The Kidney

An adult human’s kidneys filter about 180 litres of plasma a day, about four times the volume of water in the whole body. Most people pass one or two litres of urine. Where does the rest go?

Back into the blood. The kidney filters far more than it lets go, then takes most of it back. This lesson follows that fluid through one mammalian nephron, then compares kidneys across animals. It builds on osmoregulation.

One nephron, three verbs: filter, reabsorb, secrete

A nephron is the kidney’s working unit: a filter attached to a long tube. An adult human kidney holds roughly a million. Filtration pushes fluid out of the blood into the tube. Reabsorption moves substances from the tube back into the blood. Secretion moves substances from the blood into the tube.

Follow the blood first. The afferent arteriole carries blood into the glomerulus, a knot of capillaries. The efferent arteriole carries it out (afferent in, efferent out). The blood then runs through the peritubular capillaries wrapped around the tube, and, for nephrons that reach deep into the kidney, through long hairpin vessels called the vasa recta.

At the glomerulus, blood pressure pushes plasma through a barrier into Bowman’s capsule, the cup around the capillaries. The pressures involved are the Starling forces from blood and vessels. The fluid that passes is the filtrate: water and small solutes such as sodium, glucose, amino acids and urea, but not blood cells and very little plasma protein. Think of a sieve that keeps the beans and passes the broth.

The filtrate then flows, in order, through the proximal tubule, the loop of Henle (a descending limb, then an ascending limb), the distal tubule and the collecting duct, which many nephrons share. In mammals:

• The proximal tubule reabsorbs about 60 to 70% of filtered sodium, with water following so the tubular fluid stays at nearly the same concentration, and normally takes back almost all the filtered glucose and amino acids. It secretes hydrogen ions, ammonium and some drugs.
• The loop of Henle and distal tubule reabsorb more salt, and the loop builds the gradient described below.
• The collecting duct sets the final water and salt content of urine under hormonal control. Its principal cells reabsorb sodium and secrete potassium; its intercalated cells secrete acid or bicarbonate, part of acid-base balance (gas exchange).

How epithelial cells move ions is in epithelia and membrane transport.

Worked problem: clearance and filtered load

Clearance is the volume of plasma per minute that the kidneys would have to clear completely of a substance to account for its rate in urine. At steady state:

C = U x V / P, where U is urine concentration, V is urine flow and P is plasma concentration

These numbers are illustrative, for an adult human at steady state. Inulin, a plant sugar polymer, is freely filtered and neither reabsorbed nor secreted, so its clearance equals the glomerular filtration rate (GFR), the plasma volume filtered per minute.

1. Urine flow. 120 mL of urine in 60 min gives V = 2.0 mL per min.
2. Clearance. U = 30 mg per mL and P = 0.50 mg per mL. C = 30 x 2.0 / 0.50 = 120 mL per min. Check the units: (mg per mL x mL per min) / (mg per mL) = mL per min. Clearance is a plasma volume. Over a day: 120 x 1,440 = 172,800 mL, about 173 L.
3. Filtered load. Glucose is freely filtered. With plasma glucose at an illustrative 5.0 mmol per L, filtered load = GFR x P = 0.120 L per min x 5.0 mmol per L = 0.60 mmol per min, or 864 mmol per day (about 156 g). With no glucose in the urine, all of it was reabsorbed.
4. Reading other clearances. A substance with a clearance of 300 mL per min leaves faster than filtration alone could carry it, so the tubule must secrete it. For a freely filtered substance, a clearance of 1 mL per min means it is mostly reabsorbed. Creatinine, a muscle waste, is used to estimate GFR, but a little is secreted, so creatinine clearance tends to overestimate GFR.

Substance Inulin Glucose A substance with clearance 300 mL per min A substance with clearance 1 mL per min Urine concentration U 30 mg per mL Urine collected in 60 min 120 mL Plasma concentration P 0.50 mg per mL

Urine flow V

Clearance C

C over a day (1,440 min)

C against GFR (120 mL per min)

In this sketch: adult human at steady state, illustrative numbers, GFR taken as the inulin clearance of 120 mL per min, the test substances freely filtered, and the sketch holds urine glucose at 0 at every plasma glucose. Schematic tube, not to scale; the stack lists segments in the order the filtrate meets them.

Clearance C = U × V / P is a volume of plasma per minute. For a freely filtered substance, C equal to GFR fits no net handling (inulin); C above GFR means net secretion; C below GFR means net reabsorption. Creatinine, a muscle waste, is used to estimate GFR, but a little is secreted, so creatinine clearance tends to overestimate GFR.

Micropuncture: reading the fluid inside a frog nephron

Physiologists long argued over how urine forms. If the glomerulus filters, capsular fluid should resemble plasma without its protein. If tubule cells secrete urine, it should already look like bladder urine.

In 1924 Joseph Wearn and Alfred Newton Richards, at the University of Pennsylvania, sampled the fluid itself. They exposed a frog kidney under a microscope and guided a fine glass pipette into Bowman’s capsule to draw off a tiny sample. This is micropuncture. The capsular fluid held no protein their tests could find, yet it contained sugar and chloride. The frog’s bladder urine contained no sugar.

That ruled in a protein-free filtrate at the glomerulus, followed by reabsorption of glucose somewhere along the tubule, in the frog. It did not say where, it gave no mammalian rates, and frog nephrons lack a loop of Henle. Later micropuncture in rats and other mammals placed glucose reabsorption mainly in the proximal tubule.

Making urine saltier than blood: the loop and the collecting duct

Osmolality counts dissolved particles per kilogram of water (mOsm per kg). Adult human plasma is about 290 mOsm per kg; a water-deprived adult human can make urine near 1,200. That takes tissue saltier than plasma, and a controlled way for water to leave the tubule into it.

The region is the medulla, the inner part of the kidney. In 1951 Wirz, Hargitay and Kuhn measured the freezing point of rat kidney slices and found osmolality rising steadily from the outer cortex to the tip of the papilla, where urine drains out.

The loop of Henle builds that gradient. Picture two lanes of traffic running in opposite directions down and back up a hairpin road.

• The thick ascending limb transports sodium chloride out of the tubule into the surrounding tissue, mainly through a cotransporter called NKCC2, and it is nearly impermeable to water. Salt leaves; water stays. The tissue outside gets saltier and the fluid inside gets more dilute.
• The thin descending limb is permeable to water and much less to salt, at least in the rabbit. Kokko’s 1970 isolated rabbit tubules showed this. As the descending fluid passes through the salty tissue, water leaves it passively, by osmosis, and the fluid concentrates on its way to the bend.

Because the two limbs carry fluid in opposite directions (countercurrent flow), and fluid concentrated on the way down is handed around the bend to the salt-removing ascending limb, a modest salt difference at each level adds up along the loop into a large difference from cortex to papilla. This is the countercurrent multiplier. No pump moves water anywhere in it: only salt is transported, and water follows by osmosis. Not every descending segment is water-permeable, either: in rats some inner-medullary descending segments lack the water channel aquaporin-1.

The vasa recta run in hairpins too. Blood going down gains salt and loses water; blood coming up does the reverse. This countercurrent exchange removes reabsorbed water while limiting washout of the gradient.

The collecting duct uses the gradient. It runs from cortex to papilla through ever saltier tissue. When antidiuretic hormone (ADH), also called vasopressin, is present, its cells insert stored water channels, aquaporin-2, into the membrane facing the urine. Water leaves by osmosis and urine becomes concentrated. Without ADH the duct stays nearly water-tight and dilute urine flows out. In the deepest medulla, urea also leaves the collecting duct through urea transporters, by facilitated diffusion rather than an active pump, and adds to the gradient there.

Hormones that set water and salt

ADH is made in the hypothalamus and released from the posterior pituitary (endocrine system; hypothalamus and autonomic control). Plasma osmolality rising or blood volume falling increases its release. In collecting-duct cells it acts through V2 receptors and cAMP (G proteins, cAMP and PKA). In perfused rat collecting ducts, Nielsen and colleagues (1995) showed that the hormone moves stored channels into the membrane and that the channels leave again when the hormone is removed.

The renin-angiotensin-aldosterone system responds to low blood pressure and low salt. Renin is an enzyme released by juxtaglomerular cells in the wall of the afferent arteriole when renal perfusion pressure falls, when less salt reaches the nearby macula densa of the distal tubule, or when sympathetic nerves stimulate beta-adrenergic receptors. Renin cuts angiotensinogen, a plasma protein made by the liver, into angiotensin I, and angiotensin-converting enzyme turns that into angiotensin II. Angiotensin II raises sodium reabsorption in the tubule and stimulates the adrenal cortex to release aldosterone, a steroid that raises sodium reabsorption through ENaC channels in principal cells and, coupled to it, potassium secretion (nuclear receptors).

Atrial natriuretic peptide (ANP) comes from heart muscle cells in the atria when they are stretched, as after a rise in blood volume. It pushes the other way: more sodium and water in urine. In anesthetized rats it inhibited sodium reabsorption in the inner-medullary collecting duct, and in perfused rat ducts it reduced ADH-stimulated water permeability. It works through cGMP (cGMP signaling).

Kidneys compared: loops, deserts and fish without glomeruli

Loops of Henle occur only in mammals and birds, and only these vertebrates make urine more concentrated than their plasma. Amphibians and most fish have no loop, and their urine is at most about as concentrated as their plasma.

Among mammals, loop length and medullary depth vary. Across many species, Beuchat (1990) found that relative medullary thickness tracks maximum urine concentration. The kangaroo rat Dipodomys merriami, a seed-eater of North American deserts, has long loops and a thick inner medulla and makes urine several times more concentrated than an adult human can (osmoregulation tells how it balances water).

Birds have a mix of loopless and looped nephrons and concentrate urine only modestly, to a small multiple of plasma osmolality; much of their nitrogen leaves as uric acid, which takes little water.

Some marine teleosts, including the goosefish Lophius and the toadfish Opsanus, have kidneys with no glomeruli at all. They form urine by secretion instead of filtration: tubule cells move ions such as magnesium and sulfate into the tubule, and water follows. Plants move water with no filter at all (plant transport).

Design problem

Urea comes from protein breakdown (protein turnover). Fenton and colleagues (2004) made mice lacking the collecting-duct urea transporters UT-A1 and UT-A3. Predict how maximal urine osmolality compares between knockout and normal mice on a 20% protein diet and on a 4% protein diet, after water restriction. Then interpret their results: knockouts concentrated less on 20% protein, the gap almost vanished on 4%, and inner-medullary sodium and chloride did not differ significantly between genotypes.

One way to do it

Predict first. If urea leaving the inner-medullary collecting duct adds to the gradient, knockouts should concentrate less, and the gap should be large on high protein and small on low protein, which supplies little urea in either genotype.

The results fit: urea delivered by facilitated transport adds to the inner-medullary gradient when there is urea to recycle.

The salt result tests a specific model. The passive model of the inner medulla predicts that urea delivered from the collecting duct helps draw salt out of the thin ascending limb, so knockouts should have less inner-medullary NaCl. They did not, which argues against that dependency.

Keep the limits. These are mice, and one knockout does not settle how the inner medulla concentrates; that question is still open.