Hydrogen Ion Homeostasis and Physiological Buffer Systems

Hydrogen ion homeostasis connects chemical equilibrium with transport, metabolism, and excretion. A buffer limits the immediate change in free hydrogen ion activity, while lungs and kidneys alter the conditions under which buffering occurs. These functions cooperate, but they are not interchangeable: pH can improve while an acid burden remains, and acid can be retained without a proportionate change in one circulating measurement. This chapter establishes those foundations before the interpretive frameworks and individual disorders developed in Chapters 53–57.

52.1 pH Describes Hydrogen Ion Activity Rather Than Acid Inventory

Hydrogen ions influence protein charge, enzyme reactions, channel behavior, and molecular binding. Their relevant chemical tendency is described by activity, which accounts for interactions in solution rather than simply counting ions per liter. The conventional definition is:

a_H is hydrogen ion activity expressed relative to its standard state, making the logarithm's argument dimensionless. Practical pH measurement uses calibrated electrodes and measurement conventions; it is not a direct count of all protons present in a specimen. The International Union of Pure and Applied Chemistry (IUPAC) definition and measurement recommendations make this activity basis explicit. 1

For approximate physiological teaching, activity can be represented as an equivalent concentration. Ignoring the activity-coefficient distinction, hydrogen ion concentration in nmol/L is 10^(9 − pH). At pH 7.40 this is 10^1.60 ≈ 39.8 nmol/L; at pH 7.10 it is 10^1.90 ≈ 79.4 nmol/L. A fall of 0.30 therefore corresponds to approximately doubling free hydrogen ion activity. Equal numerical pH changes do not represent equal absolute concentration changes across the scale.

The free pool is very small compared with proton equivalents bound to buffers. Suppose an illustrative extracellular compartment contains 14 L at an approximate free hydrogen ion concentration of 40 nmol/L. Its free amount is 14 × 40 = 560 nmol, or 0.00056 mmol. This is not the amount of acid that would be needed to change the patient's pH, because buffers and compartment exchanges dominate that response.

Likewise, a low pH does not identify the total quantity of accumulated acid, its source, or its distribution. Different carbon dioxide tensions and buffer states can produce the same pH. Temperature, sample conditions, and the compartment measured also matter. An arterial sample describes arterial blood under the reporting conditions, not a direct average of all intracellular environments.

52.2 Weak Buffers Exchange Protons Reversibly

A weak acid can donate a proton, leaving a conjugate base capable of accepting one. For a simplified monoprotic pair:

HA denotes the protonated form and A⁻ the conjugate base. The double arrow indicates reversible exchange. When strong acid adds hydrogen ions, some A⁻ becomes HA. When base removes hydrogen ions, some HA dissociates to replace them. The free hydrogen ion activity changes less than it would in the absence of that exchange.

The acid dissociation constant describes the equilibrium under specified conditions. Its negative logarithm is pK_a. In an ideal concentration approximation, equal amounts of HA and A⁻ occur at pH equal to pK_a. Above that pH, the deprotonated form predominates; below it, the protonated form predominates. Apparent values in physiological fluids depend on temperature, ionic environment, and the chemical system being represented.

A buffer therefore does not hold pH absolutely constant. It changes the relation between an added amount of acid or base and the resulting free activity. As proton-accepting forms become depleted during acid addition, further additions cause progressively larger pH changes. The reverse applies when proton-donating forms are depleted by base.

Direction and capacity must also be distinguished. A buffer already mostly protonated has less remaining ability to accept additional acid than the same total buffer near equal protonation. A mixture that resists acid may not resist an equally large base addition to the same degree. Several buffers with different dissociation properties broaden the useful range.

Biological buffers exchange protons with one another through the same aqueous environment. They are not independent storage tanks that can each be assigned the entire acid load. Changes in protein protonation, bicarbonate, phosphate, and dissolved carbon dioxide must satisfy the coupled equilibria and conservation relationships together.

This explanation establishes the chemical mechanism. It does not imply that the acid disappears when bound, or that the amount administered to a living patient can be calculated from a single buffer pair. Transport and metabolism change both the available species and the boundaries of the system.

52.3 Buffer Capacity Is Different from the pH a Buffer Supports

Two solutions can have the same pH but resist added acid very differently. The ratio of conjugate forms helps establish pH, whereas their total concentration strongly influences the amount they can exchange. Buffer capacity describes a local slope: the strong base needed per liter for a small rise in pH, or equivalently the strong acid needed for a small fall.

For an ideal closed monoprotic buffer at constant volume, neglecting water's own buffering contribution:

β is buffer capacity in mmol/L per pH unit, C_T is total HA plus A⁻ concentration in mmol/L, and r is the dimensionless ratio A⁻/HA. The expression is a differential approximation, not an amount required to move a patient across a large pH interval.

At equal protonation, r is one. For illustrative C_T of 20 mmol/L, capacity is 2.303 × 20 ÷ 4 ≈ 11.5 mmol/L per pH unit. At the same pH and ratio, C_T of 5 mmol/L gives 2.303 × 5 ÷ 4 ≈ 2.9 mmol/L per pH unit. The pH is identical while this component of capacity is four times greater in the more concentrated solution.

Capacity changes as pH changes because r changes. Dilution, multiple dissociating groups, ionic interactions, and changing volume complicate the simple expression. Most importantly, a system allowed to exchange carbon dioxide with its surroundings behaves differently from the same liquid sealed in a container.

Watanabe and colleagues examined human whole-blood titration under defined open carbon dioxide conditions. Their experimental and mathematical treatment explicitly accounted for volume and bicarbonate changes, illustrating why the conditions of titration determine the measured buffer value. Such measurements are not one universal constant for all circulating blood or tissue states. 2

The bedside implication is interpretive: similar pH values can coexist with different buffering reserves and very different future trajectories. It does not follow that multiplying β by a desired pH change yields a safe or accurate replacement amount for the body.

52.4 Bicarbonate Participates in an Open Carbon Dioxide System

Bicarbonate buffering is linked to carbon dioxide through reversible reactions:

Carbonic anhydrase accelerates the interconversion without changing the final equilibrium. In blood and tissues, carbon dioxide can be transported to the lungs and eliminated. This exchange makes the bicarbonate system open: one reaction partner can leave rather than remaining trapped in the same fluid volume.

A commonly used plasma approximation at 37°C expresses the relationship as:

Bicarbonate concentration is in mmol/L and PCO₂ is carbon dioxide partial pressure in mmHg. The approximate solubility coefficient is 0.03 mmol/L/mmHg, so the denominator is dissolved carbon dioxide concentration in mmol/L. The apparent pK of 6.1 combines the equilibria represented here; these coefficients depend on conditions rather than being universal constants for every fluid or temperature.

With illustrative bicarbonate of 24 mmol/L and PCO₂ of 40 mmHg, dissolved carbon dioxide is 0.03 × 40 = 1.2 mmol/L. The ratio is 24 ÷ 1.2 = 20, giving 6.1 + log₁₀(20) ≈ 7.40. If bicarbonate is instead 12 mmol/L at the same PCO₂, the ratio is ten and pH is approximately 7.10.

These calculations describe specified equilibrated states, not compensation rules or the amount of acid added. In a living patient bicarbonate and PCO₂ can change together, while nonbicarbonate buffers exchange protons and compartments redistribute carbon dioxide. The framework comparison follows in Chapter 53.

Eliminating additional carbon dioxide can limit the pH fall when bicarbonate accepts protons, but it does not itself restore bicarbonate consumed by a retained nonvolatile acid. Conversely, impaired carbon dioxide elimination can weaken the expected pH benefit of a buffer-related process. Chemical buffering and adequate gas exchange must therefore be connected rather than treated as separate achievements.

52.5 Hemoglobin, Proteins, and Phosphate Share the Buffering Work

Proteins contain groups that accept or release protons as their chemical environment changes. Histidine and other ionizable groups contribute across physiological pH ranges. Hemoglobin is particularly important within erythrocytes because of its concentration and oxygenation-dependent proton binding. Plasma proteins, including albumin, contribute in their own compartment.

In systemic capillaries, carbon dioxide enters erythrocytes and is rapidly converted through carbonic anhydrase reactions. Hemoglobin accepts some of the associated protons. Deoxygenated hemoglobin generally accommodates more protons and carbon dioxide than its oxygenated form, helping blood carry the metabolic carbon dioxide load at a limited rise in its partial pressure.

The classic human-blood experiments of Christiansen, Douglas, and Haldane established the effect of oxygenation on carbon dioxide carriage. The Haldane effect links oxygen unloading with carbon dioxide transport; it does not mean that every change in blood PCO₂ directly measures a change in total carbon dioxide production. 3

Bicarbonate formed within the erythrocyte can exchange outward for chloride, preserving electrical balance while distributing carbon dioxide-derived bicarbonate into plasma. In the lungs, the sequence reverses as carbon dioxide is eliminated and hemoglobin becomes oxygenated. Gunn and colleagues' isotope studies in human erythrocytes supported a shared, carrier-mediated chloride–bicarbonate transport pathway. The experiment characterized membrane exchange, not renal chloride balance. 4

The oxygenation effects on proton binding and carbon dioxide carriage are related to, but distinct from, the Bohr effect of hydrogen ions and carbon dioxide on oxygen affinity. Their coupling helps transport; neither effect creates an unlimited buffer supply or replaces ventilation.

Phosphate supplies another reversible pair, particularly HPO₄²⁻ and H₂PO₄⁻. Its concentration and dissociation properties make its contribution compartment dependent. The circulating inorganic phosphate pool contributes less extracellular buffering than its importance in cells and urine might suggest. Intracellular organic phosphates also participate, but their measured total amount does not mean every phosphate group has the same accessibility or buffering behavior.

52.6 Intracellular and Tissue Buffering Redistribute the Burden

Cells contain proteins, phosphates, and other proton-binding groups, but their pH is not determined solely by passive equilibration with plasma. Membrane transport and metabolism establish compartment-specific conditions. Sodium–hydrogen exchange, bicarbonate transport, and other pathways can move acid equivalents across the membrane while chemical buffers limit the immediate change inside.

Carbon dioxide can redistribute rapidly across many membranes. A change in extracellular PCO₂ can therefore alter intracellular pH even when extracellular pH has been constrained experimentally. The eventual response depends on membrane properties, transport activity, local production, and perfusion, so arterial pH does not certify the pH of every tissue.

Boron and De Weer measured intracellular pH in squid giant axons exposed to carbon dioxide and ammonia. Rapid pH changes were followed by slower regulatory responses. The preparation demonstrated the distinction between membrane entry, buffering, and active regulation; its numerical responses cannot be assigned directly to human myocardium or other ICU tissues. 5

Tissue reservoirs can also buffer over longer periods. Bone mineral contains alkaline components capable of participating in acid–base exchange, while cell and connective-tissue groups can change protonation. These interactions can reduce the change visible in extracellular bicarbonate, but redistribution or tissue neutralization is not the same as excreting the added burden.

Lemann and colleagues performed 18 balance studies in 14 healthy men receiving acid or alkali loads. Positive calculated acid balance could coexist with low but relatively stable circulating carbon dioxide content and pH. The authors proposed tissue buffering or storage; bone was a plausible contributor, not a location directly established by those balance experiments. 6

This observation limits the inference drawn from an apparently stable sample. Stability can reflect balanced production and excretion, but it can also conceal continuing transfer into tissue reservoirs. Conversely, releasing previously buffered equivalents during recovery can change circulating measurements after the initiating process has improved.

Buffering therefore protects the free chemical environment while changing the state of other molecules and compartments. Its effectiveness must not be interpreted as evidence that all cells are unaffected or that cumulative balance has normalized.

52.7 Buffering, Metabolism, and Excretion Have Different Accounting Roles

Proton binding changes chemical form without requiring that acid equivalents leave the body. Moving them into another compartment changes location. Excretion crosses a body boundary, while metabolism can create or consume acid–base equivalents according to the complete reaction and the eventual fate of its products. These processes must be identified separately.

Carbon dioxide production from metabolism is handled mainly through the lungs. Calling it volatile acid is useful shorthand for its acidifying participation through hydration equilibria; the exhaled molecule is carbon dioxide, not a stream of free hydrogen ions. Nonvolatile acid associated with sulfur-containing and other substrates requires different final handling.

Metabolism of an organic anion can contribute an alkalinizing effect under appropriate conditions, but the accompanying cations, prior acid production, and subsequent losses matter. Oxidizing an anion retained with the products of an earlier acid-generating process differs from administering its sodium salt. Excreting a potentially metabolizable anion can remove a source of future base. The label lactate or citrate alone cannot settle the balance.

Loss of bicarbonate from the gastrointestinal tract or urine removes base. It can therefore produce an acidifying consequence without adding the same amount of an external acid molecule. Conversely, some gastrointestinal losses remove acid. External acid–base accounting must include these pathways rather than treating renal excretion as the only possible body boundary.

The conventional balance of nonvolatile acid can be represented conceptually as generation plus acid input minus net acid removal, with base gains and losses entered consistently. A positive retained amount describes a burden over time; it does not specify the resulting pH without buffer, gas, and distribution information.

Critical illness changes several terms together. Nutrition alters substrate delivery; cellular injury changes metabolite release; impaired perfusion changes transport and metabolism; extracorporeal therapy transfers both buffer and associated ions. A normalizing pH can result from improved ventilation even while nonvolatile acid production continues. Likewise, persistent acidemia can remain after one source has resolved if elimination and redistribution lag.

52.8 Bicarbonate Reclamation Is Different from Adding New Base

The kidney filters a large bicarbonate load and normally recovers most of it. Proximal hydrogen secretion, including sodium–hydrogen exchange, reacts with luminal bicarbonate. Carbonic anhydrase facilitates conversion to carbon dioxide, which can enter the cell and participate in re-forming bicarbonate for transport toward blood. The net result is recovery of filtered bicarbonate rather than simple intact apical uptake.

Preisig and colleagues used in vivo rat proximal microperfusion and intracellular pH measurements to show a major role for sodium–hydrogen exchange in bicarbonate absorption, with an additional sodium-independent component. This identifies transport mechanisms without assigning one fixed percentage to every human nephron or disease state. 7

The scale explains why reclamation matters. Assume illustrative glomerular filtration rate (GFR) of 100 mL/min and filterable bicarbonate approximated by plasma concentration of 24 mmol/L. GFR is 0.100 × 1,440 = 144 L/day, and filtered bicarbonate is 144 × 24 = 3,456 mmol/day. This is turnover through filtration, not that amount of newly produced metabolic acid.

When a secreted proton neutralizes filtered bicarbonate, the body avoids losing existing base. Counting every such proton as net acid excretion would grossly overstate new base generation. Net addition requires different accounting: proton equivalents leave with nonbicarbonate buffers, or renal metabolism produces bicarbonate whose paired nitrogen products are appropriately excreted.

Distal intercalated cells contribute regulated acid secretion through proton pumps and related transport pathways, with bicarbonate returned toward blood. Other intercalated-cell transport arrangements support bicarbonate secretion when appropriate. The kidney can therefore conserve base, add base equivalents, or excrete base, depending on the integrated state.

Pitts and colleagues' human experiments showed that excreted titratable acid could exceed the acid delivered in glomerular filtrate, supporting active tubular addition. Their historical model preceded modern molecular transporter identification. The finding does not make all tubular proton secretion equivalent to net acid removal. 8

52.9 Titratable Acid Depends on Urinary Buffers and Their Delivery

Tubular proton secretion faces an increasing gradient as free hydrogen ion activity rises. Urinary buffers accept protons, allowing many more equivalents to leave than could remain freely dissolved at the same pH. Phosphate is an important example: HPO₄²⁻ accepts a proton to become H₂PO₄⁻.

Titratable acidity describes the base required to titrate urine toward a defined reference pH, conventionally related to plasma pH. It depends on the amount and properties of the urinary buffers and their protonation relative to that reference. It is not simply the concentration of free hydrogen ions, nor does it include ammonium adequately in the conventional titration.

Schiess and colleagues varied urinary buffer delivery in human experiments and showed that buffer quantity and acid strength influenced titratable acid excretion. These controlled observations explain why the same urine pH can accompany different acid fluxes. Their specific experimental conditions do not establish a universal pH limit or a clinical buffer-delivery prescription. 9

The free-ion calculation shows the scale. Assume urine pH 5.0, approximate concentration 10⁻⁵ mol/L = 0.01 mmol/L, and volume 2 L/day. Freely dissolved hydrogen ion loss is only 0.01 × 2 = 0.02 mmol/day. Buffered proton equivalents can be orders of magnitude larger. This approximation ignores activity differences but is sufficient to demonstrate why free acidity cannot carry the entire daily burden.

Conventional renal net acid excretion combines the principal measured terms:

Express every term as acid or base equivalents per interval, for example mmol of proton equivalents per day. With illustrative ammonium of 50 mmol/day, titratable acid of 20 mmol/day, and bicarbonate of 2 mmol/day, net excretion is 50 + 20 − 2 = 68 mmol/day.

This renal accounting does not include every gastrointestinal, metabolic, organic-anion, or extracorporeal contribution to whole-body balance. Matched collections and appropriate analytical methods remain necessary. Urine pH identifies free activity; it cannot substitute for these excretion amounts.

52.10 Ammonium Excretion Requires Production, Recycling, and Transport

Renal ammonia handling expands acid-excretory capacity beyond the available filtered titratable buffers. Proximal metabolism of glutamine generates ammonium and bicarbonate equivalents. Ammonium is directed toward urine through coordinated transport, while bicarbonate can return toward blood. The eventual nitrogen destination matters: ammonia returned to the circulation and processed differently by the liver does not have the same whole-body base consequence as urinary excretion.

Total ammonia comprises NH₃ and NH₄⁺. Their equilibrium strongly favors NH₄⁺ at typical physiological pH, but the smaller NH₃ fraction remains important for membrane transfer. Proximal secretion, loop reabsorption, medullary accumulation, and collecting-duct secretion form an integrated pathway rather than a single final trapping step.

In isolated rat thick ascending limbs, Good, Knepper, and Burg demonstrated ammonia absorption that was inhibited by furosemide and could not be explained by simple nonionic diffusion. The findings supported transport of ammonium and a role for the loop in supplying medullary ammonia. The experiment did not directly measure all components of human whole-kidney ammonium excretion. 10

Collecting-duct proton secretion converts delivered NH₃ toward NH₄⁺, favoring retention within acidic tubular fluid. However, transfer is not adequately described as unrestricted lipid diffusion followed by passive trapping. Transport proteins, membrane properties, gradients, and luminal conditions determine the actual flux.

Biver and colleagues showed that mice lacking Rhesus C glycoprotein, or Rhcg, had impaired ammonium excretion during acid loading and reduced NH₃ transport in microperfused collecting ducts. This establishes a regulated transport contribution while preserving the distinction between a knockout preparation and an acquired human acid–base disturbance. 11

Changes in potassium balance, functioning renal mass, tubular metabolism, medullary organization, and hormonal conditions can modify the response. A low urine pH with limited ammonium supply can coexist with inadequate net acid removal. Conversely, substantial ammonium excretion need not require the lowest urine pH encountered. Production, delivery, and transport must be sufficient together.

52.11 Lungs and Kidneys Cooperate Across Different Time Scales

Chemical proton exchange begins rapidly, but whole-body adjustment requires transport between compartments and removal across organ boundaries. Ventilation changes carbon dioxide elimination, while renal transport and metabolism change bicarbonate conservation and net acid excretion. Their time courses overlap rather than forming three perfectly separated stages.

Increased carbon dioxide elimination can raise pH before a nonvolatile burden has been excreted. Reduced elimination can lower pH even while the kidney is conserving bicarbonate. Conversely, persistent renal adaptation can continue after an initiating respiratory or metabolic perturbation changes. A sample taken during transition cannot automatically be interpreted as a steady state.

Sartorius and colleagues studied human responses during ammonium chloride loading and recovery. Ammonia and titratable acid responses began early, while the larger ammonium response developed more slowly and could persist during recovery. The experiments show adaptation and timing; they do not supply one mandatory response interval or compensation slope for every patient. 12

The organs also constrain one another. Effective carbon dioxide elimination requires sufficient alveolar ventilation and delivery of tissue carbon dioxide to exchanging lung units. Increased respiratory drive may be unable to produce the required ventilation when mechanics, fatigue, dead space, or neurological function limit output. A high minute ventilation alone does not certify adequate alveolar elimination.

Renal adaptation requires functioning transport and metabolic capacity, adequate substrate delivery, and suitable tubular conditions. Reduced GFR can coexist with altered ammonium production, impaired distal acidification, or bicarbonate losses. These are different limitations, so one creatinine or urine pH value cannot summarize acid–base reserve.

Circulation links both organs to tissues. Impaired perfusion can alter local production, washout, renal function, and the arterial–tissue relationship. Improvement in arterial pH after a ventilatory change therefore does not establish normalization of tissue metabolism. Serial observations become interpretable when linked to the intervention, compartment, and expected timing of the proposed mechanism.

52.12 Integrate Chemical State with Flux and Cumulative Balance

Consider an illustrative state with bicarbonate 12 mmol/L and arterial PCO₂ 40 mmHg. Using the stated approximation, 6.1 + log₁₀(12 ÷ 1.2) = 7.10. If PCO₂ falls to 30 mmHg while bicarbonate is assumed unchanged, dissolved carbon dioxide becomes 0.03 × 30 = 0.9 mmol/L, and calculated pH is 6.1 + log₁₀(12 ÷ 0.9) ≈ 7.22.

The calculation demonstrates how gas elimination can improve pH without restoring the assumed bicarbonate concentration. It is not a prediction of the complete response, because nonbicarbonate buffers and renal or compartmental changes can alter bicarbonate. Neither PCO₂ value is a ventilator target or a compensation rule.

Now examine a separate illustrative 24-hour balance in that patient. Urinary ammonium excretion is 40 mmol, titratable acid 25 mmol, and bicarbonate loss 5 mmol. Conventional renal net acid excretion is 40 + 25 − 5 = 60 mmol/day. If the net nonvolatile acid load requiring removal is assumed to be 70 mmol/day and other routes are negligible, 70 − 60 = 10 mmol/day remains unremoved.

A rising pH can therefore coexist with continuing positive acid balance. The retained amount cannot be converted directly into a predicted next pH without specifying buffers, distribution, gas exchange, and time. Likewise, a stable bicarbonate concentration would not prove that this simplified balance is zero.

The practical interpretation combines different kinds of evidence. pH describes chemical state; PCO₂ and bicarbonate describe a coupled equilibrium; urine measurements describe renal flux; clinical context identifies production, loss, metabolism, and organ constraints. Sample quality and compartment identity remain necessary before combining them.

This sequence supports a causal explanation without prematurely selecting a treatment. Determine what changed free activity, whether acid or base crossed a boundary, and whether the cumulative burden is improving. The next chapter develops complementary ways to describe that same integrated system, while later chapters apply these foundations to individual disorders.

References

1. Buck RP, Rondinini S, Covington AK, et al. Measurement of pH: definition, standards, and procedures (IUPAC Recommendations 2002). Pure Appl Chem. 2002;74:2169–2200. doi:10.1351/pac200274112169

2. Watanabe K, Miyamoto M, Imai Y. An estimation of buffer values of human whole blood by titration experiment under the open condition for carbon dioxide gas. Jpn J Physiol. 2001;51:671–677. doi:10.2170/jjphysiol.51.671

3. Christiansen J, Douglas CG, Haldane JS. The absorption and dissociation of carbon dioxide by human blood. J Physiol. 1914;48:244–271. doi:10.1113/jphysiol.1914.sp001659

4. Gunn RB, Dalmark M, Tosteson DC, Wieth JO. Characteristics of chloride transport in human red blood cells. J Gen Physiol. 1973;61:185–206. doi:10.1085/jgp.61.2.185

5. Boron WF, De Weer P. Intracellular pH transients in squid giant axons caused by CO₂, NH₃, and metabolic inhibitors. J Gen Physiol. 1976;67:91–112. doi:10.1085/jgp.67.1.91

6. Lemann J Jr, Lennon EJ, Goodman AD, Litzow JR, Relman AS. The net balance of acid in subjects given large loads of acid or alkali. J Clin Invest. 1965;44:507–517. doi:10.1172/JCI105164

7. Preisig PA, Ives HE, Cragoe EJ Jr, Alpern RJ, Rector FC Jr. Role of the Na⁺/H⁺ antiporter in rat proximal tubule bicarbonate absorption. J Clin Invest. 1987;80:970–978. doi:10.1172/JCI113190

8. Pitts RF, Lotspeich WD, Schiess WA, Ayer JL. The renal regulation of acid–base balance in man. I. The nature of the mechanism for acidifying the urine. J Clin Invest. 1948;27:48–56. doi:10.1172/JCI101923

9. Schiess WA, Ayer JL, Lotspeich WD, Pitts RF. The renal regulation of acid–base balance in man. II. Factors affecting the excretion of titratable acid by the normal human subject. J Clin Invest. 1948;27:57–64. doi:10.1172/JCI101924

10. Good DW, Knepper MA, Burg MB. Ammonia and bicarbonate transport by thick ascending limb of rat kidney. Am J Physiol. 1984;247:F35–F44. doi:10.1152/ajprenal.1984.247.1.F35

11. Biver S, Belge H, Bourgeois S, et al. A role for Rhesus factor Rhcg in renal ammonium excretion and male fertility. Nature. 2008;456:339–343. doi:10.1038/nature07518

12. Sartorius OW, Roemmelt JC, Pitts RF, Calhoon D, Miner P. The renal regulation of acid–base balance in man. IV. The nature of the renal compensations in ammonium chloride acidosis. J Clin Invest. 1949;28:423–439. doi:10.1172/JCI102087