The Localization of Cationic Steroid Antibiotics in Escherichia coli
through Transmission Electron Microscopy
Faculty Mentor: Paul Savage, Chemistry
The growing numbers of antibiotic resistant bacteria has caused concern in the health
community. To combat this enemy, our lab has developed an arsenal of new compounds called
cationic steroid antibiotics (CSAs). To date, many of these compounds have successfully killed
bacteria resistant to common antibiotics. Although we have successfully killed these bacteria, the
question still remains as how this occurs. If we could identify the mode of action by which these
compounds kill bacteria, then scientists would be able to develop more effective CSAs.
We have theorized how CSAs likely kill bacteria. CSAs are positively charged while the
membrane of gram-negative bacteria such as Escherichia coli (E. coli) is negatively charged.
The electrostatic attraction caused by these charge differences brings the CSAs and E. coli
together. Next, the compound either binds to a receptor on the surface of the membrane causing a
molecular reaction that kills the bacteria or it passes through the membrane and causes death by
some other means. To distinguish between these two possibilities, we devised a plan to visualize
where our compound is in relation to the bacteria cell. Previously, researchers in our lab labeled
CSA treated E. coli with gold particles that could be seen by transmission electron microscopy
(TEM). They believed that the gold would interact with sulfur containing groups on the
compound and thus show whether or not the compound had entered into the cell. These results
depended on the binding of the gold to the CSA specifically and not to any other part of the cell.
They obtained several TEM images that showed the gold aggregated inside the cell. These
results were encouraging, but there remained doubt as to whether or not the gold was interacting
with the compound or a part of cell (the DNA for instance). Our goal was to find a way to prove
more convincingly that our compound was indeed passing into the cell. To that end, we
developed a new procedure based on the molecular interaction of specific molecules, biotin and
streptavidin. Biotin binds non-covalently to streptavidin with very high affinity. Streptavidin can
be conjugated with colloidal gold particles.i Chemists in our lab prepared a CSA with biotin
attached to it. We hypothesized that the streptavidin-gold conjugate, when incubated with this
biotinylated CSA, would provide a very effective marker that could be visualized using TEM.
Based on the results obtained previously, we expected to see the gold-labeled CSAs inside of the
Our methods were very similar to those used in earlier experiments done in the lab. Our first goal
was to replicate successfully what had been done previously then advance to adding the
streptavidin-gold conjugate to the samples. The basic procedures were as follows. We first grew
the E. coli for 24 hours in Mueller-hinton broth at 37° Celsius. Then we placed bacteria into
conical vials and spun them down in a centrifuge. Following centrifugation, we treated the
experimental sample with varying concentrations of the biotinylated CSA for one hour. In the
next step we fixed the bacteria with a phosphate buffered solution (PBS) containing
glutaraldehyde. Next, the glutaraldehyde was removed through several washes with PBS. Then
we suspended the pellets in warm agarose gel and allowed them to solidify while centrifuging
the mixture. This resulted in a bacteria pellet suspended in agarose gel. Then we dehydrated the
agarose/bacteria pellet with varying degrees of ethanol. Once the pellet was suspended in 70%
ethanol, we sent it to the microscopy lab for final preparation and image development.
Our first attempts failed. Based on previous images, we expected to see bleeding (vesicle like
protrusions around the cell membrane) of the bacteria cells, but there was none. To solve this
problem, we thought that it was necessary to try higher concentrations of the biotinylated CSA.
When this did not work, we had to review the procedures we had been following and decide if
changes needed to be made to them. We decided to change the order of our procedures: instead
of forming a pellet of bacteria then treating it with the compound, we treated it with the CSA and
then formed the pellet. Our rational was that we could not sufficiently break up the pellet
following centrifugation to allow the CSA easy access to all of the bacteria cells. By treating the
bacteria with CSA while in suspension, we hypothesized that this would allow all the bacteria to
be exposed to the compound. We were correct. The images obtained after making that change
showed the bleeding that we had hoped to see.
Once we obtained the desired bleeding, we prepared to label the CSA treated E. coli with the
streptavidin-gold conjugate. First we had to determine how to prepare the appropriate dilution of
the gold solution. This required that we prepare a special buffering solution. Then we had to
coordinate carefully with microscopy so they could apply the solution to the biotinylated CSA
treated samples in a timely matter. After we prepared the samples we sent them to microscopy
where they finished the preparation and did the imaging for us.
The results were not what we hoped for. Although we saw good bleeding of the bacteria cells,
we did not see specific binding of the streptavidin-gold conjugate inside the cell. Instead, the
gold particles could be seen scattered about the images with no recognizable pattern.
There may be many reasons for why the gold particles were not seen specifically inside the cell.
First of all, it may be that the gold preparation we received was not prepared well. In other
words, the streptavidin and the gold may not have actually been complexed together by the
manufacturers. However, this hypothesis is difficult to test and strict industrial standards would
suggest that this was not the case. Another possibility is that the compound does not actually
enter the bacteria cell. However, the bleeding that we observed suggests that the compound is
somehow disrupting the bacteria cell membrane. Other possibilities include not incubating the
samples and gold solution preparation long enough or waiting too long to use the preparation. Or
we could have used the wrong dilution or mishandled the preparation of the solution we used to
dilute the gold.
Future experiments should examine the effects of varying dilutions of the streptavidin-gold
conjugate and other variables involved in preparing the slides for TEM imaging.ii
Imai T, Watanabe T, Yui T and Sugiyama J. 2002. Directional degradation of β-chitin by chitinase A1 revealed by a novel
reducing end labeling technique. FEBS Let 510: 201-205.
Rich Evanson worked closely with me on the majority of this research. Dr. Savage and Bangwei Ding also contributed many
helpful insights. Mike Standing performed the final preparation of our samples and transmission electron microscopy.