Glomerulus and the filtration barrier

What it normally does

• Blood enters through the afferent arteriole and leaves through a narrower efferent arteriole, so hydrostatic pressure inside the tuft stays high — commonly quoted as 45 to 50 mmHg, with some texts using 60. Bowman capsule pressure and rising plasma oncotic pressure oppose it, leaving a net filtration pressure of only about 10 mmHg. Across the whole kidney that yields a GFR near 125 mL/min, and roughly a fifth of the plasma arriving is filtered (filtration fraction ~0.2).

  filtration is a pressure phenomenon set by two taps in series: afferent tone controls inflow, and the efferent arteriole is the outflow tap that holds pressure up inside the tuft. Relax the efferent arteriole and glomerular capillary pressure and GFR fall; constrict it and both rise (until efferent constriction is severe enough to cut renal plasma flow, when GFR falls again). Sustained high intraglomerular pressure damages the tuft over years. GFR is also autoregulated — myogenic response plus tubuloglomerular feedback, mostly through the afferent arteriole — across a mean arterial pressure of roughly 80 to 180 mmHg, so modest blood pressure swings do not change it much.

• The sieve has three layers in series: fenestrated endothelium (holes about 70 to 100 nm, covered by a glycocalyx), a type IV collagen and laminin mat (glomerular basement membrane), and the interlocking foot processes of podocytes bridged by slit diaphragms built from nephrin, podocin and NEPH1. Water, electrolytes, glucose, urea and small peptides pass freely. Restriction rises steeply with size from about 7 kDa upwards, and by the size of albumin (about 69 kDa, effective radius ~3.6 nm) the barrier is already holding back the great majority of what arrives.

  the slit diaphragm is the final and tightest layer, which is why mutations in nephrin or podocin (congenital nephrotic syndrome, familial FSGS) produce heavy proteinuria on their own. Some albumin does cross normally and is reclaimed by the proximal tubule, so albumin in the urine is a matter of degree, not an all-or-nothing sign.

• The barrier also carries fixed negative charge — the endothelial glycocalyx, heparan sulfate proteoglycans in the basement membrane, and sialoproteins such as podocalyxin on the podocyte. Albumin is negatively charged at blood pH, and classical teaching holds that this charge repels it in addition to size restriction.

  you should know the charge-barrier model because it is still examined, but hold it loosely: the endothelial glycocalyx now looks like the main charge-bearing layer, and animals engineered to lose glomerular basement membrane heparan sulfate do not develop significant albuminuria. Charge loss alone is no longer accepted as the explanation for nephrotic-range proteinuria — podocyte injury is.

• Podocytes are terminally differentiated. They wrap the capillary, hold the basement membrane in shape, and once detached or lost they are not effectively replaced in the adult human kidney (parietal epithelial cells can contribute a little, but not enough to restore a lost population).

  podocyte injury is largely a one-way street: bare basement membrane adheres to Bowman capsule, that segment scars, and the scarring can keep going after the original insult has gone. Effacement of foot processes, by contrast, is a reversible change — which is why minimal change disease remits and established sclerosis does not.

What goes wrong

• Minimal change disease← from “Podocytes are terminally differentiated. They …”

  The podocyte is injured without being lost. Foot processes retract and flatten along the basement membrane (effacement) and the slit diaphragm is disrupted, so albumin pours through. The injury has classically been attributed to an unidentified circulating factor from activated T cells; more recent work has found circulating anti-nephrin autoantibodies in a substantial proportion of patients, which fits the disease being antibody-driven and steroid-responsive. There is no structural destruction of the tuft and no immune complex deposition, so light microscopy looks essentially normal and immunofluorescence is typically negative (sometimes trace IgM or C3 in the mesangium); the effacement is seen only on electron microscopy.

  Podocyte injury with foot process effacement and nothing visible on light microscopy. Selective albuminuria, EM-only findings, steroid responsive, relapsing course. Commonest cause of nephrotic syndrome in children.

  You would find: A child, typically 2 to 6 years, with puffy eyelids on waking and frothy urine, nephrotic-range albuminuria, normal blood pressure and a bland urinary sediment. In Australia a child with this typical picture is generally treated with corticosteroids on the clinical picture alone, with biopsy reserved for atypical features (age outside roughly 1 to 10 years, hypertension, macroscopic haematuria, low complement, impaired kidney function) or for steroid resistance. Around 90% of such children remit on steroids. Adults presenting this way are usually biopsied.

• Nephrotic syndrome — the consequences of the leak← from “The sieve has three layers in series: fenestra…”

  Once the barrier fails, plasma protein is lost into the urine faster than the liver can replace it and serum albumin falls. Oedema was classically explained by the fall in plasma oncotic pressure alone (the 'underfill' model), and that fits many children with minimal change disease; in most adults, primary sodium retention by the distal nephron ('overfill') contributes at least as much — filtered proteases such as plasmin activate the epithelial sodium channel in the collecting duct. The liver increases lipoprotein synthesis and lipoprotein clearance falls, so cholesterol climbs. Antithrombin III (about 58 kDa) is lost in the urine alongside albumin, and this combines with raised fibrinogen and platelet activation to make the blood prothrombotic. Immunoglobulin is lost too.

  Proteinuria + hypoalbuminaemia + oedema + hyperlipidaemia. Antithrombin III loss (with other changes) makes it a hypercoagulable state; immunoglobulin loss makes it an immunodeficient one. Oedema is part low oncotic pressure and part primary renal sodium retention, not oncotic pressure alone.

  You would find: Periorbital oedema worst in the morning and ankle oedema by evening, frothy urine, urine protein:creatinine ratio above roughly 300 to 350 mg/mmol (about 3 to 3.5 g/day), serum albumin below 30 g/L (often below 25 in severe disease), and hyperlipidaemia. Urine shows oval fat bodies and fatty casts but no red cell casts. Sudden flank pain with haematuria in a nephrotic patient is renal vein thrombosis until proven otherwise — classically in membranous nephropathy, which carries the highest thrombotic risk.

• Focal segmental glomerulosclerosis← from “Podocytes are terminally differentiated. They …”

  Podocytes are lost — to a circulating permeability factor (primary FSGS), a virus such as HIV, high-risk APOL1 variants, drugs, obesity, or simply too few nephrons each working too hard (adaptive FSGS). Bare basement membrane adheres to Bowman capsule and that part of that tuft scleroses. Only some glomeruli are involved (focal) and only part of each tuft (segmental), so a superficial biopsy that samples mainly outer cortex can miss it. Each scarred nephron dumps more filtration onto the survivors, which lose their podocytes in turn, so the disease grinds on.

  Podocytes do not regenerate; focal and segmental sclerosis follows. One of the two commonest causes of primary nephrotic syndrome in adults (with membranous nephropathy), and the commonest in most contemporary biopsy series. Distinguish primary (nephrotic, may respond to immunosuppression) from adaptive (sub-nephrotic, normal albumin, treat the driver). A low nephron count at birth is a risk factor, which matters in populations with high rates of low birthweight.

  You would find: Nephrotic syndrome in an adult, often with hypertension, some blood in the urine and an already-reduced eGFR — frequently steroid-resistant, unlike minimal change. Adaptive FSGS (obesity, reflux nephropathy, reduced nephron mass) behaves differently: proteinuria is often sub-nephrotic, serum albumin is preserved and there is no frank oedema, and it is managed with renin-angiotensin blockade and treatment of the driver rather than immunosuppression.

• Diabetic glomerulopathy← from “Blood enters through the afferent arteriole an…”

  Chronic hyperglycaemia dilates the afferent arteriole (partly through proximal tubular sodium-glucose reabsorption blunting tubuloglomerular feedback) while angiotensin II constricts the efferent one, so intraglomerular pressure rises and the glomerulus hyperfilters for years. That pressure and the metabolic milieu thicken the basement membrane, expand the mesangium (eventually into Kimmelstiel-Wilson nodules) and cause podocyte loss and detachment. Size and charge selectivity fail together — first a trickle of albumin, then heavy proteinuria, then a falling GFR as glomeruli sclerose.

  Hyperfiltration first, albuminuria second, falling GFR third — but not in everyone. ACR plus eGFR screening catches it in the window where treatment still changes the trajectory.

  You would find: Albuminuria found on annual urine albumin:creatinine ratio screening, years before the creatinine moves — though a substantial minority follow a non-albuminuric course with falling eGFR and little or no albuminuria, so eGFR must be tracked as well. Diabetes is the single commonest primary renal disease among Australians starting kidney replacement therapy (ANZDATA), and end-stage kidney disease is several times more common in Aboriginal and Torres Strait Islander people — the gap is widest in remote communities, where diabetes, low birthweight and post-infectious glomerular damage stack on top of each other.

What we give, and how it works

Open a drug to see what it binds, what that does to the cell, and what you then see in the patient.