7.3 Equilibration on the surface.
Once trapped by the surface, two-dimensional translation of the molecule might still
occur. In the study by Rivetti et al. (Rivetti, Guthold et al. 1996) molecules
were immobilized by drying prior to imaging, leading to the possibility that molecules
were able to equilibrate in two dimensions (2D equilibration). Molecules that become
immobilized upon adsorption are ‘3D-trapped’ in as much as their conformation
represents the solution conformation projected onto a surface. (Presumably adsorption and
imaging in buffer with no intermediate drying requires 3D trapping.) These two adsorption
processes lead to quite different conformations with different end-to-end lengths, as
illustrated in Fig. 17. 3D trapping results in a projection which has an end-to-end length
which is 
of the free value for a worm-like chain of 
(in
the 
limit) where P is the persistence length and L is the contour length of the
polymer. This follows from the collapse of the dimension perpendicular to the surface,
taken to be z, so that 
. For 2D equilibration, Rivetti et al. (Rivetti, Guthold et
al. 1996) show that 
. Their study utilized DNA of 389nm length for which the
difference in end to end length for the two mechanisms is large. Table I summarizes the
results of some mica pre-treatments. In general, the more chemically active surfaces
result in 3D trapping. Thorough rinsing prior to DNA deposition appears to produce an
active surface, presumably because of protonation of the mica.
Figure 17:Showing schematically the difference in polymer conformation for
trapping from solution, resulting in a 2D projection of the solution structure (a) versus
equilibration on the surface resulting in a more extended structure (b).
Ideal and reproducible 3D trapping would be preferable in as much as the imaged
conformation would most closely reproduce a projection of the solution conformation. This
is illustrated by a comparison of bare mica (2D equilibration) and spermine-treated mica
(presumably 3D trapping) taken from the work of Tanigawa and Okada (Tanigawa and Okada
1999), reproduced in Fig. 18. The sample was a pUC19 supercoiled plasmid on bare mica
(Fig. 18a) and spermine treated mica (Fig. 18b). The supercoiling is evident in the
trapped samples but absent in the samples that have equilibrated on the mica surface.
Presumably the equilibrated sample has substituted writhe for twist in order to maximize
contacts with the surface.
Figure 18: Supercoiled plasmid imaged on untreated mica (2D trapped,
(a)) and spermine treated mica (3D trapping, (b)). The supercoiling is retained in (b) but
is removed in (a). Scale bars are 500nm. (From Tanigawa and Okada (Tanigawa and Okada
1999) with permission).
7.4 Salt effects on adsorption and conditions for weak adsorption
Muscovite mica consists of tetrahedral double sheets of (Si/Al)2O5 electrostatically
linked by potassium ions. Dissolution of these ions at a hydrated surface gives rise to a
surface charge of -0.0025C/m2at neutral pH,
equivalent to 0.015 electronic charges per nm2 (Pashley 1981). Thus, electrostatic
interactions play an important role in the adsorption and binding of charged biomolecules.
In the case of negatively charged molecules like DNA and purple membrane, it is necessary
to carry the adsorption out at high salt concentration in order to screen the
electrostatic repulsion and permit adsorption by attractive van der Waals interactions.
Muller et al. have investigated the adsorption of purple membrane (surface charge = -0.05
C/m2) as a function of salt concentration (Muller, Amrein et al. 1997). Optimal adsorption
occurred for concentrations of 1:1 electrolytes (LiCl, NaCl, KCl) above 40mM and for 1:2
electrolytes (MgCl2, CaCl2, NiCl2) above 1mM. Similar considerations apply to (undesired)
adsorption onto the SPM tip. Silicon nitride and silicon tips have a considerable amount
of surface oxide which ionizes by dissolution of surface ions to produce a surface charge
in the region of -0.032 C/m2 at pH 7 (Butt 1991). Muller et al. (Muller, Amrein et al.
1997) suggest adsorbing the target molecules onto the substrate at high salt and then
diluting the electrolyte before placing the sample into the microscope so as to reduce
subsequent adsorption onto the scanning probe. Hansma et al (Hansma and Laney 1996) have
carried out a systematic study of the effect of cation radius on DNA adsorption onto mica
finding stronger adsorption in the presence of highly charged transition metal ions.
While strong adsorption is desirable for high resolution imaging, weak attachment is
important if processes are to be studied. Thompson et al. have described methods for
modulating the ionic environment so as to produce loose enough attachment for DNA to be
visible in DFM images yet still mobile on the surface (Thomson, Kasas et al. 1996).
Zuccheri et al. (Zuccheri, Dame et al. 1998) have used similar methods to make a movie of
DNA undergoing conformational changes as the ionic strength is raised and lowered during
imaging.
7.5 Silane treatments of mica
One of the first methods for reliable AFM imaging of DNA was based on the attachment of
positive amine groups to the mica surface using aminopropyltriethoxysilane (APTES)
(Lindsay, Lyubchenko et al. 1992; Lyubchenko, Gall et al. 1992; Lyubchenko, Jacobs et al.
1992; Lyubchenko, Shlyakhtenko et al. 1993). This approach binds DNA very strongly, too
much so for studies of processes in situ (Zuccheri, Dame et al. 1998). As originally
described (Lindsay, Lyubchenko et al. 1992) it is capable of producing unacceptable
roughening of the mica surface (Tanigawa and Okada 1999). However, we have found it to be
one of the most reproducible methods for adsorbing negatively charged molecules onto mica,
the treatment of the surface being more uniform (in our hands) when carried out in
solution. A fresh aqueous suspension of APTES (10,000:1 water:APTES) is contacted to the
mica for a few minutes and then rinsed after which a drop of buffered solution of the
biomolecules is placed onto the treated mica surface.
