Monday, January 30, 2017


Cell Sorting and Discovery of Selective Adhesion Proteins

H. V. Wilson discovered that randomly rearranged cells of sponges and corals can reform their normal anatomical, histological & functional patterns.

There is a good painting (oil portrait) of Wilson across the hall from this lecture room.


video of balls of sponge cells aggregating together


The long video shown before the lecture was a compilation of time-lapse films of sponges in culture, by Calhoun Bond and Albert Harris. Some clips from this video are posted at  


living sponge that has ingested green fluorescent beads


Johannes Holtfreter discovered that frog and salamander embryos survive being separated and randomly mixed, and can rearrange by germ layer, and subdivisions of germ layers.

(But they can't reform the bodies of frogs or salamanders). However, dissociated sea urchin blastulas and gastrulas CAN re-form somewhat normal larvae! I have seen them do it!

Embryonic mammal and bird hearts, livers, retinas etc. will rearrange by cell type (demonstrated by J.P. Trinkaus, using radioactive labeling of DNA and time lapse photography)


J.P. Trinkaus with Tommy Harris at Oxford University, 1987


What can we learn from cell sorting about the normal embryological mechanisms that create anatomical arrangements?

Many researchers concluded that sorting is caused by different mechanisms than are used by embryos. (Prof Trinkaus for example) so nothing can be learned about normal mechanisms by studying sorting). He and I spent decades arguing about this question.

Wilson didn't believe that differentiated cells rearrange by cell type; He believed that cells change from one cell type to another, based on their new locations relative to each other. Later Wilson changed his mind, and argued that previously undifferentiated cells make the new sponges ("archeocytes" = "stem cells")

Wilson's preference for cell re-differentiation instead of rearrangement is analogous to Columbus' refusal to believe that America wasn't China.

Holtfreter believed sorting is caused by differences in cell-cell adhesiveness combined with amoeboid locomotion,

Holtfreter also believed that gastrulation, neurulation etc. are caused by changes in the same properties that cause sorting out.

His ideas led others to discover Cadherins, N-CAM and other cell surface adhesion proteins.

And also led to discoveries that changes in cell shapes and behavior can be caused by changes in which adhesion proteins they have. For example, the primary mesenchyme of sea urchins switch from one kind of adhesion protein to another just at the time they ingress. (Prof Rachel Fink discovered this)

Holtfreter commented on an interesting fact that embryonic cells sometimes can arrive at the same geometric arrangement by two or three (or more) pathways, in the sense of different sequences of intermediate stages.

In particular, neural tube cells:

    x) fold as cell sheets to form hollow balls
    y) sort out from random mixtures to form hollow balls,
    z) form hollow balls from solid masses of cells.

Someone should try to invent a single mechanism that can do all three.


Malcolm Steinberg invented "The Differential Adhesion Hypothesis" according to which cell aggregates behave like drops of non-living liquids, maximizing cell-cell adhesions like water molecules maximize closeness. (Supporters of this theory have a Wikipedia page:

Steinberg believed he had measured cell-cell adhesion of differentiated cells from resistance of cell aggregates to centrifugal flattening.

Part of his supporting evidence was that when he mixed pairs of differentiated cell types (for example, random mixtures of heart cells and liver cells) then liver cells always wound up at the outside surface and heart cells always clumped together inside the aggregates.

For every pair of cell types that he tried (42 pairs) their relative position of sorting was always transitive (in this sense: If cell type A sorts out to the interior position relative to cell type B, and is B sorts out interior to C, then A always sorted out interior to cell type C.)

Transitivity was predicted by the differential adhesion hypothesis. It can't be a coincidence that 42 different combinations do this.

When aggregates of single cell types were centrifuged, and them amounts of flattening were compared, then liver aggregates flattened more than heart aggregates, always in the same "transitive hierarchy" as in sorting out.

An opposing theory was that he actually was measuring strengths of contractility of cell aggregate surfaces, and of boundaries between one cell type and another.



Transitivity is predicted by any theory based on a quantitative difference. (If A contracts more strongly than B, and B contracts more strongly than C, then what can you predict about the relative strengths of A and C?)

Cell-cell adhesions by means of cadherins are qualitative differences (differences in kind rather than differences in amount)

Optional reading: some additional papers on this topic:

Green, Jeremy B.A. (2008).Sophistications of cell sorting. Nature Cell Biology 10(4), 375-377.

Maître, J.L., Berthoumieux, H., Krens, S.F., Salbreux, G., Jülicher, F., Paluch, E., and Heisenberg, C.P. (2012). Adhesion functions in cell sorting by mechanically coupling the cortices of adhering cells. Science 338(6104): 253-256.


Things to think about: Please suggest your own ideas about these combinations of phenomena:

i) Do you think gastrulation and gastrulation might have the same causes as cell sorting?

ii) Does Steinberg's evidence of transitivity of inside-outside sorting support his belief that quantitative differences (in amounts of adhesiveness) cause cell sorting, and therefore cause gastrulation?

iii) Neurulation is believed to be caused by differences in contractile differences in cell surfaces. Can you fit that in with observations that random mixtures of neural tube cells sort out to internal positions
and form hollow tubes?

iv) When the same end results can be reached by two or three or more different pathways (sequences of intermediate arrangements) does that mean that thermodynamics should be used to explain the causes?

v) Do hypotheses help biologists understand causes?

vi) Are hypotheses the best way to invent experiments?

vii) Most vertebrates form their central nervous system from sheets of cells that fold and seal their edges together. In dramatic contrast, teleosts form their nervous systems by hollowing out solid rod of cells.

What do you conclude (or guess) from this?
That the mechanisms are fundamentally different?
That the mechanism of sheet folding can also cause hollowing?
Something else?

viii) Birds form their neural tubes by folding a sheet of cells, except the posterior 15 or 20% parts of birds form their neural tubes by hollowing out ("cavitation") of a solid rod of cells.

Does this affect your opinion about whether the same mechanism can cause folding, and also cause cavitation,?
and also cause sorting out of randomly mixed cells?

ix) Neural tubes expand by pumping water into their interior. If the posterior ends of neural tubes were open, then inflation by water pressure couldn't work?

Recently, it occurred to me this may be why the posterior ends of bird neural tubes form by cavitation, and also why the entire neural tubes of teleosts cavitate instead of folding. Can you figure out a reason?
Hint: it has to do with progressive formation of the body from head to tail.


back to syllabus