Loop of Henle
The hairpin that dives from the cortex into the medulla and comes back: it strips salt out of the urine on the way up and packs that salt into the medulla around it, which is what makes a concentrated urine possible — the loop lays down the salt half of the gradient, and urea recycling supplies much of the rest.
What it normally does
The thick ascending limb drags salt out of the tubule with one pump sitting on the urine side of the cell (the sodium-potassium-two-chloride cotransporter, NKCC2), and it reclaims about a quarter of all the sodium you filter. The proximal tubule handles more (about two thirds), but nothing downstream of the loop has anything like the capacity to make up what the loop lets through. Because the pump faces the lumen, a drug that blocks it has to reach the tubular fluid first — and since loop diuretics are almost entirely albumin-bound, very little is filtered and they get there mainly by proximal secretion.
this is the biggest lever in the kidney for shedding salt and water, and a drug only works here if it actually reaches the urine.
The descending limb leaks water but not salt; the ascending limb is the reverse — waterproof, but pumping salt out. Salt leaves without water following, so the fluid climbing up gets dilute while the medulla around it gets salty. The two limbs run side by side in opposite directions, so that small difference is multiplied along the length of the loop (countercurrent multiplication), reaching roughly 1200 mOsm/kg at the papillary tip in humans. Urea recycling from the inner medullary collecting duct contributes a large share of that peak osmolality — conventionally put at about half — so the loop builds the gradient but does not build it alone.
there is a salty medulla to pull water out of the collecting duct — without that gradient, antidiuretic hormone has little to work with.
Potassium carried in on the pump leaks straight back into the lumen through a potassium channel (ROMK), leaving the tubular fluid positively charged. That positive charge pushes calcium and magnesium out between the cells rather than through them (paracellular reabsorption, via the claudin-16/19 pore).
calcium and magnesium handling is welded to the same pump — block it and both are lost in the urine.
The thick ascending limb is one of the most oxygen-hungry tubular segments, because its basolateral sodium-potassium ATPase runs flat out to keep NKCC2 supplied. It sits in the medulla, where blood arrives already stripped of much of its oxygen by countercurrent shunting in the hairpin vasa recta, and where locally produced prostaglandins keep those vessels open.
the outer medulla runs on the edge of hypoxia, so it is among the first regions injured when perfusion drops — the usual mechanism of ischaemic acute kidney injury in hospital.
What goes wrong
- Loss of the medullary gradient (washout)← from “The descending limb leaks water but not salt; …”
The corticopapillary gradient flattens toward plasma for one of two reasons. Either salt is no longer being deposited — a loop diuretic blocks NKCC2, so the multiplier has nothing to pump into the interstitium — or high tubular and vasa recta flow carries interstitial solute away faster than the multiplier can rebuild it, as in the diuresis after an obstructed bladder is catheterised, the recovery phase of acute kidney injury, or a scarred shrunken medulla in chronic kidney disease. Prolonged low protein intake removes the urea contribution as well. The collecting duct still has its aquaporins and the posterior pituitary still releases antidiuretic hormone (made in the hypothalamic supraoptic and paraventricular nuclei), but there is little osmotic gradient for water to move down.
No gradient, no concentrating power. Antidiuretic hormone can do little without the gradient the loop builds, whether the gradient was never laid down or has been washed out.
You would find: Nocturia and large volumes of pale urine. Urine osmolality stuck near plasma, around 300 mOsm/kg, however dry the patient gets (isosthenuria), and a brisk diuresis on the ward in the hours after a catheter relieves retention. A low threshold for looking for chronic kidney disease is reasonable in Aboriginal and Torres Strait Islander patients, in whom end-stage kidney disease rates run several times the non-Indigenous rate.
- Ischaemic injury of the outer medulla (acute tubular necrosis)← from “The thick ascending limb is one of the most ox…”
Sepsis, haemorrhage, or a long hypotensive anaesthetic drops medullary blood flow. Non-steroidal anti-inflammatories make it worse by removing the prostaglandins that keep the vasa recta open. The two oxygen-hungry outer medullary segments — the medullary thick ascending limb and the S3 straight segment of the proximal tubule — swell, lose their polarity and detach into the lumen, where the debris forms casts and obstructs flow. Because the cells are injured, they can no longer reabsorb sodium normally.
The outer medulla is normally hypoxic, so it is injured early. Failure to conserve sodium is what separates established tubular injury from a simple pre-renal state — but only in a patient who is not on a diuretic.
You would find: Oliguria and a creatinine rise over the 24 to 48 hours after the insult, muddy brown granular casts on microscopy, and urinary sodium above about 40 mmol/L with a fractional excretion of sodium above 2%. Those sodium indices are only interpretable off diuretics — a loop diuretic raises urinary sodium and fractional excretion of sodium regardless of the cause, and fractional excretion of urea (below about 35% suggests pre-renal) is the usual workaround.
- Hypokalaemic metabolic alkalosis with calcium and magnesium wasting← from “Potassium carried in on the pump leaks straigh…”
Block NKCC2 — with a loop diuretic, or by being born without a working transporter (type 1 Bartter syndrome) — and the lumen-positive voltage collapses, so calcium and magnesium are no longer driven out between the cells and leave in the urine. At the same time all that escaped sodium arrives at the collecting duct, where aldosterone, driven up by the volume loss, drives sodium reabsorption in exchange for potassium and hydrogen secretion. Chloride and water are lost while bicarbonate is retained and concentrated (contraction alkalosis, on top of the distal acid secretion). If magnesium does fall, low intracellular magnesium unblocks ROMK in the distal nephron and potassium pours out faster still.
One blocked pump explains the picture: lost lumen-positive charge wastes calcium (and magnesium), and distal sodium delivery plus aldosterone wastes potassium and hydrogen.
You would find: Cramps, weakness, prominent U waves with QT/QU prolongation, low potassium, low chloride and a raised bicarbonate, with a postural drop in blood pressure. Hypercalciuria is the expected finding; serum magnesium is often normal or only mildly low here (marked hypomagnesaemia with hypocalciuria points instead to the distal convoluted tubule — thiazides or Gitelman syndrome). When potassium will not come up, check the magnesium.
- Fluid overload the loop can no longer clear (diuretic resistance)← from “The thick ascending limb drags salt out of the…”
The drug has to get into the tubule to reach a luminal pump. In heart failure and chronic kidney disease renal blood flow falls, the drug is heavily protein bound, and retained uraemic anions compete for the proximal organic anion transporters that secrete it, so less of each dose reaches its target. Gut oedema and slowed transit make oral absorption erratic on top of that. Then there are two separate adaptations: within hours of a dose wearing off the nephron retains sodium avidly (post-diuretic sodium retention, the short-term braking phenomenon), and over weeks of repeated blockade the distal convoluted tubule cells hypertrophy and up-regulate NCC, mopping up the sodium the loop released (chronic distal nephron remodelling).
Resistance is a delivery problem plus a downstream compensation problem, not a receptor problem.
You would find: Rising weight, worsening orthopnoea and swelling on an unchanged oral dose, with a disappointing urine output after each one. Oral frusemide bioavailability averages roughly half and is highly variable between patients, so the same number of milligrams given intravenously delivers appreciably more drug to the kidney — which is part of why the intravenous route can work when the tablet has stopped working. Dosing decisions belong with the treating team.
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.
- Binds
- The chloride binding site of the sodium-potassium-two-chloride cotransporter (NKCC2), on the urine-facing membrane
- Which does
- Sodium, potassium and chloride stay in the tubular fluid instead of crossing into the cell. The medulla stops being fed salt, so the gradient decays, and the lumen-positive voltage disappears with the potassium recycling that generated it.
- So you see
- Up to about a quarter of filtered sodium — the whole load the thick ascending limb normally reclaims — can be lost, with water behind it, which is the largest diuresis of any class (high ceiling). In acute pulmonary oedema symptoms often ease before any urine appears; this early effect is traditionally attributed to venodilation, though the evidence is mixed and some studies show transient neurohormonal vasoconstriction instead.
- And the same mechanism causes
- Everything the pump did is undone at once: sodium delivered distally is traded for potassium and hydrogen, and the lost lumen-positive charge lets calcium and magnesium escape into the urine — so hypokalaemia, metabolic alkalosis, hypomagnesaemia and hypercalciuria (with a modest fall in serum calcium; frank hypocalcaemia is uncommon) fall straight out of the mechanism rather than being a generic side-effect list. Hypovolaemia and hyponatraemia follow from the diuresis itself. A closely related transporter (NKCC1) secretes endolymph in the stria vascularis of the inner ear, which is why high or rapidly infused intravenous doses cause tinnitus and hearing loss, usually but not always reversible: the same transporter family, a different tissue. Gout and hyperuricaemia are also mechanistic — the drug competes with urate at the same proximal organic anion transporters and volume contraction raises urate reabsorption.
- Handling
- Highly protein bound, so little is filtered and most must be secreted into the tubule by proximal organic anion transporters. Hypoalbuminaemia and poor renal perfusion both cut the fraction of the dose that ever reaches the pump, and filtered albumin can bind the drug within the tubule in heavy proteinuria.
Catches people out: It does not treat acute tubular injury. It can convert oliguric acute kidney injury into non-oliguric, which makes fluid balance easier to manage, but trials have not shown it improves renal recovery, dialysis-free survival or mortality.
- Binds
- The intracellular mineralocorticoid receptor of the principal cell (spironolactone, eplerenone), or the epithelial sodium channel itself (amiloride)
- Which does
- Fewer open sodium channels in the luminal membrane — mineralocorticoid receptor blockade reduces the number and activity of channels over hours, amiloride plugs the channel pore directly — so less sodium enters the principal cell and the lumen stays less electronegative. The electrical driving force for potassium secretion by principal cells and hydrogen secretion by neighbouring intercalated cells is reduced.
- So you see
- A small extra sodium loss, and potassium and hydrogen retained instead of dumped — the loop's hypokalaemia and alkalosis are pulled back toward normal.
- And the same mechanism causes
- Retaining potassium is the whole point, so hyperkalaemia is that same effect gone too far, and it becomes dangerous quickly as eGFR falls or when an ACE inhibitor, angiotensin receptor blocker or potassium supplement is already on board. Reducing distal hydrogen secretion can also give a mild metabolic acidosis. Spironolactone is a non-selective steroid-receptor ligand: as well as the mineralocorticoid receptor it antagonises androgen receptors and interacts with progesterone receptors, causing gynaecomastia, breast tenderness and menstrual irregularity. Eplerenone acts at the same mineralocorticoid receptor but is far more selective for it, so these effects are much less frequent; amiloride does not act at steroid receptors at all.
Catches people out: Correcting a low potassium here is difficult while magnesium is low — replace the magnesium too, or the potassium simply leaks out again through ROMK.
- Binds
- V2 vasopressin receptors on the basolateral membrane of collecting duct principal cells
- Which does
- Gs coupling raises cyclic AMP, which shuttles stored aquaporin-2 channels into the luminal membrane so water can cross the epithelium.
- So you see
- Water follows the medullary gradient out of the tubule. Urine volume falls and urine osmolality climbs.
- And the same mechanism causes
- The retained water that concentrates the urine also dilutes the blood, so hyponatraemia follows directly from the therapeutic effect — and if sodium falls quickly, headache, confusion, seizures and cerebral oedema.
Catches people out: With a washed-out medulla, or in a patient on a loop diuretic, desmopressin cannot concentrate the urine much — not because the receptor is broken but because there is little gradient for water to move down. That is a clean illustration that the loop, not the collecting duct, sets the ceiling on how concentrated urine can get.
- Binds
- The sodium-chloride cotransporter (NCC) on the luminal membrane of the distal convoluted tubule
- Which does
- Blocks the segment that had been reclaiming the escaped sodium load, so it travels on to the collecting duct and out into the urine.
- So you see
- A brisk diuresis in someone who had stopped responding, because both the loop and its compensating downstream segment are blocked at once.
- And the same mechanism causes
- Two blocked segments means far more sodium reaching the collecting duct, so potassium and hydrogen are traded away faster and harder — profound hypokalaemia and metabolic alkalosis, and a diuresis that can overshoot into hypovolaemia and pre-renal acute kidney injury. Hyponatraemia is a particular thiazide risk because thiazides impair free water excretion — the diluting segment they block is downstream of the loop — while leaving concentrating ability intact. The combination needs close monitoring of electrolytes, weight and renal function, and is not a standing script.
Catches people out: Thiazides tend to raise serum calcium while loop diuretics lower it. Same patient, opposite direction, purely because of which segment is blocked.
Calcium tells you which segment a diuretic hit. In the loop, calcium crosses between the cells, dragged by the lumen-positive voltage that potassium recycling creates — block NKCC2 and that voltage goes, so calcium is lost in the urine (historically the basis for using a loop diuretic in hypercalcaemia, though rehydration and a bisphosphonate have replaced it). In the distal convoluted tubule calcium is reabsorbed through the cell, and blocking NCC promotes that reabsorption — which is why thiazides raise serum calcium, lower urinary calcium, and are used to reduce recurrence of calcium stones. Know where the calcium crosses and you can work out the direction rather than memorising it.
Now test whether it stuck
Reading this through is not the same as being able to reconstruct it. Every question in the bank is free, with a full debrief on each option.