Tubular Transport, Segment by Segment

How can blocking a sodium transporter change calcium movement through a different pathway? You can answer by following the entry proteins along the nephron, then testing the voltages they help generate. A urine sample gives the combined result; a segment experiment separates the steps.

One pump supports different apical doors

Keep the nephron map, countercurrent mechanism and aquaporin-2 background from the kidney lesson. Here, apical means facing tubular fluid; basolateral means facing the surrounding tissue. The basolateral Na+-K+-ATPase exports sodium and maintains the inward sodium gradient used by apical entry proteins. Its membrane energetics belong to membrane transport.

Human review estimates assign roughly 60–70% of filtered sodium reabsorption to the proximal tubule, 20–25% to the thick ascending limb, about 5–7% to the distal convoluted tubule and 2–5% to the collecting duct. These are approximate ranges from different accounts, not four measurements whose endpoints should sum to 100%. The apical entry routes are proximal NHE3 (sodium-hydrogen exchanger 3) and SGLT2 (sodium-glucose cotransporter 2), thick-limb NKCC2 (sodium-potassium-chloride cotransporter 2), distal NCC (sodium-chloride cotransporter) and principal-cell ENaC (epithelial sodium channel). SGLT2 carries only about 5% of total renal NaCl reabsorption in non-diabetic conditions, so it matters far more for glucose than for sodium.

Transcellular transport crosses a cell’s apical and basolateral membranes. Paracellular transport passes between adjacent cells; epithelia explains the junctions that select this route. Aquaporins are water channels, not sodium carriers: you must specify water permeability separately from sodium entry.

Glucose saturation produces a rounded transition

Transport maximum (Tm) is the maximal reabsorption rate for a substance under specified conditions. Glucose uptake rises approximately with filtered load before carrier saturation limits it. Threshold is the plasma glucose concentration where glucose first appears in urine. Splay is the rounding of the reabsorption and excretion curves near saturation. These describe different features of a titration curve.

Worked problem: follow the glucose mass

Use round illustrative values for an adult human at rest, not measurements from a paper. The glomerular filtration rate (GFR) is the volume filtered per unit time. Assume steady state at each glucose level, freely filtered glucose, no glucose secretion, constant GFR of 125 mL per min and Tm of 350 mg per min. First assume sharp saturation without splay. Take GFR as given; its measurement belongs to glomerular filtration.

Filtered glucose load = GFR x plasma glucose concentration

Reabsorption = the smaller of filtered load and Tm; excretion = filtered load − reabsorption

At plasma glucose 100 mg/dL, divide by 100 mL per dL to obtain 1.0 mg/mL. The filtered load is 125 mL/min x 1.0 mg/mL = 125 mg/min. This is below 350 mg/min, so reabsorption is 125 mg/min and excretion is 0 mg/min.

At the model’s threshold, filtered load equals Tm:

No-splay threshold = Tm / GFR = 350 mg/min / 125 mL/min = 2.8 mg/mL = 280 mg/dL

At 400 mg/dL, or 4.0 mg/mL, filtered load is 500 mg/min. Reabsorption is capped at 350 mg/min, leaving 150 mg/min for excretion. If those conditions persist for 60 min, excretion totals 9,000 mg, or 9.0 g. Check conservation: 350 mg/min returned plus 150 mg/min excreted equals 500 mg/min filtered.

Now change only GFR to 60 mL/min. The no-splay threshold becomes 350 mg/min / 60 mL/min = 5.83 mg/mL, approximately 580 mg/dL. At 400 mg/dL, filtration is only 240 mg/min, so this model excretes no glucose. Lower filtration raises the threshold when capacity stays fixed.

Real healthy adult human curves show low-level glucose spilling around 180–200 mg/dL, roughly 80–100 mg/dL below our illustrative 280 mg/dL. Individual nephrons differ in capacity relative to their filtered loads, and finite carrier affinity rounds uptake near saturation. Tm/GFR therefore gives an idealized upper estimate of threshold, not the measured onset. With GFR 180 L/day and plasma glucose 90 mg/dL, the illustrative daily filtered load is 162 g/day; that large throughput can coexist with very little urinary glucose.

You can compare the sharp model with an illustrative splay curve by varying GFR and Tm independently; the rounded curve represents a teaching model rather than a fit to experimental data.

• filtered load
• reabsorption
• excretion

Plasma glucose 400 mg/dL GFR 125 mL/min Tm for glucose 350 mg/min

Carrier model

Sharp saturation (no splay) Splay (illustrative rounded curve)

Reset to the lesson’s worked-problem values

Filtered load

Reabsorption

Excretion

Threshold

In this sketch: steady state at each glucose level, adult human at rest, glucose freely filtered and not secreted, GFR constant. The starting values (GFR 125 mL/min, Tm 350 mg/min) are the lesson’s round illustrative values, not measurements. The slider ranges, the marker’s definition and the rounded curve’s half-saturation constant (100 mg/min) are made up for the sketch; the rounded curve (reabsorption = Tm × load ÷ (100 + load), never more than the load) is an illustrative shape, not a fit to data. The corner where the rounded curve leaves the load line is an artifact of this sketch’s formula.

In the sharp model, glucose appears in urine once the load passes what the carriers can take back. Rule to keep: filtered load = GFR × plasma glucose; the marker sits at Tm ÷ GFR, so in this sketch it moves right when GFR falls or Tm rises and left when GFR rises or Tm falls. That figure is an idealized upper estimate, not the measured onset; the rounded curve leaves zero earlier, as real curves do.

Proximal uptake couples sodium to glucose and bicarbonate

An isolated perfused tubule is a dissected segment held between glass pipettes with its lumen and surrounding bath controlled separately; Burg, Grantham, Abramow and Orloff described the rabbit method in their 1966 paper. Kokko, Burg and Orloff’s 1971 rabbit proximal-tubule study reported salt and water absorption without a change in luminal sodium concentration: nearly isosmotic reabsorption, meaning removal of fluid with approximately unchanged total solute concentration. With poorly reabsorbed raffinose present, luminal sodium instead settled 33–35 mEq/L below the bath; reported rabbit transport rates were only one-half to one-third of rat rates measured in vivo.

SGLT2 is the early proximal sodium-glucose cotransporter, localized to the mouse brush border, the microvillus-rich apical membrane. NHE3 exchanges luminal sodium for intracellular hydrogen ions; a human-and-rodent review credits NHE3, directly and indirectly, with more than half of filtered sodium reabsorption. Mice lacking NHE3 have sharply reduced proximal bicarbonate and fluid absorption (in situ microperfusion found bicarbonate and fluid absorption reduced by 54% and 63%), evidence that glucose cotransport cannot account for bulk proximal uptake alone.

For bicarbonate reclamation, secreted hydrogen ions combine with filtered bicarbonate, and luminal carbonic anhydrase facilitates formation of carbon dioxide. Carbon dioxide enters the cell, where cytoplasmic carbonic anhydrase supports bicarbonate formation; bicarbonate then exits basolaterally while hydrogen ions can be secreted again. This recovers filtered bicarbonate rather than generating net new bicarbonate. The rest belongs to acid-base balance.

Vallon and colleagues’ 2011 Sglt2 knockout experiment combined whole-kidney measurements with free-flow micropuncture, sampling fluid from accessible proximal sites in anesthetized mice. If SGLT2 supplies early glucose uptake, deletion should leave glucose in early tubular fluid. Early collections showed 78% of filtered glucose already reabsorbed in wild type versus none in knockouts; late collections showed 93% versus 21%. The knockout mice also excreted more glucose and water without a difference in plasma glucose or GFR.

That locates the missing function early and leaves residual downstream uptake consistent with SGLT1. Unchanged plasma glucose and GFR argue against increased filtration as the explanation for glucose loss. The experiment does not directly measure a sodium share, establish human percentages or reproduce acute block: lifelong deletion permits compensation.

In a glucose titration study of 12 healthy adults, investigators estimated the threshold from a model at 171 ± 57 mg/dL (mean ± SD, one outlier excluded). After 7 days of an SGLT2 inhibitor, Tm fell by about 55%, but the estimated threshold fell much further, to about 37 mg/dL, below the lowest glucose level studied. The authors concluded that the lower Tm and the changed splay could not explain the glucose that appeared in urine at ordinary plasma levels; the lowered threshold did. Such stepped glucose clamps hold successive plasma glucose levels approximately steady while renal glucose handling is measured, testing the human curve rather than importing mouse fractions.

Separately, human-and-animal reviews link SGLT2 block to greater NaCl delivery at the macula densa and feedback-mediated reduction of single-nephron GFR, through the loop covered in glomerular filtration.

NKCC2 and ROMK make salt uptake an electrical circuit

Rocha and Kokko’s 1973 isolated rabbit medullary thick-limb study reported no osmotic water flow under imposed gradients. At slow perfusion, outgoing sodium and chloride concentrations were about 0.79 and 0.77 of incoming concentrations. The lumen was positive relative to the bath, about 6.7 mV during their stable-flow measurements. Salt left while water stayed.

NKCC2 carries one Na+, one K+ and two Cl- into the cell. ROMK, a potassium channel, returns potassium to the lumen; basolateral chloride channels, including ClC-Kb, allow chloride exit. Together with cation-selective junctions, these pathways support a lumen-positive transepithelial voltage, the electrical potential difference between tubular fluid and surrounding tissue. That voltage favors paracellular calcium and magnesium reabsorption. Read this as a segment-level summary: recent work describes thick-limb cell types that lack apical potassium channels, including the cortical cells ringed by calcium- and magnesium-transporting claudins, a point returned to near the end of this lesson.

Greger and Schlatter’s 1981 rabbit cortical thick-limb experiments reduced short-circuit current from 215 to 133 microA/cm2 by removing luminal potassium. Short-circuit current is the current that would flow with transepithelial voltage held at zero. In these isolated-tubule studies it was an equivalent current, calculated from the measured voltage and resistance rather than clamped directly. Earlier data from their laboratory had suggested one sodium and two chlorides; these potassium experiments extended that model, tentatively, to a carrier interacting with all three ions. Current alone did not establish the stoichiometry.

A loop diuretic blocks apical NKCC2 from the lumen, reducing salt uptake and the positive voltage. In an intact mammalian kidney, increased sodium delivery downstream can support increased potassium secretion, while loss of the thick-limb voltage reduces calcium and magnesium reabsorption: predict increased urinary sodium, potassium, calcium and magnesium, subject to downstream compensation. Proximal organic anion transporters secrete the blocker into the lumen. Bath exposure and luminal exposure therefore test different steps.

Distal experiments distinguish NCC from ENaC

In 1987, David H. Ellison, Hector Velázquez and Fred S. Wright separately microperfused early and late surface distal segments in anesthetized rats, passing controlled fluid through each segment. If one entry mechanism served both regions, chloride removal and a thiazide-type NCC blocker should affect both. Instead, the blocker reduced sodium and chloride transport to zero early, with no significant effect late; removing chloride reduced sodium transport early but not late. This localized a chloride-dependent, thiazide-sensitive function. It neither identified a protein nor tested deep nephrons or other species.

NCC is the apical Na-Cl cotransporter of the distal convoluted tubule. Gamba and colleagues expression-cloned a thiazide-sensitive transporter in 1993 from winter flounder urinary bladder using Xenopus oocytes. Expression cloning identifies genetic material by the function it produces in a host cell. Their carrier was electroneutral, moving no net charge per cycle, and required no potassium. Mammalian renal NCC identity required evidence beyond that fish preparation.

Loss-of-function variants in the human NCC gene cause Gitelman syndrome, with low plasma potassium and magnesium and low urinary calcium; the authors proposed reduced salt reabsorption as the link. A global NCC knockout mouse study reportedly reproduced low magnesium and low urinary calcium while maintaining blood pressure on a sodium-replete diet. Loss of NKCC2 or ROMK instead disrupts thick-limb voltage, favoring increased urinary calcium.

For sustained NCC block, predict increased sodium loss, a tendency toward potassium loss, reduced urinary calcium and a greater fraction of filtered magnesium escaping reabsorption. The calcium mechanism remains unsettled: mouse micropuncture supports enhanced proximal calcium reabsorption, whereas older rat microperfusion supported a distal effect. NCC block does not remove a thick-limb positive voltage. Its magnesium effect also involves distal transport adaptation, rather than magnesium passage through NCC itself.

ENaC, the epithelial sodium channel, provides principal-cell sodium entry; alpha, beta and gamma subunits were identified using rat cDNA in Xenopus oocytes: alpha by expression cloning, beta and gamma by testing how they contributed to channel function when expressed with alpha. Entry of positive charge makes the lumen negative, favoring potassium secretion through ROMK. Under fixed delivery, ENaC block reduces sodium reabsorption and potassium secretion. ENaC block does not erase the thick-limb voltage, but it does not leave divalent handling untouched either: in rat distal microperfusion, luminal amiloride inhibited sodium reabsorption and increased calcium reabsorption in late distal segments, consistent with apical hyperpolarization favoring calcium entry. Potassium-sparing diuretics also tend to spare magnesium. The size of these effects depends on the preparation and conditions. Reported human gain-of-function variants in the ENaC beta subunit provide a complementary test of excessive sodium entry.

Extra sodium reaching ENaC after NCC loss can increase potassium secretion; during sustained salt loss, greater distal acid secretion can also raise plasma bicarbonate. Regulation belongs to water, sodium and potassium balance and acid-base balance. However, a rodent report summarized in a review found neither increased ENaC-mediated uptake nor increased potassium excretion after acute NCC block. You cannot assume that a chronic deletion and an acute blocker give identical results. Conversely, reported increases in NCC activity through variants in the WNK kinase pathway (WNK1 and WNK4, kinases that regulate NCC) can reduce downstream sodium delivery and favor potassium retention and higher arterial pressure.

Mouse and cell studies show that WNK4, a kinase (an enzyme that adds phosphate groups to proteins), is sensitive to intracellular chloride and that this sensitivity regulates NCC phosphorylation. A recent review’s abstract also suggests three thick-limb cell types besides macula densa cells, only one with apical potassium channels, so the single-cell circuit above remains a segment-level simplification.

Design problem

You have isolated rabbit cortical thick ascending limbs held on glass pipettes. You can control lumen and bath separately, measure voltage and short-circuit current, and collect outgoing fluid. Design tests of whether luminal entry requires potassium and whether a loop diuretic blocking NKCC2 acts from the lumen or the bath. Include recovery, nonspecific controls and limits.

One way to do it

Use paired baseline, test and washout measurements in each tubule. Recovery supports reversible inhibition rather than progressive damage. Measure transepithelial voltage and resistance, report the equivalent short-circuit current calculated from them, and compare incoming and outgoing sodium and chloride fluxes.

Replace luminal potassium with an impermeant cation (one that cannot cross the membrane) while matching osmolality and keeping sodium, chloride and bath composition constant. Predict reduced current, not necessarily zero: the rabbit experiment gave 215 to 133 microA/cm2. Returning potassium should restore transport. Separately block the luminal potassium conductance with barium; reduced recycling should weaken the positive voltage. Residual current does not establish potassium-independent NKCC2 transport.

Compare luminal and bath application of the NKCC2 blocker at matched concentrations in separate runs with washout, reversing run order across tubules. Expect luminal application to suppress salt uptake and positive voltage; bath application should lack that immediate effect if the blocker remains confined to that side. In an intact kidney, proximal organic anion secretion supplies luminal access, which this apparatus bypasses.

Include vehicle controls, cooling and a basolateral Na+-K+-ATPase inhibitor. Cooling and pump inhibition should reduce transport by mechanisms distinct from luminal carrier block; they do not establish the blocker’s binding site. Match solution conditions and check recovery after each reversible intervention.

The conclusions concern rabbit segments in vitro. Voltage and current describe the whole epithelium, not an NKCC2 transport current: NKCC2 itself is electroneutral. Establishing its ion ratio requires ion substitutions and flux measurements, followed by molecular identification; extrapolating rates to rats or humans requires further evidence.