Inhibiting the urokinase‐type plasminogen activator receptor system recovers STZ‐induced diabetic nephropathy

Abstract The urokinase‐type plasminogen activator (uPA) receptor (uPAR) participates to the mechanisms causing renal damage in response to hyperglycaemia. The main function of uPAR in podocytes (as well as soluble uPAR ‐(s)uPAR‐ from circulation) is to regulate podocyte function through αvβ3 integrin/Rac‐1. We addressed the question of whether blocking the uPAR pathway with the small peptide UPARANT, which inhibits uPAR binding to the formyl peptide receptors (FPRs) can improve kidney lesions in a rat model of streptozotocin (STZ)‐induced diabetes. The concentration of systemically administered UPARANT was measured in the plasma, in kidney and liver extracts and UPARANT effects on dysregulated uPAR pathway, αvβ3 integrin/Rac‐1 activity, renal fibrosis and kidney morphology were determined. UPARANT was found to revert STZ‐induced up‐regulation of uPA levels and activity, while uPAR on podocytes and (s)uPAR were unaffected. In glomeruli, UPARANT inhibited FPR2 expression suggesting that the drug may act downstream uPAR, and recovered the increased activity of the αvβ3 integrin/Rac‐1 pathway indicating a major role of uPAR in regulating podocyte function. At the functional level, UPARANT was shown to ameliorate: (a) the standard renal parameters, (b) the vascular permeability, (c) the renal inflammation, (d) the renal fibrosis including dysregulated plasminogen‐plasmin system, extracellular matrix accumulation and glomerular fibrotic areas and (e) morphological alterations of the glomerulus including diseased filtration barrier. These results provide the first demonstration that blocking the uPAR pathway can improve diabetic kidney lesion in the STZ model, thus suggesting the uPA/uPAR system as a promising target for the development of novel uPAR‐targeting approaches.


| INTRODUCTION
Diabetic nephropathy (DN) is a microvascular complication of diabetes leading to end-stage renal disease that is difficult to handle despite strict glycaemic control and targeted therapies, thus indicating the paramount importance to develop novel treatments. DN can be reproduced in the rat model of type-1 diabetes induced by streptozotocin (STZ) 1 with main alterations that are established 4 weeks after diabetes onset. 2 They include glomerular hypertrophy and altered filtration barrier that is associated to increased albumin and creatinine excretion. 3 In the medulla, increased glomerular filtration results in up-regulated levels of the main transporter proteins involved in urine concentration among which aquaporin 2 (AQP2) concurs to prevent excessive water loss. 4,5 At the glomerular level, additional alterations include an increased mesangial area and renal fibrosis induced by excessive accumulation of extracellular matrix (ECM) presumably because of inflammatory processes activated by transcriptional regulators of inflammation-related genes. 6,7 Major players in the diabetes-associated renal fibrosis include members of the plasminogen (Plg)-plasmin system that consists of the circulating zymogen Plg and its activators including the urokinase-type plasminogen activator (uPA), a secreted protease that, through the binding to its receptor (uPAR), converts Plg into plasmin that promotes ECM degradation either directly or indirectly through the activation of metalloproteinases (MMPs). 8 uPA is a secreted protease, while uPAR is expressed by glomerular cells, resident fibroblasts and cells of the collecting ducts. 9 In rodent models of DN, uPA and uPAR are both up-regulated in glomerular cells including podocytes suggesting that dysfunction of the uPA/uPAR system may be associated with kidney disease. 10 In contrast, uPA or uPAR deletion results in protective effects against kidney injuries. 9 uPAR also exists in a soluble form ((s)uPAR) that is generated by the proteolytic cleavage of the membrane anchored uPAR. 10 In particular, (s)uPAR has been recently demonstrated as a key molecule in the diseased kidney 10,11 and high (s)uPAR levels in the circulation play a massive role in diabetic kidney disease by regulating podocyte function. 12 Indeed, vast data are available on major role of uPAR in podocytes (as well as (s)uPAR from circulation) to regulate the activity of αvβ3 integrin that in turns stimulates small GTPase Rac-1 proteins, potent regulators of podocyte foot process motility and effacement. 13,14 Altered activity of αvβ3 integrin/Rac-1 pathway has been linked to podocyte dysfunction leading to proteinuria 12 and main efforts have been made to clarify mechanisms for uPAR signalling in regulating podocyte adhesion and migration. 10 At the intracellular level, the interaction between uPA and uPAR is mediated by several membrane proteins including the formyl peptide receptors (FPRs) that are G protein-coupled receptors involved in different pathophysiological processes, 8 although little is known about their possible involvement in diabetic kidney disease. The transmembrane partnership between uPA, uPAR and FPRs has been reported as an attractive target for the treatment of DN, the role of drugs potentially interacting with the uPAR/FPR pathway remains to be established. The uPAR-derived tetrapeptide Ac-L-Arg-Aib-L-Arg-L-Cα(Me)PheNH 2 (UPARANT, recently designated as Cenupatide in the International Nonproprietary Names nomenclature) has been designed to compete with N-formyl-Met-Leu-Phe peptide for binding to FPRs. 15 It is endowed with a significant anti-inflammatory and antiangiogenic activity both in vitro and in vivo [15][16][17][18][19][20] and has been shown to protect the retina from pathologic changes induced by diabetic retinopathy (DR) in animal models. 21,22 In this scenario, the possibility that dysregulated uPAR pathway participates to the pathogenic mechanisms of DN may add further value to the possible development of UPARANT as valuable therapy against diabetes complications.
Here, we evaluated the curative effects of subcutaneously administered UPARANT on diabetic kidney disease using rats with STZinduced diabetes. In the present study, UPARANT concentration in the plasma, kidney and liver was determined. Protein levels of uPA, uPAR and FPRs were measured in kidney extracts, while transcripts of FPRs were determined in isolated glomeruli, before and after UPARANT treatment. UPARANT effects on (s)uPAR were also evaluated. In addition, protein levels and activity of uPA, Plg, MMP-2/ MMP-9 were measured in kidney extracts and in the plasma. Finally, UPARANT action on αvβ3 integrin and Rac-1 was explored and its possible effects on ECM components including fibronectin, collagen I and collagen IV were considered as an indirect evidence of UPARANT | 1035 pharmacokinetics (PK) and the tissue distribution study. To this aim, rats received a single dose of UPARANT succinate dissolved in PBS (vehicle) at 20 mg/kg via subcutaneous injection according to a previous study. 22 To investigate a possible therapeutic role of UPARANT, the drug was administered at 1 or 8 mg/kg daily for 5 days in 15 rats for each concentration used. Fifteen rats were left untreated while 15 rats were vehicle-treated. In all experiments, no differences were observed between untreated and vehicle-treated rats. UPARANT dose and regimen were in line with those used in previous studies. 21,22 The rats were kept individually in metabolic cages for 24 hours to collect urine for the measurement of urine output. Systolic blood pressure was measured by Tail 2.2 | Pharmacokinetics study and UPARANT tissue distribution PK and UPARANT tissue distribution were as previously described. 21 Briefly, for plasma PK, a volume of 0.

| Western blotting
Whole kidneys homogenized using RIPA buffer supplemented with a cocktail of protease and phosphatase inhibitors were processed for protein extraction. They were used to investigate the effect of UPARANT on the uPAR pathway (uPA, uPAR and FPRs), αvβ3 integrin, the phosphorylated form of β3 integrin, Rac-1, Plg, plasmin, MMPs (MMP-2 and MMP-9), markers of fibrosis (fibronectin, collagen I and collagen IV), markers of either vascular permeability (ZO-1, occludin and VEGF) or inflammation (iNOS, ICAM-1, phosphorylated and total forms of the p65 subunit of NF-κB and of CREB, and the α subunit of HIF-1 [HIF-1α]). The effect of UPARANT on AQP2 was determined in samples of the medulla that was dissected from the renal cortex (3 medullas for each experimental condition). In all experiments, homogenates were centrifuged at 22 000 g for 15 minutes at 4°C. Protein concentration was evaluated using the Micro BCA method (Thermo Fisher Scientific). Equal amounts of proteins were separated by 4%-20% SDS-polyacrylamide gel electrophoresis gels (TGX Stain-free precast gels; Bio-Rad Laboratories, Hercules, CA, USA) and transferred onto nitrocellulose membrane using a Bio-Rad Trans-Blot Turbo System. The membranes were probed using the primary antibodies listed in Table S1. After the incubation with the appropriate horseradish-peroxidase-conjugated secondary antibody, bands were visualized using the Clarity Western ECL substrate with a ChemiDoc XP imaging system (Bio-Rad Laboratories). Bands were quantified for densitometry using the Image Lab software (Bio-Rad Laboratories) and normalized to β-actin, NF-κB or CREB, as appropriate.

| Colorimetric assay
The activity of uPA and plasmin was measured using colorimetric assays. In particular, uPA activity was assayed using the uPA Activity Assay Kit (Sigma-Aldrich) while plasmin activity was assayed using Chromozym PL, a plasmin-specific chromogenic substrate (Sigma-Aldrich).
Fluorescence was measured at an excitation wavelength of 320 nm and an emission wavelength of 405 nm.

| ELISA
The activity of Rac-1 was measured using the Rac-1 Activation Assay Kit (Cytoskeleton, Inc., Denver, CO, USA), a quantitative ELISA assay  Samples were compared using the relative threshold cycle (Ct Method). The increase or decrease (fold change) was determined relative to control mice after normalization to Rpl13a. All reactions were performed in triplicate.

| Evans blue dye leakage
UPARANT effects on vascular permeability were determined by quantifying Evans blue dye leakage extravasation. Evans blue dye was dissolved in normal saline (20 mg/mL) and filtered. Anaesthetized rats were injected with Evans blue dye through the femoral vein at 20 mg/kg. Sixty minutes later, each rat was perfused through the left cardiac ventricle with 15 mL of heparinized saline (4 U/mL) under constant peristaltic flow (10 mL/min) to purge out the circulating dye. Then, the kidneys were harvested, dissected and weighed.
The Evans blue dye was extracted with formamide overnight at 65°C and read at 620 nm using a plate reader (Microplate Reader 680 XR; Bio-Rad Laboratories).

| Immunofluorescence confocal analysis
Rat kidneys were fixed overnight with 4% paraformaldehyde at 4°C, cryopreserved in 30% sucrose for 24 hours, and embedded in optimal cutting temperature medium. Thin transverse cryosections (4 μm) were placed on Superfrost/Plus Microscope Slides (Thermo Fisher Scientific). The sections were incubated with a rabbit polyclonal antibody directed to AQP2 (1:1000 dilution) 23 and then with an AlexaFluor488-conjugated secondary antibody (Life Technologies, Carlsbad, CA, USA). Confocal images were obtained with a confocal laser-scanning microscope (TSC-SP2; Leica, Wetzlar, Germany).

| Transmission electron microscopy
The cortical part of the kidneys was processed according to standardized procedures for electron microscopy. The sample were cut into 1 mm 3 pieces, fixed in 4% glutaraldehyde (phosphate buffered,

| Data analysis
Statistical significance was evaluated using one-way ANOVA followed by Newman-Keuls' multiple comparison posttest. The results are expressed as mean ± SEM or mean ± SD of three independent measurements (Prism 5; GraphPad Software, San Diego, CA, USA).
Differences with P < 0.05 were considered significant.

| Plasma, renal and liver levels of UPARANT
Subcutaneously administered UPARANT rapidly appeared in the plasma being quantifiable at 0.25 hour, reached the C max at 2.3 hours, and declined following a monophasic profile. UPARANT was still detectable at 24 hours reaching values that, although below the lower limit of quantification, were statistically different from the blank. These values are not represented in Figure 1 and were not considered in evaluating PK parameters in agreement with the FDA Guidance on Bioanalytical Method Validation. 24 UPARANT was cleared from the plasma with a t 1/2 of 2.2 hours in the elimination phase. Control values were in line with a previous work. 21 No differences were found between control and STZ rats. Additional PK parameters as evaluated using a two-phase model equation are summarized in Table S2. In the kidney of control rats, at 24 hours postdosing, the concentration of UPARANT was 2.2-fold higher (P < 0.001; 45.9 ± 3.6 μg/g) than in the liver (21.1 ± 2.4 μg/g). In STZ rats, the liver concentration of UPARANT (20.4 ± 7.5 μg/g) did DAL MONTE ET AL.
| 1037 not differ from that in controls, whereas the UPARANT concentration in the diabetic kidney (28.2 ± 4.3 μg/g) was 1.6-fold lower than in controls (P < 0.01).

| The uPA/uPAR/FPR system
Representative blots of Figure 2A are indicative of the uPA/uPAR/FPR system activation in kidney extracts. As shown by Figure 2B-F, high glucose enhanced renal levels of uPA and uPAR as well as uPA activity by about 2.6-, 4.7-and 2.5-fold (P < 0.001; Figure 2B-D), while not affecting FPRs ( Figure 2E and F). UPARANT at 1 mg/kg was ineffective on the uPAR pathway, whereas at 8 mg/kg reduced uPA levels and activity by 1.6-and 1.4-fold (P < 0.05) without affecting uPAR levels.

| Fibrotic process
Representative blots of Figure 4A are indicative of renal levels of secreted proteases that are known to mediate ECM remodelling. As shown in Figure 4B Figure 4N-P, high glucose increased fibronectin, collagen I and collagen IV by 2.4-, 7.7-and 2.9-fold (P < 0.001). UPARANT at 8 mg/kg reduced fibronectin, collagen I and collagen IV by about 1.6-, 3.1-and 1.8-fold (P < 0.001) as an indirect evidence of its ameliorative effect on fibrotic process. In agreement with these data, the histological assessment of renal fibrosis by Masson's Trichrome staining showed that when comparing kidney sections from STZ rats with those from control rats, the glomerular fibrotic areas (stained in blue) were significantly increased in STZ rats. Of note, they were decreased by UPARANT at 8 mg/kg, but not at 1 mg/kg ( Figure 4Q-T).

| Standard renal parameters
As shown in Table 1, the bodyweight of STZ rats was 1.5-fold lower than in controls (P < 0.001). Both kidney weight/bodyweight and blood glucose significantly increased by 1.6-and 4.2-fold (P < 0.001). Treatment with UPARANT at 1 or 8 mg/kg did not affect these parameters. STZ rats also showed increased urine output, urine albumin, urine creatinine, albumin to creatinine ratio,

| Vascular permeability
A prediction of the vascular permeability integrity was performed by analysing the expression levels of (a) ZO-1 and occludin, two tight F I G U R E 1 Average plasma concentrations of UPARANT after subcutaneous administration (20 mg/kg) to control (red dots and line) or STZ rats (black dots and line). Continuous lines correspond to the best fit of the equation C t = C 0 (e −kelim•t − e −kabs•t ) where k elim and k abs correspond to the kinetic constants of the elimination and absorption phase, respectively. Values are expressed as means ± SD junction proteins participating in the glomerular filtration barrier and (b) the propermeability factor VEGF ( Figure 5A). In STZ rats, ZO-1 ( Figure 5B) and occludin ( Figure 5C) decreased by 2.7-and 3.6-fold (P < 0.001) in agreement with previous studies, 25 while VEGF increased (3.9-fold, P < 0.001; Figure 5D) in line also with previous results. 26 In addition, high glucose caused plasma extravasation as determined by an increased amount of Evan's blue (2.2-fold, P < 0.001; Figure 5E). UPARANT at 1 mg/kg resulted ineffective in DAL MONTE ET AL.
Decreased plasma extravasation would result in decreased urine output to which UPARANT-associated increase in AQP2 expression in the medulla may contribute ( Figure 6A). As shown in Figure 6B and in line with previous findings, 4 DN was associated with relative abundance of AQP2 (2.6-fold, P < 0.01). UPARANT at 1 mg/kg did not influence AQP2 up-regulation, whereas at 8 mg/kg it further increased AQP2 levels (2.8-fold, P < 0.001). In addition, in control rats, AQP2 was localized to the apical plasma membrane of cells lining the collecting ducts in both the cortex and the medulla ( Figure 6C-E). No differences between control and STZ rats were found in the cortex ( Figure 6F), while AQP2 apical expression was higher in the medulla of STZ rats ( Figure 6G and H). UPARANT at 8 mg/kg did not influence AQP2 expression in the cortex ( Figure 6I), while further increased AQP2 immunoreactivity in the medulla particularly at the apical membrane ( Figure 6J and K). showing that UPARANT at 8 mg/kg, but not at 1 mg/ kg, reduced up-regulated levels of pβ3 integrin (Tyr 773 ) and αvβ3 integrin without affecting Rac-1 levels. (E) Up-regulated Rac-1 activity was unaffected by UPARANT at 1 mg/kg, but was decreased by UPARANT at 8 mg/kg (*P < 0.05 and **P < 0.001 vs control; § P < 0.01 and § § P < 0.001 vs STZ; one-way ANOVA followed by Newman-Keuls' multiple comparison posttest). Data are presented as scatter plots (B-D) or histograms (E). Each plot or histogram represents the mean ± SEM of data from three independent samples F I G U R E 2 UPARANT effects on the uPA/uPAR/FPR system. (A) Representative blots showing protein levels of uPA, uPAR, FPR2 and FPR3 in kidney extracts. β-actin was used as the loading control. (B, D-F) Densitometric analysis showing that UPARANT at 1 mg/kg was ineffective on up-regulated levels of uPA and uPAR, whereas at 8 mg/kg reduced uPA without affecting uPAR. (C) Up-regulated uPA activity in kidney extracts was unaffected by UPARANT at 1 mg/kg, but was decreased by UPARANT at 8 mg/kg. (G) FPR transcripts in isolated glomeruli showing that UPARANT at 8 mg/kg, but not at 1 mg/kg, effectively reduced up-regulated levels of FPR2 messengers. (H-J) Up-regulated uPA levels and activity in the plasma were reduced by UPARANT at 8 mg/kg, but not at 1 mg/kg. Up-regulated (s)uPAR was unaffected at any drug concentration (*P < 0.05 and **P < 0.001 vs control; § P < 0.05 and § § P < 0.01 vs STZ; one-way ANOVA followed by Newman-Keuls' multiple comparison posttest). Data are presented as scatter plots (B, D-F) or histograms (C, G-J). Each plot or histogram represents the mean ± SEM of data from three (B-G) or seven (H-J) independent samples of NF-κB p65 at Ser 276 and CREB at Ser 133 , were higher in STZ rats than in controls (5.0-, 5.6-, 11.7-, 10.2-and 6.9-fold, respectively; P < 0.001). Inflammatory markers were not affected by UPARANT at 1 mg/kg, whereas UPARANT at 8 mg/kg reduced iNOS, ICAM-1, NF-κB p65 phosphorylation, CREB phosphorylation and HIF-1α by 3.0-, 2.5-, 3.8-, 2.4-and 3.9-fold, respectively (P < 0.001).

| UPARANT delivery and administration regimen
As shown here, the plasma concentration of UPARANT is comparable in control and STZ rats in line with previous results 21  *P < 0.001 vs control; † P < 0.001 vs STZ (one-way ANOVA followed by Newman-Keuls' multiple comparison posttest). which systemic treatment with the drug has been shown both effective and safe. 21,22 In particular, UPARANT dosage is almost 50% less than that used in the PK study and is in line with the UPARANT concentration measured in the kidney.

| Recovery of the uPA/uPAR/FPR system
Increased activation of the uPAR pathway and increased (s)uPAR levels are indicative of uPA/uPAR contribution to a proteolytic cascade that has detrimental effects on glomerular cell permeability. 33 In addition, FPR2 overexpression found here in isolated glomeruli is in line with FPR accumulation in models of impaired nephrogenesis in which FPR2, in particular, appears to participate to the fibrosis process. 34,35 As shown by the present results, UPARANT efficacy against diabetic kidney lesion is associated to decreased uPA accumulation and activity, thus presumably influencing ligand availability to its receptors and increasing drug efficacy more than if the drug acted at the postreceptor level only. In this line, there is evidence that reducing uPA-uPAR interactions results in beneficial effects on kidney lesions. 36,37 As also shown here, UPARANT does not affect (s)uPAR up-regulation in agreement with previous findings demonstrating that blockers of the renin angiotensin system commonly used to prevent or delay DN do not affect plasma (s)uPAR in diabetic patients. 38 The additional finding that the drug does not influence uPAR in podocytes is in line with the possibility that UPARANT acts downstream uPAR by presumably blocking its binding to FPRs. In this respect, UPARANT has been designed to mimic the sequence through which uPAR not only interacts with FPRs thus competing with FPR ligands, 15

| Recovery of kidney lesions
Plg is mainly synthesized in the liver, circulates in the plasma and is activated to plasmin by a finely tuned balance between activators, including uPA, and inhibitors: a dysregulation of this system is reported in renal diseases, including DN, but the mechanisms through which it contributes to diabetic kidney lesions are not still completely understood. 9 The present finding that UPARANT recov- counteracting MMPs dysregulation are found to inhibit ECM accumulation, thus preventing glomerular damage in STZ rats. 47,48 To the best of our knowledge, this is the first report indicating that inhibiting the uPAR pathway restores MMPs, thus suggesting a possible mechanism through which UPARANT may act to maintain the structural and functional integrity of the glomerulus. As also observed here, the UPARANT-induced recovery of impaired glomerular function seems to be the fundamental way to lower altered renal parameters. In particular, recovered proteinuria is in line with uPAR role in regulating the αvβ3 integrin/Rac-1 pathway and suggests the possibility that uPAR interacts with αvβ3 integrin receptors to affect podocyte function. 10 Indeed, as shown here, up-regulated levels of αvβ3 integrin and increased Rac-1 activity are recovered by UPAR-ANT indicating a major role of uPAR (as well as (s)uPAR from circulation) to regulate the αvβ3 integrin/Rac-1 pathway in podocytes. In this respect, induction of uPAR signalling in podocytes leads to foot process effacement and urinary protein loss via a mechanism that includes the activation of the αvβ3 integrin/Rac-1 pathway. 36 Thus, preventing uPAR/(s)uPAR interactions with αvβ3 integrin is likely to affect the activation of the αvβ3 integrin/Rac-1 pathway and the uPAR/(s)uPAR-mediated podocyte injury. 10 Ameliorative effects upon the inhibition of the uPAR pathway are in agreement with the finding that uPAR down-regulation restores the filtration barrier function in cultured podocytes 49 and reduces proteinuria in lipopolysaccharide-treated mice. 36,37 This is also in line with the present finding that UPARANT-induced recovery of altered vascular permeability is established through restored levels of tight junction proteins and inhibition of VEGF production. These results together suggest an additional mechanism through which the uPAR pathway negatively affects glomerular filtration and are in line with previous findings demonstrating that deletion of various components of the uPA/uPAR system is protective against vascular changes in models of renal diseases. 9 Recovered vascular permeability is associated with UPAR-ANT-induced increase of AQP2 in the medulla in which AQP2 seems to prevent excessive water loss through urine. 4 In this respect, therapies that up-regulate AQP2 expression in the inner medulla have been shown to prevent diabetic kidney disease. 50 Additional effects of UPARANT include its potent anti-inflammatory action. This is particularly interesting since inflammation has been shown to play a major role in diabetic kidney disease and the suboptimal efficacy of the current therapeutic strategies may depend on their limited impact on inflammatory processes. 51 Flavonoids, for instance, have been shown to attenuate DN more efficiently than commercial antidiabetic drugs via suppressing the activation of NF-κB and decreasing ICAM-1 and iNOS. 52,53 As shown here, UPARANT inhibits up-regulated levels of NF-κB, which, in addition to activate gene transcription of inflammatory factors, 54 promotes uPA transcription 55 and regulates ECM production. 56 In addition, NF-κB regulation of ICAM-1 positively correlates with nephropathy 57 by influencing mesangial cell proliferation. 58 Moreover, NF-κB participates to iNOS accumulation 59 and, together with CREB, regulates AQP2 gene transcription. 60 Whether UPARANT-induced inhibition of HIF-1α accumulation ameliorates kidney disease remains unclear as to some extent the activation of HIF-1, but not its inhibition, seems to exert a beneficial role in the progression of DN. 61 Suppression of inflammatory processes and ameliorated renal fibrosis results in recovered renal morphology as shown here by the reduction of glomerular hypertrophy and mesangial area increase. 62 Additional efficacy of UPARANT includes ameliorated renal filtration barrier as demonstrated by recovery from increased thickness of glomerular basement membrane and from loss of podocyte foot processes leading to proteinuria. Consistently, the observed reduction in the urine albumin level in UPARANT-treated rats is congruent with UPARANTinduced protection of the renal architecture.

| CONCLUSION
Together, the present findings support the possibility that uPA/uPAR activation in response to high glucose participates in the mechanisms causing DN and provide evidence that the uPAR pathway is a promising target for the development of novel multitarget drugs in the treatment of diabetic kidney disease. By acting on multiple pathways involved in DN pathogenesis, UPARANT constitutes a promising strategy to cure the diseased kidney although the extrapolation of these experimental findings to the clinic is not straightforward.