184,63-66
ANALYTICALBIOCHEMISTRY
(1990)
A Paper Membrane Filter Assay for Ciliate Chemoattraction Vagn Leick,l
Karin
Frederiksen,
Department of Biochemistry
Received
March
Inger
B, Panum Institute,
Lyhne,
and Per Hellung-Larsen
University
of Copenhagen, Blegdamsvej 3C, DK-2200 Copenhagen N, Denmark
27,1989
A quantitative bioassay for ciliate chemoattraction based on the Boyden assay is described with the ciliates Tetrah ymena thermophila and Tetrah ymena pyriformis as test organisms. A chamber is separated into two compartments by a Whatman 3MM filter, and a suspension of starved cells (- 10’ cells/ml) is placed in one compartment and a solution containing attractant in the other. The gradient of chemoattractant across the filter causes the cells to swim through the filter into the attractant-containing compartment where their appearance is determined by electronic cell counting. The assay described is convenient with a signal-to-noise ratio of approximately 10. It is shown here to work with the attractants proteose peptone and platelet-derived growth factor. 0 1990 Academic Press, Inc.
Boyden (1) introduced a convenient and reproducible method for assaying leukocyte locomotion by observations of the behavior of a cell population. He used a twocompartment chamber in which the cell suspension under test was separated from the chemotactic substance by a Millipore filter. Chemotactic factor diffused through the filter from below to reach the cells on the upper surface which responded by migrating through the filter by an amoebic motility. The number of leukocytes per unit area of the lower surface area of the membrane was counted microscopically as a function of time or concentration of test compound. Working with the ciliate protozoan Tetrahymena which is a free swimming ciliate we have developed various assay procedures to assay for the chemoattractant potential of a test compound (2-4). One of the methods (2) was based on the capillary principle. However, it was necessary to work at rather low cell concentrations (l-3 X lo4 cells/ml) in order to avoid high background levels 1 To whom
correspondence
should
0003.2697/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
be addressed.
Inc. reserved.
in control assays containing no attractant. The present paper describes a method where the capillary tubes are replaced by a paper membrane filter (Whatmann 3MM Chr). It is possible to work at much higher cell concentrations with this method than with the capillary technique. MATERIALS
AND
METHODS
Tetrahymenu thermophila (strain B7) or Tetrahymena pyriformis (strain GL) were grown to mid-stationary phase in PY2 medium: 1% proteose peptone (Difco) containing 0.1% yeast extract (Difco) and 0.03%0 sequestrene. For chemoattraction assay experiments, cells were collected by centrifugation at 5OOg,washed, and resuspended in 10 mM Tris-HCl, pH 7.4, to a cell concentration of 2 X lo5 cells/ml. Cells were starved in this medium for 16 h and transferred to fresh 10 mM Tris-HCl containing 0.15% (w/v) gelatin Type III (Sigma) to a final cell concentration of 1.5 X lo6 cells/ml (cell stock). The cells were preincubated in this medium for 1 h before use in chemoattraction experiments. Gelatin was included in order to serve as carrier medium for fibroblast growth factor (FGF) and platelet-derived growth factor (PDGF). However, using PY, the assay works perfectly well without the presence of 0.15% gelatin, which increases the viscosity of the medium only slightly. The two-compartment chamber used is shown in Fig. 1. The two compartments (-1.5 ml each) are separated by Whatman 3MM Chr paper (Cat. No. 3030117). Leakage out of the system through the Whatman paper was prevented by placing grease (white Vaseline) on the paper surrounding the chamber and squeezing the greased paper between the two Plexiglas blocks in order to seal the chamber. For chemoattractant experiments 0.9 ml 10 mM TrisHCl, pH 7.4, containing 0.15% gelatin (PDGF carrier) ’ Abbreviations used: PY, proteose trene; FGF, fibroblast growth factor; factor; BSA, bovine serum albumin.
peptone-yeast extract-sequesPDGF, platelet-derived growth
63
64
LEICK Whatman
3MM
FIG. 1. Paper membrane assay chamber. Whatman paper 3MM Chr separates two compartments A and B. Cells are placed in compartment A and the chemoattracting test compound in compartment B. After incubation at 2%30°C the cell number is registered by electronic cell counting using a Coulter electronic cell counter.
was placed in both compartments. Then 100 ~1cell stock was placed in compartment A immediately followed by 100 /*l test compound in compartment B. Mixing was carried out by filling and emptying the lOO-~1 pipette twice immediately after the addition of the aliquot. The
I
I
I
1
4(
Hours FIG. 2.
ET AL.
Appearance of cells (B7) in compartment B as a function of time at 28°C. Cells (1.5 X 106; strain B7) were placed initially in compartment A (total volume 1 ml). PY growth medium (final dilution 1:lO) was placed in compartment B (m). Control experiments without attractant were performed with the same number of cells (Cl). Each point represents an individual assay chamber.
0
I
I
1
2
Cells per ml x 1Oj in A FIG. 3. Number of cells (B7) transferred into compartment B as a function of the initial cell concentration in compartment A. Up to 2 X lo5 cells (strain B7) were placed in compartment A and the percentage of cells accumulating in compartment B during 1.5 h at 28°C was calculated (m). Control chambers containing no attractant (Tris) were run in parallel (0). Each point represents an individual assay chamber.
chamber was incubated at the appropriate temperature and the total cell number in the two compartments was estimated using a Coulter electronic cell counter by emptying the whole compartment (1 ml) and diluting with 9 ml counting liquid (0.1% NaN3, 0.9% NaCl); 0.5ml aliquots of this fixed cell suspension were counted in the Coulter counter. All determinations were carried out in duplicate assay chambers (see Figures). Statistical analysis was unnecessary in the individual experiment because of good reproducibility of double estimates and of signal-to-noise ratios. However, a day-to-day biological variation of cellular response was observed. The experiments described are, however, representative and carried out with cells taken from early to mid-stationary cultures (~2 to 4 days after inoculation). It should be noted that cells from the logarithmic growth phase (and late stationary phase) were not used. Porcine platelet-derived growth factor was obtained from R & D Systems Inc., Minnesota, and basic fibroblast growth factor was from Sigma. In some experiments other types of filter material (Whatman GF/C glass fiber filter or Whatman AGF 617) were used. RESULTS
When placing l-2 X lo5 cells in compartment A, approx 30% of the cells migrates through the filter in 1.5 h
CILIATE
0
10
20 Temperature
30
CHEMOATTRACTION
40
(C)
FIG. 4. Temperature dependence of the cell transfer (strain B’7). Assay chambers were incubated for 1.5 h at different temperatures using PY (dil 1:lO) as chemoattractant (m). Tris control containing no attractant (Cl). Each point represents an individual assay chamber.
at 28°C when PY chemoattractant is present in compartment B. The time course of cell transfer to compartment B is shown in Fig. 2. It also shows that the signalto-noise ratio is around 10 and sometimes even higher (results not shown) when cell attraction using PY is compared with transfer to compartment B containing no attractant. The signal-to-noise ratio does not change significantly when using the cell concentrations below 1.5 X lo5 cells/ml. However, as will appear from Fig. 3, the signal-to-noise ratio decreases as the cell concentration in compartment A goes up. It is therefore appropriate to work with cell concentrations around 1.5 X lo5 cells/ml. At this cell concentration cell transfer can be determined with high accuracy with the electronic cell counter and the quantitative reproducibility of the chemosensory response for each individual chamber is good (see figures). In the above respects, the filter membrane technique is superior to a capillary technique, described previously (2). Moreover, the technical setup of the capillary assays is more laborious than the present modified Boyden chamber, which is simple to handle. When assaying for chemoattractant activity of a solution, cell transfers at 28°C into compartment B for l-l.5 h are optimal asjudged from signal-to-noise ratios. This is evident from Fig. 2, which shows a time course dependency very similar to that observed with the capillary assay (2). Moreover, the temperature dependence of cell transfer through the filter, which is shown in Fig. 4, indicates that the response is maximal around 30°C. Figure 5 shows the chemoattraction potential at different dilutions of
65
ASSAY
three attractants (PY, PDGF, and BSA). For both PY and PDGF optimal concentrations giving rise to maximal chemosensory responses are seen. Concentrations of chemoattractant, both lower and higher than the optimal one, result in smaller cell transfers to compartment B. For PY this optimal concentration is in the range of lo-100 pg/ml, whereas for PDGF the concentration is in the range 5-10 rig/ml. It should be noted that, when assayed in this system, some chemoattractants have low or virtually no attractant activity. This is true for a mixture of the 20 L-amino acids, or n-butyrate (results not shown), which have moderate chemoattractant effects when assayed in capillary systems (3). Even proteins (like bovine serum albumin) showing moderate to high activity in capillary assays have low or no activitywhen assayed in the paper membrane filter assay (Fig. 5). The same is true for fibroblast growth factor in concentration ranges from 1 to 200 rig/ml (results not shown). Finally, it should be mentioned that diffusion of chemoattractant from compartment B to A is of the order I
1o-4
I
10”
10-2
I
10-l
I
loo
10
Dilution factor FIG. 5. Percentage of cells (strain B7) transferredat 28°C into compartment B during 1.5 h as a function of the concentration of three test compounds: PY (m), PDGF (O), and bovine serum albumin (0). Cells (1.5 X 10’) were placed in compartment A and different dilutions of the test compounds were placed in compartment B. Cell transfer to B during 1.5 h was scored at 28°C. Absolute concentrations correspond to dilution factor 1 = 10’ and were PY, 10 mg/ml; PDGF, 100 rig/ml; bovine serum albumin, 1 mg/ml. Each point represents an individual assay chamber.
66
LEICK
of magnitude of 5-10% in 2 h. When the dye methyl blue is placed on one side of the membrane, 9% of it could be recovered after 2 h on the other side as assayed by spectrophotometry. Two other types of filter materials were tested: When using glass fiber filter GF/C as a separating membrane, no transfer of cells could be seen. With Whatman paper AGF 617 the same number of the cells were transferred both in the experimental chamber and in the chamber containing no attractant. Thus, the filter must have a pore size related to that of the test organism. The pore size of Whatman 3MM paper is around 20 pm whereas it is around l-l.5 pm for the GF/C glass fiber filter. The average dimensions of T. thermophila is 20 X 40 pm.
DISCUSSION
The present data show that free swimming ciliates will penetrate a filter paper membrane if a peptide chemoattractant gradient is established over the membrane. Zigmond (6) has calculated the time required to set up a gradient of a molecule the size of albumin as 45 s across a micropore filter of about 150 pm thickness. Such a gradient of a peptide or protein across a Whatman 3MM paper membrane (about 0.5 mm thickness) will take only a few minutes longer to set up; however, it will retain appreciable steepness during the few hours necessary to measure the chemosensory response. The various chemoattractants work somewhat differently with the present paper membrane assay than with the capillary assay (2). So far the experiments have shown that only peptides or protein chemoattractants, like PDGF and PY, are active, whereas various other chemoattractants, like FGF, n-butyrate, and bovine serum albumin, which are active in other assays (3-5), do not show significant activity with the present paper
ET
AL.
membrane assay. It may reflect that only strong chemoattractants, like PDGF and PY, are able to induce altered locomotion strong enough to make the cells swim through the Whatman paper membrane. On the other hand it may also reflect a qualitative difference between two types of chemosensory responses, which can be distinguished by the paper membrane assay. Finally, it should be mentioned that the dependence of the chemosensory response on the temperature (Fig. 4) is not simply related to the swimming speed of the cells. From 16 to 28°C the swimming speed of T. thermophila only increases from 0.34 to 0.52 mm/s (HellungLarsen, unpublished). It is rather likely that the cells move through the filter by oriented locomotion, but with reduced swimming speed, similar to that observed in semisolid gelatinized media (7). Oriented swimming was recently observed in nonviscous solutions (8). ACKNOWLEDGMENTS This work was supported by a grant from The Carlsberg Foundation. We thank Mrs. Ellen Hojer and Else Uhrenfeldt for typing the manuscript, and Ivan Abramowitz for making the Plexiglas assay chamber.
REFERENCES 1. Boyden, S. V. (1962) 2. Leick, V., and Helle,
J. Exp. Med. J. (1983) Anal.
115,453-466. Biochem.
135,466-469.
3. Leick, V., and Hellung-Larsen, P. (1985) J. Protozool. 32(3), 550553. 4. Hellung-Larsen, P., Leick, V., and Tommerup, N. (1986) Biol. Bull. 170,357-367. 5. Levandowsky, M., Cheng, T., Kehr, A., Kim, J., Gardner, L., Silvern, L., Tsang, L., Lai, G., Chung, C., and Prakash, E. (1984) Biol. Bull. 167,322-330. 6. Zigmond, S. H. (1977) J. Cell. Biol. 76,606-616. 7. Leick,
V. (1988)
Eur. J. Pro&tot.
23,354-360.
8. Hellung-Larsen, P., Leick, V., Tommerup, (1990) Eur. J. Protistol., in press.
N., and Kronborg,
D.
ANALYTICALBIOCHEMISTRY
(1990)
A Paper Membrane Filter Assay for Ciliate Chemoattraction Vagn Leick,l
Karin
Frederiksen,
Department of Biochemistry
Received
March
Inger
B, Panum Institute,
Lyhne,
and Per Hellung-Larsen
University
of Copenhagen, Blegdamsvej 3C, DK-2200 Copenhagen N, Denmark
27,1989
A quantitative bioassay for ciliate chemoattraction based on the Boyden assay is described with the ciliates Tetrah ymena thermophila and Tetrah ymena pyriformis as test organisms. A chamber is separated into two compartments by a Whatman 3MM filter, and a suspension of starved cells (- 10’ cells/ml) is placed in one compartment and a solution containing attractant in the other. The gradient of chemoattractant across the filter causes the cells to swim through the filter into the attractant-containing compartment where their appearance is determined by electronic cell counting. The assay described is convenient with a signal-to-noise ratio of approximately 10. It is shown here to work with the attractants proteose peptone and platelet-derived growth factor. 0 1990 Academic Press, Inc.
Boyden (1) introduced a convenient and reproducible method for assaying leukocyte locomotion by observations of the behavior of a cell population. He used a twocompartment chamber in which the cell suspension under test was separated from the chemotactic substance by a Millipore filter. Chemotactic factor diffused through the filter from below to reach the cells on the upper surface which responded by migrating through the filter by an amoebic motility. The number of leukocytes per unit area of the lower surface area of the membrane was counted microscopically as a function of time or concentration of test compound. Working with the ciliate protozoan Tetrahymena which is a free swimming ciliate we have developed various assay procedures to assay for the chemoattractant potential of a test compound (2-4). One of the methods (2) was based on the capillary principle. However, it was necessary to work at rather low cell concentrations (l-3 X lo4 cells/ml) in order to avoid high background levels 1 To whom
correspondence
should
0003.2697/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
be addressed.
Inc. reserved.
in control assays containing no attractant. The present paper describes a method where the capillary tubes are replaced by a paper membrane filter (Whatmann 3MM Chr). It is possible to work at much higher cell concentrations with this method than with the capillary technique. MATERIALS
AND
METHODS
Tetrahymenu thermophila (strain B7) or Tetrahymena pyriformis (strain GL) were grown to mid-stationary phase in PY2 medium: 1% proteose peptone (Difco) containing 0.1% yeast extract (Difco) and 0.03%0 sequestrene. For chemoattraction assay experiments, cells were collected by centrifugation at 5OOg,washed, and resuspended in 10 mM Tris-HCl, pH 7.4, to a cell concentration of 2 X lo5 cells/ml. Cells were starved in this medium for 16 h and transferred to fresh 10 mM Tris-HCl containing 0.15% (w/v) gelatin Type III (Sigma) to a final cell concentration of 1.5 X lo6 cells/ml (cell stock). The cells were preincubated in this medium for 1 h before use in chemoattraction experiments. Gelatin was included in order to serve as carrier medium for fibroblast growth factor (FGF) and platelet-derived growth factor (PDGF). However, using PY, the assay works perfectly well without the presence of 0.15% gelatin, which increases the viscosity of the medium only slightly. The two-compartment chamber used is shown in Fig. 1. The two compartments (-1.5 ml each) are separated by Whatman 3MM Chr paper (Cat. No. 3030117). Leakage out of the system through the Whatman paper was prevented by placing grease (white Vaseline) on the paper surrounding the chamber and squeezing the greased paper between the two Plexiglas blocks in order to seal the chamber. For chemoattractant experiments 0.9 ml 10 mM TrisHCl, pH 7.4, containing 0.15% gelatin (PDGF carrier) ’ Abbreviations used: PY, proteose trene; FGF, fibroblast growth factor; factor; BSA, bovine serum albumin.
peptone-yeast extract-sequesPDGF, platelet-derived growth
63
64
LEICK Whatman
3MM
FIG. 1. Paper membrane assay chamber. Whatman paper 3MM Chr separates two compartments A and B. Cells are placed in compartment A and the chemoattracting test compound in compartment B. After incubation at 2%30°C the cell number is registered by electronic cell counting using a Coulter electronic cell counter.
was placed in both compartments. Then 100 ~1cell stock was placed in compartment A immediately followed by 100 /*l test compound in compartment B. Mixing was carried out by filling and emptying the lOO-~1 pipette twice immediately after the addition of the aliquot. The
I
I
I
1
4(
Hours FIG. 2.
ET AL.
Appearance of cells (B7) in compartment B as a function of time at 28°C. Cells (1.5 X 106; strain B7) were placed initially in compartment A (total volume 1 ml). PY growth medium (final dilution 1:lO) was placed in compartment B (m). Control experiments without attractant were performed with the same number of cells (Cl). Each point represents an individual assay chamber.
0
I
I
1
2
Cells per ml x 1Oj in A FIG. 3. Number of cells (B7) transferred into compartment B as a function of the initial cell concentration in compartment A. Up to 2 X lo5 cells (strain B7) were placed in compartment A and the percentage of cells accumulating in compartment B during 1.5 h at 28°C was calculated (m). Control chambers containing no attractant (Tris) were run in parallel (0). Each point represents an individual assay chamber.
chamber was incubated at the appropriate temperature and the total cell number in the two compartments was estimated using a Coulter electronic cell counter by emptying the whole compartment (1 ml) and diluting with 9 ml counting liquid (0.1% NaN3, 0.9% NaCl); 0.5ml aliquots of this fixed cell suspension were counted in the Coulter counter. All determinations were carried out in duplicate assay chambers (see Figures). Statistical analysis was unnecessary in the individual experiment because of good reproducibility of double estimates and of signal-to-noise ratios. However, a day-to-day biological variation of cellular response was observed. The experiments described are, however, representative and carried out with cells taken from early to mid-stationary cultures (~2 to 4 days after inoculation). It should be noted that cells from the logarithmic growth phase (and late stationary phase) were not used. Porcine platelet-derived growth factor was obtained from R & D Systems Inc., Minnesota, and basic fibroblast growth factor was from Sigma. In some experiments other types of filter material (Whatman GF/C glass fiber filter or Whatman AGF 617) were used. RESULTS
When placing l-2 X lo5 cells in compartment A, approx 30% of the cells migrates through the filter in 1.5 h
CILIATE
0
10
20 Temperature
30
CHEMOATTRACTION
40
(C)
FIG. 4. Temperature dependence of the cell transfer (strain B’7). Assay chambers were incubated for 1.5 h at different temperatures using PY (dil 1:lO) as chemoattractant (m). Tris control containing no attractant (Cl). Each point represents an individual assay chamber.
at 28°C when PY chemoattractant is present in compartment B. The time course of cell transfer to compartment B is shown in Fig. 2. It also shows that the signalto-noise ratio is around 10 and sometimes even higher (results not shown) when cell attraction using PY is compared with transfer to compartment B containing no attractant. The signal-to-noise ratio does not change significantly when using the cell concentrations below 1.5 X lo5 cells/ml. However, as will appear from Fig. 3, the signal-to-noise ratio decreases as the cell concentration in compartment A goes up. It is therefore appropriate to work with cell concentrations around 1.5 X lo5 cells/ml. At this cell concentration cell transfer can be determined with high accuracy with the electronic cell counter and the quantitative reproducibility of the chemosensory response for each individual chamber is good (see figures). In the above respects, the filter membrane technique is superior to a capillary technique, described previously (2). Moreover, the technical setup of the capillary assays is more laborious than the present modified Boyden chamber, which is simple to handle. When assaying for chemoattractant activity of a solution, cell transfers at 28°C into compartment B for l-l.5 h are optimal asjudged from signal-to-noise ratios. This is evident from Fig. 2, which shows a time course dependency very similar to that observed with the capillary assay (2). Moreover, the temperature dependence of cell transfer through the filter, which is shown in Fig. 4, indicates that the response is maximal around 30°C. Figure 5 shows the chemoattraction potential at different dilutions of
65
ASSAY
three attractants (PY, PDGF, and BSA). For both PY and PDGF optimal concentrations giving rise to maximal chemosensory responses are seen. Concentrations of chemoattractant, both lower and higher than the optimal one, result in smaller cell transfers to compartment B. For PY this optimal concentration is in the range of lo-100 pg/ml, whereas for PDGF the concentration is in the range 5-10 rig/ml. It should be noted that, when assayed in this system, some chemoattractants have low or virtually no attractant activity. This is true for a mixture of the 20 L-amino acids, or n-butyrate (results not shown), which have moderate chemoattractant effects when assayed in capillary systems (3). Even proteins (like bovine serum albumin) showing moderate to high activity in capillary assays have low or no activitywhen assayed in the paper membrane filter assay (Fig. 5). The same is true for fibroblast growth factor in concentration ranges from 1 to 200 rig/ml (results not shown). Finally, it should be mentioned that diffusion of chemoattractant from compartment B to A is of the order I
1o-4
I
10”
10-2
I
10-l
I
loo
10
Dilution factor FIG. 5. Percentage of cells (strain B7) transferredat 28°C into compartment B during 1.5 h as a function of the concentration of three test compounds: PY (m), PDGF (O), and bovine serum albumin (0). Cells (1.5 X 10’) were placed in compartment A and different dilutions of the test compounds were placed in compartment B. Cell transfer to B during 1.5 h was scored at 28°C. Absolute concentrations correspond to dilution factor 1 = 10’ and were PY, 10 mg/ml; PDGF, 100 rig/ml; bovine serum albumin, 1 mg/ml. Each point represents an individual assay chamber.
66
LEICK
of magnitude of 5-10% in 2 h. When the dye methyl blue is placed on one side of the membrane, 9% of it could be recovered after 2 h on the other side as assayed by spectrophotometry. Two other types of filter materials were tested: When using glass fiber filter GF/C as a separating membrane, no transfer of cells could be seen. With Whatman paper AGF 617 the same number of the cells were transferred both in the experimental chamber and in the chamber containing no attractant. Thus, the filter must have a pore size related to that of the test organism. The pore size of Whatman 3MM paper is around 20 pm whereas it is around l-l.5 pm for the GF/C glass fiber filter. The average dimensions of T. thermophila is 20 X 40 pm.
DISCUSSION
The present data show that free swimming ciliates will penetrate a filter paper membrane if a peptide chemoattractant gradient is established over the membrane. Zigmond (6) has calculated the time required to set up a gradient of a molecule the size of albumin as 45 s across a micropore filter of about 150 pm thickness. Such a gradient of a peptide or protein across a Whatman 3MM paper membrane (about 0.5 mm thickness) will take only a few minutes longer to set up; however, it will retain appreciable steepness during the few hours necessary to measure the chemosensory response. The various chemoattractants work somewhat differently with the present paper membrane assay than with the capillary assay (2). So far the experiments have shown that only peptides or protein chemoattractants, like PDGF and PY, are active, whereas various other chemoattractants, like FGF, n-butyrate, and bovine serum albumin, which are active in other assays (3-5), do not show significant activity with the present paper
ET
AL.
membrane assay. It may reflect that only strong chemoattractants, like PDGF and PY, are able to induce altered locomotion strong enough to make the cells swim through the Whatman paper membrane. On the other hand it may also reflect a qualitative difference between two types of chemosensory responses, which can be distinguished by the paper membrane assay. Finally, it should be mentioned that the dependence of the chemosensory response on the temperature (Fig. 4) is not simply related to the swimming speed of the cells. From 16 to 28°C the swimming speed of T. thermophila only increases from 0.34 to 0.52 mm/s (HellungLarsen, unpublished). It is rather likely that the cells move through the filter by oriented locomotion, but with reduced swimming speed, similar to that observed in semisolid gelatinized media (7). Oriented swimming was recently observed in nonviscous solutions (8). ACKNOWLEDGMENTS This work was supported by a grant from The Carlsberg Foundation. We thank Mrs. Ellen Hojer and Else Uhrenfeldt for typing the manuscript, and Ivan Abramowitz for making the Plexiglas assay chamber.
REFERENCES 1. Boyden, S. V. (1962) 2. Leick, V., and Helle,
J. Exp. Med. J. (1983) Anal.
115,453-466. Biochem.
135,466-469.
3. Leick, V., and Hellung-Larsen, P. (1985) J. Protozool. 32(3), 550553. 4. Hellung-Larsen, P., Leick, V., and Tommerup, N. (1986) Biol. Bull. 170,357-367. 5. Levandowsky, M., Cheng, T., Kehr, A., Kim, J., Gardner, L., Silvern, L., Tsang, L., Lai, G., Chung, C., and Prakash, E. (1984) Biol. Bull. 167,322-330. 6. Zigmond, S. H. (1977) J. Cell. Biol. 76,606-616. 7. Leick,
V. (1988)
Eur. J. Pro&tot.
23,354-360.
8. Hellung-Larsen, P., Leick, V., Tommerup, (1990) Eur. J. Protistol., in press.
N., and Kronborg,
D.
