Available via license: CC BY 4.0
Content may be subject to copyright.
Page 1/30
Podocyte exopher-formation as a novel
pathomechanism in membranous nephropathy
Catherine Meyer-Schwesinger
University Medical Center Hamburg-Eppendorf https://orcid.org/0000-0003-1511-5188
Karen Lahme
University Medical Center Hamburg-Eppendorf
Wiebke Sachs
University Medical Center Hamburg-Eppendorf
Sarah Frömbing
University Medical Center Hamburg-Eppendorf
Michael Brehler
University Medical Center Hamburg-Eppendorf
Desiree Loreth
University Medical Center Hamburg-Eppendorf
Kristin Surmann
University Medicine Greifswald
Simone Gaing
Chimaera GmbH
Uta Wedekind
University Medical Center Hamburg-Eppendorf
Vincent Böttcher-Dierks
University Medical Center Hamburg-Eppendorf
Marie Adler
University Medical Center Hamburg-Eppendorf
Pablo Saez
University Medical Center Hamburg-Eppendorf
Christian Conze
Leibniz Institute of Virology (LIV) https://orcid.org/0000-0002-8849-1064
Roland Thünauer
Leibniz Institute of Virology
Sinah Skuza
University Medical Center Hamburg-Eppendorf
Page 2/30
Karen Neitzel
University Medical Center Hamburg-Eppendorf
Stephanie Zielinski
University Medical Center Hamburg-Eppendorf
Johannes Brand
University Medical Center Hamburg-Eppendorf
Stefan Bonn
Institute of Medical Systems Biology, Center for Molecular Neurobiology, University Clinic Hamburg-
Eppendorf
Uwe Völker
University Medicine Greifswald https://orcid.org/0000-0002-5689-3448
Marina Zimmermann
UKE Hamburg https://orcid.org/0000-0002-2666-5997
Thorsten Wiech
Instiute of Pathology, Nephropathology Section https://orcid.org/0000-0003-4053-1474
Tobias Meyer
Asklepios Klinikum Barmbek
Lars Fester
University Bonn
Biological Sciences - Article
Keywords: THSD7A, PLA2R1, membranous nephropathy, nephrotic syndrome, podocyte, glomerulus,
exopher, extracellular vesicle, autoantibody, epoxomicin, bortezomib, proteasome
Posted Date: April 17th, 2024
DOI: https://doi.org/10.21203/rs.3.rs-4219695/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
Additional Declarations: There is NO Competing Interest.
Page 3/30
Abstract
Background
Membranous nephropathy (MN) is caused by autoantibody binding to podocyte foot process antigens
such as THSD7A and PLA2R1. The mechanisms of the glomerular antigen/autoantibody deposition and
clearance are unknown.
Methods
We explore the origin and signicance of glomerular accumulations in (1) diagnostic and follow-up
biospecimens from THSD7A+ and PLA2R1+-MN patients compared to nephrotic non-MN patients, and (2)
in experimental models of THSD7A+-MN.
Results
We discovered podocyte exophers as correlates of histological antigen/autoantibody aggregates found in
the glomerular urinary space of MN patients. Exopher vesicle formation represents a novel form of toxic
protein aggregate removal in
Caenorhabditis elegans
neurons. In MN patients, podocytes released
exophers to the urine. Enrichment of exophers from MN patient urines established them as a glomerular
exit route for antigens and bound autoantibody. Exophers also carried disease-associated proteins such
as complement and provided a molecular imprint of podocyte injury pathways. In experimental THSD7A+-
MN, exophers were formed from podocyte processes and cell body. Their formation involved the
translocation of antigen/autoantibody from the subepithelial to the urinary side of podocyte plasma
membranes. Urinary exopher-release correlated with lower albuminuria and lower glomerular
antigen/autoantibody burden. In MN patients the prospective monitoring of urinary exopher abundance
and of exopher-bound autoantibodies was additive in the assessment of immunologic MN activity.
Conclusions
Exopher-formation and release is a novel pathomechanism in MN to remove antigen/autoantibody
aggregates from the podocyte. Tracking exopher-release will add a non-invasive diagnostic tool with
prognostic potential to clinical diagnostics and follow-up of MN patients.
INTRODUCTION
Idiopathic membranous nephropathy (MN) is an autoimmune disease of kidney podocytes. Podocytes
are essential cells of the ltration barrier to blood. In MN, circulating autoantibodies are directed against
podocyte foot process proteins, such as the M-type phospholipase A2 receptor (PLA2R1)1 in 80% and
thrombospondin type-1 domain-containing 7 A (THSD7A)2 in 5% of patients. Based on clinical correlative
observations, the causes of autoimmunity are likely to be multiple in MN3,4,5,6,7, as mirrored by the
multitude of recently discovered potential MN autoantigens (reviewed in7-9). In the current concept of MN,
Page 4/30
the circulating autoantibodies mostly of the huIgG4 type cause the disease10,11 by complement-
dependent12,13 as well as independent mechanisms14-16. The ensuing autoantibody/antigen reaction
leads to their pathognomonic glomerular deposition as immune complexes, lastly leading to glomerular
ltration barrier alterations and nephrotic range proteinuria17. The underlying mechanisms and clinical
signicance of glomerular aggregate formation in MN are unknown.
In the current study we discover exophers as the pathobiological correlate of glomerular urinary space
aggregates in human and experimental MN. Exophers represent a novel type of extracellular vesicle (EV)
recently discovered in
Caenorhabditis
elegans
neurons18 and body wall muscles19. Exophers extrude
from juxtanuclear plasma membrane areas20 as long nanotubules which can reach up to 4 µm in
diameter at the distal end21. In
C.
elegans
, exopher-formation depends on cellular stressors17,24 and
mechanistically involves the delivery of cargo to aggresome-like organelles20. Exopher-like processes
have morphologically been reported in mammalian systems22-26 and are thought to represent a protective
strategy to remove proteotoxic protein aggregates21,27 and dysfunctional organelles such as
mitochondria22,24. Phagocytosis of released exophers by neighboring cells is considered
neuroprotective18. This export of proteotoxic protein aggregates may be conserved in ies28,29 and
mammalian neurons30. Exophers as such have not been identied in humans yet.
Here we hypothesize that understanding the mechanisms that govern glomerular antigen/autoantibody
deposition and clearance in MN will open new non-invasive diagnostic avenues with prognostic potential,
as the podocytes’ ability to deal with the disease initiating autoantibodies is included. In comparative
studies to nephrotic non-MN patients, we set out to establish glomerular aggregates as disease-specic
exophers in MN. Using experimental podocyte and mouse models of THSD7A+-MN we investigate the
dynamics and disease modulating effects of exopher-formation and release. The podocytes’ ability to
form and release exophers to the urine is assessed in retrospective and prospective patient cohorts, and
its additive clinical value in disease monitoring is shown.
RESULTS
Exophers are the pathobiological correlate of glomerular urinary space aggregates in MN.
Our analyses of diagnostic biopsies and diagnostic/consecutive urine samples from a THSD7A+-MN
patient (
Fig. 1
) and from three PLA2R1+-MN patients (
Fig. S1
) revealed human exopher-formation as an
antigen-overarching principle in MN, compared to nephrotic non-MN patients (two minimal change
disease, two primary FSGS, one tubulo-toxic kidney injury, and one IgA nephritis patient). The blood and
urine parameters of the patients are summarized in
Tables S1
,
S2
.
We discovered THSD7A- or PLA2R1-antigen aggregates within the glomerular urinary space that had a
membrane connection to podocytes by confocal microscopy of the diagnostic MN-patient biopsies (
Fig.
1A panel 1, Fig. S1A
). After urinary release, these antigen aggregates were found in adherence to the
Page 5/30
tubular brush border (
Fig. 1A panel 2, Fig. S1B
), reminiscent of the ‘starry night’ phenotype of burst
C.
elegans
exophers18. Additionally, these antigen aggregates localized to the cytoplasm of parietal
epithelial and tubular cells, demonstrating cellular reuptake of aggregates released from the glomerulus
(
Fig. 1A panel 3, Figs. S1C
and
S2A
). Antigen aggregates were strongly positive for huIgG4, identifying
them as immune complexes within the glomerular urinary space and tubular lumen (
Fig. 1B
and
Fig. S1D
panels 1
,
Fig. S3
). Strikingly, antigens and huIgG4 colocalized stronger in aggregates within the
glomerular urinary space than within the podocyte subepithelial space, as observed by confocal
microscopy. Demonstrating the urinary release of these MN-antigen+/huIgG4+ aggregates, a subset of
large (~5-8 µm) extracellular vesicles (EVs) was discerned among urinary EVs of MN patients by confocal
microscopy, in which the MN-antigen and huIgG4 strongly colocalized. These EVs also showed a granular
positivity for the EV marker 14-3-3 (
Fig. 1B
and
Fig. S1D panels 2
). The mean size of huIgG4+-EVs ranged
from ~580 nm in the THSD7A+-MN patient (
Fig. S2B
) and ~450 nm in a PLA2R1+-MN patient when
measured by nanoparticle tracking analyses (NTA,
Fig. S1E panel 1
), discerning them as a very large
urinary EV subset. Characterization of the EV-bound huIgG4 antibodies demonstrated that they exhibited
a specic reactivity to human THSD7A or PLA2R1 protein upon elution, establishing them as the
pathogenic MN-autoantibodies that were bound to the urinary EVs (
Fig. 1B
and
Fig. S1D panels 3
).
Demonstrating the vesicular origin of these MN-antigen+/autoantibody+ EVs, immunouorescence,
immunoblotting, and image stream analyses showed the expression of EV marker proteins, such as 14-3-
3, Annexin A1, CD63/CD81 (
Fig. 1C
and
Fig. S1E
). To this end, 14-3-3 represented the most discerning
marker in huIgG4+-EVs. This is of interest, as 14-3-3 protein was recently shown to be required for
exopher-formation in
C.
elegans
neurons20. In comparison to nephrotic non-MN patients, huIgG4+-EVs
isolated from MN-patient urines contained abundant podocyte proteins such as THSD7A, PLA2R1,
Nephrin, Podocin, and a-Actinin-4 by immunoblot or by immunouorescence (
Fig. 1D
and
Fig. S1F
)
showing their podocyte origin. Besides podocyte proteins, disease-associated proteins such as
complement (C1q, C5b9), mitochondrial (Mitofusin 2, ATPB), lysosomal (LIMP2, LAMP2), and ubiquitin
proteasome system (UPS) proteins (K48-polyubiquitinated proteins as proteasome substrates, UCH-L1
and the immunoproteasome subunit LMP7) were abundant by immunoblot or by immunouorescence in
huIgG4+-EVs from MN-patient urines (
Fig. 1E
and
Fig. S4
). Showing the glomerular origin of these
discovered urinary huIgG4+-EVs, confocal analyses localized selected disease-associated proteins to
huIgG4+ aggregates within the glomerular urinary space (
Figs. S5
and
S6
) as well as within the tubular
lumen (
Figs. S7
and
S8
) within the diagnostic biopsies of THSD7A+- and PLA2R1+-MN patients. Together,
these analyses show the formation of exophers in MN glomeruli that contain the MN antigen with bound
autoantibody, 14-3-3, and disease-associated proteins, which are released by the podocyte and passage
through the nephron for cellular reuptake or urinary removal.
The urinary exophers identied in MN-patients were unique in respect to their appearance and content
compared to urinary huIgG4+-EVs from nephrotic non-MN patients
as summarized in
Table S3
. HuIgG4+-
EVs from nephrotic non-MN patients were mostly devoid or low abundant of podocyte proteins, except for
huIgG4+-EVs from MCD patients which were abundant in Nephrin. Additionally, huIgG4+-EVs from
Page 6/30
nephrotic non-MN patients had a lower content of disease-associated proteins such as complement,
proteasome substrates, mitochondria, and lysosomes
compared to MN-patient exophers
(
Fig 1
and
Fig
.
S1
). Further, they were smaller and had a less aggregate-like huIgG4 expression pattern by
immunouorescence compared to MN exophers (
Figs
.
S9
and
S10 panels 1, 2
). Interestingly, the urinary
abundance of huIgG4+/14-3-3+-EVs strongly differed between patient groups. Primary FSGS and tubulo-
toxic kidney injury patients exhibited a low abundance (~0.6-1.2%), whereas MCD and IgAN patients
exhibited a high abundance of huIgG4+/14-3-3+-EVs ranging from ~25% in MCD and ~10% in the IgAN
patient, irrespectively of the extent of proteinuria (
Figs
.
S9
and
S10 panels 3
). No specic reactivity to
glomerular/podocyte/HEK cell proteins could be identied from the immunoglobulins eluted from
isolated huIgG4+-EVs of non-MN patients (
Figs
.
S9
and
S10 panels 4
).
In summary, these ndings identify glomerular urinary space aggregates in THSD7A+- and PLA2R1+-MN
patients as exopher vesicles formed by podocytes that contain the MN-antigen and bound autoantibody
together with disease-associated proteins. These exophers are accessible to non-invasive molecular
analyses through their urinary release as huIgG4+/14-3-3+-EVs. In the following we turned to experimental
models of MN, to establish the 1) subcellular origin and dynamics of podocyte exopher-formation and
release in response to MN-autoantibody binding and 2) to determine the clinical effect of exopher-
formation and release on MN.
Exophers form at podocyte processes and cell body in experimental THSD7A+-MN.
In mice, THSD7A+-MN was induced by intravenous injection of rabbit (rb)THSD7A-antibodies (abs)31.
Scanning electron microscopy of day 1, 3, and 7 kidneys (
Fig 2A
,
Fig. S11
) revealed that podocytes
formed long membrane extensions with distal bulges that extended into the urinary space in response to
rbTHSD7A-ab exposure. These structures were similar to exophers by morphology18 and increased in
length and abundance over time. The membrane extensions originated from the urinary side of foot and
major processes as well as from the cell body (
Fig. 2A panel 1
,
Figs. S11
and
S12A
). Additionally, large
protrusions with rough surfaces without a membrane extension were focally observed at the urinary side
of podocyte processes and cell body (
Figs. S11
and
S12A
). The immunogold detection of the bound
rbTHSD7A-abs on day 7 abundantly localized rbIgG to the distal ends of the long membrane extensions
as large, gold-labeled mulberry-like structures within the glomerular urinary space by transmission EM
(
Fig. 2A panel 2
and
3, Fig. S12B
-
D
). Additionally, extensive gold-labeled aggregates were observed that
directly protruded from the urinary side plasma membranes of podocyte foot and major processes. In
comparison to the abundant gold-labeling of urinary space aggregates, the ‘expected’ subepithelial
deposition of rbTHSD7A-abs was discreet on day 7 (
Fig. 2A panel 3, Figs. S12C
and
S13A
). In contrast,
subepithelial deposition of rbTHSD7A-abs was abundant on day 1, whereas gold-labeling of podocyte
processes within the glomerular urinary space was scarce (
Fig. S13B
). No intracellular gold-particles were
observed at the evaluated time points. Immunogold labeling of THSD7A as the corresponding antigen
also discerned gold-labeled THSD7A within and at the distal end of podocyte-derived membrane
extensions that reached into the urinary space (
Fig. 2A panel 4
). Further, gold labeled THSD7A was
Page 7/30
present within aggregates protruding from podocyte foot processes (
Fig. S12E
,
F
). This pattern was
comparable to the rbIgG immunogold aggregate pattern described above. Like in MN-patients, THSD7A+-
MN mice released glomerular urinary space aggregates as exophers to the urine. In line, vesicular
structures with gold-labeled rbTHSD7A-abs were observed contacting proximal tubular brush borders
(
Fig. S14A
). By confocal microscopy THSD7A+/rbIgG+ complexes were seen within proximal tubular and
parietal epithelial cells (
Fig. S14B
) showing their reuptake after having been released from podocytes
within the glomerulus.
Exopher-formation encompasses rapid antigen crosslinking, translocation, and release in podocytes.
Human podocytes that overexpress THSD7A in culture are specically targeted by patient11 and
rbTHSD7A-abs (
Figs. S15
and
S16A
). Morphologically,
in vitro
exposure of podocytes to patient THSD7A-
abs rapidly induced the formation of honeycomb-like autoantibody patches at the luminal cell surface,
which predominated at the cell periphery (
Fig. S15
), indicating antigen crosslinking by the autoantibody.
In addition, live-cell imaging discerned areas of plasma membrane alterations at the cell periphery. These
areas were the starting point of exopher-formation after exposure to rbTHSD7A-abs (
Fig. 2B panel 1,
Movie S1
). As such, we could observe the translocation of peripheral altered plasma membrane areas
towards juxtanuclear cell areas. These translocation processes were accompanied by the formation of
stalked, large EVs, which were subsequently released to the medium (
Fig. 2B panels 1
and
2, Movies S1
and
S2
). The abluminal side of podocytes showed no alterations in holotomographic and confocal 3D
reconstructions (
Fig. 2B panels 3, 4
). From the luminal side of podocytes, protruding EVs could be seen,
which had a high refractive index as observed using holotomography (
Fig. 2B panel 3
). Confocal
resolution demonstrated that within these protruding EVs, rbTHSD7A-abs localized to the outer aspect of
THSD7A. Further, the EV markers Annexin A1 and 14-3-3 were highly abundant, 14-3-3 in an aggregate
pattern (
Fig. 2B panel 4
). Protruding EVs were seen at peripheral and juxtanuclear plasma membrane
areas. Together, this identies the protrusions as forming exophers in cultured podocytes.
The majority of exophers were rapidly released from podocytes within 6 hours of rbTHSD7A-ab exposure
and were rbIgG+ by image stream (
Fig. S16B
). Released exophers were large (mean ~470 nm, max ~4,5
µm) by NTA measurement
(
Fig. S16C
). Molecular analyses demonstrated that already within 1 hour of
in
vitro
exposure to rbTHSD7A-abs, podocytes released substantial amounts of THSD7A within exophers to
the medium, which matched the loss of THSD7A within the corresponding podocyte lysates (
Fig. 2C
).
Exophers did not only reduce the podocyte membrane burden of THSD7A and bound autoantibody but
also substantiated a cellular exit route for disease-associated proteotoxic material. As such, they were
abundant of 20S proteasome complexes with impaired proteolytic activity (
Fig. 2D panel 1
). Additionally,
selected exophers exhibited an uptake of LysoTracker and MitoSOX dyes (
Fig. 2D panel 2
), demonstrating
the release of lysosomes and damaged mitochondria, respectively.
Together these data show that autoantibody binding to THSD7A initiates crosslinking of the antigen at
the podocytes’ peripheral membranes. Altered plasma membrane areas translocate to juxtanuclear
membrane areas to be rapidly released as stalked, large exophers. As part of the proteotoxic disease
Page 8/30
process, exophers contain THSD7A with the bound autoantibody, lysosomes, dysfunctional proteasomes,
and damaged mitochondria.
Urinary exophers are derived from podocytes in experimental THSD7A+-MN.
Within the urine of THSD7A+-MN mice the mean size of EVs was ~470 nm and as such signicantly
larger than the mean of ~390 nm in ctrl-abs injected mice (
Fig. 3A panel 1
). Similarly to MN patients, a
subset of urinary EVs were identied as exophers in THSD7A+-MN mice using the markers 14-3-3 and
rbIgG by image stream (
Fig. 3A panel 2
) and by immunouorescence (
Fig. 3A panel 3
). We proceeded to
establish the podocyte restricted nature and dynamics of urinary exopher release in THSD7A+-MN using
mT/mG
reporter mice, which express membrane bound GFP (mGFP) in podocytes and membrane bound
tdTomato (mTomato) in all other cells (
Fig. 3B panel 1
and
2
). Using this setup, the urinary release of
podocyte (GFP+) EVs and of non-podocyte (tdTomato+) EVs could be differentially tracked. As
exemplarily shown for day 10 urines, a subfraction of podocyte (GFP+) EVs was rbIgG+ and Annexin V- in
THSD7A+-MN mice (
Fig. 3B
panels 3
and
4
), the latter as expected for exophers18. Time course analyses
demonstrated that podocytes increasingly released EVs to the urine especially after day 5 following
exposure to rbTHSD7A-abs in comparison to ctrl-abs (
Fig. 3C panels 1
and
2
). Starting day 3, a small
rbIgG+ EV fraction within the podocyte EVs was discernible, which signicantly differed in abundance
between THSD7A+-MN and control urines after day 7 (
Fig. 3C panel 1
). The release of non-podocyte-EVs
was not affected by rbTHSD7A-abs or ctrl-abs (
Fig. 3C panels 3
and
4
). Together, these data substantiate
that urinary exopher-release is a reaction specic to the podocyte in experimental MN.
Balanced podocyte exopher-formation and release is protective in experimental THSD7A+-MN.
We assessed whether modulation of exopher-formation and urinary release affected clinical and
morphological THSD7A+-MN. For this purpose, the extent of albumin loss to the urine was monitored in
parallel to the release of exophers in the development of THSD7A+-MN. With beginning albuminuria 1 day
after MN induction, urinary release of exophers was discernible. Following an initial increase, exopher-
release decreased in the course of disease development, whereas albuminuria continued to worsen (
Fig.
4A panel 1
). Acceleration of exopher-release through ‘short-term’ proteotoxic stress (induced by
proteasome inhibition on day 7) was mirrored by a momentary decrease of albuminuria relative to vehicle
treated THSD7A+-MN mice (
Fig. 4A panel 2
) suggesting a protective effect. Podocyte exophers were the
only urinary EVs whose release was increased by proteotoxic stress (
Fig. S17A, B
). Strikingly, exposure to
‘long-term’ proteotoxic stress (achieved by starting proteasome inhibition prior to THSD7A+-MN induction
and by maintaining proteotoxic stress throughout the development of MN) exacerbated urinary exopher-
release especially between days 5 to 7 relative to vehicle-THSD7A+-MN mice (
Fig. 4A panel 3
). Thereafter,
urinary exopher-release diminished relative to vehicle-THSD7A+-MN, despite the persistence of circulating
serum autoantibodies (
Fig. S18
). This late decrease in exopher-release was mirrored by a strong
exacerbation of albuminuria relative to vehicle controls. Morphologically, proteasome-inhibited THSD7A+-
MN mice displayed larger glomerular aggregates in the urinary space than vehicle treated THSD7A+-MN
Page 9/30
mice, especially in the setting of prolonged proteotoxic stress when visualized by super resolution
microscopy of rbIgG aggregates in optically cleared kidney sections from
mT/mG
mice (
Fig. 4A panel 4
and
5
). Identifying them as exophers, these rbIgG+ aggregates were in connection with the podocyte cell
body through long mGFP+ membrane extensions (
Fig. 4A panel 4
). These data suggest that especially in
the setting of prolonged proteotoxic stress the decrease in urinary exopher abundance at the end of the
observation time resulted from impaired exopher-release rather than from impaired exopher-formation.
We quantied the amount of THSD7A and rbIgG protein deposition within a dened number of glomeruli
at the end of the observation period by immunoblotting (
Fig. S17C
) to assess whether exopher-release
affected the extent of glomerular antigen / autoantibody burden. Indeed, correlative analyses
demonstrated a signicant inverse dependency between the extent of urinary exopher-release and the
glomerular deposition of THSD7A protein (
Fig. 4B panel 1
) or of rbIgG in vehicle and ‘short-term’
proteotoxic stress exposed mice (
Fig. 4B panel 2
). Additionally, urinary exopher abundance exhibited an
inverse correlation with albuminuria (
Fig. 4B panel 3
), demonstrating that exopher-release is protective in
MN by reducing the glomerular burden of antigen/autoantibody deposition.
Non-invasive exopher monitoring in a patient with relapsing THSD7A+-MN.
Urinary exopher-release and as such the podocytes’ ability to remove bound autoantibodies was
prospectively monitored for 2½ years in a 62-year-old male who was diagnosed with THSD7A+-MN in
May 2021. The patient initially presented with heavy nephrotic syndrome and normal renal function. His
diagnostic histopathology was signicant for a THSD7A+-MN (stage I Ehrenreich and Churg), with
segmental thickened GBM (
Fig. 5A panel 1
), a pseudolinear positivity for huIgG (
Fig. 5A panel 2
), a ne
glomerular positivity for huIgG4 at the ltration barrier (
Fig. 5A panel 3
), and THSD7A aggregates.
THSD7A aggregates were discreet in the subepithelial space and prominent in the urinary space, some
within parietal epithelial cells (
Fig. 5A panel 4
). Ultra-structurally, foot process effacement and GBM
thickening were seen, with some long membrane extensions derived from foot and major processes (
Fig.
5A panel 5
). The clinical course of the patient (
Fig. 5B graph,Table S1
,
patient P1
) showed a relapsing
nephrotic syndrome complicated by cerebral infarctions (09/2021 and 10/2022), and a progressive
decline of renal function (eGFR 90 ml/min at diagnosis to 33 ml/min 12/2023). The patient received
several courses of rituximab as well as a cyclophosphamide pulse therapy, under which he developed a
symptomatic COVID-19 infection. Besides the diagnostic urine (U1), 7 additional urines (U2-U8) were
collected and analyzed for exopher abundance (
Fig. 5B panel 1
), for the presence of exopher-bound
autoantibodies (
Fig. 5B panel 2
), and for the presence of a molecular imprint of podocyte injury pathways
(
Fig. 5B panel 3
). Albeit the presence of high sTHSD7A-ab titers of 1:160, urinary exopher-release
progressively decreased from ~35% at diagnosis to ~3% in U3. From 10/2022 on, sTHSD7A-abs were
mostly negative, nonetheless autoantibodies were persistently eluted from urinary exophers. Four months
after the last rituximab administration, the patient presented with increasing edemas and proteinuria
(28/06/2023). Upon immunological checkup, a low but recovering CD19+ B-cell count was noted (2%; 14/
µl), sTHSD7A-abs were still not detectable. Nonetheless, analyses of U7 showed an increase in urinary
exopher abundance (4%) which contained bound autoantibodies demonstrating immunologic activity.
Page 10/30
Rituximab administration ameliorated the patients’ proteinuria and edema. U8 collected 4½ months later
still in the setting of negative sTHSD7A-abs, exhibited a low exopher abundance (~0.2%) with readily
detectable exopher-bound autoantibodies. Molecular disease patterning by proteomic analyses of
exophers from U1-U8 (
Fig. 5B panel 3
) not only substantiated the vesicular/plasma membrane origin of
exophers, but also the enrichment of disease associated pathways such as cell stress response (heat
shock proteins, chaperones), classical complement and humoral immune system, as well as proteolytic
pathways (proteasome and lysosome) shown within our biochemical analyses. Even though a molecular
imprint of the podocyte injury state was visible over all investigated time points, a certain dynamic in the
proteome prole was present.
Glomerular aggregate morphometrics and non-invasive exopher monitoring in PLA2R1+-MN.
Patients with THSD7A+-MN are rare, therefore we developed a computational image algorithm to pattern
subepithelial and urinary space PLA2R1 aggregates throughout the glomeruli, the latter as the biological
correlate of podocyte exopher-formation. The ‘density score’ of this algorithm included the size, amount,
neighborhood, and intensity of PLA2R1 aggregates throughout the glomeruli not differentiating between
urinary space and subepithelial aggregates (
Fig. S19
). A low density score related to the predominance of
urinary space aggregates and a disrupted subepithelial PLA2R1 deposition pattern, whereas a high
density score reected the occurrence of dense subepithelial PLA2R1 depositions and scarce/large
urinary space aggregates (
Fig. 6A panel 1
). PLA2R1 density differed in diagnostic biopsies of a well-
characterized retrospective PLA2R1+-MN patient cohort32 in which 18 patients were stratied to a
remission group (R; amelioration of serum creatinine, eGFR, proteinuria, albuminuria, and/or serum
albumin) within a mean follow-up period of 42,83 ± 5,52 months, and 17 patients to a non-remission (NR)
group, within a mean follow-up period of 43,24 ± 9,2 months. In the remission group, PLA2R1 density and
sPLA2R1-abs were low at diagnosis. The non-remission group exhibited a high PLA2R1 density as well as
high sPLA2R1 titers at diagnosis (
Fig. 6A panel 2
). These data suggest that the amount of antigen
deposited within the subepithelial space and/or the ability to clear the subepithelial space from the
PLA2R1 deposits differed at diagnosis between both patient groups.
Exopher-release was prospectively analyzed in a 20-year-old female nephrotic patient diagnosed 10/2021
with PLA2R1+-MN (stage I-III Ehrenreich and Churg) based on the clinical and nephropathological workup
(
Fig 6B
,
Table S1
,
patient P2
). At diagnosis sPLA2R1-ab titer was moderate (92 RE/ml). Urinary exopher
diagnostics identied ~8% exophers, and exopher-bound PLA2R1 autoantibodies were readily detectable
(
Fig 6B
panel 1
). Morphologically, PLA2R1 density was low in the diagnostic biopsy, which related to a
disrupted subepithelial PLA2R1 deposition and small glomerular urinary-space PLA2R1 aggregates (
Fig
6B
panel 2
). Together, these ndings indicated a low glomerular aggregate burden in the setting of
moderate sPLA2R1-ab titer, with dynamic exopher-formation and release at diagnosis. Within one year the
patient went into clinical remission under a supportive treatment regimen, and cyclosporin A. A relapse
01/06/2022 was treated with rituximab. Serum PLA2R1 titers were rst negative 10/08/2022 and
Page 11/30
remained negative thereafter, suggesting immunologic remission. In the last follow-up urine U2
(16/10/2023) exopher abundance was low (~0.7%) and no exopher-bound PLA2R1 autoantibodies were
detectable, substantiating immunologic remission. Based on these patient observations we propose that
simultaneously tracking the fate of pathogenic autoantibodies (serum levels, glomerular binding, and
release) enables a ne-tuned monitoring of immunologic MN disease activity, especially in the setting of
negative serum autoantibody titers.
In conclusion, our cumulative investigations in patient and experimental MN demonstrate that a balanced
process of podocyte exopher-formation and release represents a protective antigen-overarching
pathophysiologic sequel in MN. As summarized in
Fig 6C
, we propose that simultaneously tracking the
fate of pathogenic autoantibodies (serum levels, glomerular binding, and release,
Fig. 6C
panel 1
) enables
a ne-tuned monitoring of immunologic MN disease activity. Mechanistically, exopher-formation includes
the displacement of antigen/autoantibodies from the subepithelial space to the apical membranes of
foot and major processes and cell body (
Fig. 6C
panel 2
). Exopher-release to the urine ultimately leads to
a reduction of the glomerular subepithelial autoantibody and antigen burden (
Fig. 6C
panel 3
).
DISCUSSION
Podocytes continue to emerge as primary targets to autoimmunity not only in MN33. Our knowledge
about how autoantibodies affect podocytes is limited and the mechanisms of the glomerular
antigen/autoantibody deposition and clearance are unknown.Here we discover that exophers represent
the biological correlate ofaggregates within the glomerular urinary space in MNin patients and in
experimental MN models. In patients,MN-associated (autoantibody+/antigen+) EVs fullled important
features of exophers. As such,
morphological criteria
(extrusion from apical membranes especially from
the juxtanuclear area as stalked large vesicles;
expression of a characteristic set of EV
markers)18,20,21,
dynamic criteria
(enhanced release by proteotoxic stress, reuptake by cells along the
nephron)18,27,34, and
molecular criteria
(abundance of cell-specic and disease-associated proteins,
presence of organelles)18,35,22 establish them as a hitherto unrecognized podocyte-derived urinary EV
form. They are distinct from migrasomes36, a large EV-type forming at tips and intersections of retraction
bers of migrating cells37. They also differ in size and vesicular marker composition from EVs described
by Hara et al., which originate from tip vesiculation of podocyte microvilli38. Whether the membrane
extensions identied in THSD7A+-MN mice reect the well-known feature of microvillous changes of
injured podocytes39 is unclear. However, both processes, i) exopher-formation in neurons and
cardiomyocytes22, as well as ii) microvillous podocyte transformation are considered protective
mechanisms, the latter suggested to be associated with complete remission of MCD or FSGS in
patients40. The protective cellular effects of exopher-formation and release include the removal of toxic
components when proteostasis and organelle function are impaired18. Podocytes face both challenges in
experimental MN32,41-43. Additionally, protection can be attributed to clearance of the subepithelial space
from immune complexes and complement by their membrane translocation to the urinary podocyte side
Page 12/30
and release within exophers. As such, the course of albuminuria and the extent of glomerular
antigen/autoantibody deposition inversely correlated with exopher-release in experimental THSD7A+-
MN.Exopher-releasemight reduce the effects on GBM remodeling, and abolish the disease exacerbating
effect of complement activation14 thus enabling podocyte recovery.
One could speculate that an aspect of MN progression and therapy resistance could originate from high
circulating autoantibody levels that constantly target podocytes, which cannot maintain the necessary
level of exopher-formation for removal. As such, urinary exopher-release was low in the relapsing
THSD7A+-MN patient in the setting of high sTHSD7A-abs, and the histological pattern of scarce urinary
space aggregates and dense subepithelial deposition predominated in the non-remitting PLA2R1+-MN
patient group. Of note, extensive exopher-formation might also have disease aggravating aspects i) by
transfer of toxic protein aggregates to surrounding cells, promoting pathology44, ii) by exopher
phagocytosis through surrounding cells as a starting point for (cross-)presentation and immune
activation leading to secondary autoantibody formation against intracellular proteins32,45 or to epitope-
spreading of primary MN autoantibodies46, iii) by initiating lipid biosynthesis27, which when dysregulated
perpetuates podocyte injury47, and iv) by substantial loss of plasma membrane required for exopher-
formation, necessitating membrane repair.The ux balance between serum autoantibody titer, exopher-
formation and release is certainly crucial for the ultimate outcome in MN.
We identied the ubiquitously expressed 14-3-3chaperone-likeprotein48 as an exopher-identifying
protein. 14-3-3 abundance was high and ‘aggregate-like’ in patient and experimental exophers, and 14-3-3
aggregates were observed at juxtanuclear areas of exopher-formation in podocytes, corroborating the
recently published involvement of 14-3-3 in neuronal exopher-formation20. In podocytes, 14-3-3 is a
knownsynaptopodin interacting protein49, playing a role in the downstream function of cyclosporin A49.
To this end, MN-associated exophers represent a unique form of huIgG4+/14-3-3+ EVsin respect to
their appearance and molecular composition. It is, however,unclear whether the huIgG4+/14-3-3+ EVs
abundantly present in nephrotic urines from IgAN and MCD patients represent exophers or not, as they
differ in size, appearance, and molecular composition.In IgAN the binding of human antibodies to
aberrantly glycosylated and mesangially deposited IgA drives disease50, and a binding of antibodies to
podocytes has not been described. Nonetheless, the isolatedhuIgG4+/14-3-3+ EVscontained 14-3-3,
mitochondrial, lysosomal and UPS proteins as well as podocyte proteins. Also,the abundant huIgG4+/14-
3-3+ EVspresent inMCD urines are intriguing. They contained Nephrin but compared to MN-associated
exophers no strict exopher-dening proteotoxic/organellar proteins, and a different abundance of EV
markers. However, the aggregate-like pattern of huIgG4 and 14-3-3 expression on a subset of MCD EVs,
might relate to EV-bound anti-Nephrin antibodies recently identied in the serum of a subset of MCD
patients33. But as the IgG subtype of anti-Nephrin antibodies is unclear, and the detection of anti-Nephrin
antibodies remains technically challenging, further investigations are underway. Of note,huIgG4 present
Page 13/30
within nephrotic urines adheres to EV surfaces to some extent51. Therefore, larger cohorts will be needed
to dene an adequate cut-off for exopher abundance.
Following the identication of the rst MN antigen in 20091, it took several years, a substantial amount of
patients and independent observations to establish the indisputable clinical power of serum
autoantibody measurement for diagnosis, evaluation of prognosis, and for clinical monitoring of
MN7.Identication of exopher-formation as the pathobiology behind glomerular aggregate formation
adds aclinically relevant information to patient monitoring.The two prospective MN patient observations
suggest that tracking the fate of autoantibodies (1. serum levels, 2. glomerular binding, 3. urinary
exopher-release and composition) allows ne-tuned monitoring of MN disease activity, especially in the
setting of negative serum autoantibody titers. In the PLA2R1+-MN patient, the low abundance of exophers
and the absence of exopher-bound autoantibodies in the follow-up urine substantiated the immunologic
remission reected by the absence of sPLA2R1-abs. In the THSD7A+-MN patient, however, detection of
~4% exophers with bound autoantibodies in the setting of negative sTHSD7A-abs substantiated
persistent immunologic activity as the cause of nephrotic syndrome aggravation. Another interesting
aspect was the low urinary exopher-release (~0.35%) in the setting of abundant sTHSD7A-abs in March
2023 which, besides scarring, could indicate impaired exopher-formation/release by podocytes as a
possible trigger of disease persistence or even aggravation similar to the ‘long-term’ proteasome-inhibited
mice.
Therapeutic modulation of podocyte exopher-formation/release could represent an interesting approach
i.e. by proteasome modulation or by targeting renal cell exopher uptake22 and the ensuing immune
reaction. Existing therapeutic principles in MN might already modulate exopher-formation/release,
i.e.cyclosporin A through interference with 14-3-3/synaptopodin signaling49 and rituximab through off-
target effects on podocyte lipid biology52, both aspects described to modify exopher-formation in
C.
elegans
. As exophers can be non-invasively enriched in substantial amounts, they open a new avenue of
diagnostic approaches in nephrology, reaching from simple quantication and autoantibody detection
(potentially beyond MN) to deep molecular disease patterning. Exopher-enrichment from nephrotic urines
is technically elaborate, therefore the development of simple urine tests is necessary to enable routine
clinical use.In the future this might, however, represent a sensitive and non-invasive method to monitor
MN autoantibodies by clinicians and even patients themselves, allowing for an early detection of
immunologic activity. As exophers enable the molecular assessment of podocyte injury patterns/extent,
future investigations in larger prospective patient cohorts will prove their clinical and prognostic value.
MATERIALS AND METHODS
Please refer to the supplemental appendix for detailed materials and methods.
Declarations
Page 14/30
ACKNOWLEDGEMENTS
We are grateful to the staff of the FACS Sorting Core Unit and the UKE imaging Facility (UMIF)under the
DFG Research Infrastructure Portal: RI_00489for excellent technical assistance. We thank the Nikon
Center of Excellence at LIV for support
.
This work was funded by the Deutsche Forschungsgemeinschaft
(DFG, German Research Foundation,CRC 1192 project B3, B6 and ME 2108/10-1 to CMS; CRC 1328
(Project A20) to P.J.S.)and by the Human Frontier Science Program (HFSP) RGP0032/2022.The LSM980
Airyscan 2 was funded by the DFG (INST 152/952-1 FUGG to CMS).UW was nancially supported by the
Else Kröner-Fresenius-Stiftung iPRIME Scholarship (2021_EKPK.10), SS, and KN were nancially
supported by the Integrated Research Training Groupe of the CRC 1192. UKE, Hamburg. We are grateful to
Andrea Hilpert and Anita Hecht in Erlangen for excellent technical assistance in preparing the murine
immunogold EM samples. We thank Christian Hentschker, and Manuela Gesell Salazar, for excellent
support in the proteome analyses. We thank Johanna Herwig for help in the human podocyte
in vitro
assays.
AUTHOR CONTRIBUTIONS
KL established and performed all the urinary murine and patient vesicle work. WS and KL performed the
animal work comprising characterization of
mT/mG
and BALB/c THSD7A+-MN. WS performed the
THSD7A glomerular quantication. SF performed the hu-podocyte EV experiments, UW, KN, and SK the
THSD7A abundance in hu-podocytes. SF, MRA, PJS performed the phase contrast time-lapse and
holotomography experiments. LF performed the immunogold SEM and TEMs of murine glomeruli and
patient urinary exophers, SG performed the 3D-reconstruction of TEMs. KS and UV performed proteomic
analyses of urinary exophers. DL performed histologic procedures, and generated samples, and
micrographs for morphometrics. MB, MZ and SB developed the computational imaging algorithms for
rbTHSD7A-abs and PLA2R1 morphometry. SZ, JB performed excellent technical assistance. TW provided
the images of original patient pathological diagnosis. VB-D imaged exophers in patient biopsies. CC and
RT performed the 3D-reconstruction of
mT/mG
cleared kidney slice. TNM collected and provided the urine
samples and performed the clinical correlations. CMS conceived the project, designed the experiments,
performed stainings and confocal microscopy, analyzed data, supervised the project, and wrote the
paper.
CONFLICT OF INTEREST
None declared.
DISCLAIMER
SG did not receive any nancial compensation for the 3-dimensional TEM reconstructions.
References
Page 15/30
1. Beck, L.H., Jr.
, et al.
M-type phospholipase A2 receptor as target antigen in idiopathic membranous
nephropathy.
N Engl J Med
361, 11-21 (2009).
2. Tomas, N.M.
, et al.
Thrombospondin type-1 domain-containing 7A in idiopathic membranous
nephropathy.
N Engl J Med
371, 2277-2287 (2014).
3. Stanescu, H.C.
, et al.
Risk HLA-DQA1 and PLA(2)R1 alleles in idiopathic membranous nephropathy.
N
Engl J Med
364, 616-626 (2011).
4. Hoxha, E.
, et al.
A Mechanism for Cancer-Associated Membranous Nephropathy.
N Engl J Med
374,
1995-1996 (2016).
5. Wiech, T.
, et al.
Bacterial infection possibly causing autoimmunity: Tropheryma whipplei and
membranous nephropathy.
Lancet
400, 1882-1883 (2022).
. Beck, L.H., Jr. PLA2R and THSD7A: Disparate Paths to the Same Disease?
J Am Soc Nephrol
28,
2579-2589 (2017).
7. Hoxha, E., Reinhard, L. & Stahl, R.A.K. Membranous nephropathy: new pathogenic mechanisms and
their clinical implications.
Nat Rev Nephrol
18, 466-478 (2022).
. Tomas, N.M., Huber, T.B. & Hoxha, E. Perspectives in membranous nephropathy.
Cell Tissue Res
(2021).
9. Sethi, S. & Madden, B. Mapping antigens of membranous nephropathy: almost there.
Kidney Int
103,
469-472 (2023).
10. Reinhard, L.
, et al.
Pathogenicity of Human anti-PLA2R1 Antibodies in Minipigs: A Pilot Study.
J Am
Soc Nephrol
(2023).
11. Tomas, N.M.
, et al.
Autoantibodies against thrombospondin type 1 domain-containing 7A induce
membranous nephropathy.
J Clin Invest
126, 2519-2532 (2016).
12. Schiller, B.
, et al.
Inhibition of complement regulation is key to the pathogenesis of active Heymann
nephritis.
J Exp Med
188, 1353-1358 (1998).
13. Kerjaschki, D.
, et al.
Transcellular transport and membrane insertion of the C5b-9 membrane attack
complex of complement by glomerular epithelial cells in experimental membranous nephropathy.
J
Immunol
143, 546-552 (1989).
14. Seifert, L.
, et al.
The classical pathway triggers pathogenic complement activation in membranous
nephropathy.
Nat Commun
14, 473 (2023).
15. Meyer, T.N.
, et al.
A new mouse model of immune-mediated podocyte injury.
Kidney Int
72, 841-852
(2007).
1. Herwig, J.
, et al.
Thrombospondin Type 1 Domain-Containing 7A Localizes to the Slit Diaphragm and
Stabilizes Membrane Dynamics of Fully Differentiated Podocytes.
J Am Soc Nephrol
30, 824-839
(2019).
17. Fogo, A.B., Lusco, M.A., Najaan, B. & Alpers, C.E. AJKD Atlas of Renal Pathology: Membranous
Nephropathy.
Am J Kidney Dis
66, e15-17 (2015).
Page 16/30
1. Melentijevic, I.
, et al.
C. elegans neurons jettison protein aggregates and mitochondria under
neurotoxic stress.
Nature
542, 367-371 (2017).
19. Turek, M.
, et al.
Muscle-derived exophers promote reproductive tness.
EMBO Rep
22, e52071 (2021).
20. Arnold, M.L.
, et al.
Intermediate laments associate with aggresome-like structures in proteostressed
C. elegans neurons and inuence large vesicle extrusions as exophers.
Nat Commun
14, 4450
(2023).
21. Arnold, M.L., Cooper, J., Grant, B.D. & Driscoll, M. Quantitative Approaches for Scoring in vivo
Neuronal Aggregate and Organelle Extrusion in Large Exopher Vesicles in C. elegans.
J Vis Exp
(2020).
22. Nicolas-Avila, J.A.
, et al.
A Network of Macrophages Supports Mitochondrial Homeostasis in the
Heart.
Cell
183, 94-109 e123 (2020).
23. Nicolas-Avila, J.A., Pena-Couso, L., Munoz-Canoves, P. & Hidalgo, A. Macrophages, Metabolism and
Heterophagy in the Heart.
Circ Res
130, 418-431 (2022).
24. Nicolas-Avila, J.A., Sanchez-Diaz, M. & Hidalgo, A. Isolation of exophers from cardiomyocyte-reporter
mouse strains by uorescence-activated cell sorting.
STAR Protoc
2, 100286 (2021).
25. Huang, Y., Yu, M. & Zheng, J. Proximal tubules eliminate endocytosed gold nanoparticles through an
organelle-extrusion-mediated self-renewal mechanism.
Nat Nanotechnol
18, 637-646 (2023).
2. Falchi, A.M.
, et al.
Astrocytes shed large membrane vesicles that contain mitochondria, lipid droplets
and ATP.
Histochem Cell Biol
139, 221-231 (2013).
27. Cooper, J.F., Guasp, R.J., Arnold, M.L., Grant, B.D. & Driscoll, M. Stress increases in exopher-mediated
neuronal extrusion require lipid biosynthesis, FGF, and EGF RAS/MAPK signaling.
Proc Natl Acad Sci
U S A
118(2021).
2. Donnelly, K.M.
, et al.
Phagocytic glia are obligatory intermediates in transmission of mutant
huntingtin aggregates across neuronal synapses.
Elife
9(2020).
29. Ray, A., Speese, S.D. & Logan, M.A. Glial Draper Rescues Abeta Toxicity in a Drosophila Model of
Alzheimer's Disease.
J Neurosci
37, 11881-11893 (2017).
30. Davis, A.A., Leyns, C.E.G. & Holtzman, D.M. Intercellular Spread of Protein Aggregates in
Neurodegenerative Disease.
Annu Rev Cell Dev Biol
34, 545-568 (2018).
31. Tomas, N.M.
, et al.
A Heterologous Model of Thrombospondin Type 1 Domain-Containing 7A-
Associated Membranous Nephropathy.
J Am Soc Nephrol
28, 3262-3277 (2017).
32. Reichelt, J.
, et al.
Non-functional ubiquitin C-terminal hydrolase L1 drives podocyte injury through
impairing proteasomes in autoimmune glomerulonephritis.
Nat Commun
14, 2114 (2023).
33. Watts, A.J.B.
, et al.
Discovery of Autoantibodies Targeting Nephrin in Minimal Change Disease
Supports a Novel Autoimmune Etiology.
J Am Soc Nephrol
33, 238-252 (2022).
34. Wang, Y.
, et al.
Large vesicle extrusions from C. elegans neurons are consumed and stimulated by
glial-like phagocytosis activity of the neighboring cell.
Elife
12(2023).
Page 17/30
35. Banasiak, K., Turek, M. & Pokrzywa, W. Preparation of Caenorhabditis elegans for Scoring of Muscle-
derived Exophers.
Bio Protoc
13(2023).
3. Liu, Y.
, et al.
Podocyte-Released Migrasomes in Urine Serve as an Indicator for Early Podocyte Injury.
Kidney Dis (Basel)
6, 422-433 (2020).
37. Buzas, E.I. The roles of extracellular vesicles in the immune system.
Nat Rev Immunol
23, 236-250
(2023).
3. Hara, M.
, et al.
Apical cell membranes are shed into urine from injured podocytes: a novel
phenomenon of podocyte injury.
J Am Soc Nephrol
16, 408-416 (2005).
39. Andrews, P. Morphological alterations of the glomerular (visceral) epithelium in response to
pathological and experimental situations.
J Electron Microsc Tech
9, 115-144 (1988).
40. Royal, V.
, et al.
Ultrastructural Characterization of Proteinuric Patients Predicts Clinical Outcomes.
J
Am Soc Nephrol
31, 841-854 (2020).
41. Beeken, M.
, et al.
Alterations in the ubiquitin proteasome system in persistent but not reversible
proteinuric diseases.
J Am Soc Nephrol
25, 2511-2525 (2014).
42. Meyer-Schwesinger, C.
, et al.
Ubiquitin C-terminal hydrolase-l1 activity induces polyubiquitin
accumulation in podocytes and increases proteinuria in rat membranous nephropathy.
Am J Pathol
178, 2044-2057 (2011).
43. Rood, I.M.
, et al.
Increased expression of lysosome membrane protein 2 in glomeruli of patients with
idiopathic membranous nephropathy.
Proteomics
15, 3722-3730 (2015).
44. Nath, S.
, et al.
Spreading of neurodegenerative pathology via neuron-to-neuron transmission of beta-
amyloid.
J Neurosci
32, 8767-8777 (2012).
45. Prunotto, M.
, et al.
Autoimmunity in membranous nephropathy targets aldose reductase and SOD2.
J
Am Soc Nephrol
21, 507-519 (2010).
4. Seitz-Polski, B.
, et al.
Epitope Spreading of Autoantibody Response to PLA2R Associates with Poor
Prognosis in Membranous Nephropathy.
J Am Soc Nephrol
27, 1517-1533 (2016).
47. Kim, J.J., Wilbon, S.S. & Fornoni, A. Podocyte Lipotoxicity in CKD.
Kidney360
2, 755-762 (2021).
4. Fu, H., Subramanian, R.R. & Masters, S.C. 14-3-3 proteins: structure, function, and regulation.
Annu
Rev Pharmacol Toxicol
40, 617-647 (2000).
49. Faul, C.
, et al.
The actin cytoskeleton of kidney podocytes is a direct target of the antiproteinuric
effect of cyclosporine A.
Nat Med
14, 931-938 (2008).
50. Floege, J. The pathogenesis of IgA nephropathy: what is new and how does it change therapeutic
approaches?
Am J Kidney Dis
58, 992-1004 (2011).
51. Buzas, E.I., Toth, E.A., Sodar, B.W. & Szabo-Taylor, K.E. Molecular interactions at the surface of
extracellular vesicles.
Semin Immunopathol
40, 453-464 (2018).
52. Fornoni, A.
, et al.
Rituximab targets podocytes in recurrent focal segmental glomerulosclerosis.
Sci
Transl Med
3, 85ra46 (2011).
Page 18/30
Figures
Figure 1
Glomerular urinary space aggregates are exophers in a THSD7A+-MN patient. The diagnostic biopsy and
urines (U1-U8) collected from the THSD7A+-MN patient P1 were evaluated for the presence of exophers.
Page 19/30
A) High-resolution confocal micrographs to THSD7A (red) and wheat germ agglutinin (green, WGA, binds
to N-acetyl-D-glucosamine and sialic acid of the glycocalyx and demarcates plasma membrane), DNA
(blue).
Panel 1:
Glomerular THSD7A aggregates localize to the subepithelial space (white arrowheads)
and to the urinary (u) space (orange arrows). Urinary space THSD7A aggregates exhibit a WGA+
membrane connection (green arrow) to podocytes (p).
Panel 2:
THSD7A+ aggregates (orange arrows) are
found within the tubular lumen (L) adherent to the brush border or
Panel 3
taken up by tubular cells, dtc =
distal tubular cell. B) Glomerular and urinary THSD7A aggregates are huIgG4+.
Panel 1:
Confocal
analyses demonstrate that THSD7A (red) colocalizes with huIgG4 (green) in urinary space aggregates
(orange arrows). White arrowheads highlight huIgG4 in the subepithelial space.
Panel 2:
3D-reconstructed
confocal
z
-stack of a urinary exopher (from U1) stained for THSD7A (red), huIgG4 (green) and the vesicle
marker 14-3-3 (orange). Lower row exhibits individual channels in one plane.
Panel 3:
Antibodies eluted
from U2 exophers exhibit specic reactivity to THSD7A in human glomerular extract (HGE) and in HEK
cells transduced with human THSD7A but not with human PLA2R1. C) Urinary exophers express vesicle
markers.
Panel 1:
3D-reconstructed confocal
z
-stack of an exopher from U2 demonstrates expression of
the vesicle markers 14-3-3 (orange), Annexin A1 (red), CD63/CD81 (blue). MemBrite594 (light blue) stains
the plasma membrane, huIgG4 (green) the bound autoantibody.
a-e:
individual channels in one plane.
Panel 2:
Immunoblot to vesicle markers in exophers isolated from U7 and U8 in comparison to two
nephrotic minimal change disease (MCD) patients P5 and P6.
Panel 3:
Image stream quantication of
vesicle marker distribution on MemBrite594+huIgG4+ exophers in U1 discerns 14-3-3 as an abundant
marker. Note the lack of Annexin V signal (marker for apoptotic bodies). Plot also shown in
Fig. 5B
panel
1
Urinary exophers contain podocyte proteins.
Panel 1:
Immunoblot detection of podocyte-specic
proteins in exophers from U7 and U8 in comparison to the two nephrotic MCD patients.
Panel 2:
Confocal
analyses demonstrate expression of podocyte-specic proteins (protein of interest (POI) in red) within
huIgG4+ (green) exophers of U2. E) Urinary exophers contain disease associated proteins.
Panel 1:
Immunoblot detection of disease associated proteins belonging to the complement cascade (C5b9,
membrane attack complex), mitochondria (membrane proteins Mitofusin 2 and ATP synthase ATPB),
lysosomes (membrane proteins LAMP2 and LIMP2), and the ubiquitin proteasome system (UPS: K48-
polyubiquitinated proteins, the proteolytic b-subunit LMP7, the deubiquitinating enzyme UCH-L1) in
exophers isolated from U7 and U8 in comparison to the two nephrotic MCD patients.
Panel 2:
Confocal
analyses demonstrate expression of Mitofusin 2, LIMP2, or K48pUB (all proteins of interest (POI) in red)
within huIgG4+ (green) exophers of U2.
Page 20/30
Figure 2
Ultrastructure and dynamics of exopher-formation in experimental MN.
A) THSD7A+-MN was induced in BALB/c mice by injection of rabbit (rb)THSD7A-antibodies (abs) or
control-abs. Exopher-formation is visualized by ultra-structural techniques on day 7 in THSD7A+-MN
mice.
Panel 1
: Scanning EM (SEM) depicting long membrane extensions with distal bulges (orange
Page 21/30
arrowheads) which originate from major (mp) and foot processes (fp) as well as from the cell body (cb)
of podocytes; asterisks = area of fp effacement.
Panel 2
: Transmission EM (TEM) of immunogold-labeled
rbTHSD7A-abs demonstrates gold aggregates within the urinary space connected to the podocyte cell
body by long membrane extensions (orange arrows); u = urinary space, c = capillary lumen, gbm =
glomerular basement membrane, p = podocyte. Close-up: exopher.
Panel 3
: 3D-reconstruction of 48
consecutive TEM micrographs demonstrates the budding-off of gold-labeled structures from podocyte
foot and major processes; colored arrows highlight one individual gold structure; gold appears black in
the plane of cutting; white arrowheads demarcate subepithelial gold.
Panel 4
: TEM of immunogold-
labeled THSD7A depicts the antigen along (blue arrow) and at the distal end (green arrow) of membrane
extensions. B) Live-cell imaging of human podocytes overexpressing human THSD7A that were exposed
to rbTHSD7A-abs in culture.
Panel 1:
Phase contrast live-cell imaging of a forming exopher (see
Movie
S1
). The inset panels correspond to the emboxed area. Images at different time points are shown (i=
01:42 h:min, ii= 01:54, iii= 02:15, iv= 02:22, v= 02:28, vi= 02:32, vii= 02:33, viii= 02:34). Orange arrowhead
highlights exopher stalk.
Panel 2:
Phase contrast snapshots at indicated times demonstrate exopher
translocation from the cell periphery to the juxtanuclear cell area. The exophers trajectory is shown color
coded over 2 h 26 min (see
Movie S2
).
Panel 3:
Holotomographic images shown with an inverted LUT in
grayscale visualize an exopher with a high refractive index protruding from the luminal podocyte side into
the medium. The dashed square shows the abluminal (left micrograph) and luminal (middle micrograph)
side of the podocyte area, of which the 3D-reconstructed
z-
stack is shown in the right micrograph. Note
the inconspicuous abluminal side.
Panel 4:
Confocal resolution of rbIgG (red), THSD7A (green), 14-3-3
(orange), and Annexin A1 (light blue) expression in a podocyte after
in vitro
exposure to rbTHSD7A-abs.
For clarity, color overlays are shown pair wise.
Panels a
and
b
represent 3D-reconstructions of the
z-
stack
showing exophers that arise from the juxtanuclear plasma membrane area. The micrographs in a dashed
box represent single planes in the
z
-axis (red dashed box) or
x/y-
axis (white dashed box) of the 3D-
reconstructed image. Note the abundance of Annexin A1 and of 14-3-3 aggregates, as well as rbIgG along
the outer aspect of THSD7A in the protruding exophers. C) Exposure to rbTHSD7A-abs induces a rapid
decrease of THSD7A within the podocyte cell lysate and a concomitant increase of THSD7A in released
EVs; UT = untreated THSD7A overexpressing podocytes, mock = control cells without THSD7A
expression. D) Content of EVs enriched from the medium after 6 h of exposure to ctrl-abs or rbTHSD7A-
abs.
Panel 1:
Native PAGE resolution of proteasome complex activity in the cell lysate and the
corresponding EV fraction using a pan proteasomal cy5-epoxomicin activity-based probe. Proteasome
content is assessed by probing for the a3 proteasome subunit. Note the high abundance of inactive 20S
proteasome in exophers.
Panel 2
: Immunouorescent resolution of lysosomes using LysoTracker or of
dysfunctional mitochondria using MitoSOX in exophers (rbIgG+Mb594+ EVs) released to the medium
within 6 hours of rbTHSD7A-ab exposure.
Page 22/30
Figure 3
Dynamics of urinary exopher-release in experimental THSD7A+-MN
THSD7A+-MN was induced in mice by injection of rbTHSD7A-abs or ctrl-abs and urinary EVs were
analyzed. A)
Panel 1
: Size distribution of urinary EVs assessed by NTA. Violin plots indicate median, 25%
and 75% percentile; ****
p
<0.0001, unpaired t-test, pooled data of day 7 and 9 urine measurements of
N
=
Page 23/30
3 mice per group. Mean size of EVs is indicated.
The presence of exophers was assessed
panel 2
by
image stream and panel 3 by confocal microscopy using MemBrite594 (light blue) to stain the plasma
membrane, rbIgG+ (green) to detect bound autoantibodies, and 14-3-3 (red) as vesicular marker that is
abundant in exophers. Micrograph depicts 3D-reconstructed confocal
z
-stack of an exopher, individual
channels of one plane are shown in the bottom row micrographs. B)
Panel 1
illustrates experimental
setup in
mT/mG
reporter mice, which express membrane bound GFP (mGFP) in podocytes and
membrane bound tdTomato (mTomato) in all other cells. Urinary EVs were isolated from 250 µl urine and
109 total EVs were analyzed by image stream. EVs released from podocytes were discerned by their GFP
uorescence, EVs released from non-podocyte cells by tdTomato uorescence by image stream.
Panel 2
:
Kidney of a naïve
mT/mG
mouse was cleared with SCALE VIEW and the intrinsic uorescence of the
reporter proteins was assessed. Note the clear demarcation of the podocyte (p) plasma membrane with
mGFP at the cell body and foot processes (arrows).
Panel 3
: Representative day (d) -1 and day 10 image
stream plots of GFP+/rbIgG+ urinary EVs, which represent the MN-associated exophers.
Panel 4
:
Representative day 9 image stream plots of % Annexin V+ EVs from the mGFP+/rbIgG+ EV population
(MN-associated exophers) to demarcate apoptotic bodies. C) Graphs depict quantication of the 4
different EV subtypes released to the urine in the course of disease development, mean ± SEM,
N
= 11
control-abs,
N
= 14 rbTHSD7A-abs, pooled data from 2 independent experiments; *
p ≤
0.05, **
p ≤
0.01 to
ctrl mice, §§
p
≤
0.01 between ctrl-ab and rbTHSD7A-ab time course, 2-Way ANOVA with Tukey`s multiple
comparison test. Red arrow indicates time of rbTHSD7A-ab injection.
Page 24/30
Figure 4
Modulation of exopher-release in experimental THSD7A+-MN.
A)
Panel 1
: THSD7A+-MN was induced in
mT/mG
reporter and BALB/c mice on day 0 by administration
of rbTHSD7A-abs (red arrow), urine was collected in the time course, on day 10 kidney slices and
glomeruli were isolated. Abundance of rbIgG+-EVs was assessed from 109 total urinary EVs by image
Page 25/30
stream. Albuminuria was measured in parallel. Graph depicts the time course of urinary exopher release
(rbIgG+-EVs; particles/ml; left
y
axis, red round symbols) and albuminuria (right
y
axis, grey square
symbols; g/µl) in THSD7A+-MN.
Panel 2
: Additionally, short-term proteotoxic stress was achieved by
administration of a proteasomal inhibitor (epoxomicin, blue arrow) or vehicle (DMSO) every day after day
7 for 4 consecutive days. Graph depicts changes in urinary exopher-release (left
y
axis, colored round
symbols) and albuminuria (right
y
axis, grey square symbols) in short-term proteasome inhibited mice
relative to vehicle-THSD7A+-MN mice (dashed lines at 100%). Pooled values from 3 independent
experiments;
N
> 9 per group, mean ± SEM; *
p
< 0.05, 2-way ANOVA with Tukey`s multiple comparison
test.
Panel 3
: Long-term proteotoxic stress was achieved by administration of the proteasomal inhibitor
bortezomib or vehicle prior to MN induction on day -1 followed by 2 times per week for 10 days. Graph
depicts changes in urinary exopher-release (left
y
axis, blue round symbols) and albuminuria (right
y
axis,
grey square symbols) in long-term proteasome inhibited relative to vehicle THSD7A+-MN mice (dashed
lines at 100%); values from 1 experiment; N > 4 per group, mean ± SEM, ****
p
< 0.0001, 2-way ANOVA
with Tukey`s multiple comparison test.
Panel 4
and
5:
Kidney slices of
mT/mG
reporter mice that
express
membrane (m) bound GFP (mGFP, green) in podocytes were
optically cleared and stained for bound
rbTHSD7A-abs (red). In
panel 4
3D-reconstructed z-stacks of the podocyte cell body (cb) plasma
membrane is shown, in
panel 5
a 3D-reconstructed overview of a glomerulus; u = glomerular urinary
space, orange arrows highlight individual exophers. B) Spearman’s correlations of total glomerular
panel
1
THSD7A or
panel
2
rbIgG protein abundance (as detailed in
Fig. S17C
) and
panel 3
of albuminuria (g/
µl) to the respective urinary exopher abundance (particles/ml) at time of sacrice.
Page 26/30
Figure 5
Non-invasive exopher monitoring in relapsing THSD7A+-MN. A) Micrographs from the nephropathological
diagnosis (19/05/2021) of the patient P1 encompassing light microscopy with
Panel 1
: PAS staining for
general morphology; immunohistochemistry to
Panel 2
: huIgG,
Panel 3
: huIgG4, and
Panel 4
: THSD7A.
Orange arrows highlight aggregates within the urinary space, red arrows aggregates within parietal
epithelial cells.
Panel 5
: Transmission EM; white arrows highlight membrane extensions from major
Page 27/30
(
panel 5a
) and foot processes (
panel 5b
), p = podocyte, gbm = glomerular basement membrane, fp = foot
process, mp = major process, u = urinary space, c = capillary space. B) Dynamics and characteristics of
urinary exopher-release over 2 ½ years of relapsing nephrotic syndrome.
Graph
depicts on the left
y
-axis
the serum THSD7A autoantibody titer (red curve) and on the right
y
-axis urinary protein to creatinine ratio
(blue curve). The main clinical interventions, as well as the COVID-19 infection are indicated.
Panel 1
:
Image stream plots
demonstrate the relative content of huIgG4+ exophers in the urine in % exophers from
109 total EVs.
Panel 2:
Reactivity of the huIgG4 eluted from 109 exophers to THSD7A protein; hu-pod =
human podocytes with human THSD7A overexpression, mock = human podocytes without THSD7A
overexpression. Membranes were reprobed for THSD7A and b-actin to control for THSD7A
expression/height and loading. The exposure times for the huIgG4 signal are indicated, all blots originate
from one membrane exposed at the same time. Note the clear detection of THSD7A-autoantibodies
bound to released exophers in all urines, albeit a concomitant negative sTHSD7A-ab titer.
Panel 3
:
Proteomic molecular disease patterning of exophers isolated from U1-U8. For pathway analysis, only
proteins identied with ≥ 2 peptides with a Q-value < 0.001 were used. Graph displays the
p
-values
corresponding to the signicance of pathway enrichment of selected functions.
P
-values below 1E-80
were capped; recall rate = proteins identied and assigned to category / number of proteins that are
annotated to the category. Pathways reecting exopher-specic processes are highlighted in ocher.
Detailed information on the pathway proteins is provided in
Table S4
. A total list of identied proteins can
be found on PRIDE.
Page 28/30
Figure 6
Glomerular aggregate morphometrics and non-invasive exopher monitoring in PLA2R1+-MN. A) Confocal
images of PLA2R1 stainings (white) from PLA2R1+-MN patients were subjected to computational image
analyses of the glomerular PLA2R1 aggregate pattern.
Panel 1
: Scoring range of the PLA2R1 aggregate
density algorithm (D = density; arbitrary units); red areas in overlay are shown, orange arrows = urinary
Page 29/30
space PLA2R1 aggregates, white arrowheads subepithelial PLA2R1 deposition, P = podocyte nucleus, u =
urinary space, c = capillary space.
Panel 2:
PLA2R1 aggregate density was quantied in diagnostic
biopsies of a retrospective PLA2R1+-MN patient cohort32, assembled based on clinical outcome.
N
= 18
patients were stratied into the remission group R and
N
= 17 into the non-remission group NR. Boxplots
show the median, rst and third quartiles, and whiskers 1.5 times interquartile range of glomerular
PLA2R1 density (A.U. = arbitrary units) or of the serum PLA2R1 autoantibody (sPLA2R1-ab) titers in
units/ml at diagnosis, *
p
≤ 0.05, ***
p
≤ 0.001, Mann Whitney U-Test. B) Prospective analysis of a 22-yr-
old female PLA2R1+-MN patient P2 diagnosed 16/10/2021. Graph depicts clinical course of the serum
PLA2R1 autoantibody titer (red curve; left
y
-axis) and of the urinary protein to creatinine ratio (blue curve;
right
y
-axis). Diagnostic (U1) and follow-up (U2) urines were collected. The main clinical interventions are
indicated.
Panel 1:
Image stream quantication of %huIgG4+/14-3-3+ exophers present within 109
Mb594+ urinary EVs and immunoblot assessment of the specic reactivity of exopher-eluted huIgG4 to
huPLA2R1 or huTHSD7A protein in HEK cell lysates; reprobes control for PLA2R1 expression/height and
loading; exposure times for the huIgG4 signal are indicated. Note the clear detection of PLA2R1
autoantibodies in U1 and their absence in U2.
Panel 2
: Confocal analyses of PLA2R1 (red) and huIgG4
(green) expression pattern in the diagnostic biopsy. Boxplot shows quantication of PLA2R1 density, the
median, rst and third quartiles, and whiskers 1.5 times interquartile range are indicated; each datapoint
represents one glomerular section. Note the presence of small exophers in the urinary space (orange
arrows) and a disrupted subepithelial PLA2R1 pattern (white arrowheads). C) Scheme summarizing the
current concept of autoantibody fate using THSD7A as model antigen.
Panel 1:
The glomerular
autoantibody burden depends on 1. the amount of circulating autoantibody that binds subepithelially to
THSD7A and the ability of podocytes to dispose of these autoantibodies by 2. exopher-formation and 3.
exopher-release to the urine.
Panels 2
,
3
: Exopher-formation involves autoantibody-mediated THSD7A
crosslinking in the subepithelial space (1.), deposition of THSD7A/autoantibody complexes in the
subepithelial space (2a., classical concept of deposit formation; black arrowheads in the immunogold EM
micrographs to rbIgG) and/or translocation towards the apical foot process membrane (2b., new concept;
orange arrows in the immunogold EM micrographs to rbIgG). Apically displaced THSD7A/autoantibody
complexes at foot, major processes and podocyte cell body extend to the urinary space through exopher-
formation (3.). Exophers contain THSD7A and the extracellularly bound autoantibodies in addition to 14-
3-3 as discerning EV marker, disease-associated proteins, dysfunctional mitochondria, and degradative
machinery (4.). Exophers are released to the urine and can be taken up by other nephron cells (5.).
Scheme created with Biorender; ec = endothelial cell, mc = mesangial cell, p = podocytes, GBM =
glomerular basement membrane, fp = foot process, c = capillary lumen, u = urinary space, cb = cell body.
Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.
MovieS1andS2legend.docx
