The MIFE^TM system 
for non-invasive measurement of specific fluxes in solution near living plant or animal tissue

Overview
Membrane Transport & fluxes
MIFE Key Features
MIFE applications
development
references

MIFE theory
 ion flux theory
multivalent ion mobility
ionic mobility values
neutral molecule flux
Methodological isues
H+ flux in buffered media

Hardware
amplifier
microscope
manipulator
System requirements

Software
analysis

Purchasing
Univ of Tas
Eppendorf NP2

Methodological issues in flux measurement 

The cornerstone of each electrophysiological technique is microelectrode fabrication. For MIFE measurements, liquid membrane type ion-selective microelectrodes are used. Specific details of their fabrication and calibration are given in our publications (e.g. Shabala et al. 1997; Shabala and Shabala 2002; Shabala et al. 2006). Briefly, electrodes are pulled from non-filamentous borosilicate glass capillaries to tip diameter ~1 µm. Electrode blanks are then silanized with tributylchlorosilane (90796, Fluka Chemicals) to make their surface hydrophobic. Dried and cooled electrode blanks are stored under cover and may be used over several weeks. To make the electrode, the electrode tip is first broken slightly to achieve the required diameter (typically 2-3 µm). It is then backfilled with an appropriate solution (see Shabala et al (2006) for more details) and finally front-filled with the appropriate liquid ion exchanger (LIX). The column length of the LIX in the prepared electrode is usually 50 to 200 µm. Immediately after filling, electrodes are immersed in solution and kept there until use (up to 8-10 hours).

In our laboratory, we routinely use the MIFE system to measure net fluxes of H⁺, Ca²⁺, K⁺, Na⁺, Cl^-, Mg²⁺, NH₄⁺, NO₃^-, and Cd²⁺ from various systems: higher plant tissues and protoplasts; animal tissues (e.g. muscles); fungi; algae; protists; yeasts; bacterial monolayers and biofilms. Several more ions (e.g. Zn²⁺, Cu²⁺, Cs⁺, Pb³⁺, SO₄^-2) also can be measured using commercially available ionophores.  We also measure O₂ fluxes.

Despite the same principle being used for measurements of fluxes of each of these ions, there are some “specific features”, related to fabrication, calibration, and use of ion-selective microelectrodes to measure fluxes of a specific ion. The most critical ones are mentioned here. 

1.    Basic electrode characteristics. As a rule, for most physiological conditions, electrode characteristics are expected to be linear. Accordingly, each electrode is calibrated in a set of three known standards, covering the range of concentrations expected to be found in the experiment. The average responses of electrodes are about 53-54 mV/decade for monovalent ions, and 27-28 mV/decade for divalent ions, with a correlation R >0.999. If measurements are made at very low concentrations, non-linearity requires that more than 3 standards should be used.

2.    Electrode “conditioning”. Most of the prepared microelectrodes can be used immediately after preparation, while others (e.g. H⁺ and Cl^-) need some conditioning time (~1h) to ensure a stable response.

3.    Responsiveness. During MIFE measurements, the electrodes are moved back and forth at 5-sec intervals. For accurate flux calculations, the LIX must “settle” at each position. From practical experience, settling is achieved quicker if the LIX column length is relatively short. However, in this case there is a danger of a gradual leak of LIX out of the tip and electrode loosing its selectivity. The compromise is achieved by optimizing the amount of silane used for electrode fabrication and the amount of LIX used for electrode filling.

4.    Effect of ionic strength. Variations in ionic strength of solutions might significantly affect characteristics of ion selective electrodes and result in inaccurate estimates of ionic concentrations and, ultimately, net ion fluxes. The actual concentration (and, thus, flux) is overestimated for solutions with ionic strength lower than that of the standard, and is underestimated vice versa. For many ions, such a difference may be as big as a factor of 2 in a physiologically relevant range of concentrations (e.g. Na⁺ levels 200 mM). More details are available in Shabala et al. (2006).

5.    Temperature. From general knowledge, it was expected that both the slope and the intercept of the calibration curve for ion-selective electrodes might be slightly affected by changing temperatures. The theory shows that the Nernst slope is proportional to the Kelvin temperature. This effect is automatically covered if calibration and measurements are done at the same temperature. The effects of such temperature changes during an experiment are still open to question, it is unknown how significant these changes are and whether they should be taken into account. Our experiments in the 4 to 40 ^oC range suggested that in all cases the Nernst slope remained > 50 mV/decade, and the maximum inaccuracy in flux calculation did not exceed 6% (Shabala et al. 2006). Thus, no practical actions are needed. However at temperatures above 32 ^oC, the LIX often became very “noisy”. Thus, it is highly recommended that the performance of specific LIX should be tested at high temperatures if these are to be applied in an experiment, to ensure that the signal to noise ratio is acceptable, and that net flux responses can be distinguished from background noise.

6.    ignal to noise ratio. Due to the thermal electron noise in electrodes, there is some theoretical “lower limit” on the magnitude of the flux that can be measured against the background noise (Ryan et al. 1990). There are two practical measures to overcome this problem and to improve the sensitivity of the flux measurements. One is to increase the travel range of the electrode (making voltage changes larger), and another one is to increase electrode tip diameter. For more details, see Shabala et al. (2006).

7.    Confounding effect of inhibitors. Pharmacological experiments are frequent in plant electrophysiology. However, many of the channel blockers and metabolic inhibitors routinely used in patch-clamp experiments may significantly affect LIX characteristics. For example, even micromolar concentrations of CCCP completely “killed” Ca²⁺ LIX, reducing electrode slope from 27 to < 3 mV/decade. Therefore, a rigorous test of LIX performance in the presence of inhibitors should be undertaken first.