* neurobiology questions *

Joseph Strout jstrout at ucsd.edu
Mon Sep 2 00:58:23 EST 1996

On 1 Sep 1996, Raeto West wrote:

>      1. Has anyone shown neuroglial unit cell membranes adjacent to neuronal
> membranes in intact tissue by light or electron microscopy? If so, please send
> me micrographs or references in intact tissue.

I'm not sure what you mean by "intact tissue"; you have to slice it to fit
it into the microscope.  Assuming that, then the answer is yes: I have
several such pictures on my fridge.  I'm not going to send them; look in
almost any neurobiology text or journal.

>      2. What publications have shown that the characteristic morphological and
> staining properties by which neurons, astrocytes, oligodendrocytes and
> microglia, were originally described are also present in cultures which are
> believed to be the same cell types?

Any publications describing original culture techniques.  Indeed, a good
neuron culture is typically defined by the quality of neuronal markers
(staining properties) and neuronal morphology displayed.

>      3. Why does only a small proportion of the area of a section of the
> central nervous system show up with neuronal and neuroglial stains and
> antibodies?

Where did you get this idea?  If you're staining white matter, then of
course some areas are thicker with it than others.  If you stain cell
bodies, you get a different pattern.  If you stain for something else --
say, a potassium channel -- you get another pattern, which may be
concentrated in (say) dendrites.  If you had a stain which stained
everything (e.g. all neurons and glia in their entirety), you'd throw it
in the trash because you wouldn't be able to see anything.

>      4. Why does an intracellular recording electrode, penetrating the brain
> and spinal cord, record no potential difference in most of its travel, if only
> 5-25% of the volume of the central nervous system is extracellular space?

Again, I wonder where this notion came from.  If you look at (say) Hubert
and Wiesel's papers, they report recording quite a few cells on each
linear penetration.  Of course, sometimes your electrode will push aside
processes instead of penetrating.  Imagine sticking a poker into a pile of
straw -- only 5-25% of the pile is air, but how many straws will you
actually impale?  Probably none.  Most of the brain is like straw (i.e.
axons and dendrites).

>      5. Has anyone ever seen under their electron microscopes, or in
> micrographs of longitudinal or transverse sections of myelin lamellae of the
> central or peripheral nervous system, the expected appearance of 'splaying'
> in longitudinal sections not going through the middle of the axon, or in
> oblique transverse sections 

Yes, you get sections at all angles, resulting in ovals.  Many of these
are not pretty pictures, so not likely to get published, but they're

>      6. If the central nervous system is virtually full of neurons, neuroglial
> cells and their processes, how can phagocytes assemble within hours around an
> infection from other sites in the brain and spinal cord, in the process
> described as gliosis?

They probably "ooze" through the gaps between and around cells.  Hours is
a long time to travel a few centimeters.

>      7. Why are immunocytochemical markers believed to be characteristic of
> different cell types nearly always demonstrated in tissue culture and not in
> the intact nervous system?

I beg your pardon!  My lab works almost exclusevly in rat brain slices (as
opposed to culture), and its primary business is developing and using
immunocytochemical markers.  I see no basis for your assumption.

>      8. Does anyone have micrographs - rather than diagrams - of myelinated
> fibres arising from neuron somas?

Hmm, do you mean electron micrographs?  I don't have any offhand, but I
haven't looked.

>      9. Have fixatives, washings, ethanol, sera and permeabilisers been shown
> not to affect the intensity or distribution of fluorescence as seen in
> immunocytochemistry?

No, they probably do affect it -- these effects have to be minimized or

>      10. Why do some sections have to be 'permeabilised' to allow access of
> antibodies, since the cells are already cut open in thin sections, and one
> must assume that the permeabilisers have no effect on the immune reaction?

You don't just want staining at the cut surface.  Antibodies often stick
to cut edges anyway (non-specific staining), so data there is almost
useless.  You want staining a few cell layers into the tissue, and for
that you frequently need to permeabilise it.

>      11. Why is it assumed that findings across nerve-muscle junctions are
> relevant to nerve-nerve junctions (synapses)?

Because the structures and molecules involved are similar, as are the
functions; and in addition, biology often seems to make multiple use of
the same mechanisms.  As another example, many of the same ion channels
found in neurons are also found in heart muscle (both are very active
cells).  Of course, there are important differences too, especially in the
size and reliability of the PSP.  But the neuromuscular junction is easier
to study by far, so it was studied first.

>      12. Why are there no light micrographs in the literature - as opposed to
> diagrams - showing a dendrite arising from one cell body attached to the
> synapse on another cell body, since the light microscope has the resolution
> to show up such pre-synaptic fibres?

Beg pardon?  You probably mean axon.  In the confocal microscope, a
synapse appears as a tiny dot just barely visible if you specifically
stain for it.  Axons are visible (if stained).  But the odds of finding
two cells close enough to fit in the same frame, joined by a synapse, and
all in the same focal plane, are astronomically small.

>      13. Why are the diameters of synaptic knobs seen by light microscopy
> approximately 8-10 times the diameters of those seen by electron microscopy?

8-10?!?  I'd like to see YOUR references on that one.  Tissue does shrink
a bit in EM, both due to processing and irradiation.  Also, tiny
structures (like synapses) appear larger than they really are due to
spread of light in LM as you approach the diffraction limit.  As I said,
synapses are just barely resolvable under LM.

>      14. Why are the synaptic thickenings seen by electron microscopy nearly
> always normal to the plane of section, and why are they never seen obliquely
> or on face view (as circles)?

For a face view, look at freeze-fracture images.  In transmission EM, the
best (i.e. clearest/prettiest) pics are transverse, so those are the ones
that get published.  Any EM lab has cabinets full of bad pictures, if
that's what you want...

>      15. Why are receptors to transmitters, which are believed to have been
> isolated and sequenced and shown to be two to three times the diameters of the
> cell membrane - which is generally believed to be seen by electron microscopy
> - not seen by transmission electron microscopy, with the exception of the
> acetylcholine receptor  (Unwin)? 

To get a good image of receptors requires crystallographic techniques (or
something similar).  A membrane is a much easier target because it's
fairly uniform, it's large (orthogonal to the image plane), and it's
frequently stained by our EM stains.

>      16. How can one prove that synaptic vesicles hit the presynaptic membrane
> in life, when they can only be seen by electron microscopy of fixed (dead)
> tissues?

You catch them in various stages of the act (esp. by freeze-fracture).
You measure events in live tissue that appears to require quantal release.
You note that these vesicles are in just the right place and have the
right characteristics to serve this function.  You may a theory.

>      17. Why do the synaptic vesicles appear to be almost the same diameters,
> when sections made for electron microscopy must cut them in a variety of
> planes?

I think you do get a variety of diameters, but I'll have to pull out some
pics for a second look.  Perhaps vescicles which are not hit squarely by
the knife tend to move aside.

>      18. Why are miniature end-plate potentials uniform in amplitude despite
> the belief that they arise at varying distances from the recording electrodes?

The are not uniform; a typical mEPP ranges from 0.5-1.0 mV or so, in a
bell-shaped curve.  The distance variation arises from vescicles being
released at different parts of the end-plate, but I don't think this
difference is very great compared to the spread of transmitter by
diffusion and the spread of current in the postsynaptic membrane.

>      19. Since the synaptic cleft is believed to contain a high concentration
> of acetylcholinesterase, how can the acetylcholine from the presynaptic
> membrane cross the cleft without being broken down?

It hurries across.  ;)  Seriously, some of it probably does get broken
down.  Some more of it hangs around longer than average, too.

>      20. If action potentials can be conducted from an isolated nerve to a
> recording electrode, from the heart to the electrocardiogram, from muscle
> fibres to an electromyogram, from cortical neurons to an electrocorticogram,
> without the intervention of synapses, why can signals not be conducted from
> one dendrite to another?

First, there *are* dendrodendritic synapses.  Second, there are gap
junctions, which are non-synaptic joinings most analogous to electrodes.
Finally, dendrites probably do influence on another when they are
juxtaposed (as do axons).  Hopefully, this is not a very large influence
compared to synaptic inputs.

>      21. Why are there no tumours of neurons in the central nervous system?

I'm no pathologist, but are you sure this is true?  I was under the
impression that neuronal cell lines come from neuronal cancers.

|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|    jstrout at ucsd.edu           http://www-acs.ucsd.edu/~jstrout/  |

More information about the Neur-sci mailing list