The effect of size, charge, and peptide ligand length on kidney targeting by small, organic nanoparticles

Abstract Chronic kidney disease (CKD) affects 15% of the US adult population. However, most clinically available drugs for CKD show low bioavailability to the kidneys and non‐specific uptake by other organs which results in adverse side effects. Hence, a targeted, drug delivery strategy to enhance kidney drug delivery is highly desired. Recently, our group developed small, organic nanoparticles called peptide amphiphile micelles (PAM) functionalized with the zwitterionic peptide ligand, (KKEEE)3K, that passage through the glomerular filtration barrier for kidney accumulation. Despite high bioavailability to the kidneys, these micelles also accumulated in the liver to a similar extent. To further optimize the physicochemical properties and develop design rules for kidney‐targeting micelles, we synthesized a library of PAMs of varying size, charge, and peptide repeats. Specifically, variations of the original (KKEEE)3K peptide including (KKEEE)2K, (KKEEE)K, (EEKKK)3E, (EEKKK)2E, (EEKKK)E, KKKKK, and EEEEE were functionalized onto nanoparticles, and peptide surface density and PEG linker molecular weight were altered. After characterization with transmission electron microscopy (TEM) and dynamic light scattering (DLS), nanoparticles were intravenously administered into wildtype mice, and biodistribution was assessed through ex vivo imaging. All micelles localized to the kidneys, but nanoparticles that are positively‐charged, close to the renal filtration size cut‐off, and consisted of additional zwitterionic peptide sequences generally showed higher renal accumulation. Upon immunohistochemistry, micelles were confirmed to bind to the multiligand receptor, megalin, and histological analyses showed no tissue damage. Our study provides insight into the design of micelle carriers for kidney targeting and their potential for future therapeutic application.

tolvaptan is the only FDA approved treatment for ADPKD. 6 Despite enthusiasm, tolvaptan has been reported to cause several side effects including polyuria, thirst, nocturia, and drug-induced liver damage. [7][8][9][10] According to a study of 961 ADPKD patients undergoing tolvaptan treatment, the adverse event rates for thirst and nocturia were found to be approximately 55% and 29%. 9 Polyuria affected 38% of patients and was the primary reason for dose reduction and drug discontinuation due to its significant impact on patients' quality of life. 11,12 Moreover, tolvaptan is known to cause liver toxicity and elevated liver enzyme levels, which may further deteriorate liver health, as liver cysts and fibrosis are common extrarenal manifestations of ADPKD. 8,13 As a result, frequent monitoring of liver function is required for patients on tolvaptan treatment. 14,15 Like tolvaptan, many drugs for CKD lead to non-specific uptake and undesired adverse effects, 16,17 and hence, the development of a targeted drug delivery platform that can directly unload drugs to diseased kidneys has the potential to reduce systemic side effects without compromising therapeutic efficacy.
To that end, our group has recently developed small, organic nanoparticles called peptide amphiphile micelles (PAMs, 15 nm in diameter) to target the kidneys by incorporating the zwitterionic peptide ligand, (KKEEE) 3 K([Lys-Lys-Glu-Glu-Glu] 3 -Lys), which has been found to bind to megalin, a multiligand receptor highly expressed on renal proximal tubule cells. [18][19][20] Upon intravenous injection of (KKEEE) 3 K PAMs into wildtype mice, although (KKEEE) 3 K PAMs were found to accumulate in the kidneys to a high degree, they also accumulated similarly in the liver, likely due to the mononuclear phagocytic system (MPS). 19 The kidneys filter blood and each kidney is composed of approximately a million functional units called nephrons. Each nephron consists of two main parts, the renal tubule and the renal corpuscle which contains a network of blood capillaries called the glomerulus. When blood enters the glomerulus, a fraction of blood plasma will pass through the glomerular filtration barrier (GFB) and into the renal tubule system to be processed into urine ( Figure 1). 21 Specifically, the GFB is comprised of three components and includes the endothelium, glomerular basement membrane (GBM), and podocyte foot processes that collectively acts as a size-selective and charge-selective sieve during the filtration of macromolecules. [21][22][23] The pores present in each of these layers include fenestrations of the endothelium (70-90 nm), a mesh created through a network of collagen, laminin, and proteoglycans that make up the GBM (8 nm), and filtration slits of the podocyte foot processes (4-11 nm). 21,24 Regarding charge, both the endothelium and podocytes express glycocalyx on their membranes, mostly in the form of heparan sulfate (over 50%), which is negatively-charged. [25][26][27] The GBM is also negatively-charged due to the presence of anionic proteoglycans (e.g., heparan-sulfate and chondroitin-sulfate glycosaminoglycan), and hence, positively-charged nanoparticles have been reported to passage through the GFB more readily due to charge repulsion upon interaction with negativelycharged macromolecules. 28,29 While most nanoparticles are too large to cross the fenestration of the GFB and will be recognized by the MPS system, ultrasmall nanoparticles (<8 nm) have been reported to pass through the GFB and clear via renal clearance. 30 Since the majority of cyst development in CKD such as PKD occurs in the tubules, 31,32 one strategy for targeted drug delivery for PKD is to design nanoparticles that can pass through the GFB through their physicochemical properties, but be retained in proximal tubule through the addition of targeting ligands.
Hence, to study how size, charge, peptide ligand length, and surface ligand density affects nanoparticle kidney-targeting ability and to optimize the kidney-targeting potential of PAMs, we developed a library of micelles by incorporating variations of the original zwitterionic (KKEEE) 3 K peptide including (KKEEE) 2 K, (KKEEE)K, (EEKKK) 3 E, (EEKKK) 2 E, (EEKKK)E, KKKKK, and EEEEE. The size and surface charge of the micelle library was characterized by TEM, DLS, and zeta potential, and to assess nanoparticle kidney-targeting ability, micelles were intravenously administered into C57B/6J mice. Kidney targeting and micelle biodistribution was evaluated via ex vivo imaging after 24 hours. Moreover, micelle targeting to megalin was evaluated through immunohistochemistry, and kidney biocompatibility was assessed through histology, blood urea nitrogen, and urine creatinine.

| Synthesis and characterization of PAMs
The kidney-targeting peptide (KKEEE) 3 K was found to accumulate in the kidneys in part by binding to the megalin receptor expressed on proximal tubule cells and is composed of three repeats of the peptide sequence, KKEEE. 19 Previously, when (KKEEE) 3 K was incorporated into micelles and assessed in vivo, their biodistribution profile showed enhanced kidney accumulation but also liver accumulation of 35% likely due to the MPS system as (KKEEE) 3 K PAMs were 15 nm in diameter and larger than the cut off reported for passage through the GFB (8-10 nm). 19 To optimize the physicochemical properties of PAMs, herein, we synthesized micelles with fewer repeats of KKEEE to test how peptide repeat number affected kidney accumulation. Moreover, since micelles are self-assembled from monomers containing DSPE hydrophobic tails linked to polyethylene glycol (PEG), PEG molecular weight was also varied to further study the size effects on the renal targeting ability of nanoparticles. In addition to peptide repeat number and size, peptide sequences of opposite charge (e.g., EEKKK) as well as non-zwitterionic peptide sequences (e.g., KKKKK and EEEEE) were incorporated to test the charge selectivity of the GFB on micelles. Furthermore, to assess how peptide ligand surface density affects renal targeting ability, we synthesized micelles with 50% or 100% peptide surface presentation.
To characterize the micelles, PAMs were self-assembled in water or PBS at 100 μM (which is above the critical micelle concentration of 10.2 ± 1.3 nm, which were similar to PEG2000-(KKEEE)K and PEG2000-(EEKKK)E due to the shorter peptide sequence (Table 1, Figure S1). As expected, 34 increasing the PEG molecular weight to 5,000 resulted in a slight increase in diameter of nanoparticles to 17.2 ± 1.8 nm, and nanoparticles consisting of DSPE-PEG1000 had a diameter of 10.4 ± 1.8 nm (Table 1).
Interestingly, despite (KKEEE) 3 K and (EEKKK) 3 E having an opposite net charge of −2 and + 2 derived from additional negativelycharged glutamic acid (E) or positively-charged lysine (K), the zeta potential of both micelles were found to be near neutral at −0.6 ± 1.4 mV and 0.03 ± 0.1 mV, respectively (Table 1) was also the case with nanoparticles consisting of 50% KKKKK, which had a zeta potential of 0.07 ± 0.2 mV (Table 1). On the other end, when the peptide sequence was altered to 50% EEEEE, micelles consisted of a negative zeta potential of −40.9 ± 4.6 mV.

| In vivo renal targeting and biodistribution
To test the effects of nanoparticle characteristics on renal targeting, Cy7-labeled micelles were intravenously administered into 6-7 week old male and female C57B/6 J mice and after 24 hours, micelle accumulation was assessed via ex vivo imaging of the kidneys, brain, lung, heart, liver, spleen, intestine, and bladder. 10 mol% of Cy7 was included into micelles to maximize the fluorescence signal without quenching and micelles with 45:45:10 molar ratio of DSPE-PEG-methoxy: Schematic of the nephron, renal corpuscle, and the glomerular filtration barrier DSPE-PEG-peptide:DSPE-PEG-Cy7 or 90:10 molar ratio of DSPE-PEG-peptide:DSPE-PEG-Cy7 were synthesized for in vivo studies. 35 As shown in Figure S2, despite all micelles being at or above the reported renal filtration cut-off size, all micelles accumulated in the kidneys to a greater extent than all other organs with the exception of 90% PEG5000-(KKEEE) 3 K ( Figure S2). 19 90% PEG5000-(KKEEE) 3 K, which had the highest PEG molecular weight and largest diameter of 17.2 ± 1.8 nm (Table 1), mostly accumulated in the liver (1.6 × 10 9 ± 1.3 × 10 8 p/s/cm 2 /sr) likely via recognition of the MPS system ( Figure S3). However, this was not statistically significant with its accumulation in the kidneys, which indicates an ability to simultaneously pass through the GFB and accumulate in the kidneys ( Figures S2 and S3). Although the size of micelles was slightly larger than the cut-off diameter of glomerular filtration (10 nm PEG2000-(EEKKK)E had a fluorescence intensity of 1.8 × 10 9 ± 1.9 × 10 8 p/s/cm 2 /sr, which was also statistically lower compared to PEG2000-(EEKKK) 3 E (p < .05, Figure 3a). Given that KKEEE and EEKKK are zwitterionic peptides, one possible explanation for the higher kidney accumulation of micelles consisting of PEG2000-(KKEEE) 3 K and PEG2000-(EEKKK) 3 E micelles is that the additional peptide repeats provided enhanced zwitterionic characteristics and hence, resistance to nonspecific protein and opsonin absorption as well as liver uptake via the MPS system. [37][38][39][40][41][42] Additionally, given that the hydrodynamic diameter of the micelles with varying peptide repeats fell within a narrow range (Table 1 and Figure S1), peptide sequence and ligand length may play a more important role in the renal accumulation of micelles rather than size, which will be further probed in future studies.

| Biodistribution of PAMs with opposite charge
In Figure 3b, we compared the charge effects on micelle renal targeting.
The potential influence of the protein corona on micelles on GFB penetration will be further studied in the future to better understand the factors that affect nanoparticle kidney targeting.

| Micelle colocalization with megalin
As mentioned, previously, the (KKEEE) 3 K peptide was found to target the kidney in part through megalin, a multiligand receptor that is present on the plasma membrane of proximal tubule cells, and the (KKEEE) 3 K peptide showed significantly reduced uptake in megalindeficient mice. 19 Figure 5), confirming that in addition to (KKEEE) 3 K PAMs, PAMs consisting of zwitterionic peptides of opposite charge also accumulate in the kidneys in part via megalin. This corresponds with the ex vivo results that showed 45% PEG2000-(EEKKK) 3 E had similar kidney accumulation to 45% PEG2000-(KKEEE) 3 K (Figure 3a).

| Tissue morphology and kidney health
To assess the safety and biocompatibility of micelles and their eventual application as drug delivery vehicles, after ex vivo imaging, the brain, lung, heart, liver, spleen, intestine, kidney, and bladder were stained with H&E. In agreement with other PAM reports, the tissue morphology of all organs including the liver, intestine, and kidneys, where PAMs mostly accumulated, showed no tissue damage, and no significant difference was found to the PBS treatment group ( Figure 6). 19,[55][56][57][58][59] Kidney health and function of all the mice treated with micelles were further evaluated by assessing blood urea nitrogen (BUN) and urine  60 and all the groups in this study were found to have BUN levels within this healthy range (Table 2). Similarly, the creatinine concentration in the urine indicates kidney health and urine creatinine levels for all groups were measured and found to fall within the healthy range for C57BL/6 mice (4.7 ± 3.1 mg/dl, Table 2). 61

| CONCLUSION
In this study, we altered the physicochemical properties of PAMs by developing a library of micelles containing various peptide sequences and molecular weight and assessed its effect on renal targeting.
Micelles were first characterized by TEM and DLS before intravenous administration into wildtype mice. After 24 hours post-administration, ex vivo imaging was conducted to compare kidney targeting ability.
All PAMs tested in this study had high kidney accumulation, and our studies showed that nanoparticles with more zwitterionic amino acid repeats that are positively-charged and close to the renal filtration cut-off size tend to have higher renal accumulation. Furthermore, histological analyses confirmed no tissue damage, and kidney function levels of mice were within normal ranges. Overall, our studies indicate the potential of PAMs as targeting, nanocarriers for kidney applications and future work will focus on evaluating the mechanical properties and micelle concentration of nanoparticles on kidney targeting, as well as the influence of the protein corona on GFB penetration. 62 Moreover, the therapeutic efficacy of drug-loaded micelles and the contribution of nanoparticle delivery through the peritubular capillaries will be assessed in CKD mouse models.

| In vivo renal targeting and biodistribution
To test the renal targeting ability in vivo, 100 μl of 1,000 μM micelles containing 90:10 molar ratio of DSPE-PEG-peptide:DSPE-F I G U R E 4 Ex vivo images of kidney fluorescence after (a) 45% PEG2000-(KKEEE) 3

| Histology
Immediately after ex vivo imaging, harvested organs were flash frozen in 2-methylbutane and liquid nitrogen, embedded in optimum cutting temperature (OCT) compound, and sectioned into 8 μm samples via a CM3050 S Cryostat (Leica CM3050S, Leica, Wetzlar Germany). Tissue sections were then stained with hematoxylin and eosin (H&E) and imaged with a microscope (Leica DMi8, Leica, Wetzlar, Germany).

| Immunohistochemistry
To assess micelle colocalization with megalin, kidney tissue sections were first washed with Tris buffered saline (TBS) plus 0.025% Triton X-100 with gentle agitation for 10 minutes. A block buffer with 1% bovine serum albumin (BSA), 10% normal goat serum, and 0.3 m glycine in 0.1% PBS Tween was applied to the slides for 1 hour at room temperature.
Then, the slides were applied with an anti-Lrp2/Megalin antibody

| Statistical analysis
All statistical analyses were performed using GraphPad Prism 8 (San Diego, CA). Analysis of variance (ANOVA) with a Tukey's test for post-hoc analysis was used to determine statistical significance and p ≤ .05 was considered to be significant.