* unanswered.. debate with C Krasel

Cornelius Cornelius
Fri Jun 28 11:06:12 EST 1996


Margaret Fowler (101722.35 at CompuServe.COM) wrote:
> >>>Harold Hillman (101722.35 at CompuServe.COM) wrote:

[Attribution is difficult. I think the following is roughly correct:]

> >>Cornelius Krasel wrote:

> >Harold Hillman wrote:

I snipped some of the things on which we appear to agree, e.g. that
cell fractionation increases entropy or that EM cannot tell you anything
about the chemical composition of the structures visualized.

[snip]

> It is difficult to calculate the energy, therefore one has to control the
> experiments. In Hillman H (1972) 'Certainty and Uncertainty in Biochemical
> Techniques', Surrey University Press, Henley on Thames, I have listed 7
> different kinds of controls. Homogenisation, centrifugation, purification all
> change entropy therefore free energy, which drives biochemical reactions.
> Therefore it is an illegal procedure without controls!

delta G = delta H - T * delta S

delta G is the driving force. For an ATP-driven reaction, delta H is
somewhere about 9 kcal/mol if I recall correctly (sorry, no biochemistry
book at hand). How large has the entropy change to be to significantly
change delta G? I have no idea how large an entropy change of a
homogenization would be or whether there are any estimates.

> >>>> Question 5:  Does the finding that a chemical substance or activity
> >>>>     is located in the same subcellular fraction and a structure ident-
> >>>>     ified by electron microscopy mean that the same chemical activity
> >>>>     was located in that particular organelle in the living cell of the
> >>>>     intact animal or plant.
>
> >>> It probably depends on the type of "structure". AFAIK, enzymes cannot be
> >>> visualized by electron microscopy until you use their enzymatic
> >>> activity for a stain. (I know that e.g. myosin molecules *can* be
> >>> visualized, but not in a cellular context.)
>
> Please let me have references to AFAIK, firstly what it is an abbreviation
> for.

I apologize. AFAIK is net jargon for "as far as I know".

> Why can not myosin molecules be seen in cells, if they can be seen
> by electron microscopy and of they are there?

Whereas myosin filaments can be seen in cells, myosin molecules cannot
(at least I am not aware of any electron micrograph showing single
myosin molecules in a cell). Reasons are probably (I'm not an electron
microscopist) the orientation of myosin molecules in a myosin filament
and the low signal-to-noise ratio of such high-resolution pictures
(as shown, for example, in Alberts et al, Molecular biology of the cell).

> >If I find an enzymatic activity in a subcellular
> >fraction which I have identified as being Golgi previously, and I can also
> >locate the activity in the Golgi by let's say immunelectron microscopy,
> >the probability is high that the enzyme is indeed located in the Golgi.
> >If I understand the question wrong, please try to rephrase it.
>
> Subcellular fractionation is used for locating enzyme *activities* on the
> assumption that the procedure does not change *activity* or location -
> the latter assuming that diffusion does not occur.

Membrane-bound enzymes cannot diffuse freely. Enzymes located in vesicles
cannot diffuse freely unless the vesicles are leaky. If I have located
an enzyme by immunoelectron microscopy in the Golgi *and* it shows
Golgi-like cellular distribution in immunofluorescence *and* it
appears in a distinct cellular fraction in a carefully controlled
experiment, I can assume that this fraction contains Golgi. Even more
so if this is true for several enzymes.

> >You claim the cytoplasmic viscosity is low in life. It isn't.
>
> Cytoplasmic viscosity is low. See Hillman & Sartory (1980) The Living Cell,
> Packard, Chichester, pp 55-57; viscosity in cytoplasm is usually less
> than glycerol.

I've been told that cytoplasm contains a concentration of proteins which
is at least in the millimolar range, making it highly viscous. I cannot
back this up with a reference. I remember a review in TIBS dealing with
this issue, maybe one or two years ago (the title page showed a drawing
of cell interior packed with proteins and ribosomes).

> >>>> Question 7: Where do protein synthesis and acid hydrolysis occur in
> >>>>     cells in which ribosomes and lysosomes cannot be seen?
>
> >>> Are there cells which synthesize proteins and don't have ribosomes?
> >>> Examples please.
>
> >> All cells synthesise proteins, in many one cannot see ribosomes, e.g.
> >> muscle.
>
> >Is it possible to localize ribosomal proteins in these cells by cell
> >fractionation or immunoblotting? Or is it possible to isolate ribosomes
> >by cell fractionation? If yes, there are ribosomes -- you just can't
> >see them in the EM because of whatever reason (I'm not an electron
> >microscopist, so I don't know whether it is indeed impossible to
> >see ribosomes in muscle cells).
>
> 'Ribosomal' activity is believed to be protein synthesis in ribosomes.
> All cells synthesise proteins, including prokaryotes, where ribosomes
> can not be seen.

I think I've seen nice micrographs of polyribosomes in E. coli. I could
be wrong.

> >>>> Question 16: Can one know the thickness in life of any biological
> >>>>     membrane?
>
> >>> Yes -- use AFM on living objects.
>
> Please spell out 'AFM'

Sorry. Atomic force microscopy. It's a technique where you probe surfaces
by touching them with a very fine needle. By keeping the force between
the surface and the needle constant, you actually get an "image" of the
surface. AFM has been demonstrated to work down to atomic resolution
for inorganic surfaces. Proteins are quite difficult to visualize because
of their high mobility at ambient temperature. AFM has been used to
monitor enzyme dynamics in solution, see

@article{radmacher:94,
        author  = {Manfred Radmacher and Monika Fritz and Helen G. Hansma and
                Paul K. Hansma},
        title   = {Direct Observation of Enzyme Activity with the Atomic Force
                Microscope.},
        journal = {Science},
        volume  = 265,
        pages   = {1577--1579},
        year    = 1994
}

and bacteriorhodopsin (which is a membrane protein, BTW) has been
visualized at molecular resolution (as have several other membrane
proteins, for example OmpF from E. coli or the nicotinic acetylcholine
receptor). For bacteriorhodopsin, see

@article{dmueller:95a,
        author  = {Daniel J. M"uller and Georg B"uldt and Andreas Engel},
        title   = {Force-induced Conformational Change of Bacteriorhodopsin.},
        journal = {J. Mol. Biol.},
        volume  = 249,
        pages   = {239--243},
        year    = 1995
}

One of the nice things of AFM is that it can be used under water on living
cells.

> >>>> Question 20: What is transport?
>
> >Membrane transport, as you probably know, can be classified into facilitated
> >diffusion and active transport. Transport is movement against a concentration
> >gradient and needs energy to be accomplished (ATP or ion gradients).
>
> 'Transport' is a vague term, only meaning movement - not necessarily across
> membranes. Occam's Razor encourages one to consider that movement is by
> diffusion, Brownian movement, convection, *before* considering any other
> process, which - if claimed to be unique - should be faster or slower than
> all the above pout together.

So by what means do you think cells keep their internal Ca level low?

> >>>> Question 21: Why are receptors and channels, which have been character-
> >>>>     ised, sequenced and their sizes measured or calculated, not seen
> >>>>     on membranes by transmission electron microscopy?
>
> >>> Too small.
>
> >> Every week in Nature, Science, Molecular Biology etc one sees sequencing
> >> of molecules 3x the width of the cell membrane, seen by em.
>
> >It's pretty easy to visualize concentrated amounts of macromolecules.
> >Check out Unwin's paper about nicotinic acetylcholine receptors from
> >Torpedo electric organs. However, common receptors, such as most G-protein-
> >coupled ones, are just to rare to be distinguishable from noise. The
> >signal-to-noise ratio is much higher in AFM.
>
> Unwin's nicotinic ach receptor is the *only* one anyone has claimed to
> see. Where are the others?

Bacteriorhodopsin. OmpF. OmpC. LamB. Several proteins from chloroplasts
(check out papers by Werner Kuehlbrandt). Rhodopsin. F0-F1-ATPase. These
are just a few that come to mind immediately.

As I stated above, most receptors have too low a concentration to be
seen in electron microscopy.

BTW, you didn't address my point about visualizing of membrane proteins
(or whatever) in freeze-fracture electron microscopy of membranes.

> >>>> Question 26: Why is it assumed that the receptors for transmitters,
> >>>>     hormones, messengers, antibodies, drugs and toxins are on the
> >>>>     surface of the cell membrane?
>
> >>> For the beta2-AR:
> >>> 1) Evidence from use of hydrophilic ligands.
> >>> 2) Evidence from epitope mapping.
> >>> 3) Evidence from AP fusion studies.
> >>> 4) Evidence from protease accessibility.
> >>> 5) Evidence from immunoelectron microscopy.
>
> >> Are you assuming that diffusion does not occur during homogenisation,
> >> centrifugation, fixation, dehydration, embedding, etc?
>
> >Some of these experiments, e.g. the ligand binding experiments, are
> >done with whole cells. Same goes for protease cleavage and epitope
> >mapping.
>
> >As others have pointed out, there are also receptors which are not
> >located in membranes (e.g. for steroid hormones). There are also
> >receptors in internal membranes, or receptors that are cycling
> >between different compartments (e.g. transferrin receptor).
>
> There is still the question about why these large macro-molecules whose
> size is known are represented *in diagrams* as 2-3 x width of cell
> membrane are not (except for Unwin's) see by transmission electron
> microscopy.

Most G-protein-coupled receptors do not have 2-3 times the width of a
cell membrane. Others have this size, some can be seen (e.g. F0-F1-
ATPase for another example).

> Localisations are usually done by microscopy of dehydrated tissues or
> subcellular fractionation in both of which diffusion *must* occur
> therefore one can not decide localisation.

As I stated above, a lot of these experiments were done neither by
microscopy nor by subcellular fractionations. Others were, and the
results did fit together.

> >Of course you can use adrenaline to look for adrenergic receptors. However,
> >it binds fairly unspecific to several receptors. Other ligands bind
> >more specifically, and since people are usually interested in the
> >properties of one receptor, they use ligands specific for it. (But
> >adrenaline is also a ligand for these receptors.)
>
> Of course, you can use adrenaline to look for adrenaline receptors.
> Then *why* do people *not*?

Excuse me, but did you read my answer?

> >>>> Question 34:  What is the evidence that each cell of a particular
> >>>>     plant or animal contains the same quantity of DNA?
>
> >>> Honestly, I don't know.
>
> >> This is an *assumption*
>
> >Would you assume that each cell contains the same amount of chromosomes?
>
> I would not. It has not been proven. It is an assumption.

It's impossible to prove, since you would have to look at every cell
of every organism of the species in question to strictly prove it.
However, it appears that each cell (if not abnormal) contains the
same amount of chromosomes.

>
> >>>> Question 35: If the cell membrane is fluid mechanically, how can cells
> >>>>     maintain their integrity?
>
> >>> Because a fluid bilayer is intrinsically stable.
>
> >> Glass is a *solid* mechanically but a fluid physicochemically.
>
> >You lost me here.
>
> What evidence is there for that other than the *belief* that the cell
> membrane is a fluid bilayer?

Lipid membranes of certain composition are fluid bilayers. The composition
of E. coli membranes is such that they are fluid at ambient temperature,
and experiments have been done which demonstrate that E. coli actively
regulates the composition of its membrane to keep it fluid (sorry, no
reference).

> >>>> Question 36:  In immunocytochemistry, is it assumed that the fixatives,
> >>>>     dehydrating reagents, washings, and primary and secondary anti-
> >>>>     bodies, do not change the reaction of the antibody to the antigen
> >>>>     believed to be in a particular cell or part of a cell?
>
> >>> Yes. That's why many antibodies do not work in immunocytochemistry.
>
> >> Nearly all immunocytochemistry is done on fixed, dehydrated, and
> >> mounted sections.
>
> >I know. So?
>
> Therefore, one assumes that the fixative dehydrating agent and mounting agent
> do not affect the antigen-antibody reaction - an untested and very
> unlikely assumption.

I stated above that many antibodies do *not* work in immunocytochemistry.
In most papers I have read, controls are presented which show that the
antibody used for immunofluorescence recognizes the antigen in question,
and that blocking with this antigen quenches the immunofluorescence.

> >>>> Question 46:  In diseases believed to be auto-immune, either
> >>>>     organ-specific or tissue-specific, why does the body not reject
> >>>>     the specific organ or tissue, as it rejects incompatible
> >>>>     transplanted hearts, or blood of the wrong group, often
> >>>>     making the patients ill, or even killing them?
>
> >>> It does. That's where diabetes type I can come from.
>
> >> If diabetes were autoimmune, how do Islets continue to exist
> >> in that condition?
>
> >I don't know the exact mechanism of diabetes type I but there has been
> >recently a review published about it as an autoimmune disease in Cell.
> >(I haven't had time yet to read it.)
>
> I can not understand why anyone alleges a disease to be autoimmune if
> the main organs, e.g. the joints, the Islets of Langerhans, the brain
> (schizophrenia) are not *rejected*, as would incompatible blood be.

[Maybe answered later on; I have to read that review first :-) ]

> >I use BSA in my Bradford assays just because
> >it is convenient. Everybody knows that Ovalbumin gives a completely different
> >standard curve. It's just to standardize the assay to *something*.
>
> What one needs to do is a separate recovery curve with your bovine
> serum albumin for *each* fraction.

I am afraid that I don't understand what a recovery curve is.

> Dr Krasel, I will send you a copy of one of our books, 'The Living Cell'
  ^^not yet, but I'm working on it :-)
> as it is out of print.

Thank you very much.

--Cornelius Krasel.

--
/* Cornelius Krasel, U Wuerzburg, Dept. of Pharmacology, Versbacher Str. 9 */
/* D-97078 Wuerzburg, Germany   email: phak004 at rzbox.uni-wuerzburg.de  SP3 */
/* "Science is the game we play with God to find out what His rules are."  */



More information about the Cellbiol mailing list