Biomedical Engineering Reference
In-Depth Information
partial fragments. It is unclear why this was the case, although we hypothesize that
the probability of the heavy chain portion of a partial fragment binding to a quantum
dot is considerably higher than the light chain portion because there is twice the
surface area for heavy chain binding and it is a condition that is sterically favored
(since the bend in the partial chain may tend to hide the light chain from the quan-
tum dot). Another potential explanation for the lack of heavy chain is methodologi-
cal. Given the intensity of other bands in the membranes, small amounts of free
heavy chain may have gone undetected given the exposure time we used to develop
the membrane, which if had been longer may have shown the presence of heavy
chains but would have overexposed the other darker bands resulting in uninterpre-
table smearing. Additional evidence that heavy chains covalently bound to quantum
dots originating from partial fragments remained bound to the quantum dots is
inferred by a nonspecifi c colloidal blue protein stain, which labels any protein in the
gel that did not transfer to the membrane (Fig.
4b
). Since blue bands appeared at the
position in the gels that corresponded to the quantum dots, some amount of residual
protein did remain on their surface. Given that most of the light chains were cleaved,
since they transferred strongly to the membrane, this residual protein is most likely
heavy chain. Regardless, for the purposes of calculating the amount of functional
antibody on quantum dot surfaces, this is of minimal importance, since it is the
amount of available light chain that we are interested in since it is the light chain that
contains the ligand-binding epitope. Another important consideration to note is that
the amount of partial fragments initially available for binding to quantum dots fol-
lowing the initial DTT reduction was very low, as evident in the reduced unconju-
gated IgG controls (Fig.
4a
, lanes 7-9). This point is an important consideration for
why the number of available functional antibody in the covalently conjugated con-
dition was calculated to be so low. (Note that no partial fragments were visible for
the quantum dot lanes because the entire partial chain cannot be cleaved intact from
the quantum dot since the SMCC linkage cannot be broken by DTT.) If antibodies
had been electrostatically attached to quantum dots, several bands would have
shown up in nonreduced lanes (Fig.
4
, lanes 2 and 3) because the gel would have
electrostatically separated the antibodies from the quantum dots according to their
molecular size and weight. Further indirect evidence that antibodies were cova-
lently bound is implied by the fact that quantum dots in nonreduced lanes did not
travel through the gels but remained in the loading wells due to the large size of the
unreduced complex (visible as hyperintense signals in the loading wells for lanes 2
and 3 of the SDS-PAGE).
We ran the same experiments with biotinylated antibodies and streptavidin-
coated quantum dots at 2:1 and 1:1 antibody-to-quantum dot molar ratios.
Biotinylated antibodies have biotin molecules throughout the entire antibody, which
results in the IgG molecules being conjugated to quantum dots presumably in all
possible spatial arrangements (Fig.
4c
). Importantly and very differently from the
direct covalent conjugation reaction, using the biotin-streptavidin system, the entire
antibody molecule is conjugated to the quantum dot; it is not reduced into its light
chain and heavy chain fragments prior to binding. Similar to the covalent antibody
conjugation method, nonreduced conditions resulted in quantum dots remaining in
