Ion-selective microelectrodes

• Introduction
• Theory
• Types of microelectrode
• Making microelectrodes
  • Pulling of glass micropipettes
  • Silanization of the inside of surface of the ion-selective electrode or barrel
  • Backfilling
  • Protocol 1
  • Calibration
  • Protocol 2
  • Fault finding and some possible problems
  • Storage of ion-selective microelectrodes
• Intracellular recordings
• References
• Suppliers addresses

Introduction

The word microelectrode has come to mean a glass micropipette which is pulled into a fine tip at one end and filled with an aqueous salt solution. The junction between the salt solution inside the microelectrode and the input to the electrometer amplifier is provided by a half-cell. There are different types of half-cell, but usually the metal contact is AgCl-coated silver wire and the salt solution is 0.1 M KCl. The simplest microelectrodes measure voltage and when inserted into cells measure the membrane potential, in mV, between the inside and outside of the cell. An ion-selective microelectrode contains an ion-selective membrane in the tip of the glass micropipette and is responsive both to the membrane potential and the activity (not concentration) of the ion sensed by the selective membrane.

Ion-selective microelectrodes are used to measure ion gradients across membranes. These measurements can be made outside and inside cells. For example, ion fluxes at the surface of roots can be measured by using either directly ion-selective microelectrodes (1) or by using an ion-selective vibrating probe. Intracellular measurements have been used to give important information on the compartmentation of nutrients, dynamics of cellular ion activities (e.g. in intracellular signalling) and transport mechanisms, particularly the energy gradients for ion transport. To make intracellular measurements it is necessary to also simultaneously measure the membrane potential either by insertion of a second electrode or, for small cells, by combining the ion-selective and voltage-measuring electrodes into a double-barrelled microelectrode (see micrograph below).

	Micrograph of a double-barrelled microelectrode showing the twisted tip. Inset: end view showing the open ends of each barrel. The external diameter of the tip is 1 micrometer

The main criticism of intracellular measurements made with microelectrodes is that they report the ion activity at a single point within the cell. This will result in incomplete information if there are significant ion gradients within the cytoplasm of a single cell as occur in some situations. Overall, the chief advantages of using ion-selective microelectrodes are that:

• They offer a non-destructive method of measuring ions within cells
• They do not change the activity of the ion being measured
• They permit simultaneous measurement of the electrical and chemical gradients across membranes
• They are relatively cheap, when compared to other methods for measuring intracellular ions, and once purchased, the same equipment can be used to measure a range of different ions.

Theory

The theoretical background has already been described by many authors ((2), and references therein) and will only be outlined here. The properties of an ion-selective microelectrode are defined by several characteristics:

• Detection limit
• Selectivity
• Slope
• Response time

The ideal relationship between electrode output (mV) and the activity (a _i ) of the ion of interest (i) is log-linear and is described mathematically by the Nernst equation. Calibration of the electrode against a range of standard solutions should ideally, yield a slope (s) of 59 mV (at 25 C) per decade change in the activity of a monovalent ion. In practice, however, the situation is more complicated than this because no ion-selective electrode has ideal selectivity for one particular ion and under most conditions there is more than one ion present in the sample solution. Hence contributions to the overall electro-motive force (EMF) made by each interfering ion, j, must be taken into account. In this situation, the Nicolsky-Eisenman equation, a modified Nernst equation, describes the EMF:

EMF = E + s ^. log [a_i + K _ij ^pot (a _j) ^{zi / zj}]

where K_ij^pot is the so-called selectivity coefficient of the electrode for the ion i with respect to ion j. This term expresses, on a molar basis, the relative contribution of ions i and j to the measured potential.

The parameters s and K_ij^pot are the two main characteristics defining any type of ion-selective electrode. The slope should be a near ideal Nernstian response when an electrode is calibrated against ion activity, but s is temperature sensitive (see 3.4). The selectivity coefficient measures the preference of the sensor for the detected ion i relative to the interfering ion, j. It can be determined by the separate solution method, the fixed interference method or the fixed primary ion method. For ideally-selective membranes, or for samples containing no other ions with the same net charge as the ion in question, K_ij^pot must be zero. A selectivity coefficient <1 indicates a preference for the measuring ion i relative to the interfering ion j, and vice versa for a selectivity coefficient >1. The K_ij^pot values should not be considered to be constant parameters that characterise membrane selectivity under all conditions; the values are dependent on both the method used for determination, and on the conditions under which the calibrations are made. The fixed interference method is most commonly used to calculate the selectivity coefficient, and it is the method recommended by the International Union of Pure and Applied Chemistry (2). But whichever type of method is used the one used should always be quoted.

Figure 1.

A schematic representation showing an ideal ion-selective microelectrode calibration curve. The slope s, is the change in EMF per decade change in activity of a monovalent anion i, which is equivalent to 59.2 mV at 25 C; the limit of detection is defined as described in the text and is also indicated.

Another important parameter of an ion-selective microelectrode is the detection limit, which is the lowest ion activity which can be detected with confidence and is defined by the intercept of the two asymptotes of the Nicolsky response curve (see Figure 1). In practice, the detection limit seems to depend on the tip geometry and composition of the microelectrode's ion-selective membrane. Finer or smaller diameter tips have , higher detection limits, while composition affects detection in ways that can only be determined experimentally. The presence of interfering ions alters the detection limit (e.g. 3). The electrodes provide no useful information below their detection limits and for maximum benefit should be used in the linear portion of their calibration curves. The response time of ion-selective electrodes can be important when measuring changes in ion activities. This microelectrode parameter is dependent on many factors, including tip geometry, membrane composition and resistance. Response time can be measured during the calibration as the time taken for the voltage to adjust when ion activity at the tip is changed.

Types of microelectrode

There are three major types of ion-selective electrode, all of which can be minaturised for use in plant cells. These are solid state, glass, and liquid-(or fluid) membrane electrodes. Solid-state microelectrodes have been used to measure pH or Cl^- inside plants cells (4). While recessed tip glass microelectrodes have been made using pH-selective glass (5). These two types of microelectrode have largely been superceded for intracellular measurements by liquid-membrane electrodes so only the latter will be described here. Liquid membrane sensors are commercially available for a wide range of ions (see Table 1 ).

Table 1.Some examples of sensors for liquid membrane ion-selective microelectrodes and some of their properties.

Ion	Sensor molecule(s)	Detection limit	Major interfering ions in plant cells
Ca2+	ETH 129 ETH 1001	10 nM 40 nM	H+, K+, Mg2+ H+, K+, Mg2+
Cl-	Mn(III)TPPCa	1-5 mM	acetate, HCO3-, SCN-, NO3-, pH > 7.6
H+	Tridodecylamine ETH 1907	>pH 9 pH 9	K+ K+
K+	Valinomycin	100 µM	Ca2+, NH+,
Mg2+	ETH 5214	200 µM	Ca2+, K+
Na+	ETH 227b ETH 157b	3 mM 2 mM	Ca2+, H+, K+ H+, K+
NH4+	Nonactinb	2 µM	K+
NO3-	MTDDA.NO3	0.5 mM	Cl-, NO2-, SCN-

a Mn(III)TPPC = 5,10,15,20-tetraphenyl-21H,23H-porphin manganese(III) chloride (6). b data from Fluka Chemicals. MTDDA.NO3 = methyl-tridodecylammonium nitrate (3).

The detection limits quoted are from calibration in solutions approximating to cytoplasmic composition, values will also depend on the tip diameter but the values above are for tips less than 1 µm in diameter. This means that lower detection limits are possible for extracellular measurements where larger tip diameters can be used. All the above sensor molecules can be obtained from Fluka Chemicals.

To make an ion-selective microelectrode, the tip of the electrode is filled with an ion-sensing chemical cocktail which gives a voltage output of different values when placed in solutions containing different activities of the ion. Therefore when the electrode is inserted in the cell, the voltage measured gives a direct indication of the ion activity inside the cell. This situation is complicated by the voltage across the cell membrane; the ion-selective electrode will sense this in addition to voltage due to the activity of the ion of interest (see Introduction). To obtain the output for the ion alone, the cell voltage must be subtracted. This is done by using either two single electrodes or a double-barrelled electrode in which the ion-sensing electrode is combined with a cell-voltage-measuring electrode (see Figure 2). Both output voltages are measured against a reference ground electrode in the external solution. The ion activity is determined from the calibration curve after subtracting the membrane potential.

Figure 2.

Diagram of a liquid membrane double-barrelled ion-selective microelectrodes suitable for intracellular recording. The microelectrode is made by twisting together double-barrelled glass. A : Ion-sensing barrel with ion-selective cocktail in the tip (dark shaded area) B : cell voltage recording barrel. The output of B is subtracted from the output of A and converted to ion activity using a calibration curve such as that in Figure 1 .

Making microelectrodes

The preparation of ion-selective microelectrodes can be divided into four main stages:

1. pulling of glass micropipettes
2. silanization of the inside of surface of the ion-selective electrode or barrel
3. backfilling
4. calibration

The preparation of a nitrate-selective cocktail for backfilling microelectrodes is described in Protocol 1. A detailed generalized method which is suitable for all of the different types of ion-selective microelectrode is described in Protocol 2. The background to each stage is described here.

Pulling of glass micropipettes

Microelectrodes should be prepared to give dimensions suitable for impaling the target cell type. Double-barrelled microelectrodes can be prepared by twisting two single pieces of filamented borosilicate glass or using glass which is already fused together. Filamented glass has a glass fibre attached to the inner wall, this fibre assists backfilling by providing a hydraulic conduit along which the solution can flow by capillarity. This twisting is done using an electrode puller which both heats the glass and pulls it in a way pre-determined by the operator. The heating is paused for the two barrels to be twisted around one another then the heating and pulling continues. There are various different types of microelectrode puller (7). The most important feature is reproducibility, which ensures that when an optimum microelectrode shape for a particular cell type has been prepared, it can be exactly duplicated many times.

Modified Narishige Electrode Puller (inset shows puller modification to give 360 ^o twist)

Microelectrodes are usually made from borosilicate glass, although the harder aluminosilicate glass is also sometimes used. Multi-barrelled glass of varying dimensions can be purchased from suppliers (e.g. Hilgenberg). This type of glass seems to be the best for ion-selective microelectrode work. An alternative type of double-barrelled glass called "theta" glass can be used; this has a single thin glass wall between the two pre-formed barrels. Adjacent ion-selective barrels may mutually interfere because the thin glass walls at the electrode tip have an impedance that may be as low as the impedances of the liquid ion-exchangers so that the measured potential depends on the potential across the glass as well as the potential across the liquid ion-exchanger. This problem is more acute when "theta" glass is used because the final glass partition in the tip is much thinner. Both barrels of glass should have an internal filament to assist with backfilling. Identification of the different barrels can be done by marking the different barrels or cutting to different lengths (see Protocol 2). Wear safety glasses at all times when pulling and breaking glass.

Before preparing the ion-selective microelectrode it is important to determine that the glass microelectrodes filled with 0.1 M KCl can be used to impale cells and measure stable resting membrane potentials sensitive to metabolic inhibitors (in the usual range for the cell type, in the bathing solution used). An estimate of the tip geometry of the microelectrode is provided by measuring its electrical resistance when filled with KCl, larger tips having lower resistances. For tips of 2 - 0.1 µm diameter the electrical resistances of ion-selective microelectrodes are usually in the GW range, while microelectrodes filled with 0.1 M KCl have 10³ smaller resistances in the MW range. The dimensions of the microelectrodes are usually a compromise between obtaining a stable membrane potential and a good calibration response (detection limit).

Silanization of the inside of surface of the ion-selective electrode or barrel

The inside of the glass micropipettes must be given a hydrophobic coating, to allow the formation of a high resistance seal between the glass and the hydrophobic ion-selective membrane. The barrel designated to be ion-selective is silanized by placing a few drops of a solution of 2 % (w/v) silanizing agent in chloroform on its blunt open end. There are a range of different silanizing agents which can be used at this concentration but dimethyldichlorosilane or tributylchlorosilane are most common. Care must be taken to ensure that the reagent does not enter the membrane potential-measuring barrel. Beware silanizing agents are corrosive and toxic, protective glasses and gloves must be worn and glass must be treated in a fume hood. The microelectrode is then placed under a heating lamp giving a temperature of 140ºC at the micropipette surface. The silanizing solution quickly vapourises giving the ion-selective barrel a hydrophobic coating. After silanization there should be no liquid residue remaining in the microelectrode tip before backfilling.

Backfilling

There are actually two steps to backfilling, the first (a) uses a cocktail to form the ion-selective membrane in the microelectrode tip and the second (b) step, usually a minimum of 72 h later, uses an aqueous salt solution to provide contact between this membrane and the Ag/AgCl metal electrode (in the base of the microelectrode holder). Both steps are made much simpler by using filamented glass to make the microelectrodes and can be achieved using a syringe and fine needle (30 G).

1. The electrodes are back-filled with the sensor cocktail containing several different types of component:

• An ion-selective molecule, sensor or exchanger
• Membrane solvent or plasticizer
• Additives e.g. lipophilic cation/anion
• A membrane matrix to solidify the ion-selective membrane. This is essential for measurements in cells possessing turgor.

For many ions, the membrane cocktail can be purchased already mixed and it is advisable to start by using the commercial mixture. However, the individual cocktail components can be bought from chemical suppliers and preparing the cocktail one's self is cheaper. For commercially-available liquid membrane cocktails the membrane matrix is not normally included. A matrix is needed if microelectrodes are to be used in plant cells, it is needed because turgor will displace a liquid membrane from the electrode tip, thereby changing or eliminating the sensitivity to the measuring ion (8-10). The matrix used is normally a high molecular weight poly(vinyl chloride) (PVC) polymer, but can also include nitrocellulose for additional strength.

Of all of the components, the ion-selective sensor is the main factor determining the electrodes properties (e.g. slope, selectivity, limit of detection), however, the membrane solvent can, by an unknown mechanism, also alter the properties such as electrode lifetime, stability, and selectivity. Additionally, membrane additives, such as lipophilic ions, can be used to improve the performance of microelectrodes. The roles played by each component are described in detail by Ammann (2). The final optimum cocktail composition is found by trial and error (testing the performance of electrodes made using slightly altered composition). Good electrodes should have a low detection limit, a near ideal slope, and a small selectivity coefficient for physiologically-important interfering ions.

Protocol 1.

Preparation of nitrate-selective cocktail for backfilling ion-selective barrel.

• MTDDA.NO₃6% (w/w) (3)
• 2-nitrophenyl octyl ether 65% (w/w) (Fluka 73732)
• methyltriphenyl phosphonium bromide 1% (w/w) lipophilic cation (Sigma M7883)
• nitrocellulose 5% (w/w) (Whatman 7184002)
• PVC 23% (w/w) high molecular weight polymer (Fluka 81392)
• Tetrahydrofuran, THF (Sigma T5267)

Method

1. Weigh the MTDDA.NO₃(3 mg), nitrocellulose (2.5 mg), PVC (11.5 mg) and lipophilic cation (0.5 mg) into a 1 ml glass screw-topped vial using a balance accurate to 0.1 mg.

2. Add nitrophenyl octyl ether (32.5 mg) to the vial using a microcapillary on the balance pan.

3. Dissolve the cocktail in approximately 4 volumes of THF. This should be dispensed using a glass syringe and metal needle with no plastic components ^a. The cocktail takes at least 30 minutes to dissolve completely.

4. This cocktail can then be stored at 4ºC for several weeks and is ready for use in Protocol 2 and is enough to make about 70 nitrate-selective microelectrodes.

^a THF will dissolve some types of plastic disposable syringes.

(b) Backfilling with aqueous salt solution can also be done using a fine needle and syringe or a plastic disposable pipette tip can be gently heated and pulled into a fine capillary suitable for inserting inside the glass barrel. It is important to try to displace all the air from the interface between the ion-selective membrane and the backfilling solution. Sometimes a hair or cat's whisker can help to do this.

Calibration

Ion-selective microelectrodes can be calibrated using concentration or activity, the electrodes actually respond to changes in activity. Furthermore, activity is actually the important parameter for all biochemical reactions. Therefore calibrating with ion activity gives a microelectrode output which can be used directly without any assumptions of the intracellular activity coefficient for the ion. For these reasons the calibration of microelectrodes generally uses solutions which resemble the intracellular environment in terms of interfering ions, and ionic strength. Calibration of pH microelectrodes is easy because standard pH buffers can be used and simply checked with a pH meter. For other types of ion-selective microelectrode the calibration solutions may need to contain a pH buffer and a background salt solution to give an ionic strength approximately equivalent to that inside the cell. Care must be taken in the choice of these additional ions, they must not give significant interference over the range of measurements. In other words, the microelectrodes must have very small selectivity coefficients for these background ions. The calibration solutions are usually chosen to be approximately 0.14 M ionic strength. There are few examples of whole sap analysis to suggest what an appropriate figure might be, but for giant algal cells this value would seem reasonable (11). The use of computer programmes to calculate ion activity and the availability of a wide range of ion-selective macroelectrodes make it easier to prepare calibration solutions for all types of ion-selective microelectrodes. Furthermore calibration solution recipes have been published for some ion-selective microelectrodes, (Ca²⁺(12), Mg²⁺(13), NO₃^-(3)). Be aware that some calibration solutions use concentration not activity, and also the term "free" ion usually means concentration of unbound ion and not activity, particularly for Ca²⁺and Mg²⁺. The calibration of calcium-selective microelectrodes for intracellular measurements requires the use of calcium buffering agents such as EGTA because of the very low concentrations being measured (12).

Calibration of the ion-selective microelectrode can be done in the microscope chamber (see later) where intracellular measurements will be made or in a U-shaped glass funnel alongside the microscope. Finally, note that the slope of the calibration curve is temperature sensitive and both calibrations and intracellular measurements should be done at the same temperature. If the temperature of the calibration solutions is 4ºC and the cell is at 20ºC, the slope of the electrode calibration for a monovalent ion will be 55 mV per decade change in activity, not the 58 mV expected at 20ºC.

Protocol 2. Preparation of ion-selective microelectrodes

• Dimethyldichlorosilane (Fluka 40136), 2% v/v silanizing agent in chloroform
• THF (Sigma T5267)
• Ion-selective cocktail (Fluka) with PVC added or cocktail from Protocol 1
• PVC (high molecular weight polymer, Fluka 81392)
• Nitrocellulose (Whatman 7184002)

Method

1. Double-barrelled glass microelectrodes are pulled and twisted using an electrode puller.

2. Break back one barrel to give a short barrel, this will be the membrane potential recording barrel. The barrel can be broken back using square-ended pliers or a razor blade on the edge of a metal plate (wear safety glasses).

3. Microelectrodes are placed under a heating lamp for at least 30 minutes to dry before silanizing.

4. Use a disposable syringe and 25 G needle to introduce silanizing agent into the blunt end of the longer barrel which will become the ion-selective barrel. Warning: this must be done in a fume hood, under the heating lamp because the silanizing vapour is toxic and corrosive. The double-barrelled micropipette remains under the lamp for a further 30 minutes.

5. If the commercial ion-selective cocktail for backfilling does not contain PVC or nitrocellulose one or both should be added by first dissolving these matrix components in excess THF (4 volumes) and then mixing with commercial cocktail. The quantity of PVC can range from 10 to 30 % (w/w) and nitrocellulose 5 % (w/w) of cocktail when THF has evaporated (see Protocol 1).

6. Backfill ion-selective barrel of microelectrode with cocktail dissolved in THF/PVC mixture. Use a glass syringe (1 ml) and all metal 30 G needle (Scientific Laboratory Supplies Ltd).

7. The microelectrodes must then be stored with tip down in a silica-gel dried environment for at least 72 hours. During this time most of the THF evaporates leaving a solvent-cast membrane plug in the tip of the designated ion-selective barrel.

8. When the THF solvent has evaporated, the ion-selective barrel can be back-filled with an aqueous salt solution containing the ion to be measured.

9. Most types of ion-selective microelectrode require "conditioning" for a minimum of 30 minutes in an aqueous solution of the ion to be measured. This process involves immersing the tip in the solution containing a high concentration (e.g. 100 mM) of the ion to be measured.

10. Calibration of microelectrodes can be performed in the chamber built to take the plant tissue, or using a U-shaped funnel.

General practical points.

• Handle microelectrodes with forceps.
• When dispensing THF pour a few mls from the stock bottle into a clean glass beaker, after first rinsing the beaker with a little freshly-dispensed THF. Cover the beaker with parafilm, then dispense further THF using a glass syringe and needle by piercing the film cover with the needle, this helps to reduce solvent vapour and prevents contamination of THF.
• If more than one type of cocktail is used, employ a different syringe and needle for each type of cocktail. It is best to dedicate a syringe for one particular cocktail only, and thus avoid any contamination by other ion sensors.
• Calibrate starting with the highest concentration, and calibrate only in the range in which you expect to be working. There is no point in exposing electrodes unnecessarily to low ion concentrations as most types of ion-selective membrane respond badly to long exposures at very low concentrations.

Fault finding and some possible problems

The best approach is to try to solve problems by a process of eliminination. Firstly, establish whether a problem occurs in the circuitry or is specific to the ion-selective microelectrodes. The circuitry can be tested by putting a broken-tipped KCl-filled microelectrode in place of the ion-selective microelectrode. The broken-tipped electrode should give a stable zero output. It may be necessary to recoat AgCl-silver contact in the half-cell or there may be a wiring problem. Noisy recordings can be caused by poor earthing, or air bubbles in backfilling solutions. If the circuitry has no problems then the ion-selective microelectrode must be the cause. When the ion-selective microelectrode does not respond to the calibration solutions then the membrane can be checked by, deliberately breaking the tip to expose a larger area of ion-selective membrane. Breaking the tip can displace the ion-selective membrane from the tip so it is important to measure the resistance to check it is still in the GW range. If the broken tip gives a good response to changes in ion activity then the problem is independent of the composition of the membrane. When the microelectrode tip diameter becomes too fine the output from the ion-selective electrode will no longer respond to changes in ion activity.

Storage of ion-selective microelectrodes

For long term storage, the ion-selective microelectrodes should be stored without backfilling, in a silica-gel dried sealed container in the dark. This can be done in a screw-cap glass jar containing dry silica gel, with the microelectrodes attached to the inner wall using plasticine or Blu-tack. Ion-selective microelectrodes stored for several years in this way can still give a reasonable performance when back-filled.

Intracellular recordings

The equipment needed for intracellular measurements is similar to that for normal electrophysiological measurements in plant cells and general details have been described in reviews on electrophysiological methods (7). The equipment needed is listed below.

• Electrometer with high input impedance
• Micromanipulator
• Microscope with stage mounted perfusion chamber
• Faraday cage
• PC microcomputer with A/D interface
• Chart recorder.

The photograph below shows the equipment used in our lab for electrophysiological studies

Only a few special features of equipment for ion-selective microelectrode work will be described here and most of these are needed because ion-selective microelectrodes have high impedances. The microscope must be the fixed-stage type with a fibre optic light source. The lamp is located outside the Faraday cage to avoid interference which would be caused by including any mains voltage cable inside the Faraday cage. The chamber to hold tissue is mounted on the microscope stage and designed so that the tissue can be continously perfused with the experimental nutrient solution throughout the measurement to avoid a concentration gradient developing at the surface of the tissue. A high impedance electrometer amplifier (e.g. World Precision Instruments, model FD223, pictured below) is required because the ion-selective microelectrodes have very high resistances.

The amplifier must have a high input impedance at least 1000 times higher than the ion-selective electrode e.g. 10¹⁵W. Furthermore the input leakage current from the electrometer must be low so that no significant offset voltage (> 1 mV) is produced across the ion-selective electrode. Other useful facilities include GW range resistance tester and a difference-voltage output so that a direct output equivalent to cell ion activity can be obtained.

The electrometer output can be passed to a chart recorder and also via an A/D converter to a microcomputer. Data handling and processing is made much easier using software such as that developed by I.R. Jennings at the Biology Department, University of York, UK (3).

Several criteria for acceptable measurements can be defined. Firstly, after impalement the ion-selective microelectrode should be recalibrated and should give a very similar response to that shown before the cell impalement particularly at activities similar to those measured in vivo.

Sometimes the recalibration shows a displacement up or down the Y mV output axis. More often the detection limit of the ion-selective microelectrode has changed but provided the measurement was on the linear response range of the electrode calibration curve this is not usually a reason to disregard the result. Sometimes the performance of the ion selective microelectrode can even improve with the detection limit actually becoming lower. For this reason, it may be best to quickly impale a cell with a new tip before calibrating prior to measuring the activity in the cell. A comparison between the electrical resistance of the ion-selective microelectrode before and after an impalement provides a good indicator of whether the tip will recalibrate. If the resistance decreases below 1 GW, the ion-selective membrane has probably been displaced during impalement and the electrode will not recalibrate. Throughout the recording the state of the cell can be assessed by monitoring the membrane potential (which should remain stable unless deliberately perturbed) or processes like cytoplasmic streaming.

When measuring changes in intracellular ion concentrations, artifacts can be caused by the differential response times of the two barrels; the ion-selective barrel usually has a slower response time than the membrane potential-sensing barrel. This can be corrected for when the response time of the electrode is known (5). The electrode response time can limit detection of rapid changes in ion activity.

In plant cells, identifying in which internal cell compartment (cytoplasm or vacuole) the microelectrode tip is located can be a problem for some ions and it may be necessary to grow the plant under conditions in which two populations of measurements can be identified. Alternatively, a triple-barrelled microelectrode can be used where one barrel is pH or Ca²⁺ selective. Large gradients of these two ions are known to exist across the tonoplast, with the cytoplasm maintained at relatively constant values (pH 7.2, Ca²⁺ 100 nM) so compartment identification is possible. Another approach is to use tissues where the two major cell compartments can be identified under the microscope e.g. root hairs, or cell cultures which have no large vacuole. However, identifying which compartment the electrode is in can still be problematic, particularly if the electrode indents the tonoplast but does not penetrate it.

Leakage of salts from the tip of the membrane potential-sensing barrel has been reported (14), this may particularly be a problem in small cells. Diffusion of ions from the membrane potential-sensing barrel could give high local gradients of ions at the tip of a double-barrelled microelectrode. It may be important to try measurements where different types of backfilling solution are used in the reference barrel. Large leaks should affect membrane potential and monitoring this should indicate possible problems.

A further possible problem can arise when using ion-selective microelectrodes with inhibitors. Some inhibitor chemicals are highly lipophilic and will readily dissolve in the ion-selective membrane. These chemicals can poison the membrane but this will be demonstrated during the recalibration of the ion-selective microelectrode.

One last point regarding statistical analysis of data concerns the calculation of means. These should be calculated using the data which is distributed normally, that is using the log activity or output voltages not the actual activities (15). Therefore when mean activity value is used it can only be expressed with 95% confidence limits, whereas -log [activity] can be given standard errors or standard deviations.

References

1. Henriksen, G.H., Bloom, A.J. and Spanswick, R.M. (1990). Plant Physiol. 93, 271.
2. Ammann, D. (1986). Ion-selective microelectrodes, principles, design and application. Springer-Verlag, Berlin Heidelberg, Germany.
3. Miller, A.J. and Zhen, R.-G., (1991).Planta 184, 47.
4. Coster, H.G.L. (1966). Aust. J. Biol. Sci.19, 545.
5. Sanders, D. and Slayman, C.L. (1982). J. Gen. Physiol.80, 377.
6. Kondo, Y., Bührer, T., Seiler, K., Frömter E., and Simon, W. (1989)Pflügers Arch.414, 63.
7. Blatt, M.R. (1991). Methods in Plant Biochemistry 6, 281.
8. Felle, H. and Bertl, A. (1986). J. Exp. Bot. 37, 1416.
9. Sanders, D. and Miller, A.J. (1986). In Molecular and Cellular Aspects of Calcium in Plant Development (ed. A.J. Trewavas), p. 149 Plenum Press, New York and London.
10. Reid, R.J. and Smith, F.A. (1988).J. Exp. Bot.39, 1421.
11. Okihara, K. and Kiyosawa, K. (1988).Plant Cell Physiol.29, 21.
12. Tsien, R.Y. and Rink, T.J. (1981).J. Neuroscience Method.4, 73.
13. Blatter, L.A. and McGuigan, J.A.S. (1988).Magnesium7, 154.
14. Blatt, M.R. and Slayman, C.L. (1983).J. Memb. Biol. 72, 223.
15. Fry, C.H., Hall, S.K., Blatter, L.A. and McGuigan, J.A.S. (1990).75, 187.

Suppliers addresses

Fluka Chemicals Ltd., The Old Brickyard, New Road, Gillingham, Dorset SP8 4XT, UK

Hilgenberg GmbH, Postfach 1161, D-34 321 Malsfeld, Germany

World Precision Instruments (UK), Astonbury Farm Business Centre, Aston, Stevenage, Herts., SG2 7EG, UK

Scientific Laboratory Supplies Ltd., Wilford Industrial Estate, Nottingham, NG11 7EP, UK