Cell Differentiation: How is it caused, controlled, and self-perpetuated?

The human body is made of approximately 250 differentiated cell types.
Examples include, cardiac muscle cells, Schwann cells, osteocytes, chondrocytes, macrophages, etc.
Lower animals have fewer cell types; about 10 or 12 in sponges & hydra; 20+ in flatworms.

Anatomical structures consist of certain geometric arrangements of differentiated cell types.
Embryonic development is produced by spatial arrangement of cell types, plus rearrangements.
"Positional Information" theory regards differentiation as controlled by "global" gradients.
Alternative explanations include chains of many local signals; also the use or forces as "signals".

Cell differentiation results from selective transcription of a specific set of genes for each cell type.
Nearly all cell types continue to contain and duplicate their complete original set of genes.
Exceptions include lymphocytes, with their special antibody genes; & red blood cells expel their nucleus.

At the time ofDriesch, scientists expected that genes would be portioned out during development;
And each cell type would contain only that minority of genes used by that kind of cells.
That would have been an explanation why cell differentiation is usually irreversible (in animals, not plants).
Driesch's discovery of embryonic regulation disproved that explanation.
Nuclear transplantation "cloning" of frogs in 1952, and then of mammals in the 1990s, proved that (nearly) all differentiated cell types contain the whole genome.

Differentiation is gene suppression (of most genes), in addition to gene activation (of "luxury genes");
For example, all differentiated cell types contain the same genes for hemoglobin, dystrophin, etc.

A reasonable guess is that gene activation is by transcription factor proteins, that bind selectively to promoter regions located just upstream of "structural" genes (that code for regular proteins), with the special DNA sequences in promoter regions of all genes transcribed in a given cell type.
Thus, inserting a different structural gene downstream of such a promoter would cause that gene's protein to start being synthesized in all the cells of some one particular differentiated cell type.
This is the basis of "reporter genes"; and their effectiveness confirms this explanation of differentiation.

Chromosome translocations in which the promoter of a luxury gene become relocated just upstream of an oncogene are a major cause of cancer, and cause all or nearly all forms of non-Hodgkins lymphoma.

Little or no attention is given to two other important sets of questions:

#1) Self-perpetuation of each differentiated cell type: "Once a macrophage; Always a macrophage?"
"Once a cardiac muscle cell; Always a cardiac muscle cell"
Once a given subset of luxury genes are turned on; Then this subset will keep tuning itself on?
(And will also continue "turning off" (inhibiting transcription of) all other subsets of luxury genes?)
If you could force cells to switch differentiated states, that would be better than embryonic stem cells

. #2) Mutual exclusiveness of differentiated cell types: no cell can simultaneously belong to 2 cell types.
"Once a cardiac muscle cell; Then ONLY a cardiac muscle cell"
Something prevents cardiac muscle cells from also making the luxury proteins of a macrophage;
And likewise for ALL other combinations of two or more differentiated cell types.

An interesting exception is that some (or many?) cancer cells synthesize other luxury proteins, in addition to those proteins that are normally made in their original differentiated cell type.
For example, some cancers get detected because they secrete hormones (normally only made by other cells).
Also, more advanced cancers of oligodendrocytes and astrocytes become indistinguishable from each other.
"Dedifferentiation" is often a symptom of cancer. Maybe leucocyte proteins make epithelia invasive?

But you can fuse tissue culture cells with each other, by using certain inactivated viruses Or certain chemicals that damage plasma membranes.
And a low percentage of tissue culture cells will fuse with each other spontaneously.
Fusing cancerous lymphocytes with antibody-producing B cells gives "mono-clonal antibodies"

WHAT WOULD YOU EXPECT TO HAPPEN, IF...? What other results would be possible?
How would you interpret each alternative result, in terms of what it would imply about normal cells?

A) What if you fused (in tissue culture) a liver cell with a skin cell? Which genes activated, or suppressed?

B) What if you fused a human liver cell with a liver cell of a chicken? Which genesÉ? If any.

C) What if you fused a human liver cell with a red blood cell of a chicken? (bird r.b.cs have inactive nuclei)

D) What if you fused a mouse liver cell with a (very multinucleate) bird skeletal muscle cell?

E) What if you fused a cancerous mouse cell with a cancerous chicken cell? (if both the same cell type?)

F) Would you expect the result to differ if one was a cancerous liver cell and the other a lymphoma cell?
(The result differs very much from what happens in E!)

G) Why does it make sense that when individual cells of liver, blood, skin etc. are fused with skeletal muscle cells, these other nuclei quit transcribing their previous luxury genes, and switch to expressing muscle proteins?

H) How could you take advantage of this fact (in question G) to improve "gene therapy" of muscular dystrophy? Transform tissue culture cells & inject them instead of gene-carrying viruses.

I) Would your method just work for Duchenne Muscular Dystrophy, or for all forms equally?

J) Figure out how to take advantage of each of the following processes, to treat Muscular Dystrophy:

    1) Regeneration of killed skeletal muscle cells continues as long as uni-nucleate myoblasts remain.
    2) Skin, intestinal, and blood-producing stem cells do not use up their telomeres in mitosis.
    3) Such stem cells can be grown in tissue culture for 50 or more mitotic generations.
    4) Transformation of DNA (genes) into cells works better in tissue culture than inside the body.
    5) In tissue culture, you can find & subculture ("clone") 1 in a million cells that has a given change.
    6) Imagine culturing a million cells from a Muscular Dystrophy patient's skin!
    (How many mitotic cycles would that be?)
    7) Add DNA sequences for normal dystrophin protein to the medium of these million skin cells.
    8) Identify (how?) and subculture the tiny minority of cells that have incorporated this DNA.
    9) Let the subculture divide until there are a million cells (containing normal dystrophin genes)
    10) Inject most of these million cells into the patient's blood.
    11) Count on the ability of skeletal muscle cells to fuse with uninucleate cells of other types,
    12) And change the differentiated states of fused nuclei to transcription of muscle proteins.

K) Cancer cells, in tissue culture, fused either with non-cancerous cells, or fused with cancer cells originating from another differentiated cell type, stop being malignant. This was not what anyone expected, and is not what subsequent knowledge of oncogenes would lead you to expect (is it?).
Probably it results from the still mysterious "mutual exclusiveness" of any pair of differentiated states.

It also suggests radically new kinds of treatments for cancer, that may never be tried.












0   1
1   2
2   4
3   8
4   16
5   32
6   64
7   128
8   256
9   512
10   1024
11   2048
12   4096
13   8 thousand
14   16 thousand
15   32 thousand
16   64 thousand
17   128 thousand
18   256 thousand
19   Half a million
20   Just a bit more than a million
21   Two million
22   Four million
23   Eight million
24   Sixteen million
25   Thirty two million
26   Sixty four million
27   A hundred and twenty eight million
28   Two hundred and fifty million
29   Five hundred million
30   One billion
31   Two billion
32   Four billion
33   Eight billion
34   Sixteen billion
35   Thirty two billion
36   Sixty Four Billion
37   A Hundred and twenty eight billion
38   Two hundred and fifty billion
39   Five hundred billion
40   One trillion
41   Two trillion
42   Four trillion
43   Eight trillion
44   Sixteen trillion
45   Thirty two trillion
46   Sixty four trillion
47   A hundred and twenty eight trillion
48   Two hundred and fifty-odd trillion
49   Really an awful lot.
50   Even more than that.




back to index page