The Self-assembly of Microspheres


University of Bath Department of Mechanical Engineering

Technical Report 51/97

I.A. Gyepi-Garbrah

Oct 1996 - Sept 1997
 
 

0. Introduction

This report describes a series of experiments designed to test the feasibility of constructing microscopic structures from micron-scale polymer beads. The beads were coated to encourage them to link together in aqueous suspension, and thereby to self-assemble into simple geometric shapes.
 
 

1. experiments

1.1. Coating of Tosyl activated Dynabeads (2.8 and 4.5 m m) with concanavalin A and ovalbumin

The purpose of this experiment was simply to see if Dynabeads coated in con-A would stick to beads coated in ovalbumin.

The Dynabeads were thoroughly re-suspended by vortexing vigorously for one minute. Beads (1 ml) were immediately removed and placed in the magnetic particle concentrator (MPC) and left for one minute. The supernatant was removed and the beads were washed briefly in 1 ml of PBS by vortexing before being placed in the MPC. The supernatant was removed and the pellet was re-suspended in 50 m l of PBS using a Gilson pipette and then added to 1 ml of protein solution (5 mg ml-1), gently re-suspended, and left rotating overnight at 37oC. The beads were placed in an MPC for 1 minute, excess protein was removed, and the beads were re-suspended in 1 ml of PBS and left on a Dynal mixer for 5 minutes. The washing procedure was repeated a further 5 times and the beads were finally re-suspended in 1 ml of PBS and stored for one month at 4oC.

Binding conditions:

Con-A beads (2 m l) + ovalbumin beads (2 m l) + 20 m l PBS buffer

Uncoated beads (2 m l) + con-A beads (2 m l) + 20 m l PBS buffer

Uncoated beads (2 m l) + ovalbumin beads (2 m l) + 20 m l PBS buffer

Uncoated beads only (4 m l) + 20 m l PBS buffer

All beads were thoroughly re-suspended by vortexing and the required volume was removed immediately. 1.2. Results

Initial experiments of coating magnetic Dynabeads with ovalbumin and concanavalin A proved inconclusive. There was no apparent difference between coated beads and the control uncoated beads (Figures 1A, B and C). A major problem with Dynabeads is that the beads tend to form aggregates whether protein coated or uncoated (possibly due to residual magnetism). Various solvents and solutes were used to try to reduce the non-specific binding (0.5 M urea, diamino benzoate, nicotinamide, Triton, Tween and non IdentP, DMF, acetone, dioxan, THF and DMSO). Tween worked but made the solution far too viscous to be useful as a solvent.

A)

B)

C)

Figure 1: A) Con-A and ovalbumin beads, B) Con-A beads and uncoated beads, C) control uncoated beads (all beads 4.5 m m).
 
 
 
 

1.3. Binding of Con-A PVA-polystyrene beads (~55 m m) to yeast cells (4.5 m m).

As the previous experiment was inconclusive, this one was intended to create a strong binding between the con-A and sugar residues on the yeast. Yeast cells were chosen as they are roughly spherical, roughly the right size, and might be sufficiently flexible to allow a slight degree of conformality with the surfaces of the beads.

PVA-polystyrene beads (~5 g, see Appendix, Section 1.3) were activated using cyanuric chloride. The activated beads were separated into two equal portions and added to different coupling buffers: PBS pH 7.3 (10ml) and sodium acetate buffer pH 5 (10ml) which contained Con-A at a concentration of 15 mg ml-1. The beads were mixed for 24 hours at room temperature. Approximately 2.5g of Con-A beads were re-suspended into 5 ml of Tris buffer which contained 1 mM Ca2+, 1 mM Mn2+, 0.5 M NaCl, and 0.05% (w/v) azide.

Dried baker's yeast (~300 mg) was suspended in 20 ml of Tris buffer pH 7.3 which contained 1 mM Ca2+, 1 mM Mn2+, 0.5 M NaCl, and 0.05% (w/v) sodium azide.

Binding conditions:

Tris Buffer (320 m l, 1 mM Ca2+, 1 mM Mn2+, 0.5 M NaCl) and 20 m l of Con-A beads + 20 m l of yeast suspension. The suspension was placed on a Dynal rotating mixer overnight and viewed under a microscope. 1.4. Results

The yeast cells were attached in great numbers to the con-A PVA-polystyrene beads. The interaction is due to the sugar residues on the yeast cells binding to the immobilised con-A. The strength of the binding is such that the larger PVA-polystyrene beads (~55 m m) are held together by yeast cells (~ 4m m) (Figure 2).

A)

B)

C)

Figure 2: A), B), and C) Con-A PVA-polystyrene beads (~ 55 m m) bound by yeast cells (~4 m m).
 
 

1.5. Binding of con-A- coated Dynabeads with yeast cells.

Con-A-coated Dynabeads (Section 1.1) were used.

Binding conditions:

Tris Buffer (320 m l, 1 mM Ca2+, 1 mM Mn2+, 0.5 M NaCl) and 20 m l of Con-A coated Dynabeads (2.8 or 4.5 m m) + 20 m l of yeast suspension. Controls of uncoated beads were mixed with yeast cells. The Eppendorf tubes were placed on a Dynal rotating mixer overnight and viewed under a microscope. A)

B)

Figure 3: A) control: uncoated Dynabeads (4.5 m m) + yeast cells and B) Con-A Dynabeads (4.5 m m)+ yeast cells.
 
 
 
 

A)

B)

Figure 4: A) Con-A Dynabeads (2.8 m m) + yeast cells and B) Con-A Dynabeads (4.5 m m) + yeast cells.
 
 
 
 

1.6. Results

Con-A Dynabeads bound to the yeast cells quite strongly. There was a clear distinction between uncoated beads that did not interact with the yeast cells and coated beads (Figures 3A and 3B). The uncoated beads tend to aggregate with other uncoated beads (Figure 3A).
 
 

1.7. Binding of ADH Dynabeads with blue dextran PVA-polystyrene beads

The previous experiment having succeeded with the binding of beads and yeast, this experiment was an attempt to get beads to bind to beads by allowing longer molecules in the attachment chemistry. The idea was that this would make attachment more robust even though a sphere-sphere contact is not conformal.

ADH Dynabeads (Section 1.1) were used with blue dextran PVA-polystyrene beads (Appendix Sections 1.1 and 1.2).

Binding conditions:

PBS buffer (320 m l) + 20 m l of blue dextran beads + 20 m l ADH beads (2.8 m m). The suspension was rotated on a Dynal mixer overnight and viewed under a microscope. Control: PBS buffer (320 m l) + 20 m l dextran PVA-polystyrene beads + 20 m l ADH Dynabeads. 1.8. Results

Figure 5 shows that the control dextran PVA-polystyrene beads do not interact with the ADH Dynabeads.

Figure 5: Control dextran PVA-polystyrene beads (~55 m m)+ ADH Dynabeads (4.5m m).
 
 

Figure 6: (3 images) Blue dextran PVA-polystyrene beads (~55 m m)+ ADH Dynabeads (4.5 m m).

In Figure 6, the ADH beads show a strong interaction with the blue dextran PVA-polystyrene beads. The strength of the interactions is such that the smaller beads are able to hold larger beads together despite the very small area of contact.
 
 

1.9. Gold sputtering of blue dextran PVA-polystyrene beads

The previous experiment having succeeded in getting bead-bead binding working, this experiment was an attempt to make active patches on the surface of the beads, so that they would only bind at certain positions on their surface.

PVA Polystyrene beads (~2.5g) were gold sputtered at various stages of synthesis of blue dextran PVA-polystyrene in an attempt to mask some of the binding sites with a hemispherical coating of gold. However, the best results were obtained with blue dextran PVA-polystyrene beads that had already been synthesised. The beads were sputtered with gold for 7 minutes with a 1 minute gap between each sputtering burst (3 + 4 minute bursts). The beads were re-suspended in 1.5 ml of PBS buffer.

Binding conditions:

PBS buffer (320 m l) + 20 m l of gold sputtered blue dextran beads + 20 m l ADH beads (2.8 m m). The suspension was rotated on a Dynal mixer overnight.
  1. Results
The gold hemispheres were difficult to view under a microscope. The ADH Dynabeads appear to coat the surface of the blue dextran PVA-polystyrene beads (Figures 7 and 8). It was difficult to achieve a monolayer on the sputtering aluminium stub. As a result only a small proportion of the beads were actually sputtered with gold. The experiment was inconclusive.

Figure 7: Gold-sputtered blue dextran PVA-polystyrene beads (~55 m m) and ADH beads (4.5 m m)

Figure 8: (2 images) Gold sputtered blue dextran PVA-polystyrene beads (~55 m m) and ADH beads (4.5 m m)
 
 

1.11. Polarisation studies

The purpose of this experiment was similar to the previous one.

This experiment was an attempt to see if the gold sputtering had prevented the binding of ADH beads to the gold hemispheres, thus restricting binding to the 'non gold' region of the beads. A method was developed so that unbound ADH beads were removed.

Binding conditions:

PBS buffer (320 m l) + 60 m l of gold sputtered blue dextran beads + 60 m l ADH beads. The suspension was rotated on a Dynal mixer overnight and viewed under a microscope. The entire suspension was placed onto 1 ml of saturated sucrose solution in an Eppendorf that was placed in an MPC. This was left for 3-4 hours, and then the beads were carefully removed from the top of the interface between buffer and the sucrose solution (Figure 9 below).

Figure 9: Removal of excess ADH Dynabeads from gold-sputtered blue dextran PVA-polystyrene beads (~55 m m).

The beads were placed onto an aluminium electron microscope stub and allowed to dry in a laminar flow hood overnight. The sample was viewed on the electron microscope using 5 kV.
 
 

1.12. Results

The electron micrographs (Figures 10 and 11; made, of course, without the usual electron microscope gold film) show that there are still ADH beads attached to the blue dextran PVA-polystyrene beads despite the beads? being subjected to a magnetic field. However, it not possible to see any gold shadowing on the beads? surfaces. The beads displayed bright patches, which are due to charging of the polystyrene bead surface, because there is not a route for conducting electrons away.


 
 





Figure 10: (3 images) Gold sputtered blue dextran PVA-polystyrene beads (~55 m m) and ADH beads (4.5 m m)

Figure 11: (2 images) Gold sputtered blue dextran PVA-polystyrene beads (~55 m m) and ADH beads (4.5 m m)
 
 

1.13. Synthesis of biotin beads

This experiment was intended to try out a different attachment chemistry with the gold shadowing.

Aminocaproic acid (12g) was dissolved in 30 ml of 0.1 M MES buffer pH 4.7. The pH was readjusted after the acid had dissolved. Hydrazide beads (1.5 m m) were thoroughly re-suspended before removing 100 m l of the suspension. The beads were washed (x3) with MES buffer pH 4.7 and finally re-suspended in 4 ml of MES buffer. Aminocaproic acid (4 ml) was added to the bead suspension followed by 3g of 1-ethyl-3-(3-dimethylaminopropyl carbodiimide, EDC). The suspension was left rotating for 2 hours at room temperature. The beads were spun down and the supernatant was removed followed by washes with water, 1 M NaCl, and finally water again.

The beads were re-suspended in 10 ml of d-biotin (12 mg) dissolved in 0.1 M MES buffer pH 5.8. EDC (3g) was added to the suspension and was left rotating overnight at room temperature. The beads were spun down and the supernatant was removed and the beads were washed with: water, 1M NaCl, water and PBS. The beads were finally re-suspended in 3 ml of PBS.
 
 

1.14. Gold shadowing using biotin beads and colloidal gold

Biotin beads (10 m l) were re-suspended by vortexing in 1 ml of colloidal gold (diluted 4 fold in PBS and left to equilibrate for 1 hour on a Dynal mixer). The suspension was left for 3 hours at room temperature on a Dynal mixer; the suspension was spun at 13,000 rpm for 5 minutes. The supernatant was removed and 1 ml of PBS buffer was added. The suspension was placed on a Dynal mixer for 5 minutes. The process of washing and resuspension was repeated 5 times to remove any remaining glycerol. This step was necessary otherwise the sample would not dry in the fume hood. The suspension was placed on an aluminium electron microscope stub and left overnight in a laminar flow hood.

1.15. Results

Due to excessive charging, no SEM pictures were taken of the biotin beads. The gold deposition was carried out under a much higher vacuum. As a result the gold should be immobilised as distinct patches or hemispheres; depending on the surface area of bead subjected to the gold beam. Thus it was not possible to view the gold-shadowed hemispheres.
 
 

1.16. Assembly of tetrahedra based on the interaction of streptavidin and biotin beads.

Gold sputtering having not proved useful in the previous experiments as a way of creating patches on the beads, the next tests were directed at trying to make shapes by controlling bead diameter. The idea here was to form, for example, tetrahedra by using beads of two diameters, the smaller of which will just fit in the void in a tetrahedron formed by the larger.

In order to form tetrahedra one requires a bead diameter ratio of approximately 7:1. Streptavidin beads (10.4 m m) and biotin beads (1.53 m m) fitted this criterion.

Binding conditions:

Streptavidin beads (10.4 m m) were thoroughly re-suspended by vortexing and 60 m l of the suspension was removed. The beads were washed three times with PBS (1 ml). The beads were finally re-suspended in 400 m l of PBS. Biotin beads (5 m l) were re-suspended in 60 m l of PBS and added to the streptavidin beads the suspension left rotating overnight at room temperature. 1.17. Results

The complexes formed are varied: there are strings, tetrahedra, dimers and trimers as well as much larger complexes (Figures 12-18). The strength of the interactions is strong enough to hold the complexes together. Comparison of streptavidin beads only (control) and biotin and streptavidin beads shows that the presence of biotin beads increases the number of complex formation (Tables 1 and 2, Figure 19). Strength was not so great as to maintain one of the tetrahedra when an attempt was made to move it by gently pushing its top bead sideways with a micromanipulator.
 
 

Figure 12: (2 images) Control experiment - Shapes formed from streptavidin beads only.
 
 

Figure 13: (2 images) Strings and other shapes formed from streptavidin and biotin beads.

Figure 14: (2 images) Shapes formed from streptavidin and biotin beads.

Figure 15: (2 images) Shapes formed from streptavidin and biotin beads.

Figure 16: (2 images) Tetrahedra formed from streptavidin and biotin beads.
 
 

Figure 17: (2 images) Tetrahedra formed from streptavidin and biotin beads.
 
 

Figure 18: Tetrahedron formed from streptavidin and biotin beads.
 
 
 
 

FORMATION OF COMPLEXES

Table 1 - Control: streptavidin beads only
 
image
I
II
III
IV
V
VI
VII
VIII
IX
X
single
97
80
83
87
80
82
72
71
64
69
double
14
12
17
8
20
10
15
24
12
11
treble
5
7
7
6
6
1
3
4
4
5
four+
8
4
7
6
3
6
5
2
7
3

FORMATION OF COMPLEXES

Table 2 - Streptavidin and biotin beads
 
image
I
II
III
IV
V
VI
VII
VIII
IX
X
single
67
62
66
62
65
69
61
58
72
59
double
9
13
16
13
15
13
14
14
12
14
treble
2
4
5
8
6
11
10
8
7
7
four+
10
11
14
13
13
9
4
11
13
13

Figure 19 A: Number of complexes formed with streptavidin beads only

Figure 19 B: Number of complexes formed with streptavidin and biotin beads.
 
 
 
 
 

1.18. Fluorescence experiments

Having successfully constructed tetrahedra and demonstrated the formation of other complexes, the next step was to try to flag the binding sites in order to be able to distinguish between true binding and mere chance alignment. Fluorescence was used for this.

Binding conditions:

Streptavidin beads (10.4 m m) were thoroughly re-suspended, 60 m l of the suspension was removed, and the beads were washed three times with PBS (1 ml). The beads were finally re-suspended in 400 m l of PBS. Biotin beads (5 m l) were re-suspended in 60 m l of PBS and added to the streptavidin beads. The suspension was left rotating overnight at room temperature.

The beads were pulsed in a minifuge to the maximum speed and then the spinning was stopped. The supernatant was removed and 1 ml of anti-streptavidin FITC antibody (100-fold dilution) was added. The Eppendorf was covered in foil and the suspension was left on a Dynal mixer for 3 hours. Then the beads were spun down as described previously, the supernatant was removed, and the pellet was re-suspended in 1ml of PBS. The suspension was left on a Dynal mixer for 5 minutes and spun down; the process was repeated three times to remove excess FITC antibody.
 
 

 1.19. Results

Figures 20-22 clearly indicate that when biotin beads are bound to streptavidin beads there is strong fluorescence. Conversely, the phase images show that neighbouring biotin and streptavidin beads do not show any fluorescence.

Figure 20: Phase-contrast image and corresponding fluorescence image of streptavidin and biotin beads.
 
 

Figure 21: Phase- contrast image and corresponding fluorescence image of streptavidin and biotin beads.
 
 


Figure 22: Phase contrast image and corresponding fluorescence image of streptavidin and biotin beads.
 
 
 
 

1.20. Dilution effects

The anti-streptavidin FITC antibody was diluted 400 fold. The reason for this experiment was to reduce the intensity of the FITC signal. The beads were viewed with a confocal microscope as before. Figure 23 shows that the resulting signal is dispersed in four directions from the single biotin bead in the middle of the tetrahedron.

Figure 23: Phase-contrast image and corresponding fluorescence image of streptavidin and biotin beads forming a tetrahedron.
 
 

1.21. Double labelling experiment

The previous attempts to create binding patches using gold sputtering having proved inconclusive, it was decided to try a different approach. Beads were bound to a flat membrane, the remaining active sites on them blocked, and then they were separated from the membrane thus leaving - it was hoped - a single active patch where the binding to the membrane had occurred. Fluorescence would again be used to clarify what was happening.

An Immunodyne ABC (Pall) pre-activated membrane (20 x 10 cm) was placed in 100 ml of 0.2M sodium phosphate buffer pH7.5 that contained 10g of 1,6-diaminohexane dihydrochloride. The membrane was shaken for 24 hours at room temperature and washed with copious amounts of water. D-biotin (120 mg) was dissolved in 100 ml of MES buffer pH 5.8 and the membrane was placed in the solution. 1g of EDC was added and the membrane was shaken overnight at room temperature. The membrane was washed with copious amounts of water and stored under water with 0.05% (w/v) sodium azide at 4oC.

The following six steps were then taken:

1) Streptavidin beads were bound to the iminobiotin membrane in sodium carbonate buffer pH 11, containing 0.5 M NaCl (24 hours).

2) Biotinylated concanavalin A was immobilised onto streptavidin beads.

3) Excess concanavalin A was washed away carefully.

4) The pH was adjusted to pH4 with sodium acetate buffer, containing 0.5M NaCl. The change in pH protonated the iminobiotin molecule and reduced the binding, such that the streptavidin beads were released from the membrane.

  1. Streptavidin beads were mixed with biotinylated beads (green).
  2. Complexes are incubated with anti-biotin Lissamine Rhodamine antibody.

 
 
 
 

1.21. Results

Figure 24 shows the larger streptavidin beads have been double-labelled. The red fluorescence on the surface of the beads is due to the biotinylated con-A. Although biotin molecules of the con-A are attached to the streptavidin molecules on the bead surface there are ?free? biotin molecules on the con-A to interact with the anti-biotin Lissamine Rhodamine antibody. Part of the streptavidin bead that was bound to the membrane surface will not have any bound con-A. Thus this ?bare? patch is able to bind biotin beads which will also bind to the anti-biotin Lissamine Rhodamine antibody. The biotin beads appear as large bright red spots on the streptavidin bead surface.

Figure 24: Double-labelling experiment
 
 

2. Conclusions

These experiments have demonstrated that:

    1. Con-A and ovalbumin did not give good bead binding;
    2. Con-A-coated beads could be easily bound to the sugar residues on yeast;
    3. ADH Dynabeads could be bound to blue-dextran PVA-polystyrene beads forming chains and other shapes;
    4. Gold sputtering did not give good results in attempts to create active patches on the beads;
    5. Streptavadin and biotin gave good bead binding, and could be used to create tetrahedra if the bead diameters were chosen correctly. Chains and other shapes were also formed;
    6. Fluorescence gave good indications of active bead binding; and
    7. Active patches could be created by binding beads to a flat membrane, blocking the remaining sites on the beads, and then separating the beads from the membrane; smaller beads could then be attached to those active patches.
Appendix 1 - Synthesis

1.1. Synthesis of PVA-polystyrene beads

This method is adapted from the methods of Garvey (1974) and Tuncel et al (1993). Polystyrene beads (100g, copolymer of 80% styrene and 20% divinylbenzene) were added to a stirred solution of aqueous polyvinyl alcohol (PVA, Av. Mw 27,000, 98% hydrolysed) (10 mg ml-1, 1L). The PVA was allowed to adsorb for 16h at 20oC, terephthalaldehyde (470 mg) in 100 ml dioxan was added, and the suspension was then acidified with 16.7 ml of 6 M HCl and left to crosslink overnight. The beads were washed on a sintered funnel with distilled water (2 L), hot water (60oC, 1 L) methanol (200 ml) and distilled water (2 L). The support was dried on a sintered funnel and left for two days at 37oC, and stored in the dark at room temperature. The amount of PVA adsorbed onto the support could be determined by analysis of the supernatant taken before and after adsorption.

PVA-polystyrene beads (10 g) were placed in 100 ml of 1 M NaOH and heated to 60oC. 2-Hydroxyethyl methacrylate (6.1 ml, 0.5 M) was added to the suspension and stirred at 60oC for 2 hours. The beads were washed with 200 ml of water and suspended in 17 ml of 2 M NaOH containing sodium borohydride (2 mg ml-1). 8.5 ml of epichlorohydrin was added dropwise and the suspension was stirred overnight. The epoxy-activated beads were washed sequentially with water (500 ml), acetone (100 ml) and water (500 ml).

1.2. Synthesis of Blue Dextran PVA-polystyrene beads

Epoxy activated PVA-polystyrene beads (10 g) were prepared as described above. The beads were added to 100 ml of 0.25 M NaOH that contained dextran (250 mg ml-1) and sodium borohydride (2 mg ml-1) and stirred for 5 hours at 50oC. Cibacron blue was immobilised using the revised method of Lowe and Pearson (1984). Cibacron blue (200 mg) was dissolved in 100 ml of water; dextran PVA-polystyrene (10 g) was added and stirred for 10 minutes. NaCl (5.9 g, 1M) and Na2CO3 (21.2 g, 2 M) was added and the suspension was stirred at 60oC for 4 hours. The beads were washed with water until the washings were virtually colourless. The beads were packed into a 25 ml column and washed with 200 ml of 1 M NaCl, 400 ml of 50 % (v/v) ethanol/water, 100 ml of 1 M NaCl and finally with 100 ml of 50 mM sodium phosphate buffer pH 7.9. The beads were then stored in sodium phosphate buffer with a few crystals of sodium azide at 4oC.

1.3. Cyanuric chloride activation of Dextran Pva-Polystyrene Beads

This method was developed by Pittfield, (1993). PVA-polystyrene (5g) was placed in 50 ml of 1 M NaOH and was dried on a glass sinter. The beads were added to a solution of cyanuric chloride (100 mM, 50 ml) in dry acetone for 10 minutes, and washed with 80 ml each of: acetone, 50% (v/v) acetone/water, and water. The support was re-suspended in coupling buffer.
 
 

Appendix II - Suppliers






2.1. Biochemicals

Alcohol dehydrogenase, concanavalin A, biotinylated concanavalin A, dextran and ovalbumin were all purchased from the Sigma Chemical Company, Poole, Dorset.

Anti-streptavidin FITC antibody was purchased from Vector Laboratories Ltd, Peterborough.

Anti-biotin Lissamine Rhodamine antibody was purchased from Stratech Scientific Ltd, Luton.

2.2. Chemicals

All reagents were analytical grade unless otherwise stated. Epichlorohydrin, terepthalaldehyde, 6-aminocaproic acid, 2-hydroxyethyl methacrylate, cyanuric chloride and MES were all purchased from the Aldrich Chemical Company, Gillingham, Dorset.

Cibacron blue, iminobiotin, biotin, sodium borohydride, 1-ethyl-3-(3-dimethylaminopropyl carbodiimide and 1,6-diaminohexane dihydrochloride were all purchased from the Sigma Chemical Company, Poole, Dorset.

Methanol, ethanol, acetone, hydrochloric acid, sodium hydroxide, sodium dihydrogen phosphate, disodium hydrogen phosphate were all purchased from FSA laboratory supplies, Loughborough, Leics.

Immunodyne ABC (Pall) pre-activated membrane was a kind gift from Pall Europe Ltd, Portsmouth.

The chromatographic support (average diameter 55 µm) was a copolymer of 80% styrene and 20% divinyl benzene and was purchased from Fluka Chemicals, Dorset.