A new form of chamber for studying chemotaxis, similar in principle to the Zigmond chamber, allows the behaviour of the cells in a linear concentration gradient to be observed directly. The chamber was developed mainly for studying chemotaxis in fibroblasts using interferometric microscopy and the main design criteria were that it should have better optical characteristics, a higher dimensional precision and better long-term stability than the Zigmond chamber. It is made entirely from glass by grinding a blind circular well centrally in the counting platform of a Helber bacteria counting chamber. This procedure leaves an annular ‘bridge’, approximately 1 mm wide, between the new inner circular well and the original outer annular well. This bridge fulfils the same function as the linear bridge of the Zigmond chamber but the preciseconstruction of the counting chamber ensures that a gap of 20/nn between bridge and coverslip can be accurately and repeatedly achieved when the chamber is assembled. It is envisaged that the improved optical clarity, dimensional accuracy and long-term stability of the new chamber will be advantageous in other applications, particularly in studies requiring critical microscopy or a precise knowledge of the gradient and in studies of cells, such as fibroblasts, that move much more slowly than neutrophils.

The Boyden chamber has become established as the routine method for studying chemotaxis in leucocytes (Boyden, 1962) and fibroblasts (Postlethwaite et al. 1976). While it is particularly advantageous for screening large numbers of cells and putative chemotactic factors, it has serious limitations for investigating the mechanism of chemotaxis or even for confirming the occurrence of chemotaxis. One problem is that the exact nature of the concentration gradient is unknown: the cells themselves may severely modify their local gradient by obstructing the pores in the filter membrane and the gradient around the entrances and exits of the pores may be dependent on unknown flow conditions within the two wells. Another problem is that the behaviour of the cells cannot be observed but only deduced from their final distribution and this has led to some false claims of chemotaxis, although such deductions can be improved by using a full checkerboard analysis to take account of chemokinesis (Zigmond and Hirsch, 1973). Although the Boyden chamber is still useful as a convenient and sensitive method for identifying possible chemotactic responses, unequivocal confirmation of chemotaxis requires direct observation of the cells.

The currently available methods for directly observing chemotaxis include simply applying the cells and the source of chemotactic factor to a slide and covering them with a coverslip (Zigmond and Hirsch, 1973; Allan and Wilkinson, 1978), the under agarose assay (Nelson et al. 1975), and the Zigmond chamber (Zigmond, 1977). Of these methods, the Zigmond chamber has better optical properties than the under agarose assay and has the advantage that the concentration gradient approaches a linear steady state whereas, with the other two methods, the gradient never reaches a steady state.

In Zigmond’s description of the construction of her chamber, ′ Plexiglass slide was cut to have two wells 1 mm deep and 4 mm wide separated by a 1-mm bridge. A 22 mm×40 mm coverslip over the bridge and wells was held firmly in place with a brass clip screwed into the Plexiglass at each end’ (Zigmond, 1977). After assembling the chamber and filling the wells with the appropriate solutions, Zigmond found that the layer of fluid over the bridge was generally very thin and she recommended that only those assembled chambers with a gap over the bridge in the range 3–10 μm should be used for experiments. But predicting the decay of the gradient that forms across this bridge requires knowing the exact dimensions of the chamber, particularly the critical gap between the bridge and the coverslip. This is by no means easy, since the gap over the bridge is not determined by the geometry of the chamber but by factors such as flexure of the chamber, caused by the coverslip clamp springs, which cannot be duplicated reliably. The range of gaps is so wide that each must be measured individually, not only at one point but all along the bridge, if the decay of the gradient is to be predicted. The theory of prediction becomes difficult and unreliable if the gap varies along the bridge or, even worse, if it varies with time.

While the Zigmond chamber has acceptable optical properties for uncritical forms of microscopy such as low-power phase-contrast, its thickness of greater than 3 mm and construction from polymethy] methacrylate, which is a highly photoelastic polymer, render it unsuitable for critical microscopy Ln particular, for certain quantitative methods of microscopy such as microinterferometry, it is necessary to have the chamber made of an optically isotropic material, free from strain and with the optical surfaces worked to a high standard of flatness. Glass is an obvious choice but construction of a Zigmond chamber from glass is difficult, since, if strain must not be used to create the gap, the bridge must be optically polished to specify an exact gap.

A new chamber designed by one of us (Dunn) circumvents these problems by relying on the precise geometry of a commercially made counting chamber to specify the gap. Being made of glass only 1 mm thick, this has ideal optical properties for microscopy. Further advantages accrue from the concentric layout of the bridge and wells, which makes the chamber less prone to flexure, provides a more positive seating for the coverslip and avoids the end effects associated with a linear bridge. The fact that the central well is completely ‘blind’ in this design, rather than being sealed with wax or left open as both wells are in the Zigmond chamber, helps to prevent any forced flow over the bridge even if the chamber is flexed; but this is at the expense of not being able to change the medium in the inner well during an experiment.

Modifying the chambers

Helber bacteria counting chambers, type Z3, can be purchased from the manufacturers: Weber Scientific International Ltd., 40 Udney Park Road, Teddington, Middlesex, TW11 9BG, UK. The unmodified chamber consists of a 1 mm thick microscope slide with a 0.5 mm deep annular well ground into one face to leave a central platform approximately 7.6 mm in diameter. The face of this central platform is optically polished so as to lie precisely 20/an below the face of the slide and is ruled for counting purposes. These precision chambers are generally useful for observing the behaviour of living cells under ideal optical conditions. For this they can be obtained without rulings (Z3 special unruled) but this is not necessary for the current application, since the standard Thoma ruling is removed completely during modification of the chambers.

The modification consists of grinding a circular well centrally in the counting platform to leave an annular ‘bridge’ approximately 1 mm wide that fulfils the same function as the linear bridge of the Zigmond chamber. In order to prevent chipping of the bridge edge during machining of the well, a circular coverslip of approximately 8 mm diameter is cemented onto the central platform using Glassbond (Loctite UK), a u.v.-curing methacrylate adhesive. The chamber is then accurately centred on a rotating machine table using a dial test indicator and clamped with card packing. The well is excavated to a depth of 0.5–0.6 mm and a diameter of approximately 5.6 mm using a cylindrical sintered diamond burr, 2 mm in diameter, mounted in a high-speed engraving machine with water as a coolant. Concentricity of the well is ensured if the rotating table drive is used to feed the final cut round the edge of the well. The coverslip is then removed by soaking in acetone for an hour or two and the chamber is cleaned and sterilised ready for use.

The dimensions of the commercial chambers are variable between batches but we aimed to leave a bridge width of approximately 1 mm. Typical volumes for the inner and outer wells are 14/d and 30μl, respectively. The manufacturers are willing to perform the modification but, unless a large batch is required, laboratories with workshop facilities will find it much more economical to make their own. Fig. 1 is a photograph of a finished chamber shown in comparison with an unmodified Helber chamber and a Zigmond chamber.

Fig. 1.

An example of a Zigmond chamber (top left), an unmodified Helber bacteria counting chamber (bottom left) and the new chemotaxis chamber (bottom right).

Fig. 1.

An example of a Zigmond chamber (top left), an unmodified Helber bacteria counting chamber (bottom left) and the new chemotaxis chamber (bottom right).

Setting up the chambers

In most applications, it is best to set up the chamber with the experimental factor contained in the outer well. If the factor is contained in the inner well, it is difficult to avoid subjecting the cells in the bridge region to the maximum concentration and thus possibly saturating their receptors before the gradient becomes established.

The best way that we have found of setting up the chambers is first to fill both wells with control medium and then to cover the chamber with a 24mmx24mm no. 3 coverslip carrying the adherent cells. The chamber should be entirely free from bubbles and the coverslip should be offset from centre in order to leave a gap just large enough to admit a fine syringe needle at the edge of the outer well. The coverslip is then carefully pressed at the supported edges, using a medical wipe to soak up surplus medium, until low-order interference fringes are visible round all the supported edges. The upper surface of the coverslip and chamber is washed and dried and the coverslip is sealed into place using a hot paraffin wax/beeswax/vaseline mixture (1:1:1, by wt) round all the edges except for the gap. The outer well is then drained, using a piece of sterile filter paper, to leave the inner well and the region over the bridge filled with medium and a meniscus around the bridge. With a little practice, the outer well can now be refilled quickly and completely with medium containing the experimental factor by directing a syringe tangentially as m Fig. 2. If performed properly, the meniscus around the bridge protects the region over the bridge from becoming contaminated with the experimental factor during this procedure. Finally, the gap is sealed with the hot wax mixture and the chamber is ready for microscopy.

Fig. 2.

Showing how the outer well of the chemotaxis chamber is filled with medium containing the experimental factor.

Fig. 2.

Showing how the outer well of the chemotaxis chamber is filled with medium containing the experimental factor.

Visualising the gradient

The formation and decay of the gradient were visualised using a rhodamine B isothiocyanate-dextran 20S (Sigma R 9006) fluorescent dye of average molecular weight 17 200. This molecular weight was chosen to give a similar diffusion coefficient to that of human platelet-derived growth factor (11301), which we intend to use for studying fibroblast chemotaxis. Visualising the gradient in this way serves to check the efficacy of the chamberfilling procedure and to test the assumptions that the only mechanism of mass transport in the 20 um gap over the bridge is diffusion, whereas convection currents keep the bulk contents of the two wells stirred.

The inner well was filled with control medium and the outer well with medium containing approximately 1.5μgml−1 fluorescent dye as described above. A sector of the bridge region was examined using a Zeiss (Oberkochen) LM35 inverted microscope set up for rhodamine epifluorescence. The microscope stage was maintained at 37 °C by means of an air-curtain incubator, which was directed at the underside of the chamber in order to stimulate convection in the wells. Image intensity was quantified using a Sony CCD camera (set to gamma=l) with Peltier cooling attachment and frame integrator (Oggitronics Ltd) interfaced to a Compaq Deskpro 386/20 computer. The camera is very sensitive in the infrared region and extra infrared barrier filters were needed to reduce the background signal.

The short-term formation of the gradient was quantified at 5 min intervals from 5 min to Ih after setting up the chamber. Using a 6.3 × objective, each image of the bridge region was integrated over 16 video frames and then processed by subtracting a background image taken from a non-fluorescent region of the chamber. Fluorescence intensity as a function of position across the bridge was computed by averaging grey levels within blocks of 19×50 pixels. A strip, 50 pixels wide, composed of 12 adjacent blocks, spanned the bridge from the inner to outer wells. These intensity measurements were normalised to a maximum value of 100 and taken directly as measures of dye concentration in arbitrary units.

Because of the poor long-term stability of the arc source, it was necessary to use a different protocol based on measuring intensities in the wells to study the long-term decay of the gradient. This was done in a separate experiment using a 2.5 × objective to include a portion of each well in the image. At the end of the experiment, intensities obtained from the chamber filled uniformly with fluorescent dye at half concentration were used to compensate for any difference in well depth. Fluctuations in lamp brightness were compensated by assuming that the total amount of fluorescent dye in the two wells of known volume is conserved. For the backgound subtraction we used the intensity of the inner well at time zero and also compensated this for lamp fluctuations. Compensated intensities were taken as measures of dye concentration in the wells in arbitrary units.

Trial experiments with neutrophils

Human peripheral blood neutrophils were isolated as described by Forrester et al. (1983), except that acid citrate dextrose instead of heparin was used as the anticoagulant. The isolated neutrophils were resuspended in Hepes-buffered balanced salt solution (HBSS) containing 2% foetal calf serum (Flow) and 10 000 cells contained in this solution were allowed to settle on the central region of a 24mm×24mm no. 3 coverslip for 20 min at room temperature. The chemotactic chamber was filled with HBSS and the coverslip with adherent neutrophils was rinsed in HBSS and inverted onto the chamber as described above. The outer well was then drained and refilled with HBSS containing 4×10−8M N-formylmethionylleucylphenylalanine (FMLP, Miles) and the chamber sealed with wax. Some chambers were maintained at 37 °C on the stage of the Zeiss IM35 microscope equipped with a 40 × oil-immersion apochromat and regions of the bridge were photographed in phase-contrast 30 min later. Others were used for time-lapse video recording on a Zeiss Standard microscope equipped with a 10× phase-contrast objective.

Theory of formation and decay of the gradient

One advantage of precise and stable chamber dimensions is that the establishment and decay of the gradient can be accurately predicted for substances of known diffusion coefficient. It is assumed that diffusion is the only mechanism of mass transport in the bridge region whereas convection currents keep the bulk contents of the two wells stirred. The diffusion in the bridge region is then theoretically equivalent to diffusion within the wall of a hollow circular cylinder and solutions for the steady and non-steady states are given by Crank (1956). In the steady state, the concentration as a function of distance, r, from the centre of the chamber is:
formula
where C1 and Co are the concentrations in the inner and outer wells, respectively, and a and b are the inner and outer radii of the bridge. Because the bridge is annular, this is not a strictly linear concentration gradient but is slightly convex if the higher concentration is in the outer well. For the actual dimensions of the chamber, however, the deviation from linearity is very small.
If the inner well and the bridge region have initial concentrations of zero, the concentration profile as a function of time, t, for a molecule with a diffusion coefficient, D, is given by:
formula
where
formula
and the αn values are the positive roots of:
formula
and α0(x) are Bessel functions of order zero of the first and second kinds, respectively.
Equation (2) is only valid if the concentrations in the two wells remain constant during time t. For the actual dimensions of the chamber, it is a very good approximation during the approach to the steady-state condition, but estimating the long-term decay of the gradient requires that we take account of the flux across the bridge and the subsequent changes in the concentrations of the wells. Since very little of the total material passes from the outer to the inner well during the approach to the steady state, it is adequate to consider only the steady-state flux. The flux of diffusing substance, dQ/dt, from outer to inner well through the gap of height h over the bridge is given by:
formula
This equation can be written as a first-order differential equation in terms of the steady-state gradient G(t)= (Co—C1)/(b—a) and solved to give:
formula
where:
formula
and vt and v0 are the volumes of the inner and outer wells. The steady-state gradient therefore decays exponentially and the rate of decay can be characterised by the half-life ln(2)/kD. The constant k is determined by the geometry of the chamber and so it is only necessary to calculate the characteristic area, ln(2)/&, once for the chamber in order to predict the half-life of the steady-state gradient of any substance of known diffusion coefficient.

Mathematica (Wolfram Research Inc., P.O. Box 6059, Champagne, IL 61826, USA), a computer algebra program, was used to find the first ten positive roots of equation (4) for the values a=2.8mm, 6=3.9mm corresponding to the dimensions of one of the chambers. Equation (2) was then evaluated numerically for D= 13.3×10−6mm2s−1 at 12 points across the bridge and at 5 min intervals for the period 5 min to 1 h. This value for D was chosen as the diffusion coefficient in water at 37 °C of a typical globular protein with a molecular weight of 17 200, which is the molecular weight of the fluorescent dye that we used to test the performance of the chamber. The theoretical evolution of the gradient profile during the period 5 min to 1 h is shown in Fig. 3. At the end of 1 h, the gradient has practically reached its steady-state profile and a slight convexity can be seen in the diagram.

Fig. 3.

Theoretical evolution of the gradient during the approach to the steady state for a hypothetical chemotactic factor with a diffusion coefficient of 13.3×10−5mm2s−1. Concentration of the factor (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.9 mm) and time elapsed since setting up the chamber (t in min) with all the factor initially in the outer well. The evolution during the first 5 min is not shown because the numerical approximation rapidly loses accuracy as time zero is approached.

Fig. 3.

Theoretical evolution of the gradient during the approach to the steady state for a hypothetical chemotactic factor with a diffusion coefficient of 13.3×10−5mm2s−1. Concentration of the factor (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.9 mm) and time elapsed since setting up the chamber (t in min) with all the factor initially in the outer well. The evolution during the first 5 min is not shown because the numerical approximation rapidly loses accuracy as time zero is approached.

The decay of the steady-state gradient was observed experimentally in a chamber with slightly different dimensions and so we used these dimensions (a=2.8mm, 6=3.8 mm, bi=14μl, oo=30 μl, 6=0.02 mm) for the theoretical prediction of gradient decay. The characteristic area of this chamber is 16.08 mm2 and so the half-life of the gradient for a substance with D=13. 3×10 6mm2s 1 is approximately 120 000 s or 33.6 h. Fig. 4 shows the theoretical decay of the steady-state gradient over the period 1 to 96 h calculated from equations (1) and (6).

Fig. 4.

Theoretical decay of the steady-state gradient for a hypothetical chemotactic factor with a diffusion coefficient of 13.3×10–5 mm2s−1. Concentration of the factor (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.8 mm) and time elapsed since setting up the chamber (i in h) with all of the factor initially in the outer well. The approach to the steady state during the first hour, which is shown in detail in Fig. 3, is omitted from this figure.

Fig. 4.

Theoretical decay of the steady-state gradient for a hypothetical chemotactic factor with a diffusion coefficient of 13.3×10–5 mm2s−1. Concentration of the factor (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.8 mm) and time elapsed since setting up the chamber (i in h) with all of the factor initially in the outer well. The approach to the steady state during the first hour, which is shown in detail in Fig. 3, is omitted from this figure.

Provided that the chamber being used does not differ very much in dimensions from the above, Figs 3 and 4 can be used to estimate the formation and decay of the gradient of any substance with known diffusion coefficient. It is only necessary to rescale the time axes so that the ratio of old to new times is the reciprocal of the ratio of old to new diffusion coefficients. Thus, for a substance with a diffusion coefficient of 53.2×10−5mm2s−1 (i.e. 4×13.3×10−6), Fig. 3 shows gradient formation during the period 75 s to 15 min and Fig. 4 shows gradient decay during the period 15 min to 24 h. This diffusion coefficient is probably not very different from that of a chemotactic tripeptide such as FMLP. If the diffusion coefficient is unknown but the substance is a rigid spherical molecule, then D can be estimated from the molecular radius rm using:
formula
where η is the viscosity of the solvent, k is the Boltzman constant (1.38×10−23JK−1) and T is the absolute temperature. For a globular protein, rm may be estimated from the molecular weight M using:
formula
where N is Avogadro’s number and v, the partial specific volume, is given as differing little from 0.730 for a wide range of proteins by Squire and Himmel (1979). Applying these two equations to FMLP (M=437.6) gives an estimate of D=45.2×10−5 mm2s−1 although the method may not be very reliable for such small peptides.

Formation and decay of the fluorescent dye gradient

Fig. 5 shows the formation of a gradient of the fluorescent dye during the first hour after setting up the chamber. This is very close to the predicted evolution of the gradient profile in Fig. 3 except that the high end of the dye gradient takes approximately 20 min to reach the maximum concentration. A likely explanation is that the dye in the outer well takes some time to mix into the meniscus that was left surrounding the bridge. The assumption of rapid and thorough mixing by convection in the wells is therefore not entirely valid but the gradient nevertheless evolves as predicted after the first 20 min.

Fig. 5.

Evolution of a gradient of the rhodamine/dextran dye during the approach to the steady state. Concentration of the dye (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2 8 to 3.9 mm) and time elapsed since setting up the chamber (t in min) with all of the dye initially in the outer well.

Fig. 5.

Evolution of a gradient of the rhodamine/dextran dye during the approach to the steady state. Concentration of the dye (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2 8 to 3.9 mm) and time elapsed since setting up the chamber (t in min) with all of the dye initially in the outer well.

Fig. 6 shows the decay of the dye gradient during the first 4 days after setting up. As described in Materials and methods, this is based on measurements of the intensities in the two wells and the gradient profiles in the figure have been calculated using equation (1) on the assumption that the gradient is in a steady state after 1 h. The observed behaviour is again very close to the predicted behaviour shown in Fig. 4 and this demonstrates that the assumption of mixing in the wells is valid in the long term. The observed flux allowed us to calculate an experimental estimate of the diffusion coefficient using equation (5) and this turned out to be 11.0×10–5mm2s−1 compared with the theoretical value of 13.3×10–2mm2s−1 based on the molecular weight of the dye. This small discrepancy can be explained by any hydration or lack of sphericity of the molecule or a slightly different partial specific volume from the one that we assumed.

Fig. 6.

Decay of a gradient of the rhodamine/dextran dye. Concentration of the dye (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.8 mm) and time elapsed since setting up the chamber (t in h) with all the dye initially in the outer well. This figure is based on measurements of the concentrations in the inner and outer wells and the gradient profile is calculated theoretically.

Fig. 6.

Decay of a gradient of the rhodamine/dextran dye. Concentration of the dye (C in arbitrary units) is shown as a function of radial distance from the centre of the inner well (r in mm, the bridge extends from 2.8 to 3.8 mm) and time elapsed since setting up the chamber (t in h) with all the dye initially in the outer well. This figure is based on measurements of the concentrations in the inner and outer wells and the gradient profile is calculated theoretically.

To summarise the performance of the chamber, a protein with a molecular weight of the order of 10 000 to 20 000 will form a close approximation to a linear gradient within about 30 min of setting up the chamber and the half-life of the gradient (i.e. the time for its slope value to halve) will be about 30 h whereas a peptide of molecular weight 350–750 will form a linear gradient within 10 min and decay to half its initial value in 10 h. The times of gradient formation and decay are approximately proportional to the cube root of the molecular weight of the substance.

Neutrophil chemotaxis to FMLP in the chamber

Fig. 7 shows several neutrophils close to the outer edge of the bridge 30 min after setting up the chamber with 4×10–8M FMLP in the outer well and control medium in the inner well. In a cursory examination we found more than half of the total number of non-rounded cells in the bridge region to be oriented in the up-gradient direction quadrant (only 25 % would be expected in this quadrant on the hypothesis of random orientation). But, since chemotaxis is more properly defined in terms of the motility rather than the orientation of cells, we decided on a dynamic analysis of cell displacements in order to establish whether chemotaxis occurs in the new chamber.

Fig. 7.

Several neutrophils close to the outer edge of the bridge (which can just be seen in the upper left) m a gradient of FMLP 30 min after filling the outer well with 4×10−8M FMLP. Zeiss 40 × phase-contrast apochromat. All the nonrounded neutrophils can be seen to be oriented approximately in the direction of the gradient. Bar, 50/on. Direction of gradient is indicated by arrow.

Fig. 7.

Several neutrophils close to the outer edge of the bridge (which can just be seen in the upper left) m a gradient of FMLP 30 min after filling the outer well with 4×10−8M FMLP. Zeiss 40 × phase-contrast apochromat. All the nonrounded neutrophils can be seen to be oriented approximately in the direction of the gradient. Bar, 50/on. Direction of gradient is indicated by arrow.

Fig. 8 is a vector scatter diagram of the displacements of 22 motile neutrophils (from a total sample of 50 of which 28 were non-motile) taken from a time-lapse video recording during the lh period from 30 to 90 min after setting up a gradient of FMLP as before. The displacements are oriented and translated so that each dot in the diagram represents where a cell would be after 90 min if it had started at the origin at 30 min and the gradient had been directed vertically upwards through the origin. The + represents the vector mean migration and it is clear that the large majority of cells have tended to migrate in the up-gradient direction. Even in such a small sample, the significance of this tendency to migrate up-gradient is very high and the results of two-tailed i-tests on X- and E-components are given in the figure legend. Only one of the 22 motile cells shows a displacement down the gradient and similarly high significance levels were obtained using the non-parametric Signs test and Wilcoxon signed rank test.

Fig. 8.

Vector scatter diagram of displacements of 22 motile neutrophils (from a total sample of 50) during the 1 h period from 30 to 90 min after setting up a gradient of FMLP as in Fig. 7. Direction of gradient is vertically upwards. Mean displacement is indicated by + and the mean y-component is significantly greater than zero at the 0.001% level (two-tailed i-test: t=5.89, d.f.=21, 0.000001<P<0.00001) indicating a highly significant positive chemotaxis. As a control, the mean x-component was tested and found to be insignificantly different from zero (t=0.89, d.f. = 21, 0.3<P<0.4).

Fig. 8.

Vector scatter diagram of displacements of 22 motile neutrophils (from a total sample of 50) during the 1 h period from 30 to 90 min after setting up a gradient of FMLP as in Fig. 7. Direction of gradient is vertically upwards. Mean displacement is indicated by + and the mean y-component is significantly greater than zero at the 0.001% level (two-tailed i-test: t=5.89, d.f.=21, 0.000001<P<0.00001) indicating a highly significant positive chemotaxis. As a control, the mean x-component was tested and found to be insignificantly different from zero (t=0.89, d.f. = 21, 0.3<P<0.4).

The optical properties of the new chamber are close to those of the ideal microscope slide/coverslip combination and permit practically all forms of light microscopy. In very critical work such as microinterferometry (Brown and Dunn, 1989), it may be found that the slightly turned down edge at the outer rim of the bridge, which is left by the manufacturer’s polishing procedure, needs to be removed during the modification of the chamber. As with a haemocytometer, the covers] ip needs to be rigid in order to benefit from the precision of the chamber and we find it best to use a no. 3 (0.25–0.35 mm thick). This dictates that high-power, non-immersion objectives with a fixed compensation for the standard no. li coverslips (0.16-0.18 mm) should not be used for critical work.

It is apparent from our fluorescent dye experiments that the gradient in the new chamber can be predicted quite reliably from diffusion theory. This is in contrast to Zigmond’s experiences with her chamber using fluorescein isothiocyanate to visualise the decay of the gradient (Zigmond, 1977). In Figure 2 of her paper, the fluorescein gradient at room temperature appears to have reached its maximal value after 15 min and to have fallen to about half its maximal value after 105 min. She comments that, ‘Variations among experiments indicated that it was not possible to predict the exact shape of the gradient at any given time; nevertheless, between 30 and 90 min the gradients were usually steep and stable’ This rapid decay of the gradient is remarkable, since basic diffusion theory predicts that the half-life of the gradient should be at least 40h! The volume of each well is 100 μI and initially the two wells contain 10nmol and Inmol. The bridge is 25mm long by 1 mm wide and, for the maximum recommended gap of 10 μm, the maximum flux across the bridge should be no more than 0.05 nmol h−1 for a molecule of about 400 MR such as fluorescein isothiocyanate. Even if this initial maximum flux were to continue indefinitely instead of decaying as the gradient decays, only 2 nmol would have passed over the bridge in 40h: the first well would still contain 8 nmol and the second well would contain 3 nmol.

Reliable prediction and long-term stability were not essential for the excellent work on the mechanism of leucocyte chemotaxis done by Zigmond and others using the Zigmond chamber but they are necessary in other areas of research. Our chamber was designed for studying chemotaxis in fibroblasts and long-term stability of the gradient is essential for this, since these cells move much more slowly than neutrophils, often moving much less than their own length during one hour. Being able to predict the gradient accurately also opens up new areas of study on the mechanism of neutrophil chemotaxis, particularly with regard to such important and controversial issues as whether a stable gradient can evoke chemotaxis in neutrophils (Vicker et al. 1986). For these reasons we tried to determine the potential sources of unpredictability in the Zigmond chamber and to take particular precautions to eliminate them from the new design and mode of operation.

Small or transient convective flows in the gap over the bridge would distort or destroy the gradient and it seems from Figure 2 in Zigmond’s paper that some distortion as well as decay of the gradient has occurred after 105 min. Boundary layer theory dictates that it is unlikely that thermal convection currents occur in the narrow gap over the bridge but convective flow could arise for a variety of mechanical reasons. Bulk flow across the bridge of the Zigmond chamber could arise simply by drainage if the ends of the wells are left open and the chamber is not kept perfectly horizontal. Even sealing the wells with wax may not entirely eliminate this problem, since wax is not very stable mechanically and a change of only 0.25% in the volumes of the wells could completely destroy the gradient over the bridge. The new chamber avoids these problems by relying on the blindness of the inner well and the incompressibility of medium to prevent bulk flow across the bridge. Only one well need be blind for this and so it does not matter that the outer well is sealed with wax. But it is important to guard against trapping bubbles in the chamber, particularly in the inner well, since the contents of the inner well would not be incompressible if bubbles were included.

Small changes in the gap height could also have dramatic (but temporary) effects on the profile of the gradient by causing surges of medium to and from the gap region. These changes could result from thermal expansion and contraction of any bubbles trapped in a sealed chamber. They could also arise from flexion of the chamber resulting from differential thermal expansion or mechanical creep, or from handling. Our chamber is likely to be less susceptible to these changes because it has a larger gap than the Zigmond chamber. Nevertheless, it is designed to avoid differential thermal expansion by using glass throughout and to avoid mechanical creep by not using spring clips or similar devices that subject the chamber to continual stress. We find that it maintains the plane of focus much better than a Zigmond chamber, which is evidence that we have succeeded in improving the mechanical stability.

We investigated the susceptibility to handling of the two types of chamber by assembling them without medium and viewing a diffuse monochromatic light source reflected from their upper surfaces. A shift of one fringe width in the interference fringes observable over the bridge indicates a change in the gap of about 0.27 μm. Even gentle handling of the Zigmond chamber, such as lifting the horizontal chamber by one end, resulted in a change in gap of 0.5 gm. With the chamber supported at its ends, light finger pressure near the coverslip could result in a change of 2 μm or more. Gentle handling could therefore be sufficient to distort the gradient significantly, especially if the initial gap is set as low as 3 gm. No effect was noticeable when applying the same treatments to our chamber. Even so, any procedure that applies considerable stress to the chamber, such as using strong spring clips to clamp it to a microscope stage, should be avoided.

Another possible explanation of the rapid decay of the gradient that Zigmond observed is that the ambient thermal conditions in her experiments were such that the wells were not continually being mixed by convection currents. This seems unlikely, since she remarks that, if the chamber is set up with Trypan Blue (961 Mr) in one well and water in the other, a difference in concentration between the two wells is still noticeable after 72 h at room temperature. If there were no mixing in the wells, the difference in concentration would be more than just noticeable, it would be almost unchanged except near the bridge after 72 h. Yet it is clearly important to be sure that mixing is occurring in the wells of either type of chamber and, in our experiments, we directed the air flow of the incubator towards the lower surface of our chamber in order to encourage convection. Our observations indicated, however, that mixing will occur even without taking such precautions.

In our preliminary experiments with neutrophils and FMLP in the new chamber, we found levels of orientation consistent with Zigmond’s observations (Zigmond, 1977). Our analysis of cell displacements leaves no doubt that chemotaxis can occur in the new chamber and our theoretical and experimental analysis of gradient formation in the chamber strongly suggests that the gradient was practically linear and stable during the period of observing displacements. This analysis would seem, therefore, to present evidence against the controversial assertion that a stable spatial gradient of FMLP will not induce chemotaxis in neutrophils (Vicker et al. 1986) but there is also the possibility that the rapidly changing gradient during the first 10 min could still influence the cells over the period 30 to 90 min. This delayed effect might conceivably arise from a persistence in cell motility or in receptor occupancy. Further experiments with the new chamber and a detailed dynamic analysis will be needed to decide this issue.

Daniel Zicha is supported by a Wellcome Trust Fellowship. Graham Dunn and Alastair Brown are members of staff of the Medical Research Council.

Allan
,
R B
and
Wilkinson
,
P C.
(
1978
).
A visual analysis of chemotactic and chemokinetic locomotion of human neutrophil leucocytes. Use of a new chemotaxis assay with Candida albicans as gradient source
.
Expl Cell Res
111
,
191
203
.
Boyden
,
S
(
1962
).
The chemotactic effect of mixtures of antibody and antigen on polymorphonuclear leucocytes
J. exp Med
.
115
,
453
466
Brown
,
A F.
and
Dunn
,
G. A.
(
1989
).
Microinterferometry of the movement of dry matter in fibroblasts
J. Cell Sci
.
92
,
379
389
.
Crank
,
J
(
1956
)
(2nd edn. 1975) The Mathematics of Diffusion
.
Clarendon Press
,
Oxford
.
Forrester
,
J. V.
,
Lackie
,
J. M
and
Brown
,
A F
(
1983
)
Neutrophil behaviour in the presence of protease inhibitors
.
J. Cell Sci
.
59
,
213
230
.
Nelson
,
R. D.
,
Quie
,
P. G.
and
Simmons
,
R. L.
(
1975
)
Chemotaxis under agarose: a new and simple method for measuring chemotaxis and spontaneous migration of human polymorphonuclear leukocytes and monocytes
.
J. Immun
.
115
,
1650
1656
Postlethwaite
,
A. E.
,
Snyderman
,
R.
and
Kang
,
A. H.
(
1976
)
The chemotactic attraction of human fibroblasts to a lymphocyte derived factor
.
J exp Med
144
,
1188
1203
Squire
,
P. G.
and
Himmel
,
M. E.
(
1979
).
Hydrodynamics and protein hydration
.
Archs Biochem. Biophys
.
196
,
165
177
.
Vicker
,
M. G.
,
Lackie
,
J. M
and
Schill
,
W.
(
1986
).
Neutrophil leukocyte chemotaxis is not induced by a spatial gradient of chemoattractant J
.
Cell Sci
.
84
,
263
280
.
Zigmond
,
S. H.
(
1977
).
Ability of polymorphonuclear leukocytes to orient in gradients of chemotactic factors
.
J. Cell Biol
.
75
,
606
616
Zigmond
,
S. H
and
Hirsch
,
J G.
(
1973
)
Leukocyte locomotion and chemotaxis. New methods for evaluation, and demonstration of a cell-derived chemotactic factor
.
J. exp. Med
.
137
,
387
410
.