFU) (Fig. 1b). To examine whether
infected organoids can excrete MPXV, we quantified viral DNA and determined infectious
titers of the released viruses into the culture medium beneath the transwell insert.
A pronounced increase of viral DNA levels from 6.8 to 9.2 log10 copies was observed
during the first 48 h, after which viral DNA excretion stabilized (Fig. 1c). Plaque
assay showed that infectious viral titers increased from 1.4 to 3.5 log10 PFU/mL within
48 h, and then gradually decreased to 2.5 log10 PFU/mL at day 7 post inoculation (Fig.
1d).
Fig. 1
MPXV infection in human kidney organoids.
a Quantification of viral DNA levels and (b) infectious viral titers in organoids
(n = 4). c Quantification of viral DNA levels and (d) infectious viral titers in the
culture medium (n = 4-6). e Visualization of MPXV by immunostaining organoids with
an antibody against the virions (green). Blue is DAPI staining of nuclei. Scale bar:
200 μm. f Transmission electron microscopy visualized the MPXV particles in organoids.
Scale bar: left, 2 μm; right, 1 μm. g The assembly steps of MPXV start from crescents
(Cs), proceed to immature virions (IVs), and finally form the mature virions (MVs).
Scale bar: 200 nm. h The percentages of mapped MPXV transcripts in infected organoids.
i MPXV transcripts mapped to the locations in viral genome. j The top 30 significantly
regulated genes by MPXV infection. k Top 30 significantly regulated pathways by gene
ontology analysis at day 7 post-inoculation, compared with the uninfected group. Red:
upregulated; blue: downregulated. l Immunohistochemical staining of kidney organoid
sections for glomerular structures (PODXL), proximal tubular structures (Villin-1),
distal tubular structures (E-cadherin) and endothelial structures (PECAM-1) at 1 h,
48 h, 96 h and 7 d after inoculation with MPXV. m Quantification of MPXV DNA level
in organoids (n = 4) and (n) culture medium (n = 4) 7 days after tecovirimat treatment.
o Quantification of infectious titters in culture medium 7 days after tecovirimat
treatment. Infectious titers were undetectable in the treatment groups (n = 4). p
Cell viability of kidney organoids after treatment by tecovirimat for 7 days (n = 4–5).
h: hour post-inoculation; d: day post-inoculation. Data are presented as means ± SEM.
*P < 0.05, Mann–Whitney test.
We visualized the infection by immunostaining organoids with an antibody against the
virions. The anti-MPXV fluorescence signal was absent in uninfected organoids and
at 1-h post-inoculation, but occurred at 48 h in a subset of cells (Fig. 1e; Supplementary
Fig. S2). After 96 h, nearly all cells were infected and at day 7 there was evidence
of disruption of organoid structures (Fig. 1e). Transmission electron microscopy visualized
the intracellular MPXV particles and captured the three successive stages of MPXV
assembly in kidney organoids including the emergence of crescents (Cs), procession
to immature virions (IVs), and finally the formation of mature virions (MVs) (Fig.
1f). At high magnification, crescents appear as single bilayer with opened membranes.
The surfaces of circular IVs are covered by a honeycomb lattice, which disappear after
transition to MVs (Fig. 1g). RNA sequencing analysis revealed abundant expression
of viral genes in infected organoids since 48 h post-inoculation (Fig. 1h), as well
as the patterns of temporal expression mapped to the MPXV reference genome (Fig. 1i).
Further analysis quantified the (putative) viral transcripts at different time points
post-inoculation (Supplementary Fig. S3). These results collectively demonstrated
that human kidney organoids effectively support the full life cycle of MPXV infection.
Genome-wide RNA-seq analysis revealed rewiring of host transcriptome in a time-dependent
manner shown by principal component analysis (Supplementary Fig. S4a). Inoculation
of MPXV for only 1 h, during which active replication likely has not been initiated,
had no major effect on host gene transcription, but from 48 h onwards major changes
in gene expression were observed (Supplementary Fig. S4a; Fig. 1j). The most prominently
regulated genes by MPXV infection include PAPPA2, SLC7A1, SLC1A4, SLC1A5, APOB, which
are primarily associated with nutrient transportation and metabolism, but not the
classical antiviral effectors such as interferon-stimulated genes (Fig. 1j). Gene
ontology analysis indicated that several upregulated pathways at day 7 post-inoculation
are associated with DNA biogenesis, processing, and interactions with proteins, as
well as “response to unfolded protein” (Fig. 1k). Interestingly, significantly downregulated
pathways include “renal system development” and “kidney development” (Fig. 1k). Gene
set enrichment analysis identified transcriptional signatures of MAPK signaling pathway,
apoptosis, necroptosis, ferroptosis in infected kidney organoids (Supplementary Fig.
S4b).
MPXV infection had a striking differential effect on kidney organoid structures. Glomerular
structures, observed by PODXL staining, showed a loss of integrity at 96 h, which
became more pronounced at day 7 post-inoculation. Villin-1 staining for proximal tubular
structures showed a strong increase at 96 h and was followed by disintegration of
the structures at day 7. Distal tubular structures, however, were unharmed by MPXV
infection, as evidenced by E-cadherin staining. Endothelial structures (PECAM-1) were
fragmented at 96 h and 7 days post-inoculation (Fig. 1l; Supplementary Fig. S5). To
explore the viral tropism to different cell types of kidney organoids, we co-stained
the representative markers of different kidney structures with MPXV virions. Glomerular
and proximal tubular structures were broadly infected since 48 h post-inoculation
(Supplementary Fig. S6a, b). In contrast, limited infection occurred in distal tubular
structures at 48 h, which was slightly increased at 96 h (Supplementary Fig. S6).
This finding was consistent with the observation that glomerular and proximal tubular
structures but not distal tubular structures were disintegrated by the infection (Fig.
1l).
Since tecovirimat is an FDA-approved antiviral drug for treating smallpox, it is currently
being explored for compassionate use of treating mpox patients, but the clinical efficacy
is far from conclusive
1,9
. Here, we tested serial concentrations of tecovirimat, covering the clinically relevant
range of blood concentrations (1–5 µM)
10
, in human kidney organoids infected with MPXV. After treatment of tecovirimat ranging
0.1 to 25 µM for 7 days, a 2 to 3 log10 reduction in viral DNA load was observed for
both the intracellular (Fig. 1m) and extracellular (Fig. 1n) compartments. Importantly,
infectious viral titers became undetectable in culture medium after 7 days treatment
(Fig. 1o). Cell viability assay showed that tecovirimat had no clear cytotoxicity
on kidney organoids at 1 μM, although very mild inhibitory effects were observed at
5 and 25 μM (Fig. 1p). Tecovirimat is an inhibitor of the orthopoxvirus VP37 envelope
wrapping protein to prevent infectious virus production
11
. In our kidney organoids, the fact that no infectious virus was detected (even at
0.1 and 1 μM) in culture medium after treatment is consistent with the mechanism-of-action
of tecovirimat
12
.
During the 2022 mpox outbreak, acute kidney injury has been reported in severely infected
patients who required hospitalization
1
. It is yet unclear how the virus mediates these kidney problems. One possibility
would be that direct infection of the kidney by MPXV causes tissue injury, and this
actually has been observed in nonhuman primates infected with MPXV
13,14
. However, no autopsy kidney tissue was available to confirm this hypothesis in infected
patients
1
. Here, we demonstrated that human kidney organoids are highly permissive for MPXV
infection. The viral DNA levels and infectious titers in the organoids are dramatically
increased overtime after infection with MPXV. Importantly, we observed the three successive
stages of MPXV assembly in infected cells by transmission electron microscopy. Interestingly,
we found that kidney organoids secrete infectious MPXV particles. This is in line
with clinical observations that MPXV DNA can be frequently detected in urine samples
from infected patients
3,4,15
. However, the infectivity of MPXV DNA positive urine samples has not been determined,
which would be important for understanding whether this route plays a role in human-to-human
transmission particularly in sexual networks
4
.
In summary, human kidney organoids support the full life cycle of MPXV infection,
and provoke active virus-host interactions and recapitulate tissue injury. Tecovirimat
treatment inhibited MPXV infection by preventing infectious virus production in human
kidney organoids. This innovative model system may provide a useful tool for future
MPXV research.
Supplementary information
Supplementary Information