Animal experiments were approved by the Massachusetts General Hospital Subcommittee on Research Animal Care, in accordance with the National Institutes of Health, Department of Agriculture, and AAALAC requirements. 12 to 48 week old male Sprague Dawley rats were anesthetized with pentobarbital sodium (50 mg/kg body wt i.p, Nembutal, Abbott Laboratories, Abbott Park, IL) and perfused through the heart with phosphate-buffered saline (PBS, 0.9% NaCl in 10 mM phosphate buffer, pH 7.4) followed by paraformaldehyde (4%) lysine (75 mM) periodate (10 mM) fixative in 0.15 M sucrose, 37.5 mM sodium phosphate (modified PLP) as previously described 
, with a measured osmolarity of 935 mOsm/kg, or by 2 or 4% glutaraldehyde (GA) in 0.1 M sodium cacodylate (pH 7.4). The flow rate of perfusion was 20 ml/min and was performed for about 5–7 min with PBS and subsequently for about 5 min with the fixative. Tissues were post fixed overnight in modified PLP or GA at 4°C, washed in PBS, and stored at 4°C in PBS containing 0.02% NaN3
After fixation and removal from the animal, 500 µm thick tissue sections were cut under PBS using a TPI PELCO 101 series 1000 vibratome (Technical Products International, Inc., St. Louis, MO). The 500 µm thickness was chosen for ease of handling and to maintain the slice integrity during further processing steps. Slices were then returned to PBS containing 0.02% sodium azide at 4°C.
To label the surface glycocalyx of the kidney tissue, Triticum vulgare (WGA, EY Laboratories, San Mateo, CA) lectin (or agglutinin) conjugated to either 40 or 15 nm colloidal gold particles was diluted 1
50 in PBS and incubated with the tissue overnight at 4°C, then washed three times in PBS. To label megalin on the PT cell membrane, the kidney tissue was incubated overnight at 4°C with a mouse monoclonal anti-megalin (1H2) primary antibody 
100 in PBS, washed three times in PBS and then incubated with a 40 nm colloidal gold conjugated goat anti mouse IgG secondary antibody (EY Laboratories). After labeling, samples were washed thoroughly in PBS and post fixed for 2 h in 2% GA in 0.1 M sodium cacodylate (pH 7.4), and then washed again in PBS prior to methanol replacement and critical point drying.
Kidney slices were placed into metal baskets and incubated with a mixture of 25% MeOH and 75% aqueous 0.1M (NH4)2CO3 for 2 hours at 2°C, while the temperature was lowered to 0°C using the Leica AFS freeze-substitution apparatus (Leica Microsystems Inc., Buffalo Grove, IL). The methanol-aqueous buffer was replaced twice with a fresh batch of 25/75 mixture, while the temperature was lowered to −10°C over a period of 4 h. A 40/60 methanol/aqueous buffer was employed two times over a period of 4 hrs, while the temperature was lowered to −20°C. It was followed by four 60/40 MeOH/aqueous mixture incubations over a period of 8 h while lowering the temperature to −40°C. We then used 80% MeOH in pure water for incubation (8 h) to drop the temperature to −60°C, followed by 100% MeOH (8 h, −80°C) and 100% MeOH (8 hrs, −90°C), replacing the MeOH at each step. Alternatively, a rapid series of graded methanol solutions in PBS was applied over a 4 h period at 4°C with the following schedule and MeOH dilutions: 25% for 60 min, 40% for 45 min, 60% for 45 min, 80% for 45 min, 100% for 45 min. For each gradation, the MeOH solution was refreshed halfway through the incubation.
Critical point drying
Following the final MeOH replacement, the temperature of the samples in pure MeOH was raised to 0°C (at a rate of 3°C/min), the baskets were closed and placed into an Erlenmeyer flask containing ice-cold MeOH for transportation. Once MeOH and baskets were placed and secured in the critical point drying apparatus (Samdri-795, Tousimis Research Corp., Rockville, MD), the samples were purged with cold liquid CO2 (2°C) at elevated pressure, and then brought to supercritical pressure and temperature (1200 psi, 42°C) for incubation and equilibration (>4 min). The pressure was slowly reduced (<100 psi/min), while maintaining supercritical temperatures (>32°C), and after the bleeding process was completed, dried samples were mounted onto placeholders with sticky pads and stored under desiccant at room temperature. Storage under these conditions for more than one week did not result in obvious sample modifications.
Helium ion microscopy
Helium ion microscopy (HIM) was carried out on an Orion helium ion microscope (Carl Zeiss Microscopy, Peabody, MA) at 35 keV beam energy, with a probe current ranging from 0.1 to 1.5 pA. No conductive coatings were applied to the samples prior to imaging, in order to preserve the sample surface information. Samples were transferred into the HIM via a load-lock system and were maintained at a vacuum of 2–3×10−7
torr during the imaging session. Charge control was maintained through the use of a low energy electron flood gun, which was applied in a temporally interlaced fashion with the imaging beam. Images were formed by collecting the secondary electrons elicited by the interaction between the helium ion beam and the sample with an Everhart-Thornley Microchannel plate 
. This detector is also widely used in SEM, and consists of a scintillator placed inside a Faraday cage, that draws the elicited low voltage secondary electrons towards it. The resulting photons are collected, turned into electrons and amplified by a photomultiplier tube (PMT). The PMT signal is then digitized using an A/D converter and displayed as a grey value in a given pixel of the resulting image. The scanning of the helium ion beam and the formation of the image are synchronized so that for any given coordinate there is a corresponding signal or grey value in the resulting image. No post-processing procedures were applied to the digital images besides brightness and contrast adjustment 
. The image signal was acquired in a line-averaging mode, with either 32 or 64 lines being integrated into each line in the final image. Charge neutralization was applied after each individual line pass of the beam.
Scanning electron microscopy
Conventional scanning electron microscopy (SEM) was performed for comparison purposes. Wherever indicated, sputter coating was carried out at the Harvard University Center for Nanoscale Systems (Cambridge, MA) using a Cressington HR208 sputter coater (Cressington Scientific Instruments, Watford, England) and a Pt at 40 mA. The SEM was performed with a Merlin field emission scanning electron microscope (Carl Zeiss Microscopy) using either an SE2 or an in-lens detector.
Measurement of cellular and cell membrane features
Images were imported into ImageJ software version 1.42q (NIH, Bethesda, MD), and the scale was set based on the scale bar in the annotated image file. To measure features in the image, the line bar was used, and measurements were expressed in micrometers. To measure pore size the line tool was used to determine the diameter of the pore at its largest axis.
The mean feature size and standard deviation from the sample mean were calculated in Excel version 12.3.4 (Microsoft Corp., Redmond, WA).