Acid-Base Basics

Summary:

Although students initially learn of ionic buffering in basic chemistry, buffering and acid-base transport in biology often is relegated to specialized classes, discussions, or situations. That said, for physiology, nephrology, pulmonology, and anesthesiology, these basic principles often are critically important for mechanistic understanding, medical treatments, and assessing therapy effectiveness. This short introductory perspective focuses on basic chemistry and transport of buffers and acid-base equivalents, provides an outline of basic science acid-base concepts, tools used to monitor intracellular pH, model cellular responses to pH buffer changes, and the more recent development and use of genetically encoded pH-indicators. Examples of newer genetically encoded pH-indicators (pHerry and pHire) are provided, and their use for in vitro, ex vivo, and in vivo experiments are described. The continued use and development of these basic tools provide increasing opportunities for both basic and potentially clinical investigations.

The hydrogen ion (H⁺) (ie, a proton) is the smallest ion, its control in biological systems is critical for life. Because biologic [H⁺] vary between 10 nmol/L and 10 mmol/L, pH (ie, -log[a_H+], where a_H+ is H⁺ activity) is used for easier reference. Bacteria live and thrive in part because of their ability to maintain a H⁺ gradient across their cell membrane. This is accomplished by using H⁺ pumps (adenosine triphosphatases) to move H⁺ from the intracellular compartment to the outside world, generating the proton motive force (PMF), which has its theoretical framework in Mitchell’s¹ chemiosmotic theory. Bacterial uptake of nutrients frequently is coupled to the proton motive force (PMF) via H⁺-coupled transporters and voltage-sensitive membrane transporters and channels. Mitchell’s¹ chemiosmotic theory, developed to describe membrane permeability to H⁺, is generalized to the Gibbs free energy relationship, also known as the electrochemical potential:

$$\mathrm{\Delta }{\mu }_{\text{ion}}=\text{\hspace{0.17em}RT}\cdot \text{ln}\left\{{[\text{ion}]}_{\text{inside}}/{[\text{ion}]}_{\text{outside}}\right\}+{\text{z}}_{\text{ion}}\text{F}\mathrm{\Delta }\Psi$$ (equation 1)

F or equation 1, R is the gas constant, T is temperature (Kelvin), ln is the natural log, z is the net particle charge, F is the Faraday constant, and ΔΨ = Ψ(inside) – Ψ(outside) (ie, membrane potential [V_m]). This same electrochemical potential (ie, PMF) allows mitochondria and chloroplasts to convert chemical and voltage gradients to usable cellular energy in the form of adenosine triphosphate. For the specific case of no free energy (ie, Δμ_ion ≡ 0), this relationship may be rearranged to the Nernst potential:

$${\text{V}}_{\text{ion}}=-\left\{\text{RT}/{\text{z}}_{\text{ion}}\text{F}\right\}\cdot \text{ln}\left\{{[\text{ion}]}_{\text{inside}}/{[\text{ion}]}_{\text{outside}}\right\}$$ (equation 2)

WHY MEASURE PH OR INTRACELLULAR pH?

Although quantifying pH in cells, tissues, and organisms might appear as merely a cerebral exercise, there are explicit metabolic and physiologic reasons for this attention. Regulation of intracellular and extracellular pH (acid-base transport) maintains a particular H⁺ gradient across cell membranes. Normal cell function is a balance between inward and outward movements of these ions, often varying in response to intracellular pH (pH_i). This is especially true in the central nervous system, digestive tract, heart, respiratory tract, and urinary system.

Many cellular events are pH-sensitive,² and some particularly so. Metabolic enzymes, such as the rate-limiting enzyme in glycolysis and phosphofructokinase,³ and a critical ribosomal protein, S6,⁴ move from being fully active to fully inactive with a pH drift of approximately only 0.1. This of course means that if pH_i is not controlled, both cellular energy metabolism and new protein synthesis will stop. A sufficiently alkaline pH_i is required for proliferation in response to several growth factors.^5–7 With so many key processes being pH_i-sensitive, organisms and cells have evolved acid-base transporters, located in the plasma membrane, to regulate pH_i. Not surprisingly, acid-base transporters are controlled by hormones, growth factors, cell volume, intracellular signaling molecules, and phosphorylation.^8–24 In most cells, the most robust and effective of these acid-base transporters carry HCO₃⁻. A notable exception are cardiac myocytes for which non-HCO₃⁻. transporters control the majority of cellular acid-base flux.^25–27

ACID-BASE TRANSPORTERS

When thinking of pH and acid-base transporters, most scientists focus on H⁺ transport (eg, Na⁺-H⁺ exchangers; NHE, solute leak carrier 9 [SLC9] gene family). Certainly, these H⁺ transporters are very important. However, for most cells, HCO₃⁻ transporters carry more acid-base equivalents and are more active in a CO₂/HCO₃⁻ environment. Molecular information had been limited to the Cl⁻-HCO₃⁻ exchanger (AE1-AE3), despite rich physiological documentation of HCO₃⁻ transporters. Since cloning the salamander Ambystoma’s electrogenic Na⁺/HCO₃⁻ cotransporter (NBCe1/Slc4a4),²⁸ modern molecular biology tools have begun an explosive revisiting of HCO₃⁻ transporter identification, localization, and physiology, and many novel SLC4-HCO₃⁻ transporters^29,30 and SLC26-HCO₃⁻ transporters^31–38 have now been functionally and genomically identified (for more recent reviews, see SLC4 by Romero et al³⁹ and SLC26 by Alper and Sharma³¹).

In addition to the most common acid (H⁺) and base (OH⁻ and HCO₃⁻), there are several other ions and solutes that may accept H⁺ (base) or release H⁺ (acid) (Table 1). These compounds typically are considered buffers, that is, compounds that can either accept a H⁺ or give up a H⁺ to maintain pH. These solutes also are considered weak acids or weak bases. Biological H⁺ acceptors bases include NH₃, HPO₄⁻, lactate⁻, pyruvate⁻, and deprotonated organic acids (eg, nicotinate⁻, butyrate⁻, propionate⁻). From the cellular transport side, these substrates fall within several SLC-families (recent reviews and details of SLC families are available at http://slc.bioparadigms.org/) (Table 1): monocarboxylates, SLC5A8 and SLC5A12 (Na⁺ coupled), SLC16; dicarboxylates and sulfate, SLC13; phosphates, SLC20 and SLC34; and NH₃/NH₄⁺, SLC34. In general these compounds follow the reaction:

Table 1.

SLC Transporter Families That Move Buffer Species

SLC Family	Family Name: Transported Buffer Substrate
SLC1	Glutamate and neutral amino acids113: H+, glutamate−, aspartate−, glutamine+, asparagine+
SLC4	Bicarbonate transporters39: HCO3−, CO32− (borate)
SLC5	Na+ glucose cotransporters114: SLC5 and SLC12: monocarboxylates, short-chain fatty acids, lactate−, pyruvate−, acetoacetate−
SLC9	Na+/H+ exchangers115: H+, NH4+
SLC11	H+-coupled metal ion transporters116: H+
SLC12	Electroneutral cation-coupled Cl cotransporters117: NH4+
SLC13	Na+ sulfate/carboxylate cotransporters118: dicarboxylates (eg, succinate, citrate)
SLC15	H+-coupled oligopeptide cotransporters119: H+, charged peptides, β-lactam antibiotics
SLC16	Monocarboxylate transporters: H+, monocarboxylates
SLC20/SLC34	Na+ phosphate cotransporters120: H2PO4−, HPO42−
SLC21	Organic anion transporters121
SLC22	Organic cation/anion/zwitterion transporters122
SLC26	Multifunctional anion exchangers31: HCO3−, formate−, SO42−
SLC36	H+-coupled amino acid transporters123: H+
SLC38	System A and system N Na+-coupled neutral amino acid transporter family123: H+, charged amino acids
SLC42	Rh ammonium transporters43: NH3, NH4+
SLC46	Folate transporters124: H+
SLC47	Multidrug and toxin extrusion family125: tetraethylammonium

Note. Current SLC tables are available at http://slc.bioparadigms.org.

B⁻+ H⁺ ↔ BH, where the total buffer concentration ([BH]) equals [B⁻] + [H⁺]. The association/dissociation constant (K) is given by a general equation:

$$\begin{gathered} \text{K\hspace{0.17em}}=\left\{\text{products}\right\}/\left\{\text{reactants}\right\}=\left\{\text{p}\right\}/\left\{\text{r}\right\} \\ =\left\{[\text{BH}]\right\}/\left\{\left[{\text{B}}^{-}\right]\cdot \left[{\text{H}}^{+}\right]\right\} \end{gathered}$$ (equation 3)

Similarly, this buffer’s contribution to pH of a solution or cellular compartment is as follows:

$$\text{pH\hspace{0.17em}}={\text{pK}}_{\text{buffer}}+\text{\hspace{0.17em}log}\left\{[\text{BH}]\right\}/\left\{\left[{\text{B}}^{-}\right]\cdot \left[{\text{H}}^{+}\right]\right\}$$ (equation 4)

For CO₂/HCO₃⁻, this equation becomes the Henderson-Haselbach equation:

$$\begin{gathered} \text{pH\hspace{0.17em}}={\text{pK}}_{\text{CO}2}+\text{\hspace{0.17em}log}\left\{\left[{\text{HCO}}_3^{-}\right]/\left[{\text{CO}}_2\right]\right\} \\ =6.1+\text{\hspace{0.17em}log}\left\{\left[{\text{HCO}}_3^{-}\right]/\left(\text{s}\cdot {\text{pCO}}_2\right)\right\} \end{gathered}$$ (equation 5)

where pK_CO2 = 6.1 and s = CO₂ solubility.

Because pH buffers have different pKs (Table 2), rather than having a single small pH range of buffering, a system of buffers broadens the buffering pH range.⁴⁰ This means that total solution or compartment buffering (B_total) is as follows:

$${\text{B}}_{\text{total}}=\mathrm{\Sigma }\left({\text{B}}_1+{\text{B}}_2+{\text{B}}_3+\dots {\text{B}}_{(\text{n}-1)}+{\text{B}}_{\text{n}}\right)$$ (equation 6)

Table 2.

Biologically Important Buffering Reactions and Their pKas

Reaction	pKa
H2O ⇌ OH− + H+	14.0
H2O + H2O ⇌ OH− + H3O+	14.0
CO2 + H2O ⇌ H2CO3 ⇌ H+ + HCO3−	6.1
CO2 + H2O ⇌ H2CO3	3.6
H2CO3 ⇌ H+ + HCO3−	6.3
HCO3− ⇌ H+ + CO32−	10.32
NH3 + H+ ↔ NH4+	9.25
H-lactate (CH3CH(OH)CO2H) ⇌ H+ + CH3CH(OH)CO2	3.86
H-pyruvate (CH3COCOOH) ⇌ H+ + pyruvate− (CH3COCOO−)	2.50
H-butyrate (CH3CH2CH2COOH) ⇌ H+ + butyrate− (CH3CH2CH2COOH−)	4.82
H-propionate (CH3CH2COOH) ⇌ H+ + propionate− (CH3CH2COOH−)	4.88
H-acetate (CH3COOH) ⇌ H+ + acetate− (CH3COOH−)	4.76
H3PO4 ⇌ H2PO4− + H+	2.14
H2PO4− ⇌ HPO42−+H+	7.20
HPO42− ⇌ PO43− + H+	12.37
	6.0–6.5

This relationship of chemical H⁺ buffering also is known as the isohydric principle. For calculating the pH of a compartment, this transforms to:

$$\text{pH\hspace{0.17em}}=\mathrm{\Sigma }\left({\text{pK}}_{\text{B}1}+{\text{pK}}_{\text{B}2}+{\text{pK}}_{\text{B}3}+\dots .{\text{pK}}_{\text{Bn}}\right)+\text{\hspace{0.17em}log}\left\{\mathrm{\Sigma }\left({\text{p}}_1+{\text{p}}_2+{\text{p}}_3+\dots {\text{p}}_{\text{n}}\right)\right\}/\left\{\mathrm{\Sigma }\left({\text{r}}_1+{\text{r}}_2+{\text{r}}_3+\dots {\text{r}}_n\right)\right\}$$ (equation 7)

In other words, the pH and the solution buffering is determined by the collective contribution of all the solution buffers. Practically, the major buffers in a compartment are used to calculate pH. More frequently, the pH and knowing the specific buffers are used to calculate the ionized or total buffer. Importantly, knowing the buffers in a particular compartment and their respective pKas allows one to determine if pH is effectively controlled or if there is disequilibrium in the system.

pH MEASUREMENT

Because pH regulates critical cellular and systemic processes, being able to accurately and precisely measure pH allows a researcher or clinician to determine what components are at work in a given system (ie, subcellular, cellular, blood, interstitial, or systemic). Early measurements of pH relied on distribution of a membrane-permeant molecule, such as NH₃, CO₂, 5,5-dimethyloxazolidine-2,4-dione (DMO), or amines, which as either weak acids or weak bases may accept or lose a H⁺ (for review, see Roos and Boron⁴¹). DeVris⁴² first illustrated this permeant-weak base device by exposing beet slices to NH₃. However, these distributions of permeant molecules are difficult to calibrate to actual pH values.

Perturbing the extracellular environment to elicit a pH_i change is a very useful experimental technique. The most obvious way to change pH_i would be to transport a buffer (eg, NH₄⁺, HCO₃⁻) across the plasma membrane of a cell (Fig. 1 illustrates NH₄⁺ effects, Fig. 2 illustrates HCO₃⁻ effects). Transport of each buffer can be accomplished by one of the previously mentioned SLC transporters (HCO₃⁻ by SLC4³⁹ and SLC26³¹, and NH₃ and NH₄⁺ by SLC^42,43). In the mid-1970s, Boron and De Weer⁴⁴ described pH_i changes in squid giant axon resulting from the presence of NH₃ or CO₂ in their extracellular solutions.

Figure 1.

Model intracellular pH (pH_i) responses to an ammonium prepulse. (A) Chemical and cellular models illustrating the buffering reaction of ammonium dissociation and reassociation [NH₄⁺ ↔ NH₃ + H⁺] as discussed in the text. Cellular models indicate the cellular chemistry and transport involved at each of the curve phases indicated in panel B. (B) A model experiment measuring pH_i is shown. The red line denotes an acid recovery in which there is either no transporter, an inactive transport, or an inhibited transporter.

Figure 2.

Model pH_i responses form the addition of CO₂/HCO3⁻. (A) Chemical and cellular models showing the buffering reaction of CO₂ hydration and dehydration in the presence of a carbonic anhydrase: [CO₂ + H₂O ↔ HCO₃⁻ + H⁺] as discussed in the text. Cellular models indicate the cellular chemistry and transport involved at each of the curve phases indicated in panel B. (B) A model experiment measuring pH_i is shown with the acute addition of 5% CO₂/33 mmol/L HCO₃⁻ (pH 7.5). The red line denotes an acid recovery in which there is either no transporter, an inactive transport, or an inhibited transporter. Note that in the red trace (a’), pH_i decreases more quickly and to a more acidic pH_i because there is little cellular buffering. Similarly, without cellular HCO₃⁻ or H⁺ transport, there is no pH_i recovery (b’) (ie, alkalinization). Removal of CO₂/HCO₃⁻ returns pH_i to almost the initial pre-CO₂ pH_i.

NH₃ AND NH₄^± TRANSPORT

In the case of NH₃, dissolving NH₄Cl in solution releases NH₄⁺ into solution that is in a steady-state with NH₃ and H⁺ (Fig. 1). Despite a pKa of 9.2 (Table 2), a 10 mmol/L NH₄Cl solution contains approximately 10 μmol/L NH₃. Figure 1 shows a model response of a cell to the addition of NH₄Cl. This addition elicits a two-phase, pH_i response (Fig. 1): phase one is a fast NH₃ permeation of the cell membrane (likely a channel) causing a fast pH_i increase (Fig. 1A–a), and phase two is a slower transporter-mediated acidification (decrease in pH_i; Fig. 1A–b). The initial NH₃ channel^45–50 has not been explicitly determined for every cell type. The concept was developed with the finding by Kikeri et al⁴⁵ that the apical membranes of mammalian thick ascending limb were not able to pass NH₃. Later, Boron’s and Geibel’s laboratories showed that gastric glands and colonic crypts also have virtually no apical NH₃ permeability.^46,51 Later, the molecular path for NH₃ was shown in several proteins: AmtB⁴⁷ and certain aquaporins⁴⁸ and RhCG proteins.^48–50 The most striking example from Khademi et al⁴⁷ shown using a 1.35Å resolution AmtB crystal structure, is that NH₄⁺ is too big for the pore and that only NH₃ can fit. It should be noted that Xenopus oocytes, a commonly used protein expression system, lack proteins that act as NH₃ channels.⁴⁸ Consequently, in these cells, NH₄Cl addition only results in an acidification (phase two).

The second phase is NH₄⁺ transport (Fig. 1A–b). When NH₄⁺ moves into a cell at this increased pH_i (ie, closer to the pKa), NH₄⁺ will dissociate to NH₃ and H⁺ (acidification or decreasing pH_i). In the continued presence of NH₄Cl, this acidification will continue until a new steady-state is reached. Phase one and phase two together often are referred to as an ammonium prepulse.⁴⁴

Removal of NH₄Cl

Once a cell has been loaded with this additional NH₃/NH₄⁺ buffer, acute removal of the NH₄Cl-containing solution results in a reversal of these two phases, albeit with different transport implications. Pha’se one is again the rapid transit of NH₃ (Fig. 1A–c), which results in an increase of intracellular H⁺ (rapid acidification) as NH₄⁺ dissociates. Phase two (compare Fig. 1A–d with Fig. 1A–d’) with NH₄Cl removal is often the most useful from the ion transport perspective. Once the cell has moved to an extreme acidification, the recovery (alkalization) is the composite of all of the acid-extruding systems (H⁺ efflux or HCO₃⁻ influx, see later). In this second phase, replacement of ions, removal of coupled substrates, or addition of inhibitors are used to fingerprint physiologic mechanism of acid extrusion (Fig. 1A–d’). For example, if Na⁺ is replaced by an impermeant cation, the Na⁺/H⁺ exchanger shown likely would stop because more H⁺ could not be brought into the cell easily. Similarly, addition of amiloride or ethyl-isopropyl amiloride would inhibit Na⁺/H⁺ exchangers such that the cell would not recover from the acidification.

CO₂/HCO₃⁻ ADDITION TO MAMMALIAN CELLS

Another experimental maneuver, which shows the magnitudes of acid or base fluxes more appropriately, is the abrupt addition of CO₂/HCO₃⁻ to solution bathing a cell whose pH_i is being monitored. This, of course, is more physiologic because CO₂ is one of the products of cellular respiration. Similar to the NH₄⁺ prepulse, addition of a CO₂/HCO₃⁻ equilibrated solution elicits an initial rapid pH_i change. However, as CO₂ enters the cell, it hydrates to form H₂CO₃, which then quickly dissociates to HCO₃₋ and H⁺ (acidification, acid loading, or base extrusion) (Fig. 2A–a and B–a). In the presence of a carbonic anhydrase, CO₂ and H₂O are bound and converted enzymatically to HCO₃⁻ and H⁺, which typically would increase the rate, but not amount, of acidification.

In the presence of CO₂/HCO₃⁻, transport systems that require HCO₃⁻ become active (Fig. 2A–b and B–b) (eg, a Na⁺ bicarbonate cotransport: NBCe1, SLC4A4). In this case, HCO₃⁻ directly enters the cell as HCO₃⁻, causing an alkalization (increased pH_i; base-loading = acid extrusion). If we assume that this is a 5% CO₂ solution at room temperature (25 mmol/L HCO₃⁻) (Fig. 2B), [HCO₃⁻] at the a-b (pH 7.0; ~10 mmol/L) and b-c (pH 7.4; ~26 mmol/L) junctions can be calculated by rearrangement of the Henderson-Haselbach equation. This means that the HCO₃⁻ loading (intracellular buffering) in Figure 2B (ie, Δ[HCO₃⁻]) is 16 mmol/L. If there are no HCO₃⁻ transporters to allow the HCO₃⁻ ion to enter the cell, then the chemistry follows the scheme illustrated in Figure 2A–a’ and A–b’ (Fig. 2B, red lines). The initial decrease in pH_i is much larger (0.4 pH units), and b’ has a slope of zero. The steady-state pH_i of 6.8 means that intracellular [HCO₃⁻] is 6.5 mmol/L rather than 10 mmol/L. The HCO₃⁻ loading in Figure 2B–b versus Figure 2B–b’ shows that this active base-loading system increases [HCO₃⁻] at b-c versus b’-c’ by more than 16 mmol/L.

When the CO₂/HCO₃⁻ is removed from the solution, any HCO₃⁻ formed by CO₂ hydration or transported into the cell will be reunited with H⁺ to form H₂O and CO₂. The CO₂ then quickly exits the cell (Fig. 2A–c, A–c’ and B–c and B–c’). Once again, the wave forms differ owing to the presence of one or more HCO₃⁻ transporters, which may not completely reverse on the same short time scale.

Cellular and subcellular pH also has been measured using a variety of pH buffers that take advantage of color changes, absorbance, or fluorescence. Classic pH measurement techniques have been reviewed previously.⁴¹ The gold standard for pH measurement (solution pH or pHi) is a pH electrode because this measurement technique shows high sensitivity over greater than 6 decades of [H⁺].

pH ELECTRODES

Electrodes to measure pH in biological solutions fall into three broad classes: blackened Pt wire, pH-sensitive glass, or a resin-encapsulated protonophore in a micropipette. The basic principle is that for every 10-fold change in [H⁺], pH unit, the voltage measured by the electrode changes approximately 60 mV. This voltage change per pH unit is the Nernst potential for protons (from equation 2):

$$\begin{gathered} {\text{V}}_{\text{H}}=-\left\{\text{RT}/{\text{z}}_{\text{H}}\text{F}\right\}\cdot \text{ln}\left\{{\left[{\text{H}}^{+}\right]}_{\text{inside}}/{\left[{\text{H}}^{+}\right]}_{\text{outside}}\right\} \\ \approx 2.3\cdot {\text{log}}_{10}({\left[{\text{H}}^{+}\right]}_{\text{inside}}/{\left[{\text{H}}^{+}\right]}_{\text{outside}} \end{gathered}$$ (equation 8)

$${\text{V}}_{\text{H}}\approx 2.3\cdot \left({\text{pH}}_{\text{inside}}-{\text{pH}}_{\text{outside}}\right)$$ (equation 9)

Although pH electrodes are relatively easy to calibrate and mV differences are absolute, this measurement technique requires skill in both manufacturing the microelectrode and maneuvering the electrode into the cell of interest. This is moderately easy for a cell such as a barnacle muscle or squid axon,^52–55 or Xenopus oocyte.^28,56–60 Nevertheless, vertebrate epithelial cells^61,62 and neurons^63–70 require special instrumentation and equipment similar to perfused tubule experiments. Extracellular pH measurements also may use colorimetric indicators (Table 2), pH indicator dyes (Table 3), microelectrodes,^71,72 or vibrating microelectrodes.^73–79

Table 3.

Colorimetric pH Indicators

Name	Acid Color	pH Range of Color Change	Base Color
Alizarin yellow R	Yellow	10.1–12.0	Red
Thymolphthalein	Colorless	9.4–10.6	Blue
Phenolphthalein	Colorless	8.2–10.0	Pink
Thymol blue	Yellow	8.0–9.6	Blue
Bromothymol blue	Yellow	6.0–7.6	Yellow
Litmus	Red	5.0–8.0	Red
Methyl red	Red	4.8–6.0	Yellow
Bromocresol green	Yellow	3.8–5.4	Blue
Methyl orange	Red	3.2–4.4	Yellow
Thymol blue (#2)	Red	1.2–2.8	Yellow
Methyl violet	Yellow	0.0–1.6	Blue

pH DYES

As indicated earlier, there are multiple compounds that can affect pH in solution. To be an effective dye that responds to pH, a compound must fundamentally be a pH buffer. However, these dye buffers have the unique properties that protonation or deprotonation of the compound results in some spectral shift. The earliest of these types of pH indicators used were those that changed visible color over a defined pH range (ie, colorimetric pH indications) (Table 3).

There also are absorbance dyes. These dyes also are pH buffers that rather than changing color, change intensity (absorbance) at defined light wavelengths.^80–82 Soon thereafter, fluorescent dyes (Table 4) such as 2’,7’-bis-(2-carboxyethyl)-fluorescein (BCECF) became the intracellular pH dye of choice^83–85 owing to the ease of loading small mammalian cells and more quantitative measurements between preparations enabled by the ratio of a pH-sensitive emission to the pH-insensitive emission. This later property allows relative, and calibrated, intracellular pH to be compared across preparations regardless of dye loading efficiency. Use of the acetoxy-methyl ester of BCECF typically allows very efficient dye uptake at room temperature and 37°C. Once the acetoxy-methyl ester of BCECF is transported into cells (apparently via carboxylate or organic anion transporters), the acetoxy-methyl ester is cleaved by cellular esterases to trap BCECF in the cell. Nevertheless, no method is perfect, and with BCECF measurements it is critical to monitor both emission wavelengths to ensure that cells are healthy.⁸⁶

Table 4.

Indicator Dyes

pH range	pKa	Backbone Fluorophore	Ex, nm	Em, nm
6.0–8.0	7.5	SNARF	488	580 (iso) 640 (pH)
7.0–8.0	7.3	HPTS (pyranine)	410 (iso) 460 (pH)	511
6.5–7.5	6.98	BCECF	440 (iso) 490 (pH)	535
6.0–7.2	6.5	Fluorescein and carboxyfluorescein	492	514
4.5–6.0	5.2	LysoSensor green DND-189	443	505
4.2–5.7	4.7	Oregon green dyes	496	524
3.5–6.0	4.2	LysoSensor Yellow/blue DND-160	329	440
4–9*	6.8	pHrodo Red106	566	590

^*pHrodo succinimidyl ester shows a complex pH titration profile. Decreasing pH (from pH 9 to pH 2) produces a continuous (but nonlinear) fluorescence increase. This pH response profile typically changes upon conjugation of the dye to proteins and other biomolecules.

Abbreviations: BCECF, 2′,7′-Bis-(2-Carboxyethyl)-5-(and-6)-Carboxyfluorescein; HPTS, 8-hydroxypyrene-1,3,6-trisulfonic acid; iso, isobestic pH; pH, wavelength most sensitive to pH changes; SNARF, seminaphtorhadafluor.

GENETICALLY ENCODED pH SENSORS

With the discovery of green fluorescent protein (GFP) and other naturally fluorescent proteins (FPs), investigators have explored more of the nuanced chemistry of these FPs. Notably, GFP fluorescence intensity has an endogenous pH dependence.⁸⁷ This property was noted soon after the discovery of GFPs in the 1960s and later was exploited by Shimomura et al as a genetically encoded ion sensor recognized by the 2008 Chemistry Nobel Prize (https://www.nature.com/news/2008/081008/full/news.2008.1159.html). However, through the work of Nobel Laurette Roger Tsien, PhD (http://www.tsienlab.ucsd.edu/), and his laboratory, GFP was mutated and altered to produce the spectrum of living colors now available (http://www.clontech.com/US/Products/Fluorescent_Proteins_and_Reporters/Fluorescent_Proteins/Fluorescent_Proteins_Selection_Tool).

In 1999, Verkman and colleagues⁸⁸ reported variants of enhanced yellow fluorescent protein (eYFP) that change fluorescent intensity with halide (I⁻ and Cl⁻) concentration. This eYFP still remained pH sensitive, but the group was able to use eYFP-stably transfected Fischer rat thyroid (FRT) epithelial cells to search for cystic fibrosis transmembrane conductance regulator inhibitors and activators.^89,90 Several modified versions of GFP, including pHluorin⁹¹ and super ecliptic pHluorin (SEpH),⁹² have been used as genetically encoded pH indicators (GEpHIs). In contrast to pH-sensitive dye, these GEpHIs may be easily modified with targeting sequences so that pH of membrane-bound cellular compartments can be measured.^93,94

The next step in general fluorescent protein evolution was to increase brightness further. GFP and many of the other initially used fluorescent proteins tended to aggregate within cells. For tracking proteins this characteristic is moderately annoying, but for fluorescent sensor proteins (GEpHI) this tendency creates an experimental shortcoming. In particular, the aggregating GEpHIs are no longer sampling just the membrane compartment but also have their excitation and/or emission affected by the protein aggregation. The newer, fruit-named proteins (eg, mApple, mCherry, mNectarine) are engineered so that they are monomeric rather than dimeric in nature.^95–98

Recently, GEpHI sensors have moved to mimic the ratiometric pH dyes such as BCECF. Rossano et al^99,100 developed pHerry, which is a tandem dimer of superecliptic pHlorin tethered to mCherry. Rather than being a dual excitation probe as BCECF, pHerry is a dual excitation with dual-emission sensor (Figs. 3–5; discussed later). Similar to BCECF, Zagaynova et al¹⁰¹ developed a dual-excitation pH_i indicator named SypHer2. Finally, Dendra2 is a Kaede-like, monomeric, GFP-like protein which is a photoconvertible fluorescent protein (changing from green to red emission).¹⁰² Because pHerry has been used for both in vitro and in vivo applications,^{99,100,103,104} we focus on this GEpHI. pHerry initially was developed as a probe to accurately and quickly measure pH_i in Drosophila nerve terminals.^103,105

Figure 3.

Intracellular pH (pH_i) response of pHerry with NH₄Cl pulse in renal epithelia. pHerry is a genetically encoded and ratiometric pH sensor expressed in anterior Malpighian tubules (MTs) of Drosophila.¹⁰⁵ (A) Fluorescent images of pHerry (super ecliptic pHluorin [SEpH] [470/510 nm Ex/Em] and mCherry] 556/630 nm ex/em]) of UAS-pHerry driven by the capaR-GAL4 (principle cells of MT) in healthy anterior MTs. The region of interest (ROI) is marked. The background (BG) region is indicated. Scale bar: 50 μm. (B) Relative fluorescence changes of pHerry (SEpH and mCherry signals) of pHerry after 20 seconds of 40 mmol/L NH₄Cl. The mCherry signal does not vary, it is stable, yet the SEpH signal increases fluorescence with alkalization (ie, increased pH_i) and decreases fluorescence with NH₄Cl washout (acidification; ie, decreased pH_i). (C) The ratio of fluorescent signals (SEpH/mCherry) is calculated from data in panel B after calibration (30-min incubation in calibration iPBS: 10 μmol/L nigericin, 130 mmol/L K⁺, pH 7.4 and 9.0). (D) Calibration curve of the absolute pHerry ratio (SEpH/mCherry) after setting pH_i during exposure to calibration insect PBS (iPBS) at eight pH values. Gray circles are individual values from 8 preparations, and the black squares and bars are means ± SD. The curve is fit to Boltzmann distribution. (E) Same data as in panel D but normalized so that pH 7.0 has a ratio of 1.0. Reprinted with permission from Rossano and Romero.¹⁰³

Figure 5.

In vivo pHerry fluorescence. The four panels show brightfield, super ecliptic pHluorin (SEpH) fluorescence, mCherry fluorescence, and a merge, respectively, of a living Drosophila. The top panels are a low magnification of the fly abdomen, which shows significant autofluorescence in the green and red channels. The dotted white box (merge panel) shows the Malpighian tubule (renal tubule epithelium, bottom images), which shows specific fluorescence, indicated by the yellow in the merged image. Note that these images were observed with the intact and anesthetized fly.

Figure 3 shows that the UAS-pHerry Drosophila line also may be expressed selectively in renal epithelia (ie, the Malpighian tubules [MT] of Drosophila).¹⁰³

Figure 3A shows the dual emission fluorescence in selected MT regions. The same NH₄⁺ prepulse elaborated in Figure 1 with green and red emissions is shown in Figure 3B, followed by the fluorescent ratio in Figure 3B.

Figure 3D shows the ratio response calibrated between external pH 5.0 to 9.0, while Figure 3E shows the normalized fluorescent response over the same range. These results indicate that pHerry and the UAS-pHerry fly is generally useful as an experimental tool to quantitatively follow up pH_i in animal tissues.

To explore this further, UAS-pHerry flies were used with different MT promotors (Fig. 3A). These ex vivo experiments (dissected MTs) using a principle cell driver (capaR-GAL4) or a stellate cell driver (c724-GAL4), make it clear that the same NH₄⁺ prepulse (Fig. 1B) results in cell-type–specific pH_i changes. The acid recovery phase (illustrated in Fig. 1B–d and B–d’) then may be used to quantify the acid-extrusion rate (Fig. 4C) and rate per tubule area (Fig. 4D). This quantification makes it clear that even adjacent cells in an isolated epithelial tube may have quite different transport and especially acid-base transport properties.

Figure 4.

Acid flux determined from pHerry responses to NH₄Cl pulse. By using pHerry, its calibration, and the rates of recovery in selected regions, a quantification of acid flux may be calculated.¹⁰³ (A) pHerry fluorescence ratio in anterior MTs: principal cells (left, driven by capaR-GAL4) and stellate cells (right, driven by c724-GAL4). Depending on MT location, stellate cells have different morphologies: cells in initial and transitional segments are bar-shaped and cells in the main segment have cellular projections. Scale bar = 100 μm. (B) pH_i changes in response to 20 seconds of 40 mmol/L NH₄Cl (in specific regions of A) are calibrated. Single exponential fits are shown as dashed curves in the acid recovery phase (withdrawal of NH₄Cl solution). The numeric fit allows a decay constant (τ) value to be derived. (C) J_H+ (acid extrusion rate or H⁺ flux) can be plotted against the calculated pH_i. (D) J_H+ (H⁺ flux) then may be transformed as a flux per unit area. Reprinted with permission from Rossano and Romero.¹⁰³

Figure 5 illustrates that the utility of these GEpHI (ie, pHerry) are not limited to in vitro/ex vivo experiments. Under the proper conditions, pHerry may be used for in vivo pH_i imaging (Fig. 5D and H).

Even though ratiometric dyes and sensors inherently increase the fluorescent signal-to-noise ratio, they are not without experimental shortcomings. For example, in many cell and tissue types, intracellular [Ca²⁺] ([Ca²⁺]_i) as well as pH_i, are both intracellular signals triggering cellular responses. In several cases, it would be ideal to track both [Ca²⁺]_i and pH_i; however, the optimal Ca²⁺ sensors also are based on green fluorescence emission (eg, GCaMP5 or GCaMP6). This of course means that the optimal Ca²⁺ sensors overlap with the optimal pH_i sensors. Consequently, there has been some additional effort to develop red-shifted pH_i-sensitive fluorescent dyes (eg, pHrodo Red¹⁰⁶) and red-shifted GEpHI (eg, pH_ire^107,108).

Although there have been early versions of red-shifted GEpHIs (eg, pHuji¹⁰⁹), nevertheless, the fluorescent yield is only a fraction of that measured with the green-emission, super ecliptic pHluorin.⁹² The RFP-based pHire has the significant advantage that the fluorescent yield is similar to that of super ecliptic pHluorin.¹⁰⁷ This enhanced fluorescence works well in transfected mammalian cells.^107,108 and easily can be used in conjunction with spectrally distinct genetically encoded sensors or dyes. Figure 6 illustrates one such experiment with mammalian cells transfected with pH_ire (Fig. 6A) and voltage sensitive fluorescent protein (VSFP) blue (Fig. 6B, membrane potential^110–112). The experiment in Figure 6C is the same CO₂/HCO₃⁻ protocol as detailed in Figure 2. TM5 cells are changed from a HEPES-buffered solution to 5% CO₂/25 mmol/L HCO₃⁻(pH 7.4), which elicits acidification (red, Fig. 6C–a) and depolarization (blue, Fig. 6C–a’). To test for the presence of a Na⁺ bicarbonate cotransporter,^28,61 Na⁺ is replaced (0 Na⁺) in the continued presence of CO₂/HCO₃⁻. This change further acidifies (red, Fig. 6C–b) and hyperpolarizes the cell (blue, Fig. 6C–b’). This particular result indicates that either an electroneutral Na⁺ bicarbonate cotransporter or is a Na⁺/H⁺ exchanger is present in the cell. Demonstrating either HCO₃⁻ dependence or inhibition (amiloride for Na⁺/H⁺ exchanger or a stilbene for a Na⁺ bicarbonate cotransporter) would allow this diagnosis.

Figure 6.

Genetically encoded pH sensors in mammalian cells. The two trace lines (blue and red) illustrate relative fluorescent responses of TM5 (normal human trabecular meshwork) cells transfected with two genetically encoded sensors. Blue is VSFP blue (lower inset) and tracks membrane potential.^110–112 Red is pH_ire (upper inset) and tracks pH_i.¹⁰⁷ he TM5 cells on a glass coverslip were exposed to a 5% CO₂/25 mmol/L HCO₃⁻ (pH 7.4 at room temperature), followed by Na⁺ removal (0 Na⁺, replacement by choline) in the continued presence of 5% CO₂/25 mmol/L HCO₃⁻. This maneuver is designed to test for the presence of a Na⁺ bicarbonate cotransporter,^28,61 but also could indicate a Na⁺/H⁺ exchanger if HCO₃⁻ is not required. The callout boxes indicate the movement of ions or charge, which in turn elicit the fluorescent changes.

As a final note, genetically encoded sensors and dyes are tools that do not need to be used in isolation. Over the past several decades, investigators studying Ca²⁺ signaling have used Ca²⁺ dyes in combination with electrophysiology. Although this is a bit more unusual when studying pH_i, the mixing of techniques allows experimental validation as well as additional parameter evaluation while optimizing signal-to-noise ratios for the combined approaches.

PERSPECTIVES

Manipulating buffer species or making use of buffers with optical changes in response to pH_i changes allows investigators to interrogate the intracellular environment. Coupling these special buffers (dyes and GEpHI) to ion replacement experiments ± inhibitors or ± other sensors, can be used to diagnose which membrane transport proteins, channels, or pumps are involved in cellular pH_i control. Currently, the only experimental limitations are how to best couple multiple experimental tools to study multiple cells simultaneously or how to best use and develop tools for in vivo assessment.