Structure and function of the glomerulus

This chapter is vaguely relevant to the aims of Section H3(i) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "describe the functional anatomy of the kidneys". Specifically, the functional anatomy of the glomerular filtration barrier is the topic. Though this material has never appeared in a CICM First Part written paper, it is an appetising target, because it offers the opportunity to torture exam candidates by asking them to draw and label diagrams. If you think it could never happen (because the cruelty), think again. With this foreboding, this chapter answers hypothetical future questions, on the theme of  "outline the structure and function of the glomerulus", or "describe the functional anatomy of the glomerulus", or "explain the physiological importance of the glomerulus". It is only a matter of time before trainees are hit with something like this.

In short:

It would be dangerous to recommend published reviews to guide the reading of a time-poor candidate, as that candidate should definitely detour around this topic altogether. A better use of their precious time would be Pollak (2014), which at least contains some factual physiology. The structure of the glomerulus is far less examinable than its function. 

The renal corpuscle

The renal corpuscle is the term used to describe the tiny spherical complex which consists of the glomerulus and Bowman's capsule.  The original term for these things was "Malpighian corpuscles", after Marcello Malpighi, a widely celebrated microscope hero and Papal physician to the surprisingly moral Pope Innocent XII. The eponym has since fallen into disuse (because archaic), somehow leaving behind the equally archaic term "corpuscle" which basically means "little body".

Within this little body, the term glomerulus ("little ball of string") now describes only the tuft of filtering capillaries. The rest of the sphere is Bowman's capsule, named after William Bowman. As is usual in such scenarios, he did not name it after himself (in his book he refers to these structures as "the Malpighian tuft" and "the Malpighian capsule"). The original artwork and description from Bowman's 1857 publication are presented here:

"The Malpighian body itself consists of a rounded bunch of capillaries derived from the afferent and terminating in the efferent vessel, the former dividing over the surface, the latter emerging from the interior. This vascular tuft lies within a clear and perfectly transparent capsule, lined at its lower part with epithelium"

Skipping ahead by over a hundred years, the scanning electron microscope became available, meaning that we could rely on something more accurate than sketches, talented though they might be. These vascular corrosion casts of glomerular capillaries were created and photographed by Dr. Fred Hossler (J.H. Quillen College of Medicine, East Tennessee State University).

The size values were not recorded in the original image, and they were added speculatively.  The first to measure these structures was again Bowman, and he measured them in a range of animals (man, dog, rat, horse, parrot, tortoise and boa), but unfortunately he presented his findings in fractions of an English inch, so these measurements are adapted from Samuel et al (2008). They are not definitive, as glomerular sizes are obviously going to differ from person to person (Samuel et al found they vary within a range of 2.5 magnitudes). 

Moving in closer, here is a beautiful colourised  SEM microphotograph from L. Mesnard and P.Callard (Hôpital Tenon, France), stolen from them via the use of Flickr:

Try as one might, it was impossible to trace the original publication where one might find tthis beautiful image. Here you can see the capsule de-roofed, exposing Bowman's space and the tuft of capillaries inside. The feathery tentacles wrapping around the glomerular capillaries are the foot processes of podocytes that live in Bowman's space.  

Cellular structure of the glomerulus

Slicing open this convoluted tangle of capillaries gives some impression of the complexity of the filtration surface. Here is a suitable image from a series by Terasaki et al (2020), who tried to assess the 3D structure of the glomerular tuft by scanning electron microscopy. The original image was garishly recoloured and sliced to make it easier to label, or something.

So, the three main cell types of interest in the glomerulus are the glomerular podocytes, the fenestrated capillary endothelial cells and the mesangial cells. One might be tempted to regard the mesangial cells as some sort of pointless filler, and in many ways one would be correct. They are highly specialised modified smooth muscle cells that are capable of contraction and which are both the destination for, and the origin of, some important vasoactive mediators.  Schlondoff (1987) reported that they contract and relax in response to all the substances you'd normally expect to affect vascular smooth muscle (noradrenaline, vasopressin, eicosanoids), and are also capable of secreting prostaglandins. In spite of these properties, their influence on glomerular blood flow appears to be minimal, and most authors attribute them with mainly janitorial duties (eg. removal of macromolecules).

Anyway; as you can see, the capillaries of the tuft are rolled up into a mess. Terasaki et al used 3D mapping to trace the path of all these capillaries and found that they do seem to branch, but the afferent and efferent branches do not connect by bridging capillaries, as that would give blood the opportunity to pass through this structure without being filtered. In other words, they are just like almost every other capillary bed, just curled up in a ball. If you were to lay them out flat, they would look something like this:

Without degenerating into Amazing Facts, there is some physiological relevance to mentioning that the total length of all the capillaries in each glomerulus is about 0.95cm, and the total length of all glomerular capillaries is about 2000m, which gives a total capillary surface area of around 6000 cm² and a total glomerular filtration surface area of around 500 cm² (Bohle et al, 1998). The last of these measurements plugs into the equation which describes glomerular filtration, and is therefore is probably worth knowing about.

To illustrate the fine structure of the cells in the glomerulus and how they relate to one another, here is some beautiful pencil/charcoal art by an unacknowledged artist from Yamada (1955), which was defaced with garish colours below: 

The feathery foot processes of the podocytes interdigitate to wrap around the glomerular capillaries. Between the thin capillary wall and the podocyte mesh lays the glomerular basement membrane. These three structures form the glomerular filtration surface.

The glomerular filtration surface

The business end of the glomerulus is the membrane that does all the filtering. This is a topographically complex surface that consists of the glycocalyx, endothelial cell fenestrations, glomerular basement membrane and filtration slits between podocyte foot processes. This excellent image from Farquhar (2006) clearly illustrates all the major players:

As Farquhar points out, there remains some uncertainty as to which of these major players is the main protagonist. "Yet there is still no consensus concerning which component... represents the primary glomerular filtration barrier. Over the intervening years, the pendulum has swung back and forth according to the interpretation of the evidence available", she observed overviewing fifty years of nephrology research. Something in this membrane allows the glomerulus to be highly selective about what it lets through and what it keeps in the capillary lumen, but there is still a lot of disagreement as to what exactly does the sieving.

The endothelial cell fenestrations seem essential for glomerular filtration (as in, there would be no filtration if they were absent), but what role do they play in filter selectivity? Their size, certainly, is a barrier to erythrocytes and other cells, but proteins should pass right through. Borrowing some electron microscope images from de Souza et al (2016) and albumin measurements from Tojo & Kinugasa (2012), an individual handy with Adobe products could easily throw together a scaled image like this one:

Satchell & Braet (2009) pose that our impression of these as gaping holes is entirely inaccurate. The capillary endothelium, everywhere as well as in the glomerulus, is covered by a layer of endothelial glycocalyx, a 200-nm-thick coat which probably also covers these fenestrations, but which is eroded by the brutal process of preparing samples for electron microscopy. When special glycocalyx-preserving techniques are used to prepare the sample, a fine coat of filamentous proteins can be seen, covering the fenestrations and presumably acting as a much finer size barrier. Here, a labelled TEM image (83,000 times magnification) from Rostgaard & Qvortrup (2002) illustrates this fuzz, sometimes referred to as “fascinae fenestrae”.

That this layer is essential for macromolecular retention was demonstrated by Singh et al (2007), who used neuraminidase to remove the glycocalyx from some in vitro glomeruli and produced a 200% increase in the transmembrane flux of albumin. Clearly, we would miss it if it was gone. Endothelial injury associated with denudening (denudement? Denudation? Nudery?) of the glomerular glycocalyx is blamed for the proteinuria associated with preeclampsia (Butler et al, 2020).

The glomerular basement membrane

The glomerular basement membrane is the next barrier to molecular transit through the glomerular filtration surface. It is a protein-based matrix layer,  a bit thicker than your usual basement membrane, and contains unique variants of the usual protein components  (laminin, type IV collagen, nidogen, heparan sulfate). These seem to be essential: knocking out the special laminin-521 or collagen α3α4α5(IV) genes seems to produce proteinuria in mice. 

Carrying on with the obsession over the measurements of these tiny things, let us revel in the work of Ilya Kamenetsky and colleagues (2010), who used a transmission electron microscope to interrogate thin slices of human GBM. The mean thickness of a normal GBM was about 370 nm (with diabetic nephropathy patients having monstrously thickened membranes measuring up to 1000 nm). These images are not from Kamenetsky et al (they were dredged out of the bowels of the Renal Fellow Network by Google image search), but they are TEM images, and they do serve the purpose of illustrating the dimensions:

In the image, three layers are visible.  This level of organisation is almost always mentioned by textbooks, and usually they even name the layers (lamina densa, lamina rara interna and lamina rara externa), but it is hard to tell what the significance of this is. The deeper you dig, the more detail there is, and fine TEM investigation has revealed all sorts of organised elements (Laurie et al, 1984). For example, in the thick electron-dense layer in the middle (hence densa) there are all sorts of mysterious structures ("cords", "basotubules", "double pegs") the functional role of which remains to be established. From the perspective of the time-poor exam candidate (who should have given up on this chapter long ago), their relevance is nil. 

Apart from acting as a trap for Goodpasture's syndrome antibodies, the GBM has a clear role to play in glomerular filtration, as becomes readily apparent whenever it is damaged. However, there is some disagreement with respect to what exactly it does, and how it does it. For example, Naylor et al (2020) fill pages upon pages with elaborately detailed discussions of each individual protein component, and at the end are forced to admit that it's probably some kind of size barrier, similar to the endothelial glycocalyx.

Now, the attentive reader will one might have noted that nowhere in the images or literature is there any reference to any sort of pores in this thing, and from all accounts it seems to be a solid sheet of protein, 360 nm thick. However molecules can still get through by diffusion, in a manner described by Ogston (1958) as a property of gel suspensions of fibres. In this model, to quote Naylor et al,

"GBM determines the permeation of macromolecules through the filter in a size-dependent manner through diffusion, whereas water and ions would pass through the filter by flow generated from hydraulic pressure"

In short, the GBM acts as a size barrier purely because molecular size is one of the determinants of molecular diffusion through a gel membrane. It is certainly gel-like, being composed of about 93% water (Comper et al, 1993). For a real-world comparison, that is close to the water composition of a jellyfish. If one had a handful of glomerular basement membrane material, the mound of it would wobble with poking. Anyway: from these gel-like properties, it follows that if one were observing the glomerular function closely enough, one should be able to detect the slow passage of large molecules through the gel by finding them stuck halfway. This, in effect, is what Lawrence et a (2017) did when they perfused some mouse kidneys with gold nanoparticles. "Albumin-sized particles permeate extensively into the lamina densa", they noted with surprise. 

Podocyte filtration slit diaphragms

The foot processes of neighbouring podocytes interdigitate without toughing, and between foot processes, a weblike structure sits, which is generally referred to as the "slit diaphragm". It is a thin electron-dense layer seen stretching between adjacent processes on TEM sections, as on the image below (from Ichimura et al, 2015):

This thing is a specialised intercellular junction with a weird zipper-like internal structure. In case the reader is interested (and let's face it, if they have made it this far, they clearly are), Rodewald & Karnovsky (1974) were able to get a clear look at this structure:

The image on the right is the podocyte slit diaphragm of a rat, magnified 153,000 times. Measurements in this 1974 paper were reported in Ångström (1.0 Å = 0.1 nm), i.e. the total thickness is about 40 nm. The protein nephrin, though to be the main structural component of the teeth in this zipper, is clearly important because proteinuria results when it is missing. Over the timeframe during which the importance of the glomerular basement membrane was gradually deprioritised in the literature, much research was published to suggest that this podocyte cell junction is the main size barrier behind glomerular filtration (Kawachi et al, 2006).

Ultrafiltrate in Bowman's space

After it has defeated the challenges of the glomerular filtration barrier, the ultrafiltrate will collect in Bowman's space, the cavity formed inside the spherical collection bowl made of translucent simple squamous epithelial cells. These cells are aggressively boring, insofar as they appear to serve a purely structural purpose, though there are rumours that they can differentiate into podocytes if properly motivated. The ultrafiltrate is probably more interesting, but to go into its properties in any great depth would trespass dangerously close to a discussion of glomerular filtration, which is a completely separate chapter.