ABSTRACT
Although most previous research has not been successful in getting mature,
differentiated neurons to express exogenous DNA, this has reportedly been done in a
variety of neuron types in Aplysia californica using a RSV promoterlenhancer sequence.
This study attempted to express the lacZ gene from Escherichia coli in the giant fiber
lobe (GFL) neurons of squid using this promoterlenhancer. No expression of the B-
galactosidase enzyme was detected in GFL cells or the Aplysia neurons when the cells
were injected with the plasmid, but ß-galactosidase activity was assayed in
neuroblastoma cells. A probable reason for non-expression of the plasmid in the Aplysia
may be ineffective injection techniques. The non-expression in the GFL cells may be due
to either the injection process or the inability of the cell to be induced by the RSV
promoter. However, one additional factor involved in expression of exogenous DNA was
found. In looking at another study done on GFL cells, it was found that heat shock
increased expression of Na channels. The exact mechanism that is affected to produce
this effect from endogenous genes is unknown, but the effect heat shock may have on
exogenous DNA expression cannot be ignored. And given the differences between the
Aplysia and the squid systems, more studies on the effect of heat shock on these
systems must be done before any definite conclusions may be drawn.
Until recently, expression of exogenous DNA in mature, non-dividing cells, such
as differentiated neurons, has not been possible. The DNA, when it is placed in the
nucleus of such cells, remains untranslated and is eventually degraded. However, in
February 1992, Bong-Kiun Kaang, Eric Kandel, and others reported expression of
exogenous DNA in the abdominal ganglia of Aplysia californica by using a Rous
sarcoma virus (RSV) promoterlenhancer sequence (Kaang, et al, 1992). They were able
to get expression of this vector not only in the abdominal ganglion, but also in the buccal
ganglion, the pleural ganglion, the pedal ganglion, the cerebral ganglion, motoneurons,
interneurons, and sensory neurons (Kaang, 1992).
Iwas able to get from the Kandel laboratory a copy of one of the plasmids they
used, the pNEX-lacz, which contained the RSV sequence and the Escherichia coli lacz
gene as the reporter gene. (Fig. 1) Through microinjecting, have been able to get
expression of this vector in mouse sympathetic neuroblastoma cells (NIE-115), but
have not been successful with its expression in either the abdominal ganglia of Aplysia
or the giant fiber lobe (GFL) cells of the squid Loligo opalescens.
MATEBIALS AND METHODS
Preparation of the Plasmid. The pNEX-lacZ plasmid was transformed into
E.coli and grown overnight in DYT and ampicillin broth. The DNA was then isolated
using the Oiagen midi preparation procedure and cleaned in a phenolschloroform
extraction.
Microinjection of Plasmids into Giant Fiber Lobe Cells. From Loligo
opalescens the giant fiber lobe was dissected and plated as previously described. (Gilly
et al, 1990) Within 1-48 hours after plating, the neurons were injected using a
compressed gas system (Eppendorf Microinjector 5242 and N2 gas) with a DNA solution
consisting of 0.5 ug/uI DNA, 0.05-0.5% fast green, 10 mM Tris-HCl at pH 7.3, and 100
mM KCI. The injection microelectrodes were made from a glass with similar properties to
the Narishige IM-CFS borosilicate 1 mm OD glass capillaries with filler fibers using either
a Livingston puller (F.S. Hockman) (Gilly, 1978).
Microinjection of Neuroblastoma cells. Differentiated mouse sympathetic
neuroblastoma cells (NIE-115, kindly provided by Sam Wang and Chris Mathes) were
received ready to be injected on glass cover slips, and they were injected with the DNA
solution using the Eppendorf Microinjector. The cells had been differentiated as will be
described. (Mathes et al, 1992) The microelectrodes used for injection were the same as
that for the GFL cells.
Microinjection of Plasmids into Aplysia neurons. The Aplysia californica
abdominal ganglion was dissected, pinned on a Sylgard plate and desheathed in L15
medium (Schacher and Proshansky, 1983) containing an equal volume of artificial sea
water (ASW) of the following composition: 475 mM Nacl, 10 mM KCI, 10 mM CaCl2, 50
MgCl2, and 10 mM HEPES. These neurons were injected using a Picospritzer Il
(General Valve Corporation) with the same DNA solution as with the GFL cells, with the
exception that KCl was replaced with Nacl. The injection microelectrodes were made
from World Precision Kwik-Fil glass capillaries (approx. 2 mm ÖD) using a Narishige PE¬
2 electrode puller.
Detection of ß-Galactosidase. The cells were washed 24 hr after
microinjection with Ca-free ASW (the above ASW composition without any CaClo), then
fixed and stained as previously described. (Kaang et al, 1992)
RESULTS
When the GFL cells were injected with the pNEX-lacz gene, no expression of the
ß-galactosidase activity was detected. Öther injection procedures were attempted by
varying the time between the plating of the cells and the injections, the DNA injections
solution, the glass and the glass pullers used to make the microelectrodes, and the wash
solutions, but positive results were not obtained.
To determine whether the plasmid received was indeed the pNEX-lacz, a series
of tests were run to verify this: First, a BamH1 restriction digest was done on the plasmid,
which resulted in fragments of appropriate sizes.
Next, the plasmid was transformed into E. coli and grown on agar with X-gal.
Because the bacteria is able to absorb the X-gal through their cell membranes and use
it, the blue colonies that resulted were a positive indication that the bacteria was
producing the ß-galactosidase enzyme. Thus the negative result with the GFL cells was
not due to a fault with the lacZ gene.
To test the integrity of the RSV promoterlenhancer sequence, the plasmid was
injected into mouse sympathetic neuroblastoma cells. Because the Rous sarcoma virus
is a mammalian virus, the neuroblastoma cells were expected to express the plasmid.
Staining these injected cells showed a mixture of very well stained and partially stained
cells. (see Table 1) Dense staining of cells was only seen in the case of DNA-injected
cells. 24 cells similar to the example in Fig. 2 were seen with 950 cells injected. Mock
injections of the cells were done with a solution of 0.05% fast green, 10 mM Tris-HCI
and 100 mM KCl, as a control for possible expression of a ß-galactosidase gene in
physically stressed cells, but no dense staining was evident. Similarly, bathing the cells
in the DNA containing solution or no treatment at all failed to produce dark staining,
which I took to indicate positive ß-galactosidase expression. With this definition, it can be
concluded that the injected neuroblastoma cells expressed the pNEX-lacZ vector. Thus,
the promoter sequence was not faulty.
In addition to the 24 darkly stained cells, there were many faintly stained cells.
These cells had either small, light blue circles of stain about 2-10 um in diameter or a
faint blue stain throughout the cell. This result could be due to background staining from
the technique or it could be due to the cells producing small amounts of the enzyme.
From these results, which it is cannot be determined. Because of this, the partially
stained cells were deemed as unusable as a positive indicator of plasmid expression.
As a final control, injections were done on Aplysia abdominal ganglion, but with
no detection of ß-galactosidase expression. This result was confusing, because the
plasmid was supposed to be functional in the Aplysia. In looking over the procedure
used by Kandel to determine where any procedural deviations may have occurred, it
was tound that in his preparation of the abdominal ganglion, the ganglion was dissected,
then treated with protease IX (Sigma) for 1 hr at 34.5°0. The normal temperature at
which Aplysia cells are cultured is 17°0, so this would subject the cells to extreme
physical stress. Nonetheless, injections were repeated using heat proteased ganglia, but
again, no expression of ß-galactosidase activity was found.
In the proteasing step of the GFL cells, the cells are maintained at room
temperature, and heat was not necessary. To determine whether the 34.5°0 treatment
was a necessary part of the plasmid expression, the heat treatment was attempted with
the GFL cells. But before this, survivorship experiments were run to determine whether
the squid cells would live through the heat shock. In Figure 4A, with the 1.5 and 2 hour
heat shocked cells, a rapid decrease in the cells surviving can be seen. In the 1 hr heat
shocked cells, an initial increase in the cell death can be seen, but after 48 hr this rate
levels off to a survival rate similar to that of the control. In Figure 4B, these two trends can
be seen in cells that were treated in the same 1 hr heat shock. In one plate the cells died
rapidly (open hexagons). In the other plate (open circles), a rapid initial decrease in cell
survival was halted after approximately 48 hrs, when the cell survival rate returned to the
control level. Thus some cells seem to survive heat shock better than others.
With the knowledge that some GFL cells can withstand 1 hr heat shock, this
duration was used in all future heat shock experiments. DNA was injected into cells 1-8
hours after heat shocking, but no ß-galactosidase activity was detected. ß-galactosidase
assays were also done on cells that were heat shocked 1 hour after injection, but these
likewise resulted in negative staining.
DISCUSSION
No expression of the pNEX-lacZ plasmid was successfully detected in the GFL
cells, and the controls that were run indicate that the lack of expression was not due to
an error in the plasmid. This leaves two possibilities why the expression didn’t work in
the squid cells. One possibility is that the GFL cells cannot be induced by the RSV
promoter to express the exogenous DNA under any circumstances, and a second
possibility is that the injections were not properly done, resulting in low efficiency.
In looking at a study done in parallel with this expression project, however,
another possible factor in expression of exogenous DNA was found. In this study (Gilly,
1992), whole cell voltage clamp experiments were performed on the control and heat
shocked GFL cells of the survivorship experiment. In cells that were heat shocked 48
hours after plating, it was found that over the 24 hours following the temperature stress,
the heat shock induced a four-fold increase in the density of Na channel conductance as
compared to the control plates. Because the GFL cells in the squid normally do not
express Na channels in the somata, (Brismar and Gilly, 1987) thus resulting in cells that
in primary cell culture initially do not contain any Na channels, and because the Na
conductance is proportional to the number of functional Na channels in the cell
membrane, any Na conductance that is detected in these cells is a result of
mappropriately expressed Na channels. Thus, the heat shock in effect caused a four-fold
increase in Na channel protein expression, a dramatic increase.
The mechanism through which this increase is produced is unknown. Morimoto,
lissières, and Georgopoulos have reported a wide ranging effect of heat shock in
protein synthesis, from the transcriptional to translational effects as well as effects on
post-translational protein stability, any of which may lead to increased levels of proteins,
any of which may have taken place in the GFL cells.
This finding cannot directly be applied to a possible effect of heat shock on the
expression of exogenous DNA, but given that heat shock has an effect on the expression
of endogenous genes, the possibility that heat shock would have a similar effect on
exogenous DNA expression cannot be excluded. In order to resolve the exact role heat
shock is playing in gene expression, further experiments must be done to understand
more completely its effect on these organisms.
thank Dr. W. Gilly for his patience, support, insights, and especially his pushing me to
get done what needed to get done so that it got done; Dr. M. Perri for her patience,
support, insights, and compassion upon a lowly undergraduate; Dr. E. Kandel for the
PNEX-lacZ, the various information relayed, both on paper and on phone, and his assent
of my visit to his laboratory; Dr. P. Pfaffinger for his information on the details of the
Aplysia project; Dr. S. Thompson for the information on and the dissections of the Aplysia
ganglia, for the use of the Picospritzer, and for his interest; C. Mathes and S. Wang for
the neuroblastoma cells and for their kindness; and J. Lucero for the dissection of the
GFL cells.
APPENDIX A: DNA MICROINJECTION TECHNIQUE
In injecting the GFL cells, the Eppendorf Microinjector and Micromanipulator
were used to insert the DNA into the cells. However, this process is not as simple as it
sounds. First, the optimal placement of the injection is in the nucleus of the cell. If the
injection is made into the cytoplasm, there is no cause to believe that the DNA would be
imported into the nucleus. Thus, to get the DNA into the nucleus, injection directly within
the nuclear membrane is the only way of ensuring this. However, this is easier said than
done. In the GFL cells, the nucleus seems to be of various sizes and visibility, depending
on the condition of the cell. In the healthier cells it is more difficult to see the nucleus,
whereas in the more stressed cells the nuclear membrane is easily detectable. When
they are visible, unless the cell is extremely unhealthy, the nucleus seems small, about
20 to 50 um in diameter. Thus, in healthy cells the nucleus is hard to detect, and in cells
where the nucleus is visible, not only is it small, but the cell isn’t as healthy. Even if the
nucleus is visible, the placement of the injection into it is yet another feat: On the
micromanipulator, the lower limit to which the microelectrode is to penetrate is set and
the glass tip placed directly over the spot at which the injection should be placed. Given
that the cell is only visible from the top and the level at which the nucleus is present is
difficult to establish, the microelectrode, when the injection is made, could come above
or below the nucleus, leaving it unaffected. To complicate the matter further, when the
nucleus is definitely penetrated, some nuclear membranes have been seen to stick to
the glass, coming out with it as the microelectrode exits the cell, killing the cell.
In the neuroblastoma and the Aplysia cells, the cell is composed primarily of the
nucleus and it's difficult to miss the nucleus, which explains the more successful
injections into the neuroblastoma cells, but brings to question the non-expression in the
Aplysia neurons. Note also that if the injection set up is coupled with an electrical
reading of the microelectrode, as had been done in this set up, just by reading the action
potentials that are detected, being in the cell is ensured.
The gist of this is that even in the Aplysia cells the injections are challenging to
perform. Eric Kandel and Paul Pfaffinger, have likened the injection technique to art, that
with more practice some people get better at it and that some people never get better at
all. The common agreement among the people who perform injections into cells is that
the injections are extremely difficult to master.
APPENDIX B: DNA CLEANING
In my protocol, the DNA used for injection was cleaned by running it through a
Oaigen column. When the Kandel lab was contacted, it was found that when they started
to do the research they also used a Oaigen column, but that they switched over to
centrifuging the DNA in a CsCllethidium bromide gradient. Whether this does get the
DNA cleaner is not known, but the Oaigen-cleaned DNA turned out to be extremely
sticky, clogging the microelectrode very easily. Whether this stickiness is an innate
property of the DNA or whether it is due to contaminants is unknown, and how this affects
the success of the injections is likewise unknown.
Literature Cited
Brismar, T. and Gilly, W. (1987) Synthesis of sodium channels in the cell bodies of squid
giant axons. Proc. Natl. Acad. Sci. USA 84:1459-1463.
Gilly, W. (1992) Personal correspondence.
Gilly, W. (1978) Contractile activation in slow and twitch muscle fibers of the frog. PhD
thesis.
Gilly, W., Lucero, M., and Horrigan, F. (1990) Control of the spatial distribution of sodium
channels in giant fiber lobe neurons of the squid. Neuron 5:663-674.
Kaang, B. (1992) Studies of long-term facilitation using gene transfer methods. PhD
thesis.
Kaang, B., Pfaffinger, P., Grant, S., Kandel, E., and Furukawa, Y. (1992) Overexpression
of an Aplysia Shaker K“ channel gene modifies the electrical properties and
synapfic efficacy of identified Aplysia neurons. Proc. Nati. Acad. Sci. USA
89:1133-1137.
Mathes, C., Wang, S., Vargas, H., and Thompson, S.H. (1992) Intracellular calcium
release in NIE-115 neuroblastoma cells is mediated by the Mj muscarinic
receptor subtype and is antagonized by MCN-A-343. Brain Research 00 (in
press).
Morimoto, R., Tissières, A., and Georgopoulos, C. (1990) Stress Proteins in Biology and
Medicine. Cold Spring Harbor Laboratory Press.
Schacher, S. and Proshansky, E. (1983) Neurite regeneration by Aplysia neurons in
dissociated cell culture: Modulation by Aplysia hemolymph and the presence of
the initial axonal segment. The Journal of Neuroscience 3:2403-2413.
Figure 1.
Figure 2.
Figure 3.
Construction of the pNEX-lacz.
A photo of a positively stained neuroblastoma cell for B-galactosidase
assay.
(A) The percentage of cells surviving in three different heat shock times.
The filled circles are the control, the open circles the 1 hr heat shocked
cells, the triangles the 1.5 hr heat shocked cells, and the squares the 2 hr
heat shocked cells. Heat shock was delivered to the cells 24 hr after
plating, as indicated by the arrow.
(B) The percentage of cells surviving 1 hr heat shocks. The filled circles
indicate the control plate, and the open markers indicate two different
plates of 1 hr heat shocks. These plates were heat shocked 48 hr after
plating, again as indicated by the arrow.
a.
ooop
0
2 8
25
oo
.
RSV
AXAP-1
AATRAA
MCS
PNEX
Bam HI
T4 DNA ligase
lac 2
RS
4P
A
DNEX - lac 2
PNASSB
JAATAAA
Fig. 1.
lac 2

U
8-
A
V
8-
CELLS SURVIVING
(%)





—
88