Biomedicines 1841156

Article 1
Isthmin-1 is a novel mediator of experimental glomerulone- 2
phritis through a pro-apoptotic action in podocytes 3
Virgilia Sahiri1,2, Jonathan Caron1,2, Elena Roger1,2, Christophe Desterke9,10, Khalil Ghachem1,2, Inna Mohamadou1,2, 4
Justine Serre1,2, Niki Prakoura2, Soraya Fellahi4, Sandrine Placier1,2, Sahil Adriouch5, Lu Zhang6, Christos 5
Chadjichristos1,2, Christos Chatziantoniou1,2 , Hans Kristian Lorenzo7,8,9,# and Jean-Jacques Boffa1,2,3,#,* 6
7
1 Sorbonne Université, UMR_S 1155, Paris, France; 8
2 Institut National de la Santé et de la Recherche Médicale UMR_S 1155, Paris, France 9
3 Département Néphrologie et Dialyses, Tenon Hospital, AP-HP, Paris, France 10
4 Sorbonne Université, Inserm UMR_S 938, Centre de Recherche Saint-Antoine, Institut Hospitalo- 11
Universitaire de Cardio-métabolisme et Nutrition (ICAN), Paris, France. 12
AP-HP, Hôpital Tenon, Service de biochimie et hormonologie, UF Bio-marqueurs inflammatoires et 13
métaboliques, Paris, France. 14
5 Normandie University, UNIROUEN, INSERM, U1234, Pathophysiology, Autoimmunity, Neuromuscular 15
diseases and regenerative THERapies (PANTHER), 76000, Rouen, France. 16
6 Division of Nephrology, Affiliated Hospital of Nanjing University of Chinese Medicine, Jiangsu Province 17
Hospital of Chinese Medicine, Nanjing, China 18
7 Department of Nephrology, Bicêtre Hospital, AP-HP. 94270 Le Kremlin-Bicêtre, France 19
8 Université Paris Saclay, INSERM UMR_S 1197. 94803 Villejuif, France 20
9 Université Paris-Saclay, Faculté de Médecine. 94270 Le Kremlin-Bicêtre, France 21
10 Université Paris Saclay, INSERM UA/09 UMR-S 935, Villejuif, France 22
23
* Correspondence: hans.lorenzo@universite-paris-saclay.fr and jean-jacques.boffa@aphp.fr 24
# Co-senior authors 25
Abstract: Focal segmental glomerulosclerosis (FSGS) is a major cause of end-stage renal 26

disease and remains without specific treatment. To identify new events during FSGS 27
progression, we have used an experimental model of FSGS associated with 28
nephroangiosclerosis in rats injected with L-NAME (Nω-nitro-L-arginine methyl ester). After 29
Citation: Lastname, F.; Lastname, F.;
transcriptomic analysis we focused our study on the role of Isthmin-1 (ISM1, an anti- 30
Lastname, F. Title. Biomedicines 2022,
angiogenic protein involved in endothelial cell apoptosis. We have studied the renal 31
10, x. https://doi.org/10.3390/xxxxx
expression of ISM1 in L-NAME rats and other models of proteinuria, particularly at the 32
Academic Editor: Firstname Last- glomerular level. In the L-NAME model, withdrawal of the stimulus partially restored 33
name
basal ISM1 levels, along with an improvement in renal function. In other four animal 34
Received: date models of proteinuria, ISM1 was overexpressed and localized in podocytes while the 35
Accepted: date renal function was degraded. Together these facts suggest that the glomerular expression 36
Published: date of ISM1 correlates directly with the progression-recovery of the disease. Further in vitro 37
Publisher’s Note: MDPI stays neu- experiments demonstrated that ISM1 co-localized with its receptors GRP78 and integrin 38
tral with regard to jurisdictional αvβ5 on podocytes. Treatment of human podocytes with low doses of recombinant ISM1 39
claims in published maps and institu- decreased cell viability and induced caspase activation. Stronger ISM1 stimuli in 40
tional affiliations. podocytes dropped mitochondrial membrane potential and induced nuclear 41
translocation of apoptosis-inducing factor (AIF). Our results suggest that ISM1 42
participates in the progression of glomerular diseases and promotes podocyte apoptosis 43
Copyright: © 2021 by the authors. in two different complementary ways: one caspase-dependent and one caspase- 44
Submitted for possible open access independent associated with mitochondrial destabilization. 45
publication under the terms and
conditions of the Creative Commons
Keywords: Glomerular diseases, Proteinuria, Chronic kidney disease progression, FSGS, Podo- 46
Attribution (CC BY) license
cytes, apoptosis 47
(https://creativecommons.org/license
48
s/by/4.0/).
Biomedicines 2022, 10, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/biomedicines

Biomedicines 2022, 10, x FOR PEER REVIEW 2 of 41
1. Introduction 49
Focal segmental glomerulosclerosis (FSGS) is the histological expression of a set of 50
pathologies characterized by marked proteinuria. It is one of the leading causes of end- 51
stage kidney disease worldwide [1]. Although the mechanisms of proteinuria are better 52
understood, FSGS still lacks effective treatment, so the discovery of therapeutic targets is 53
of utmost importance. 54
Podocytes are the main target cell for injury in FSGS. Together with endothelial cells, 55
they are essential for the formation of the glomerular filtration barrier. The podocyte is 56
particularly vulnerable due to its highly differentiated postmitotic phenotype. Structural 57
changes involved in the progression of FSGS include alterations of the cytoskeleton, 58
effacement of the foot of podocytes, and cell death. All this implies the alteration of the 59
glomerular basement membrane and the leakage of nonspecific plasma proteins into the 60
urinary space. The sources of podocyte injury are varied (genetic abnormalities, 61
circulating factors, viral infections, medications, etc.), however the effect on podocytes is 62
similar [1]. Also some vascular injuries (e.g., due to ischemia, hypertension) can cause 63
FSGS [2]. Thus, the appearance of proteinuria has been observed in patients undergoing 64
antitumor treatment with antiangiogenic drugs, probably resulting from an imbalance 65
between proangiogenic/antiangiogenic factors [3]. 66
Isthmin-1 (ISM1) is an anti-angiogenic protein involved in the apoptosis of 67
endothelial cells. It is a 60 kDa secreted protein that contains a centrally located 68
thrombospondin type 1 repeat (TSR) and a C-terminal AMOP domain (adhesion- 69
associated domain in MUC4 and other proteins) [4]. The rat gene shares a strong 70
homology (89,52% of identity) with the human gene. In adult mammals, ISM1 is expressed 71
in many tissues including lymphocytes, heart, lung and kidneys [5] but its physiologic 72
role remains poorly studied. Inded, ISM1 had never been involved in renal physiology 73
and/or pathophysiology. Recently, Jiang and colleagues reported that mature adipocytes 74
secrete ISM1 and trigger an insulin-like signaling cascade [6]. In endothelial cells, secreted 75
ISM1 potently inhibits VEGF/basic fibroblast growth factor (bFGF)-induced angiogenesis 76
[7]. Other studies demonstrated that ISM1 acts as a vascular permeability factor in 77
lipopolysaccharide (LPS) or hypoxia models in mouse lung [8][9]. In addition, ISM1 binds 78
two membrane receptors, integrin avb5 and GRP78 triggering apoptosis in tumor and 79
endothelial cells. Interestingly, both receptors are expressed in glomeruli and have been 80
linked to glomerular diseases. [10][11][12][13]. GRP78 is an endoplasmic reticulum- 81
resident chaperone that maintains protein homeostasis and regulates the unfolded protein 82
response. GRP78 is upregulated in glomerular diseases and contributes to diabetic 83
nephropathy [11]. Inhibition of integrin avb5 attenuates vascular permeability and 84
protects against renal ischemia-reperfusion injury and decrease mortality in septic mice 85
[14][15]. 86
The present study aims to elucidate whether ISM1 is involved in the mechanisms 87
leading to glomerular damage during renal disease progression particularly focused in 88
podocytes. For this purpose we used several models of proteinuria including L-NAME- 89
treated rats (Nω-nitro-L-arginine methyl ester), a non-specific nitric oxide (NO) synthase 90
inhibitor, an experimental model of hypertension and secondary FSGS of vascular origin 91
[16][17]. Similarly, we have tested the involvement of ISM1 in other in vivo models of 92
proteinuria, as well as in cultured podocytes. 93
2. Materials and Methods 94

2.1. Reagents and Antibodies 95
Recombinant ISM1 (endotoxin free) was produced by myBiosource (MBS1208159) 96
expressed in yeast and purified near homogeneity. Antibodies for immunofluorescence: 97
anti-ISM1 was provided from Thermo Fischer (PA5-24968, 1:100 dilution); anti-β5 from 98
R&D systems (AF8035, 1:20 dilution); anti-GRP78 from Genetex (GTX102567, 1:20 dilu- 99
tion); anti Cytochrome c from BD Pharmingen (556432, 1:400 dilution); anti-AIF from 100
Santa Cruz (sc-13116, 1:200 dilution). Antibodies for immunoblotting and immunoprecip- 101
itation: anti-ISM came from Biolegend (622201, 1:500 dilution); anti-AIF from Sigma 102
(A7549, 1:100 dilution); anti-cytochrome c from BD Pharmingen (556433, 1:100 dilution); 103
anti-synaptopodin (sc-21537), anti-caspase-8 (sc-5263, 1:200 dilution); anti-caspase-3 (sc- 104
7272, 1:200 dilution); anti-GRP78 (sc-166490, 1:100 dilution) and anti-αvβ5 (sc-13588, 1:100 105
dilution) were provides by from Santacruz. 106
107
2.2. Animal models of proteinuria 108
All animals were kept in an air-conditioned room with food and tap water ad libitum. 109
All these studies were authorized by an Ethics Committee (#12215). All animal procedures 110
were in accordance with European Union Guidelines for the Care and the use of labora- 111
tory animals and were approved the local ethic committee (Comité National de Reflexion 112
Ethique sur l’expérimentation Animale #05). 113
114
2.2.1. Hypertensive nephropathy with FSGS lesions (L-NAME model) 115
L-NAME (NG-nitro-L-arginine methyl ester) is a competitive inhibitor of NO syn- 116
thases, that induces high blood pressure, hypoxia, and renal fibrosis [17]. L-NAME (15 117
mg/kg per day) was administrated in drinking water to male Sprague–Dawley rats (250 118
g, Envigo. Gannat, France) as previously described [16]. We added NaCl (6 g/l) in the 119
drinking water to accelerate the renal injury as previously detailed [18]. Weekly measure- 120
ment of proteinuria allowed us to set a threshold of 1 g/mmol of urinary creatinine beyond 121
which severe nephroangiosclerosis lesions were present (mortality rate ~20%) [19]. At this 122
level of urinary protein excretion ratio (UPER), we discontinued L-NAME administration. 123
Then we proceeded to monitored the renal repair at different times after interruption of 124
L-NAME as previously reported [20]. Four groups of animals were studied: 125
- Group W0, sacrificed at the peak of renal damage (UPER> 1 g/mmol) (n = 6). 126
- Group W1, animals with UPER> 1g/mmol were sacrificed one week after the inter- 127
ruption of L-NAME (n = 5). 128
- Group W5, animals with UPER> 1 g/mmol were sacrificed 5 weeks after the inter- 129
ruption of L-NAME (n = 4). 130
- CTL group, which did not receive L-NAME and was sacrificed alongside other 131
groups to obtain animals of the same age (n = 4). 132
133
2.2.2. Puromycin model 134
Sprague-Dawley rats were purchased from Envigo (Gannat, France). Puromycine 135
aminonucleoside (PAN) was obtained from Santa Cruz Biotechnology (California, USA). 136
Rats were divided into two groups: PAN-treated and a control group of non-treated rats. 137
Then, eleven 120-150 g male rats from 7 to 8 weeks old were injected subcutaneously with 138
daily dose of PAN (120 mg/kg) for 10 days. The control untreated group included 7 rats 139
injected with the same volume of saline (0.9% NaCl). Ten days after the last injection, an- 140
imals were sacrified and urine, blood and kidney tissues collected. 141
142
2.2.3. Doxorubicin model 143
Twelve Sprague-Dawley rats received a single penis-vein injection of a solution of 6 144
mg/kg Doxorubicin (DOXO). As control group, 9 rats received a single injection of saline 145
into penis-vein. The rats were placed in metabolic cages for urine collection weekly and 146
30 days after the treatment, then rats were sacrificed and urine, blood and kidney tissues 147
collected for analysis. 148
149
2.2.4. Diabetic nephropathy model 150
This model is used to analyze the role of hyperinsulinemia and obesity in the patho- 151
genesis of non-insulin-dependent diabetes and diabetic nephropathy. C57BL/6J/ob/ob 152
mice were purchased from Jackson Laboratories (Bar Harbor, ME) and maintained in a 153
specific pathogen–free facility with a 12-hour light cycle and free access to standard diet 154
and water. Wild type C57BL/6 (n=5) mice were used as control. Mice were killed at 18 155
weeks of age (n= 4), and urine, blood and kidney tissues were collected. 156
157
2.2.5. Lipopolysaccharide (LPS) model 158
Female wild-type BALB/c mice (8–10 weeks old) were purchased from Charles River 159
Laboratories (Lyon, France), subsequently housed and given free access to food and wa- 160
ter. After two weeks, 8 mice were injected with LPS i.p. (10µg/g of weight; Sigma Aldrich). 161
A control group of 5 mice received an equal volume of saline (i.p.). Blood, urine and kid- 162
ney tissues were collected 24h after LPS or saline treatment. For all these models, blood 163
and urines samples were immediately spun for 10 minutes at 4000 rpm and frozen 164
at−80°C. The subsequent biochemical measurements (albuminuria, creatininuria) were re- 165
alized on the Architect system (Abbott Biagnostic) by the Department of Biochimie–Hor- 166
monologie, Tenon Hospital (Paris, France). 167
168
2.3. DNA microarrays analysis 169
The hybridization protocol and the computer analysis were carried out in collabora- 170
tion with the genomics platform of the Cochin Institute (Genom'IC). One quarter of each 171
animal's kidney was thawed and ground, and the RNA was extracted using Trizol (Invi- 172
trogen, Camarillo, CA, USA). The quantity, purity and quality of the RNAs were verified 173
using Nanodrop (Thermo Fisher Scientific, Waltham, MA., USA) and a Bioanalyzer (Ag- 174
ilent Technologies, Santa Clara, CA, USA). 175
The RNAs of interest were diluted to a concentration of 100 ng/µl before starting the 176
synthesis of the first strands of cDNA. The cDNAs were synthesized using a NuGEN kit 177
(San Carlos, CA, USA), the 2nd cDNA strands from the previous ones; the double- 178
stranded cDNA thus obtained was purified by magnetic method. Linear amplification 179
(SPIA technology) of these double-stranded DNAs was then performed to obtain a suffi- 180
cient amount of hybridization material on the chips while maintaining the respective pro- 181
portions of each RNA of the original sample. After purification, the amplified cDNAs 182
were enzymatically fragmented and associated with a fluorescent label. Finally, the sam- 183
ples were denatured at 99°C and hybridized on the chip (Affymetrix, Santa Clara, CA, 184
USA) for 17 h at 45°C. At the end of the hybridization, the chips were washed and placed 185
in a high resolution scanner in order to obtain the expression values corresponding to each 186
probe. The fluorescence values obtained for each of the probes were normalized using the 187
RMA algorithm (Robust Microarray Analysis) [21]. This approach allows to reduce differ- 188
ences in hybridization between the chips and to facilitate the identification of true differ- 189
ences of expression. The list of transcripts differentially expressed in the three groups 190
compared to the control group was then imported in the software IPA (Ingenuity Systems) 191
for subsequent pathway analysis. The raw data were deposited in the gene expression 192
omnibus (GEO) database of NCBI with the accession number GSE151690. 193
194
2.4. Immunohistochemistry analysis 195
Kidneys collected from animals were snap-frozen in liquid nitrogen, and fixed in ac- 196
etone. For tissue staining, frozen cryostat sections (4-µm) were treated by BSA 3% in PBS 197
during 30 min, incubated with ISM-1 antibody and detected by secondary antibodies pur- 198
chased from Nichirei-Histofine Simple Stain rabbit MAX PO (Tokyo, Japan). Revelation 199
was achieved with 3-amino-9-ethylcarbazole AEC (Dako, Carpinteria, CA). For examina- 200
tion in fluorescent conditions, Alexa Flour Plus secondary antibodies were used (Ther- 201
moFisher, France). 202
For transmission electron microscopy (TEM) analysis, Kidneys were cut into small 203
pieces and immersed in 2.5% glutaraldehyde containing 1% tannic acid in 0.1 M PBS for 2 204
hours at 4°C. Samples were post fixed with 1% OsO4, dehydrated and embedded in epoxy 205
resin. Ultrathin sections were stained with uranyl acetate and lead citrate and then exam- 206
ined under a Philips CM10 electron microscope. Ultrathin frozen sections were processed 207
for indirect immunogold labeling, as described [22]. 208
209
2.5. Isolation of glomeruli and qPCR analysis 210
Decapsulated glomeruli were isolated as described previously [23]. Briefly, freshly 211
isolated renal cortex was mixed and digested by collagenase I (2 mg/ml; Gibco) in RPMI 212
1640 (Life Technologies) for 2 minutes at 37°C, then collagenase I was inactivated with 213
RPMI 1640+10% FCS (Abcys). Tissues were then passed through a 100-µm cell strainer 214
and 40-µm cell strainer for rat tissues and through 70-µm cell strainer and 40-µm cell 215
strainer (BD falcon) for mouse’s tissues in PBS (Life Technologies) + 0.5% BSA (Eu- 216
romedex). Glomeruli, adherent to the 40-µm cell strainer, were taken from the cell strainer 217
with PBS+0.5% BSA injected under pressure, then washed twice in PBS. Isolated glomeruli 218
were then resuspended in RLT extraction buffer (Qiagen) and frozen at −80°C until total 219
RNA extraction. The glomerular mRNA was extracted by using the EZ-10 Spin Column 220
Total RNA Kit (BD Life Science) according to the manufacturer's instructions. RNA con- 221
centration was measured by using NanoDrop1000 spectrophotometer (Thermo Fischer 222
Scientific Biosciences, Germany). RNA were reverse transcribed using Maxima First 223
Strand cDNA Synthesis Kit (Thermo Fischer Scientific Biosciences, Germany) and PCR 224
was performed using SYBR green (Roche Diagnostics, Meylan, France) and specific pri- 225
mers (Table SI) purchased from Euronfins Scientific (Paris, France) on a Light Cycler 480 226
(Roche Diagnostics, Meylan, France). Expression levels were normalized to the house 227
keeping gene, HPRT (Hypoxanthine-guanine phosphoribosyl transferase) or GusB (beta- 228
glucuronidase) using Light Cycler® advanced relative quantification program (Roche). 229
230
2.6. Podocytes cell culture 231
Conditionally immortalized transgenic human podocytes enabled via a thermosen- 232
sitive variant of SV-40 were provided by Dr. Moin Saleem [24]. Briefly, podocytes were 233
cultured in RPMI-1640 with ITS (insulin (10 µg/ml), transferrin (5.5 µg/ml), selenium (5 234
ng/ml Na selenite)), heat-inactivated fetal bovine serum (10 % v/v), 100 U/ml penicillin 235
and 0.1 mg/ml streptomycin (Invitrogen, Carlsbad, CA) at 33 °C under a humidified at- 236
mosphere of 95 % air and 5 % CO2 with change of the medium every 2 days. Under these 237
conditions of culture, podocytes exhibit a proliferative phenotype (permissive condi- 238
tions). Podocytes at 50–60 % of confluence were thermo-switched from 33 °C to 37 °C for 239
differentiation under non-permissive conditions (same medium without ITS) for 14 days. 240
Differentiated podocytes were used in all experiments. 241
242
2.7. Cell viability and TUNEL assay 243
The viability of cultured podocytes was assessed using the MTT assay. The tetrazo- 244
lium salt 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay 245
was purchased from Sigma-Aldrich (M5655). Cells were cultured in 96-well plates (20000 246
cells/well) and incubated in 200 µl of non-permissive medium (37°C) in different condi- 247
tions (ISM1, inhibitors). After incubation, MTT was added to the medium (10 µg/ml), and 248
the plates incubated at 37°C for 2 hours and the MTT solution removed. The MTT-forma- 249
zan produced was solubilized with dimethyl sulfoxide (DMSO) and measured at 590 nm 250
in a microplate reader (Thermo Fisher). 251
Apoptosis was measured using a TUNEL assay kit obtained (from Abcam. CA, USA). 252
Podocytes were fixed for 15 minutes using 1% paraformaldehyde and after fixation, the 253
cells were rinsed with PBS, treated with 70% ethanol and incubated for 30 minutes on ice. 254
After incubation, the cells were washed with the provided buffer and 50 µl of DNA label- 255
ing solution was added to the cells, for 1 hour at 37°C. The cells were then washed with 256
buffer and re-suspended in DAPI solution for 5 minutes and incubated in the dark. The 257
cells were then observed under with a Leica DM-RXA23D microscope (Wetzlar, Ger- 258
many). 259
260
2.8. Immunofluorescence 261
Podocytes were seeded onto uncoated coverslips (Marienfeld GmbH, Germany) and 262
incubated with recombinant ISM1 in different conditions. After washing in PBS, cells were 263
fixed in paraformaldehyde (3%) for 15 min, washed with PBS, and incubated for 5 min in 264
100 mM NH4Cl in PBS. Cells were permeabilized with 0.1% saponin, saturated with fetal 265
bovine serum 3% in PBS and incubated for 1 h with primary antibody. After washing, 266
cells were incubated with an appropriated secondary antibody conjugated with Alexa 267
Fluor 594 or 488, (Invitrogen). Cells were mounted in Mowiol medium and fluorescence 268
observed in a Leica DM-RXA23D microscope (Wetzlar, Germany). 269
270
2.9. Fluorimetric determination of Mitochondrial Membrane Potential (ΔΨm) 271
The measurement of the mitochondrial membrane potential (ΔΨm) was assessed by 272
using the lipophilic cationic dye JC-1 (5,5, 6, 6’-tetrachloro-1,1’, tetraethylbenzimidazolo- 273
carbo-cyanine iodide; Sigma) which is accumulated within the mitochondria. When ex- 274
cited at 490nm, monomers of JC-1 fluoresce at 530nm. Higher concentrations of JC-1 forms 275
aggregates which emit maximally at 590nm. Consequently, cells with high ΔΨm fluoresce 276
orange, while cells with low ΔΨm fluoresce green. Loss of ΔΨm is considered as a marker 277
of the onset of apoptosis. As described elsewhere (Roy SS. 2009), podocytes were seeded 278
into 96-well plates or coverslips at a density of 20000 cells/well in different conditions. 279
Then JC-1 (20 µM) was added to growth medium and incubated 30 min at 37 °C. The 280
fluorescence in each well was measured in a fluoroskan® (Thermo Fisher) microplate flu- 281
orometer (excitation, 490 nm; emission, 530 nm for JC-1 monomer and 590 nm for JC-1 282
aggregates). The mitochondrial membrane potential (ΔΨm) from each group was calcu- 283
lated as the fluorescence ratio of red to green and expressed as a percentage of the control. 284
285
2.10. Western blotting and Immunoprecipitation assays 286
Cultured podocyte were lysed in RIPA buffer (Thermo Fisher), containing a protease 287
inhibitor cocktail (Thermo Fisher). Equal amounts of protein (25–50 µg), were separated 288
by SDS-PAGE, transferred onto a PVDF membrane and blocked (3% BSA in TBS-Tween). 289
Membranes were subsequently probed with the indicated primary antibodies. Appropri- 290
ate HRP-conjugated secondary antibodies were used and signal was detected with an Im- 291
mobilon Western kit (Millipore, Molsheim, France). Densitometric analysis of visualized 292
bands was performed using Image J software. 293
For immunoprecipitation experiments, podocytes were lysed in IP lysis buffer 294
(Thermo Fisher). Equal amounts of protein were immunoprecipitated with saturating 295
amounts of anti-GRP78 or anti-αvβ5 antibodies overnight at 4°C, followed by incubation 296
with 20 µl of protein G-Sepharose beads for 1 hour at 4°C. The beads were then washed, 297
boiled, and the supernatants were used to immunoblot using an anti-ISM antibody. All 298
blots were performed in three independent experiments. 299
300
2.11. Statistical Analysis 301
All in vitro experiments were repeated at least 3 times. Values are expressed as the 302
mean ±SEM. Student's t test was used to observe differences between two groups. 303
ANOVA test following by t test was used when more than two groups were present. All 304
statistical analyses were performed using GraphPad Prism 5.0. Differences were consid- 305
ered statistically significant when the p value was <0.05. 306
307
2.12. In silico analysis of ISM1-related pathways of the L-NAME model. 308
Transcriptome analysis was performed in R software environment 4.1.0. Expression 309
heatmap were drawn with pheatmap R-package version 1.0.12. Alluvialplot were drawn 310
with ggalluvial P-package version 0.12.3. ISM1 supervised expression profile was deter- 311
mined with Pavlidis Template Matching algorithm [25] during time course experiment of 312
the GEO dataset GSE151609. Functional enrichment was performed with with Toopgene 313
web suite application [26]. Functional enrichment networks were drawn with Cytoscape 314
standalone application version 3.6.0 [27]. Text mining related gene-function associations 315
was performed with GeneValorization [28] application on NCBI database. 316
317
3. Results 318
3.1. L-NAME induced nephroangiosclerosis with focal segmental glomerulosclerosis 319
Administration of L-NAME in the drinking water caused hypertension, high pro- 320
teinuria and typical lesions of nephroangiosclerosis (Figure 1A and B). The mean UPER 321
of all L-NAME-treated hypertensive rats was 2170 ± 183 mg/mmol at the time of L-NAME 322
removal, 27-fold higher compared to the control group (80 ± 19 mg/mmol p <0.001). The 323
blood pressure of the W0 group was 206 ± 9 mmHg compared to 131 ± 7 mmHg for the 324
control animals (p <0.01). L-NAME administration induced renal failure, with plasma 325
urea nitrogen of 22 ± 5 mmol/L, compared with 5 ± 0.7 mmol/L in the control animals (p 326
<0.01). Serum creatinine was 118 ± 19 µmol/L, compared with 29 ± 3 µmol/L for control 327
animals (p <0.01; Figure 1A). As described, L-NAME induced nephroangiosclerosis with 328
focal segmental glomerulosclerosis (FSGS) lesions when UPER was greater than 1 g/mmol 329
[19]. As expected, W0 group was significantly more affected than the control group when 330
considering arteriolosclerosis, interstitial fibrosis and tubular dilatation (Figure 1B). 331
One week after interruption of L-NAME treatment (W1), UPER significantly de- 332
creased from 2170 ± 183 mg/mmol to 582 ± 123 mg/mmol (p<0.001), without returning to 333
normal levels (p<0.05, W1 vs CTL). After 5 weeks without treatment (W5), proteinuria was 334
further decreased to 301 ± 39 mg/mmol. Systolic arterial hypertension was maintained 335
despite withdrawal of L-NAME, with W1 and W5 values of 183 ± 7 mmHg and 180 ± 8 336
mmHg, respectively, compared to 131 ± 7 mmHg in control animals (p <0.01). Plasma urea 337
and serum creatinine returned to normal levels after stopping L-NAME, with respectively 338
5.0 ± 0.4 mmol/L and 46 ± 8 µmol/L in W1 (p <0.01 compared with the W0 group). Five 339
weeks after L-NAME withdrawal, values returned to normal (Figure 1A) and renal lesions 340
improved markedly: glomerulosclerosis, tubular dilatation, arteriosclerosis and intersti- 341
tial fibrosis scores decreased (data not shown). 342
343
344
345
346
347
348
349
350
351
352
353
354
355
a) 356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
b) 375
376
377
378
379
380
381
382
383
384
385
386
387
Figure 1: L-NAME impaired the renal function in rats and induced renal injuries. (a) Blood 388
pressure and renal function in L-NAME rat model. Parameters studied included UPER, uremia and 389
creatinemia in CTL, W0, W1 and W5 groups. At the peak of the disease (W0), the rats developed an 390
elevated UPER accompanied by high blood pressure and high levels of urea and serum creatinine. 391
After stopping L-NAME treatment (W1 and W5), renal function recovered to normal (*** p <0.001 392
and ** p <0.01). (b) Histological evaluation of renal tissues. Representative examples of renal samples 393
stained with Masson’s trichrome stain in Control (CTL) and L-NAME-treated rats. 394
3.2. ISM1 gene expression is associated to the intensity of renal injury in L-NAME-induced 395
nephropathy and its repair. 396
The above series of experiments allowed the formation of 4 groups showing different 397
degrees of progression and/or regression of renal disease: a control group (CTL), a group 398
with severe glomerulonephritis (W0), a group at early phase of recovery (W1) and a group 399
with an almost complete reversion of renal disease (W5, see Figure 2A). We performed an 400
unbiased transcriptomic analysis in the kidneys of the 4 groups in order to identify genes 401
whose expression level is associated with the phases of progression/regression of renal 402
disease (Figure 2B). We selected ISM1 not only because of the intensity of its expression 403
but above all because of its variability, which was correlated to the recovery of renal func- 404
tion. This protein had never before been reported to be involved in renal pathophysiology 405
and, furthermore, rodent genes share strong homology with humans (89.52% identity for 406
rats). 407
We found 424 genes with at least a two-fold change in expression during the progres- 408
sion of the disease (CTL versus W0 group, p<0.05). The 25 genes with the highest increased 409
and decreased expression are listed in Table SII. Although ISM1 was not among the most 410
highly expressed genes in W0, its expression correlated well with disease progression/re- 411
gression. In this model, ISM1 was expressed in the W0 group and gradually returned to 412
normal in the W1 and W5 groups. This pattern of ISM1 expression became more evident 413
when the study was focused on isolated glomeruli, rather than total kidney tissue (micro- 414
arrays). Thus, quantitative RT-PCR assays in isolated glomeruli, revealed that the W0 and 415
W1 groups had significantly higher ISM1 expression than the control group (Figure 2C). 416
a) 417
418
419
420
421
422
423
424
b) 425
426
427
428
429
430
431
c) 432
433
434
435
436
437
Figure 2: ISM1 gene expression in L-NAME model. (a) Experimental protocol of L-NAME model. 438
When urinary protein excretion ratio was over 1g/mmol of creatinine, after 6 to 24 weeks, NG-nitro- 439
L-arginine methyl ester (L-NAME) was removed. (b) Left panel. Segregation of transcriptomic data 440
of each animal by principal component analysis: Each animal is represented by a circle in the 441
corresponding group (CTL in black, W0 in red, W1 in blue and W5 in green). Right panel. 442
Hierarchical grouping of transcriptomic data from each animal: Each line corresponds to a gene and 443
each column to an animal. The red color indicates an overexpression and the green color an under 444
expression. This representation clearly shows a variation of gene expression from one group to 445
another. (c) ISM1 is overexpressed in glomeruli of the L-NAME rat model. mRNA expression of 446
ISM1 from kidney of CTL rats, L-NAME treated rats (W0), 1 week L-NAME removal rats (W1) and 447
5 weeks L-NAME removal rats (W5). mRNA of ISM1 is significantly overexpressed in the kidney of 448
W0 and W1 L-NAME rats (* p<0.05 vs control group ). 449
3.3. ISM1 and receptors are expressed in podocytes of different models of glomerulopathy. 450
We analyzed the renal expression of ISM1 in rat, mouse and human kidneys in phys- 451
iological conditions. A clear expression of ISM1 was observed in renal glomeruli, more 452
pronounced in rats than in humans kidneys (Figure 3A). In rat kidney tissue, ISM1 co- 453
localized with nephrin (Figure 3B) but not with endothelial markers such as CD31 (Figure 454
3C) or RECA-1, a rat pan-epithelial marker (Figure S1a). These results suggest a podocytic 455
localization of ISM1 that was confirmed by immunogold electronic microscopy (Figure 456
3B, right). In addition, the two known FIGURE 3
ISM1 receptors, integrin αvβ5 and GRP78, were 457
also expressed in glomeruli under basal conditions (Figure S1b). 458
a)
ISM1 (rat) +blocking peptide (rat) ISM1 (human)
c
b)
ISM1 NEPH1 Merge TEM (anti-ISM1)
c)
ISM1 CD31 Merge
Figure 3: Physiological expression of ISM1 in rodents and human kidneys. (a) Glomerular 459
expression of ISM1 in kidney from rats (a, b) and human (c). ISM1 is expressed exclusively in 460
podocytes. (b) ISM1 expression is localized in podocytes. Immunofluorescence showing 461
colocalization between ISM1 and nephrin. Immunogold electron microscopy: ISM1 is localized in 462
the cytoplasm (circled in red) and in the foot process of podocytes (red arrow). (c) ISM1 colocalized 463
poorly with the endothelial cell marker CD31. 464
ISM1 expression was also explored in different rodent models of glomerulopathy: L- 465
NAME, Puromycin aminonucleoside (PAN), doxorubicin in rats and diabetes (ob/ob) and 466
LPS in mice (Figure 4). All those models developed severe proteinuria associated with net 467
hypoalbuminemia (Table SIII). RT-qPCR experiments showed a significant overexpres- 468
sion of ISM1 mRNA in the glomeruli of all models studied (Figure 4A). Of note, in PAN- 469
treated rats, mRNA levels of avb5 integrin and GRP78 (both described as ISM1 receptors) 470
increased concomitantly with ISM1 (Figure 4B). Compared to controls, integrin b5 subunit 471
mRNA was significantly overexpressed in aged diabetic mice, while no significant 472
changes were observed in the rest of the other models (Figure 4B). 473
474
475
476
a) 477
4785
479
4,5 ***
ISM1 expression (ratio to CTL)
4
480
3,5
**
4813 **
482
2,5
*
FIGURE 4 *
4832
a)
484
1,5
4851
486
0,5
4870 CTL L-NAME PAN DOX ob/ob LPS

488
b) 489
490
491
10 6
9 *** ***
!5 integrin (ratio to CTL)
5
GRP78 (ratio to CTL)
b) 7 4
6 ***
5 3 *
!5 integrin expression (ratio to CTL)
4 6
2 ***
3 5
2
14
1
3 *
0 0
2
CTL PAN DOXO ob/ob CTL PAN DOXO ob/ob
1
492
0
CTL PAN DOXO ob/ob 493
c) 494
c) 495
496
497
498
499
500
501
502
503
504
505
506
Figure 4. Expression of ISM1 and its receptors in different models of glomerulopathies. Estimation 507
by RT-qPCR (n=6): (a) ISM1 expression in different models of proteinuria (L-NAME (W0), doxoru- 508
bicin, puromycin, obesity and LPS. Values normalized to CTL; (b) Expression de GRP78 (left) and 509
integrin-b5 (right) in doxorubicin, puromycin and obesity models. *, p<0,05 vs CTL **, p<0,01 vs CTL 510
***, p<0,001 vs CTL (c) Interaction between ISM1 and its GRP78 and integrin αvβ5 receptors in 511
cultured podocytes. Co-Immunoprecipitation (IP) of GRP78 and integrin αvβ5 using podocyte cells 512
and then, western blot of ISM1. The ISM1 interacts with its receptors in podocytes, in basal condi- 513
tions, (n = 2). 514
515
Given the presence of ISM1 receptors on glomeruli, we proceeded to explore its bind- 516
ing to the surface of cultured podocytes. In vitro experiences revealed that ISM1 and its 517
receptors were expressed by cultured podocytes. Co-immunoprecipitation experiments 518
demonstrated the interaction of soluble ISM1 to human podocytes via αvβ5 integrin and 519
GRP78 (Figure 4C). 520
521
3.4. Recombinant ISM1 induced podocyte injury in vitro. 522
To determine the degree of damage to podocytes, we examined the effect of ISM1 on 523
the structure of the actin cytoskeleton, as well as on cell viability. To determine the degree 524
of damage to podocytes, we examined the effect of ISM1 on the structure of the actin cy- 525
toskeleton, as well as on cell viability. Loss of F-actin stress fibers in cultured podocytes 526
mimics changes in podocytes in vivo that are primarily characterized by a peripheral dis- 527
tribution of F-actin and a delocalization of synaptopodin (SYNPO), an actin-associated 528
protein [29]. In our study, the distribution of F-actin and SYNPO was significantly modi- 529
fied by the strongest stimuli at high doses of ISM1 (1 µM) and longer incubations (72 h). 530
Under these conditions, actin adopted a subcortical distribution just beneath the plasma 531
membrane, while the presence of F-fibers was considerably reduced, which is compatible 532
with destabilization of the podocyte cytoskeleton (Figure 5A). 533
On the other hand, the addition of recombinant ISM1 to podocytes induced a dose- 534
and time-dependent decrease in their viability (Figure 5B). This ISM1 toxicity was signif- 535
icantly reduced after the addition of QVD-oph, a pancaspase inhibitor (Figure 5C). Thus, 536
podocytes were treated with ISM1 (5 and 35 µg/ml) for 24 and 72 hours in the presence or 537
absence of QVD-oph. Under milder conditions (5 µg/ml, 24 h), QVD-oph prevented ISM1- 538
induced mortality, suggesting that caspases are activated and responsible for this effect. 539
However, at higher doses or longer incubations, ISM1 toxicity was only partially inhibited 540
by QVD-oph (Figure 5B right) pointing towards an additional mechanism of caspase-in- 541
dependent killing. This suggests that ISM1-induced apoptosis in the podocyte operates 542
through both caspase-dependent and caspase-independent pathways that appear to act 543
in concert. 544
545
FIGURE 5
a)
546 MTT 24-72h

b) 24h
b) 547 48h
24h
100 72h
48h
548 100 72h

Cell viability (%)
***
Cell viability (%)
75
*** ***
549
***
50
***
55050
55125 ***
0
CTL 5 µg/ml 35 µg/ml
552
c)
MTT 24h
0 24h 72h
553 ISM1
ISM1 + QVD.oph
100
c) d) 554
Cell viability (%)
555
50 ISM1
ISM1 + QVD.oph
556 # ISM1
100 ## ISM1 + QVD.oph
100
** **
557
Cell viability (%)
0
CTL 5 µg/ml 35 µg/ml ##
Cell viability (%)
75
*** 75 ***
558
50
***
50 ##
559 5
***
25
25
560 ***
0 0
561 CTL 5 µg/ml 35 µg/ml
562
563
564
565
Figure 5. Recombinant ISM1 decreased podocyte viability. (a) Staining with phalloidin-FITC 566
showed that the F-actin cytoskeleton was disorganized after strong stimuli with ISM1 in podocytes. 567
Synaptopodin (SYNPO) lost its co-localization with actin upon addition of high concentrations of 568
ISM1 to podocytes. (b) Podocytes were treated with recombinant ISM1 (5 and 35 µg/ml) for 24, 48 569
and 72 hours. ISM1 decreased significantly the podocyte viability. ***, p<0,001 vs CTL. (c) Podocytes 570
were treated with ISM1 (5 and 35 µg/ml) for 24 and 72h in the presence or absence of QVD-oph, a 571
pan-caspase inhibitor. QVD-oph prevented ISM1-induced decrease in viability under mild 572
conditions (5 µg/ml, 24h) (C left).. At higher doses or longer incubations of ISM1, QVD-oph only 573
partially inhibited podocyte mortality (C right). **, p<0,01 vs CTL ***, p<0,001 vs CTL; ##, p<0,01 574
treated vs untreated QVD, #, p<0,05 treated vs untreated QVD. 575
576
3.5. Recombinant ISM1 induced podocyte apoptosis, mitochondrial membrane depolarization 577
and release of pro-apoptotic proteins. 578
This proapoptotic activity of ISM1 was also confirmed by TUNEL assay (Figure 579
6A,B), as well as by the concomitant decrease of procaspase-3 and procaspase-8 due to 580
their cleavage and activation of both (Figure 6C-D). In addition, we analyzed ROS pro- 581
duction or calcineurin activation, both events described as destabilizing podocyte struc- 582
ture and viability [29]. Their inhibition, respectively by N-acetylcysteine or FK-506, had 583
no effect on podocyte structure or viability (data not shown). 584
Previous works in endothelial cells demonstrated that ISM1 can target mitochondria 585
and trigger apoptosis [30] . Similarly, we monitored mitochondrial health using the dye 586
JC-1, which forms aggregates based on mitochondrial membrane potential (ΔΨm). Green 587
fluorescence indicates the monomeric form of the probe, which turns red with the for- 588
mation of "J-aggregates". Then, mitochondrial depolarization is estimated by a decrease 589
in the red/green fluorescence intensity ratio. As shown in Figure 6E, ISM1 addition in cul- 590
tured podocytes induced a significant decrease of the ΔΨm. This suggests, together with 591
the loss of cell viability (described above), the involvement of a mitochondria-dependent 592
apoptosis. 593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
e) f) 611
612
613
614
615
616
617
618
Figure 6: ISM1 induces apoptosis in podocytes. (A) Estimation of ISM1-induced apoptosis by 619
TUNEL assay after un strong stimulus ISM1 (35 µg/mL) for 72 H. Cells were observed under 620
fluorescence microscopy. (B) Apopotic cells were quantified. ISM1 induced an emergence of 621
TUNEL positive cells corresponding to apoptotic cells (** p< 0,01 vs control group; n = 3). (C) and 622
(D) Evidences of caspases activation. Representative immunoblots with quantitative data of 623
densitometry are shown in (C) for caspase 8 and in (D) for caspase 3. ISM1induced the decrease of 624
pro-caspase 8 and 3 (** p< 0,01; * p<0,05 vs control group; n = 4 blots). (E) ISM1 decreases the 625
mitochondrial membrane potential of podocytes (ΔΨm). Podocytes were treated with ISM1 (35 626
µg/mL) and quantification of ΔΨm was assessed by the JC-1 probe. ISM1 decreased significantly 627
the ΔΨm of podocytes from 48H to 72 hours (*** p< 0,01; ** p< 0,01 vs control group). (F) AIF was 628
released from mitochondria to the cytosol and the nucleus after ISM1 treatment (arrows). In contrast, 629
cytochrome c remained confined in mitochondria after ISM1treatment. 630
631
The release of mitochondrial intermembrane space proteins such as cytochrome c or 632
Apoptosis Inducing Factor (AIF) are unequivocally the central event in mitochondrial- 633
dependent apoptosis. The mechanism by which these proteins are released into the cyto- 634
sol presumably depends on the cell type and the nature of the stimuli [31]. Therefore, we 635
have examined the translocation of cytochrome c and AIF to the cytosol and nucleus. Un- 636
der basal conditions, AIF and cytochrome c are distributed exhibiting a punctate profile 637
due to their exclusively mitochondrial localization in podocytes (Figure 6F). Upon strong 638
stimulation of podocytes with ISM1, AIF diffused and translocated to the nucleus differ- 639
ently than cytochrome c (Figure 6F). These data suggest that strong ISM1 stimuli induce 640
mitochondrial destabilization in podocytes with mitochondrial membrane permeabiliza- 641
tion and nuclear translocation of AIF, a caspase-independent apoptosis-inducing protein. 642
643
FIGURE 7
644
Figure 7. Mitochondria and peroxisome regulation during ISM1 time course model during 645
kidney regeneration. A) Functional enrichment on gene ontology cellular component database of 646
genes found correlated to ISM1 during kidney regeneration (fdr: False discovery rate): barplot of 647
negative log10 FDR adjust p-values; B) Functional enrichment network performed with 648
mitochondria and peroxisome related genes found correlated to ISM1; C) Expression heatmap of 649
best 75 genes found correlated to ISM1 expression kidney regeneration: blue gene cluster defined 650
genes with regulation closed to ISM1; D) Alluvialplot built on gene-function found in NCBI 651
database with genes highlighted closely regulated with ISM1 during kidney regeneration 652
(quantitative colored links represent the numbers of NCBI-Pubmed articles connected) 653
7
654
3.6. Hypothetical roles of mitochondria: in silico analysis of ISM1 regulation in the L-NAME 655
model. 656
Whole transcriptome experiments were performed on total kidney organ during time 657
course of kidney regeneration with distinct time points: week 0, week1 and week5. 658
Palvidis template matching algorithm was used to found expression profile supervised 659
on ISM1 expression as predictor because ISM1 is known to be an anti-angiogenic actor 660
well expressed in cortical region of kidney comprising high proportion of glomeruli. After 661
correction of enriched genes with False Discovery Rate adjustment still 283 genes were 662
found significantly co-regulated to ISM1 expression during kidney regeneration (Supple- 663
mental Table SIV). Functional enrichment performed with these co-regulated genes on 664
Gene Ontology Cellular Component database highlighted an important implication of the 665
mitochondria and peroxisome cellular compartments (Figure 7a). Among this expression 666
profile associated to ISM1 regulation during kidney relation we could highlighted 48 mol- 667
ecules implicated in mitochondria compartment and 13 ones in peroxisome compartment: 668
9 molecules were found shared between these two compartments (Acox1, Hmgcl, Acsl1, 669
Slc27a2, Nudt19, Ech1, Crat, Mlycd, Ehhadh) (Figure 7b). Expression heatmap of best 75 670
genes correlated to ISM1 regulation during kidney regeneration revealed cluster of 13 671
genes closely regulated to ISM1 (blue cluster, Figure 7c). This blue cluster contained 672
Gadd45g particularly known to be implicated in apoptosis and caspase related function- 673
alities, also Neu1 well known in kidney and apoptosis literature, Acot4 associated to pe- 674
roxisome citations, follow by Mlh3 and Mapk4 associated to apoptosis function. These 675
results suggest that during time course kidney regeneration main genes co-regulated with 676
ISM1 are implicated in mitochondria and peroxisome compartments and associated to 677
caspase especially Gadd45g, Neu1, Mapk4, ISM1 and Mlh3. 678
4. Discussion 679
To date, most published studies on ISM1 have focused on its role during develop- 680
ment. ISM1 was first identified in Xenopus laevis gastrula embryos with prominent expres- 681
sion in the isthmus region of the brain [4], hence its name. ISM1 has been described as a 682
target for WNT/b-catenin and NODAL signaling as well as for hematopoiesis in zebrafish 683
development [32][33]. In mammals, using a microarray database, Valle-Rios reported the 684
first pattern expression of ISM1 in adults, in both human and mouse tissues [5]. Con- 685
trasting to embryo, ISM1 is not expressed in the central nervous system and is strongly 686
associated with barrier tissues, skin, mucosa and selected lymphocyte populations. In that 687
study, the whole kidney expression was very low. Under physiological conditions, we 688
demonstrated that ISM1 was expressed at the mRNA and protein level in adult rodents 689
and humans. This expression is compartmentalized and specific of podocytes, as indi- 690
cated by two complementary approaches, electron microscopy and colocalization im- 691
munocytochemistry using nephrin as a podocyte marker. 692
Very little is known about ISM1 and its eventual physiological role. ISM1 has two 693
identified receptors: GRP78, a protein usually activated during endoplasmic reticulum 694
stress [12][34], and integrin αvβ5, which promotes extracellular matrix stabilization and 695
peritubular capillary permeability [13][14]. In our study, both receptors were present in 696
rat glomeruli under normal conditions and ISM1 co-localized with GRP78, and partially 697
with the β5 subunit of αvβ5 integrin (Figure S1b). A previous study reported that immo- 698
bilized ISM1 can act as a pro-survival factor in cultured endothelial cells by promoting 699
cell adhesion and migration [35]. However, in our study, ISM1 did not appear to be sig- 700
nificantly expressed in glomerular endothelial cells. 701
In our transcriptomic study of the L-NAME model, we observed a strong variability 702
of ISM1 expression during progression of the disease. Thus, our results showed that ISM1 703
expression varied according the degree of renal damage, and were confirmed by RT-qPCR 704
analysis of isolated glomeruli. To ensure that this association was not a peculiarity of the 705
L-NAME model, we checked the expression of ISM1 in four additional renal disease mod- 706
els: doxorubicin-induced nephropathy and diabetes, PAN, and LPS. In all these models, 707
ISM1 was overexpressed related with proteinuria and in some cases was associated with 708
the expression of its receptors. In this sense, a previous study reported increased expres- 709
sion of GRP78 in podocytes after PAN treatment [10]. The fact that this expression is spe- 710
cifically localized in podocytes suggests that the interaction of ISM1 with its receptor 711
could promote cell damage. 712
Alteration of the cytoskeleton is a common feature of podocytopathies and consti- 713
tutes a classic readout for their study [29][36]. Our experiments with cultured podocytes 714
revealed that weak ISM1 stimuli did not apparently affect the structure of the cytoskele- 715
ton. On the contrary, at high doses and prolonged incubations, ISM1 promoted the pro- 716
gressive loss of the fibrillar pattern of F-actin which ultimately was accumulated at the 717
subcortical level (Figure 5A). Similarly, synaptopodine (SYNPO), an actin-associated pro- 718
tein, significantly lost its fibrillar distribution. This effect may be due to the dephosphor- 719
ylation of SYNPO by a calcineurin-dependent mechanism and its subsequent degradation 720
[29]. However, in our model, the inhibition of calcineurin by FK506 did not prevent this 721
effect (not shown), so this mechanism seems to be ruled out. Alternatively, it is possible 722
that the cytoskeleton collapses in some other way, as occurs, for example, during the exe- 723
cution of apoptosis [37]. 724
Indeed, we found that ISM1 significantly decreased podocyte viability in a dose-time 725
dependent manner (Figure 5B). At low doses, ISM1 caused impairment of podocyte via- 726
bility that was abrogated by Q-VD-oph, a pan caspase inhibitor, suggesting the involve- 727
ment of a caspase-dependent apoptosis process. However, upon strong ISM1 stimulation, 728
Q-VD-oph only partially inhibited lethality in podocytes, suggesting the additional pres- 729
ence of a caspase-independent form of death (Figure 5C). These results were supported 730
by the TUNEL assays, as well as by observing the disappearance of the zymogen forms of 731
procaspase-8 and procaspase-3, suggesting their activation by cleavage. Other studies re- 732
ported a proapoptotic effect of ISM1 but on endothelial cells (EC) [7][35]. In these studies, 733
it was shown that ISM1 induced EC apoptosis through a caspase-dependent pathway, 734
since selective inhibitors of caspase-3 and -8 prevented this effect. They also showed that 735
activation of caspase-3 and -8 by ISM1 was dependent on its binding to αvβ5 integrin. In 736
contrast to our study, no involvement of caspase-independent activity was shown. 737
Classically, two main apoptotic pathways have been described: i) the extrinsic or 738
death receptor pathway and ii) the intrinsic or mitochondrial pathway. The mitochondrial 739
membrane potential (ΔΨm) is a key indicator of mitochondrial activity, as it reflects the 740
process of electron transport and oxidative phosphorylation, essential for ATP production 741
[38]. The drop of ΔΨm is considered an early event in the apoptotic cascade, as it occurs 742
before nuclear degradation. In our study we show that ISM1 triggers the drop of ΔΨm, 743
suggesting mitochondrial apoptosis. This event is consistent with previous studies per- 744
formed in endothelial cells where the cell death was mediated via mitochondrial through 745
GRP78. The authors described two different pathways: i) activation of p53, synthesis of 746
proapoptotic proteins (e.g. BH3-only proteins) and subsequent mitochondrial damage or, 747
ii) through an endosomal pathway in which ISM1, transported by SNAP25, could directly 748
destabilize mitochondria [30]. In our experimental setting we did not observe activation 749
of p53 (not shown), so we are tempted to privilege a direct interaction of ISM1 with mito- 750
chondria via endosomes. 751
The outer mitochondrial membrane permeabilization is frequently associated with a 752
drop of ΔΨm. Depending on the system, this phenomenon causes the release of proapop- 753
totic proteins from the intermembrane space of the mitochondria. The two main proapop- 754
totic components of this compartment are the cytochrome c and AIF (Apoptosis Inducing 755
Factor) [39]. In our study we have shown that ISM1 induced the translocation of AIF from 756
the mitochondria to the cytosol and the nucleus, contrary to cytochrome c whose distri- 757
bution remained apparently unchanged. AIF is a mitochondrial oxidoreductase with a 758
dual role in cell life/death [40]. Under normal conditions, AIF exhibited a punctate distri- 759
bution in control podocytes, typically associated with its physiological confinement in mi- 760
tochondria After ISM1 treatment, AIF adopted a more diffuse cytosolic profile, as well as 761
a nuclear translocation (Figure 6F, arrows) suggesting a caspase)independent apoptosis 762

[40]. In contrast, in the same conditions, we did not observe a similar pattern for the cyto- 763
chrome c. Like AIF, cytochrome c acts as an apoptotic protein, but in a caspase-dependent 764
manner through activation of caspase-9 [41]. This different release of mitochondrial 765
proapoptotic proteins has already been reported [42], but not very well explained. In con- 766
clusion, ISM1 acts on podocytes by activating two different death pathways (caspase-de- 767
pendent and caspase-independent) that seems to act in concert. 768
In a more speculative basis, we analyzed in silico the role of ISM1 in our L-NAME 769
model of glomerulopathy progression-recovery. Interestingly, we observed a relevant role 770
of the mitochondrial and peroxisomal compartments during the period of post-injury re- 771
nal regeneration (Figure 7). This would be consistent with the mitochondrial destabiliza- 772
tion described above. The so-called "peroxisome-mitochondria connection" involves close 773
cooperation in the cellular metabolism of lipids and reactive oxygen species [43]. This 774
functional interrelationship has obvious implications for disease development. Although 775
this hypothesis is highly speculative, it provides interesting information on the involve- 776
ment of these organelles in the development of chronic kidney disease, a hypothesis that 777
remains to be confirmed and explored. 778
In the light of our findings, we have to underline that, at the glomerular level, ISM1 779
would be produced and secreted by podocytes. Subsequently, it would exert a proapop- 780
totic effect on the same podocytes but also on endothelial cells, according to other previ- 781
ous work. This suggests that ISM1 exerts a pro-apoptotic role through autocrine (podo- 782
cytes) or paracrine (endothelial cells) regulation. As a final result, the destruction of the 783
structures that integrate the glomerular filtration barrier is ensured, proteinuria appears 784
and the disease occurs. 785
786
Figure 8: Proposed mechanism of the deleterious action of ISM1 on podocytes. (Left) Caspase- 787
dependent pathway. At low concentration, extracellular ISM1might bind avb5 and/or GRP78 at the 788
cell surface, triggering activation of procaspase-8 and the downstream effector caspase-3, which 789
would activate a caspase-dependent DNAse (CAD) to produce DNA fragmentation (left). At higher 790
concentrations, ISM1 is internalized by endosomal trafficking and would induce the activation of a 791
caspase-independent proapoptotic pathway. Thus, ISM1 potentially interacts with GRP78 on the 792
cell surface and would trigger its endocytic transport into the mitochondria. Subsequently, ISM1 793
triggers the permeabilization of the outer mitochondrial membrane and the specific release of AIF 794
into the cytosol. Ultimately, AIF translocates to the nucleus where it contributes to DNA fragmen- 795
tation and chromatin condensation. 796
797
5. Conclusions 798
In the present work, we describe for the first time to our knowledge a pathophysio- 799
logical role of ISM1 in the kidney. This interest in ISM1 emerged because of the variability 800
of its expression according to disease progression. This provided a first indication that 801
ISM1 was associated with renal dysfunction. Furthermore, our results showed that ISM1 802
was overexpressed in five different models of glomerular disease. This abnormal expres- 803
sion occurs specifically in podocytes where it produces apoptosis and glomerular damage. 804
805
Supplementary Materials: The following supporting information can be downloaded at: 806
www.mdpi.com/xxx/s1, Figure S1: Immunohistochemical analysis of co-localization of ISM1; Table 807
SI: Gene-specific primer sequences used for RT-qPCR, Table SII: Top 25 up- and down-regulated 808
genes in L-NAME model between W0 and CTL groups, Table SIII: Biochemical parameters in ani- 809
mal models of glomerulopathy, Table SIV: Genes related to ISM1 regulation during renal regenera- 810
tion 811
Author Contributions: JC has performed the transcriptomic study and developed the hypertensive 812
model of nephropathy. KG, VS, IM, JS developed the different models of glomerulopathy. SF per- 813
formed the biological measurements. SP, SA, CC provided scientific input. VS completed the entire 814
in vitro study. HKL and JJB designed the study and wrote the manuscript. All authors approved the 815
final version of the manuscript. 816
Funding: Sahiri V. had a PhD grant from the Ministry of Education and Scientific Research of Ivory 817
Coast. The recurring annual funding of Inserm, Sorbonne Université and the AP-HP grant 818
(AORC17) supported this work. The in vitro research was funded by “Fondation du Rein”, France, 819
grant number RAD16099LLA (HKL). 820
Acknowledgments: The authors thank the animal facility of INSERM UMRS 1155 for their valuable 821
help with animal breeding, but also Pr G. Touchart and N. Quellard for the immuno-electronic mi- 822
croscopy analysis. 823
Conflicts of Interest: The authors declare no conflict of interest. 824
825
References 826
827
1. Rosenberg, A.Z.; Kopp, J.B. Focal Segmental Glomerulosclerosis. Clin. J. Am. Soc. Nephrol. CJASN 2017, 12, 502–517, 828
doi:10.2215/CJN.05960616. 829
2. Costantino, V.V.; Gil Lorenzo, A.F.; Bocanegra, V.; Vallés, P.G. Molecular Mechanisms of Hypertensive Nephropathy: 830
Renoprotective Effect of Losartan through Hsp70. Cells 2021, 10, 3146, doi:10.3390/cells10113146. 831
3. Cai, A.; Chatziantoniou, C.; Calmont, A. Vascular Permeability: Regulation Pathways and Role in Kidney Diseases. Nephron 832
2021, 145, 297–310, doi:10.1159/000514314. 833
4. Pera, E.M.; Kim, J.I.; Martinez, S.L.; Brechner, M.; Li, S.Y.; Wessely, O.; De Robertis, E.M. Isthmin Is a Novel Secreted 834
Protein Expressed as Part of the Fgf-8 Synexpression Group in the Xenopus Midbrain-Hindbrain Organizer. Mech. Dev. 2002, 835
116, 169–172, doi:10.1016/s0925-4773(02)00123-5. 836
5. Valle-Rios, R.; Maravillas-Montero, J.L.; Burkhardt, A.M.; Martinez, C.; Buhren, B.A.; Homey, B.; Gerber, P.A.; Robinson, 837
O.; Hevezi, P.; Zlotnik, A. Isthmin 1 Is a Secreted Protein Expressed in Skin, Mucosal Tissues, and NK, NKT, and Th17 Cells. 838
J. Interferon Cytokine Res. Off. J. Int. Soc. Interferon Cytokine Res. 2014, 34, 795–801, doi:10.1089/jir.2013.0137. 839
6. Jiang, Z.; Zhao, M.; Voilquin, L.; Jung, Y.; Aikio, M.A.; Sahai, T.; Dou, F.Y.; Roche, A.M.; Carcamo-Orive, I.; Knowles, 840
J.W.; et al. Isthmin-1 Is an Adipokine That Promotes Glucose Uptake and Improves Glucose Tolerance and Hepatic Steatosis. 841
Cell Metab. 2021, 33, 1836-1852.e11, doi:10.1016/j.cmet.2021.07.010. 842
7. Xiang, W.; Ke, Z.; Zhang, Y.; Cheng, G.H.-Y.; Irwan, I.D.; Sulochana, K.N.; Potturi, P.; Wang, Z.; Yang, H.; Wang, J.; et al. 843
Isthmin Is a Novel Secreted Angiogenesis Inhibitor That Inhibits Tumour Growth in Mice. J. Cell. Mol. Med. 2011, 15, 359– 844
374, doi:10.1111/j.1582-4934.2009.00961.x. 845
8. Venugopal, S.; Chen, M.; Liao, W.; Er, S.Y.; Wong, W.-S.F.; Ge, R. Isthmin Is a Novel Vascular Permeability Inducer That 846
Functions through Cell-Surface GRP78-Mediated Src Activation. Cardiovasc. Res. 2015, 107, 131–142, 847
doi:10.1093/cvr/cvv142. 848
9. Li, J.; Xia, Y.; Huang, Z.; Zhao, Y.; Xiong, R.; Li, X.; Huang, Q.; Shan, F. Novel HIF-1-Target Gene Isthmin1 Contributes to 849
Hypoxia-Induced Hyperpermeability of Pulmonary Microvascular Endothelial Cells Monolayers. Am. J. Physiol. Cell Physiol. 850
2021, 321, C671–C680, doi:10.1152/ajpcell.00124.2021. 851
10. Min, S.-Y.; Ha, D.-S.; Ha, T.-S. Puromycin Aminonucleoside Triggers Apoptosis in Podocytes by Inducing Endoplasmic 852
Reticulum Stress. Kidney Res. Clin. Pract. 2018, 37, 210–221, doi:10.23876/j.krcp.2018.37.3.210. 853
11. Inagi, R.; Kumagai, T.; Nishi, H.; Kawakami, T.; Miyata, T.; Fujita, T.; Nangaku, M. Preconditioning with Endoplasmic 854
Reticulum Stress Ameliorates Mesangioproliferative Glomerulonephritis. J. Am. Soc. Nephrol. JASN 2008, 19, 915–922, 855
doi:10.1681/ASN.2007070745. 856
12. Van Krieken, R.; Mehta, N.; Wang, T.; Zheng, M.; Li, R.; Gao, B.; Ayaub, E.; Ask, K.; Paton, J.C.; Paton, A.W.; et al. Cell 857
Surface Expression of 78-KDa Glucose-Regulated Protein (GRP78) Mediates Diabetic Nephropathy. J. Biol. Chem. 2019, 294, 858
7755–7768, doi:10.1074/jbc.RA118.006939. 859
13. Henderson, N.C.; Arnold, T.D.; Katamura, Y.; Giacomini, M.M.; Rodriguez, J.D.; McCarty, J.H.; Pellicoro, A.; Raschperger, 860
E.; Betsholtz, C.; Ruminski, P.G.; et al. Targeting of Αv Integrin Identifies a Core Molecular Pathway That Regulates Fibrosis 861
in Several Organs. Nat. Med. 2013, 19, 1617–1624, doi:10.1038/nm.3282. 862
14. McCurley, A.; Alimperti, S.; Campos-Bilderback, S.B.; Sandoval, R.M.; Calvino, J.E.; Reynolds, T.L.; Quigley, C.; Mugford, 863
J.W.; Polacheck, W.J.; Gomez, I.G.; et al. Inhibition of Αvβ5 Integrin Attenuates Vascular Permeability and Protects against 864
Renal Ischemia-Reperfusion Injury. J. Am. Soc. Nephrol. JASN 2017, 28, 1741–1752, doi:10.1681/ASN.2016020200. 865
15. Su, G.; Atakilit, A.; Li, J.T.; Wu, N.; Luong, J.; Chen, R.; Bhattacharya, M.; Sheppard, D. Effective Treatment of Mouse Sepsis 866
with an Inhibitory Antibody Targeting Integrin Αvβ5. Crit. Care Med. 2013, 41, 546–553, 867
doi:10.1097/CCM.0b013e3182711b1e. 868
16. Caron, J.; Michel, P.-A.; Dussaule, J.-C.; Chatziantoniou, C.; Ronco, P.; Boffa, J.-J. Extracorporeal Shock Wave Therapy Does 869
Not Improve Hypertensive Nephropathy. Physiol. Rep. 2016, 4, e12699, doi:10.14814/phy2.12699. 870
17. Boffa, J.-J.; Lu, Y.; Placier, S.; Stefanski, A.; Dussaule, J.-C.; Chatziantoniou, C. Regression of Renal Vascular and Glomerular 871
Fibrosis: Role of Angiotensin II Receptor Antagonism and Matrix Metalloproteinases. J. Am. Soc. Nephrol. JASN 2003, 14, 872
1132–1144, doi:10.1097/01.asn.0000060574.38107.3b. 873
18. Fujihara, C.K.; Michellazzo, S.M.; de Nucci, G.; Zatz, R. Sodium Excess Aggravates Hypertension and Renal Parenchymal 874
Injury in Rats with Chronic NO Inhibition. Am. J. Physiol. 1994, 266, F697-705, doi:10.1152/ajprenal.1994.266.5.F697. 875
19. Guerrot, D.; Dussaule, J.-C.; Mael-Ainin, M.; Xu-Dubois, Y.-C.; Rondeau, E.; Chatziantoniou, C.; Placier, S. Identification of 876
Periostin as a Critical Marker of Progression/Reversal of Hypertensive Nephropathy. PloS One 2012, 7, e31974, 877
doi:10.1371/journal.pone.0031974. 878
20. Placier, S.; Boffa, J.-J.; Dussaule, J.-C.; Chatziantoniou, C. Reversal of Renal Lesions Following Interruption of Nitric Oxide 879
Synthesis Inhibition in Transgenic Mice. Nephrol. Dial. Transplant. Off. Publ. Eur. Dial. Transpl. Assoc. - Eur. Ren. Assoc. 880
2006, 21, 881–888, doi:10.1093/ndt/gfk004. 881
21. Bolstad, B.M.; Irizarry, R.A.; Astrand, M.; Speed, T.P. A Comparison of Normalization Methods for High Density 882
Oligonucleotide Array Data Based on Variance and Bias. Bioinforma. Oxf. Engl. 2003, 19, 185–193, 883
doi:10.1093/bioinformatics/19.2.185. 884
22. Regele, H.M.; Fillipovic, E.; Langer, B.; Poczewki, H.; Kraxberger, I.; Bittner, R.E.; Kerjaschki, D. Glomerular Expression of 885
Dystroglycans Is Reduced in Minimal Change Nephrosis but Not in Focal Segmental Glomerulosclerosis. J. Am. Soc. Nephrol. 886
JASN 2000, 11, 403–412, doi:10.1681/ASN.V113403. 887
23. Chatziantoniou, C.; Boffa, J.J.; Ardaillou, R.; Dussaule, J.C. Nitric Oxide Inhibition Induces Early Activation of Type I 888
Collagen Gene in Renal Resistance Vessels and Glomeruli in Transgenic Mice. Role of Endothelin. J. Clin. Invest. 1998, 101, 889
2780–2789, doi:10.1172/JCI2132. 890
24. Saleem, M.A.; O’Hare, M.J.; Reiser, J.; Coward, R.J.; Inward, C.D.; Farren, T.; Xing, C.Y.; Ni, L.; Mathieson, P.W.; Mundel, 891
P. A Conditionally Immortalized Human Podocyte Cell Line Demonstrating Nephrin and Podocin Expression. J. Am. Soc. 892
Nephrol. JASN 2002, 13, 630–638, doi:10.1681/ASN.V133630. 893
25. Pavlidis, P.; Noble, W.S. Analysis of Strain and Regional Variation in Gene Expression in Mouse Brain. Genome Biol. 2001, 894
2, RESEARCH0042, doi:10.1186/gb-2001-2-10-research0042. 895
26. Chen, J.; Bardes, E.E.; Aronow, B.J.; Jegga, A.G. ToppGene Suite for Gene List Enrichment Analysis and Candidate Gene 896
Prioritization. Nucleic Acids Res. 2009, 37, W305-311, doi:10.1093/nar/gkp427. 897
27. Cline, M.S.; Smoot, M.; Cerami, E.; Kuchinsky, A.; Landys, N.; Workman, C.; Christmas, R.; Avila-Campilo, I.; Creech, M.; 898
Gross, B.; et al. Integration of Biological Networks and Gene Expression Data Using Cytoscape. Nat. Protoc. 2007, 2, 2366– 899
2382, doi:10.1038/nprot.2007.324. 900
28. Brancotte, B.; Biton, A.; Bernard-Pierrot, I.; Radvanyi, F.; Reyal, F.; Cohen-Boulakia, S. Gene List Significance At-a-Glance 901
with GeneValorization. Bioinforma. Oxf. Engl. 2011, 27, 1187–1189, doi:10.1093/bioinformatics/btr073. 902
29. Faul, C.; Donnelly, M.; Merscher-Gomez, S.; Chang, Y.H.; Franz, S.; Delfgaauw, J.; Chang, J.-M.; Choi, H.Y.; Campbell, 903
K.N.; Kim, K.; et al. The Actin Cytoskeleton of Kidney Podocytes Is a Direct Target of the Antiproteinuric Effect of 904
Cyclosporine A. Nat. Med. 2008, 14, 931–938, doi:10.1038/nm.1857. 905
30. Chen, M.; Qiu, T.; Wu, J.; Yang, Y.; Wright, G.D.; Wu, M.; Ge, R. Extracellular Anti-Angiogenic Proteins Augment an 906
Endosomal Protein Trafficking Pathway to Reach Mitochondria and Execute Apoptosis in HUVECs. Cell Death Differ. 2018, 907
25, 1905–1920, doi:10.1038/s41418-018-0092-9. 908
31. Wang, C.; Youle, R.J. The Role of Mitochondria in Apoptosis. Annu. Rev. Genet. 2009, 43, 95–118, doi:10.1146/annurev- 909
genet-102108-134850. 910
32. Osório, L.; Wu, X.; Wang, L.; Jiang, Z.; Neideck, C.; Sheng, G.; Zhou, Z. ISM1 Regulates NODAL Signaling and Asymmetric 911
Organ Morphogenesis during Development. J. Cell Biol. 2019, 218, 2388–2402, doi:10.1083/jcb.201801081. 912
33. Berrun, A.; Harris, E.; Stachura, D.L. Isthmin 1 (Ism1) Is Required for Normal Hematopoiesis in Developing Zebrafish. PloS 913
One 2018, 13, e0196872, doi:10.1371/journal.pone.0196872. 914
34. Hosoe-Nagai, Y.; Hidaka, T.; Sonoda, A.; Sasaki, Y.; Yamamoto-Nonaka, K.; Seki, T.; Asao, R.; Tanaka, E.; Trejo, J.A.O.; 915
Kodama, F.; et al. Re-Expression of Sall1 in Podocytes Protects against Adriamycin-Induced Nephrosis. Lab. Investig. J. Tech. 916
Methods Pathol. 2017, 97, 1306–1320, doi:10.1038/labinvest.2017.69. 917
35. Zhang, Y.; Chen, M.; Venugopal, S.; Zhou, Y.; Xiang, W.; Li, Y.-H.; Lin, Q.; Kini, R.M.; Chong, Y.-S.; Ge, R. Isthmin Exerts 918
Pro-Survival and Death-Promoting Effect on Endothelial Cells through Alphavbeta5 Integrin Depending on Its Physical State. 919
Cell Death Dis. 2011, 2, e153, doi:10.1038/cddis.2011.37. 920
36. Kachurina, N.; Chung, C.-F.; Benderoff, E.; Babayeva, S.; Bitzan, M.; Goodyer, P.; Kitzler, T.; Matar, D.; Cybulsky, A.V.; 921
Alachkar, N.; et al. Novel Unbiased Assay for Circulating Podocyte-Toxic Factors Associated with Recurrent Focal Segmental 922
Glomerulosclerosis. Am. J. Physiol. Renal Physiol. 2016, 310, F1148-1156, doi:10.1152/ajprenal.00349.2015. 923
37. Povea-Cabello, S.; Oropesa-Ávila, M.; de la Cruz-Ojeda, P.; Villanueva-Paz, M.; de la Mata, M.; Suárez-Rivero, J.M.; 924
Álvarez-Córdoba, M.; Villalón-García, I.; Cotán, D.; Ybot-González, P.; et al. Dynamic Reorganization of the Cytoskeleton 925
during Apoptosis: The Two Coffins Hypothesis. Int. J. Mol. Sci. 2017, 18, E2393, doi:10.3390/ijms18112393. 926
38. Zorova, L.D.; Popkov, V.A.; Plotnikov, E.Y.; Silachev, D.N.; Pevzner, I.B.; Jankauskas, S.S.; Babenko, V.A.; Zorov, S.D.; 927
Balakireva, A.V.; Juhaszova, M.; et al. Mitochondrial Membrane Potential. Anal. Biochem. 2018, 552, 50–59, 928
doi:10.1016/j.ab.2017.07.009. 929
39. Lorenzo, H.K.; Susin, S.A. Mitochondrial Effectors in Caspase-Independent Cell Death. FEBS Lett. 2004, 557, 14–20, 930
doi:10.1016/s0014-5793(03)01464-9. 931
40. Susin, S.A.; Lorenzo, H.K.; Zamzami, N.; Marzo, I.; Snow, B.E.; Brothers, G.M.; Mangion, J.; Jacotot, E.; Costantini, P.; 932
Loeffler, M.; et al. Molecular Characterization of Mitochondrial Apoptosis-Inducing Factor. Nature 1999, 397, 441–446, 933
doi:10.1038/17135. 934
41. Ow, Y.-L.P.; Green, D.R.; Hao, Z.; Mak, T.W. Cytochrome c: Functions beyond Respiration. Nat. Rev. Mol. Cell Biol. 2008, 935
9, 532–542, doi:10.1038/nrm2434. 936
42. Muñoz-Pinedo, C.; Guío-Carrión, A.; Goldstein, J.C.; Fitzgerald, P.; Newmeyer, D.D.; Green, D.R. Different Mitochondrial 937
Intermembrane Space Proteins Are Released during Apoptosis in a Manner That Is Coordinately Initiated but Can Vary in 938
Duration. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 11573–11578, doi:10.1073/pnas.0603007103. 939
43. Fransen, M.; Lismont, C.; Walton, P. The Peroxisome-Mitochondria Connection: How and Why? Int. J. Mol. Sci. 2017, 18, 940
E1126, doi:10.3390/ijms18061126. 941
942
943
944
FIGURE 1 945
946
947
948
a) 949
950
b) 951
952
953
FIGURE 2 954
a) 955
956
957
b) 958
c)
959
FIGURE 3 960
a) 961
ISM1 (rat) +blocking peptide (rat) ISM1 (human)
962
b) 963
ISM1 NEPH1 Merge TEM (anti-ISM1)
964
c) 965
ISM1 CD31 Merge
966
967
FIGURE 4 968
a) 969
4,5 ***
ISM1 expression (ratio to CTL)
3,5
**
**
3
2,5
* *
2
1,5
0,5
0
CTL L-NAME PAN DOX ob/ob LPS 970
b) 971
972
10 6
9 *** ***
!5 integrin (ratio to CTL)
5
GRP78 (ratio to CTL)
8
7
4
6 ***
5 3 *
4
2
3
2
1
1
0 0
CTL PAN DOXO ob/ob CTL PAN DOXO ob/ob
c) 973
974
975
FIGURE 5 976
a) 977
978
b)
24h
48h
100 72h
Cell viability (%)
***
75
*** ***
***
50 ***
25 ***
0
c) 24h 72h
ISM1
ISM1
ISM1 + QVD.oph
ISM1 + QVD.oph
# 100
100 ##
##
Cell viability (%)
** **
Cell viability (%)
75 ***
75
***
***
50 ##
50
***
25
25 ***
0 0
CTL 5 µg/ml 35 µg/ml CTL 5 µg/ml 35 µg/ml
FIGURE 6 979
980
981
982
983
984
985
e) 986
987
f) 988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
FIGURE 7 1004
1005
1006
1007
1008
1009
FIGURE 8 1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
SUPPLEMENTARY DATA 1022
1023
1024
1025
1026
TABLE SI 1027
1028
1029
1030
1031
Gene name Direction Primer sequence (5′-3′)

β5 (rat and mouse) fwd: ACCTGCCAAGATGGCATATC
rev: CACGGACACTTCAAAGGATG
GRP78 (rat) fwd: CCGTAACAATCAAGGTCTACGA
rev: AAGGTGACTTCAATCTGGGGTA
GRP78 (mouse) fwd: CTGAGGCGTATTTGGGAAAG
rev: TCATGACATTCAGTCCAGCAA
ISM1 (rat and mouse) fwd: TCCAGATCTTTCCAAAGCTGAT
rev: GCCATCAACCACCTCTATGG
IL6 (mouse) fwd: CCAGGTAGCTATGGTACTCCAGAA
rev: TGCCTTCATTTATCCCTTGAA
HPRT (mouse) fwd: GGAGCGGTAGCACCTCCT
rev: CTGGTTCATCATCGCTAATCAC
HPRT (rat) fwd: GACCGGTTCTGTCATGTCG
rev: ACCTGGTTCATCATCACTAATCAC
Gusb (Mouse) fwd: CTCTGGTGGCCTTACCTGAT
rev: CAGTTGTTGTCACCTTCACCTC
1032
1033
1034
1035
TABLE SII 1036
1037
1038
1039
Up-regulated genes Down-regulated genes

Symbol Fold Change p-value Symbol Fold Change p-value
HAVCR1 36,082 7.42E-08 Slc7a12 -9,996 1.35E-03
SERPINA3 23,150 1.28E-03 RGD1563294 -6,425 3.16E-05
FGB 18,605 1.92E-07 Akr1c12 -5,578 1.76E-03
GPNMB 8,954 7.32E-06 RGN -5,501 7.61E-04
HMOX1 8,562 2.21E-03 Slco1a1 -5,164 1.06E-02
IGFBP1 7,673 1.06E-05 CACNG5 -5,028 5.68E-03
ALOX15 7,133 1.35E-03 HNMT -5,010 5.81E-04
PDK4 6,936 1.05E-06 SLC7A13 -4,980 1.50E-03
ADAMTS1 6,929 1.53E-06 PRIMA1 -4,795 4.32E-03
GPX2 6,715 1.43E-03 HRG -4,737 1.98E-03
ANGPTL4 6,471 4.85E-07 Olfr136 -4,648 1.83E-02
S100A8 6,372 2.49E-02 GC -4,339 6.88E-03
SPP1 6,313 1.48E-06 Cml1 -4,216 7.29E-04
HMGCS2 6,023 3.00E-03 Ces1e -4,207 5.15E-04
SERPINE1 5,975 6.62E-06 Slco1a6 -3,863 5.63E-03
FGA 5,805 4.59E-05 Olr1408 -3,808 1.29E-02
TIMP1 5,453 3.26E-06 Kap -3,754 3.25E-02
CLU 5,208 7.98E-06 CYP2C9 -3,724 2.40E-02
SLC34A2 5,155 1.61E-05 GUCY1B2 -3,631 7.38E-05
FGG 5,067 1.06E-05 CHTF18 -3,572 4.34E-03
LAMC2 4,874 4.02E-05 SPATA22 -3,511 1.25E-04
FAM129A 4,755 1.03E-08 SLCO4C1 -3,498 2.77E-04
CDKN1A 4,742 6.29E-05 AFM -3,472 1.59E-03
RASD1 4,539 3.07E-05 MLC1 -3,426 3.18E-03
BTG2 4,533 2.45E-03 SLC22A25 -3,407 1.28E-03 1040
1041
1042
1043
1044
1045
TABLE SIII 1046
1047
1048
1049
1050
1051
1052
UAER UPER Albuminemia

CTL model p CTL model p CTL model p
PAN 0.00±0.000 2.42±0.15 <0.001 0.06±0.004 12.71±0.75 <0.001 14.9±0.65 4.38±0.35 <0.001
DOXO 0.00±0.000 3.02±0.12 <0.001 0.15±0.01 9.18±0.51 <0.001 16.47±0.57 4.25±0,16 <0.001
ob/ob 0.002±0.001 0,056±0.008 <0.001
LPS 0.04±0.01 0.27±0.04 <0.001
1053
1054
1055
1056
1057
TABLE SIV 1058
1059
1060
1061
1062
(DOWNLOAD Excel file) 1063
1064
1065
1066
1067
1068
Figure
Figure
Figure20: 20:
20:(En
(En
(Enhaut
haut
hautàààgauche)
gauche)
gauche)Immunofluorescence
Immunofluorescence
Immunofluorescencemontrant montrant
montrantune une
unecolocalisation
colocalisation
colocalisationentre
entre l’ISMetet
entrel’ISM
l’ISM etlala
la
néphrine.
néphrine.
néphrine. (En
(En
(En haut
haut
haut
Biomedicines 2022, 10, x FOR PEER REVIEWàààdroite):
droite):Expression
droite): Expression
Expression podocytaire
podocytaire
podocytaire dede
del’ISM
l’ISM
l’ISM en
en
enmicroscopie
microscopie
microscopie électronique
électronique
électronique par
par
par 37 of 41
«««immunogolod
immunogolod
immunogolod»:»: »:L’ISM
L’ISM
L’ISMestest
estexprimée
exprimée
expriméeparpar
parlele
lepodocyte.
podocyte.
podocyte.Elle
Elle
Elleest
est
estretrouvée
retrouvée
retrouvéedans
dans
danslele
lecytoplasme
cytoplasme
cytoplasmeetet
etégalement
également
égalementdans dans
dans
les
les
lespédicelles.
pédicelles.
pédicelles.(En(En
(Enbas):
bas):
bas):Immunofluorescence
Immunofluorescence
Immunofluorescencemontrant
montrant
montrantaucune
aucune
aucunecolocalisation
colocalisation
colocalisationentre
entre
entrel’ISM
l’ISM
l’ISMetet
etCD31.
CD31.
CD31.
FIGURE S1 1069
Par
Par
Parailleurs
ailleurs
ailleursl’expression
l’expression
l’expressionde
de
del’ISM
l’ISM
l’ISMn’a
n’a
n’apas
pas
pasété
été
étéretrouvée
retrouvée
retrouvéedans
dans
dansles
les
lescompartiments
compartiments
compartimentstubulaires
tubulaires
tubulaires 1070
1071
(figure
(figure
(figure21).
21).
21).
A) 1072
ISM1 RECA-1 Merge
1073
Figure
Figure 21:Immunofluorescence
Figure21:
21: Immunofluorescence
Immunofluorescencemontrant
montrant
montrantaucune
aucune
aucunecolocalisation
colocalisation
colocalisationentre
entre
entrel’ISM
l’ISM
l’ISMetet
etRECA
RECA
RECAdans
dans
danslele
leglomérule.
glomérule.
glomérule.
1074
B) 1075
1076
Les
Les
Lesrécepteurs
récepteurs
récepteursde
de
del’ISM
l’ISM
l’ISMdécrits
décrits
décritsdans
dans
danslala
lalittérature
littérature
littératuresont
sont
sontl’intégrine
l’intégrine
l’intégrineαvβ5
αvβ5
αvβ5etet
etlele
leGRP78.
GRP78.
GRP78.Une
Une
Une
interaction
interaction
interactionentre
entre
entrel’ISM
l’ISM
l’ISMetet
etses
ses
sesdeux
deux
deuxrécepteurs
récepteurs
récepteursaaaété
été
étédémontrée
démontrée
démontréedans
dans
danslele poumon.117
lepoumon.
poumon. 117
117
Cependant
Cependant
Cependant
au
au
auniveau
niveau
niveaurénal,
rénal,
rénal,nous
nous
nousne
ne
nedisposons
disposons
disposonsactuellement
actuellement
actuellementd’aucune
d’aucune
d’aucunedonnée
donnée
donnéedans
dans
danslala
lalittérature.
littérature.
littérature. Nous
Nous
Nous
54
54
54
1077
1078
1079
1080
1081
1082
LEGENDS OF TABLES AND FIGURES 1083
1084
Figure 1: L-NAME impaired the renal function in rats and induced renal injuries. (a) Blood pressure 1085
and renal function in L-NAME rat model. Parameters studied included UPER, uremia and creatinemia in CTL, 1086
W0, W1 and W5 groups. At the peak of the disease (W0), the rats developed an elevated UPER accompanied 1087
by high blood pressure and high levels of urea and serum creatinine. After stopping L-NAME treatment (W1 1088
and W5), renal function recovered to normal (*** p <0.001 and ** p <0.01). (b) Histological evaluation of 1089
renal tissues. Representative examples of renal samples stained with Masson’s trichrome stain in Control 1090
(CTL) and L-NAME-treated rats. 1091
Figure 2: ISM1 gene expression in L-NAME model. (a) Experimental protocol of L-NAME model. When 1092
urinary protein excretion ratio was over 1g/mmol of creatinine, after 6 to 24 weeks, NG-nitro-L-arginine 1093
methyl ester (L-NAME) was removed. (b) Left panel. Segregation of transcriptomic data of each animal by 1094
principal component analysis: Each animal is represented by a circle in the corresponding group (CTL in black, 1095
W0 in red, W1 in blue and W5 in green). Right panel. Hierarchical grouping of transcriptomic data from each 1096
animal: Each line corresponds to a gene and each column to an animal. The red color indicates an 1097
overexpression and the green color an under expression. This representation clearly shows a variation of gene 1098
expression from one group to another. (c) ISM1 is overexpressed in glomeruli of the L-NAME rat model. 1099
mRNA expression of ISM1 from kidney of CTL rats, L-NAME treated rats (W0), 1 week L-NAME removal 1100
rats (W1) and 5 weeks L-NAME removal rats (W5). mRNA of ISM1 is significantly overexpressed in the 1101
kidney of W0 and W1 L-NAME rats (* p<0.05 vs control group ). 1102
1103
Figure 3: Physiological expression of ISM1 in rodents and human kidneys. (a) Glomerular expression of 1104
ISM1 in kidney from rats (a, b) and human (c). ISM1 is expressed exclusively in podocytes. (b) ISM1 1105
expression is localized in podocytes. Immunofluorescence showing colocalization between ISM1 and nephrin. 1106
Immunogold electron microscopy: ISM1 is localized in the cytoplasm (circled in red) and in the foot process 1107
of podocytes (red arrow). (c) ISM1 colocalized poorly with the endothelial cell marker CD31. 1108
Figure 4. Expression of ISM1 and its receptors in different models of glomerulopathies. Estimation by 1109
RT-qPCR (n=6): (a) ISM1 expression in different models of proteinuria (L-NAME (W0), doxorubicin, puro- 1110
mycin, obesity and LPS. Values normalized to CTL; (b) Expression de GRP78 (left) and integrin- 5 (right) 1111
in doxorubicin, puromycin and obesity models. *, p<0,05 vs CTL **, p<0,01 vs CTL ***, p<0,001 vs CTL 1112
(c) Interaction between ISM1 and its GRP78 and integrin αvβ5 receptors in cultured podocytes. Co-Immuno- 1113
precipitation (IP) of GRP78 and integrin αvβ5 using podocyte cells and then, western blot of ISM1. The ISM1 1114
interacts with its receptors in podocytes, in basal conditions, (n = 2). 1115
1116
Figure 5. Recombinant ISM1 decreased podocyte viability. (a) Staining with phalloidin-FITC showed that 1117
the F-actin cytoskeleton was disorganized after strong stimuli with ISM1 in podocytes. Synaptopodin 1118
(SYNPO) lost its co-localization with actin upon addition of high concentrations of ISM1 to podocytes. (b) 1119
Podocytes were treated with recombinant ISM1 (5 and 35 µg/ml) for 24, 48 and 72 hours. ISM1 decreased 1120
significantly the podocyte viability. ***, p<0,001 vs CTL. (c) Podocytes were treated with ISM1 (5 and 35 1121
µg/ml) for 24 and 72h in the presence or absence of QVD-oph, a pan-caspase inhibitor. QVD-oph prevented 1122
ISM1-induced decrease in viability under mild conditions (5 µg/ml, 24h) (C left).. At higher doses or longer 1123
incubations of ISM1, QVD-oph only partially inhibited podocyte mortality (C right). **, p<0,005 vs CTL ***, 1124
p<0,001 vs CTL; ##, p<0,01 treated vs untreated QVD, #, p<0,01 treated vs untreated QVD. 1125
1126
Figure 6: ISM1 induces apoptosis in podocytes. (A) Estimation of ISM1-induced apoptosis by TUNEL 1127
assay after un strong stimulus ISM1 (35 µg/mL) for 72 H. Cells were observed under fluorescence microscopy. 1128
(B) Apopotic cells were quantified. ISM1 induced an emergence of TUNEL positive cells corresponding to 1129
apoptotic cells (** p< 0,01 vs control group; n = 3). (C) and (D) Evidences of caspases activation. 1130
Representative immunoblots with quantitative data of densitometry are shown in (C) for caspase 8 and in (D) 1131
for caspase 3. ISM1induced the decrease of pro-caspase 8 and 3 (** p< 0,01; * p<0,05 vs control group; n = 1132
4 blots). (E) ISM1 decreases the mitochondrial membrane potential of podocytes (ΔΨm). Podocytes were 1133
treated with ISM1 (35 µg/mL) and quantification of ΔΨm was assessed by the JC-1 probe. ISM1 decreased 1134
significantly the ΔΨm of podocytes from 48H to 72 hours (*** p< 0,01; ** p< 0,01 vs control group). (F) AIF 1135
was released from mitochondria to the cytosol and the nucleus after ISM1 treatment (arrows). In contrast, 1136
cytochrome c remained confined in mitochondria after ISM1treatment. 1137
1138
Figure 7. Mitochondria and peroxisome regulation during ISM1 time course model during kidney 1139
regeneration. A) Functional enrichment on gene ontology cellular component database of genes found 1140
correlated to ISM1 during kidney regeneration (fdr: False discovery rate): barplot of negative log10 FDR 1141
adjust p-values; B) Functional enrichment network performed with mitochondria and peroxisome related 1142
genes found correlated to ISM1; C) Expression heatmap of best 75 genes found correlated to ISM1 expression 1143
kidney regeneration: blue gene cluster defined genes with regulation closed to ISM1; D) Alluvialplot built on 1144
gene-function found in NCBI database with genes highlighted closely regulated with ISM1 during kidney 1145
regeneration (quantitative colored links represent the numbers of NCBI-Pubmed articles connected) 1146
1147
Figure 8: Proposed mechanism of the deleterious action of ISM1 on podocytes. (Left) Caspase-dependent 1148
pathway. At low concentration, extracellular ISM1might bind v 5 and/or GRP78 at the cell surface, trig- 1149
gering activation of procaspase-8 and the downstream effector caspase-3, which would activate a caspase- 1150
dependent DNAse (CAD) to produce DNA fragmentation (left). At higher concentrations, ISM1 is internal- 1151
ized by endosomal trafficking and would induce the activation of a caspase-independent proapoptotic path- 1152
way. Thus, ISM1 potentially interacts with GRP78 on the cell surface and would trigger its endocytic transport 1153
into the mitochondria. Subsequently, ISM1 triggers the permeabilization of the outer mitochondrial membrane 1154
and the specific release of AIF into the cytosol. Ultimately, AIF translocates to the nucleus where it contributes 1155
to DNA fragmentation and chromatin condensation. 1156
1157
SUPLEMENTARY DATA 1158
1159
Table SI. Gene-specific primer sequences used for RT-qPCR 1160
1161
Table SII: Top 25 up- and down-regulated genes in L-NAME model between W0 and CTL groups. 1162
1163
Table SIII. Biochemical parameters in animal models of glomerulopathy. UAER (g/mmol), UPER 1164
(g/mmol) and albuminemia (g/L) in PAN and DOX models. UAER (g/mmol) in ob/ob model and UAER 1165
(g/g) in LPS model. . UAER: Urinary albumin excretion ratio and UPER: Urinary protein excretion ratio. 1166
1167
TABLE SIV. Genes found to be link to Ism1 regulation during kidney regeneration: this table describes 1168
statistics identified by Pavlidis template matching algorithm supervised on Ism1 expression during liver 1169
regeneration. For each respective significant genes columns present: correlation coefficient rvalues, raw p- 1170
values, and False Discovery Rate adjust qvalues. (DOWNLOAD EXCEL FILE) 1171
1172
1173
Figure S1. Co-localization of ISM1with its receptors, GRP78 and beta5 integrin. 1174
A) ISM1 does not colocalize with RECA-1, a rodent pan-endothelial cell marker. B) Co-localization of 1175
ISM1with its receptors, GRP78 and beta5 integrin in rats. 1176
1177
1178
1179
1180
1181

Biomedicines 1841156

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Biomedicines 1841156

Uploaded by

Copyright:

Available Formats

Article 1

Isthmin-1 is a novel mediator of experimental glomerulone- 2

phritis through a pro-apoptotic action in podocytes 3

Chadjichristos1,2, Christos Chatziantoniou1,2 , Hans Kristian Lorenzo7,8,9,# and Jean-Jacques Boffa1,2,3,#,* 6

Abstract: Focal segmental glomerulosclerosis (FSGS) is a major cause of end-stage renal 26

Biomedicines 2022, 10, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/biomedicines

2. Materials and Methods 94

ISM1 CD31 Merge

4870 CTL L-NAME PAN DOX ob/ob LPS

546 MTT 24-72h

548 100 72h

Biomedicines 2022, 10, x FOR PEER REVIEW 17 of 41

a nuclear translocation (Figure 6F, arrows) suggesting a caspase)independent apoptosis 762

Conflicts of Interest: The authors declare no conflict of interest. 824

ISM1 (rat) +blocking peptide (rat) ISM1 (human)

SUPPLEMENTARY DATA 1022

Gene name Direction Primer sequence (5′-3′)

TABLE SII 1036

Up-regulated genes Down-regulated genes

TABLE SIII 1046

UAER UPER Albuminemia

TABLE SIV 1058

(DOWNLOAD Excel file) 1063

ISM1 RECA-1 Merge

LEGENDS OF TABLES AND FIGURES 1083

(CTL) and L-NAME-treated rats. 1091

kidney of W0 and W1 L-NAME rats (* p<0.05 vs control group ). 1102

interacts with its receptors in podocytes, in basal conditions, (n = 2). 1115

cytochrome c remained confined in mitochondria after ISM1treatment. 1137

to DNA fragmentation and chromatin condensation. 1156

SUPLEMENTARY DATA 1158

Table SI. Gene-specific primer sequences used for RT-qPCR 1160

ISM1with its receptors, GRP78 and beta5 integrin in rats. 1176

You might also like