The bearing of shape changes on the problem of red-cell structure is so great that we cannot afford to omit studying the transformations which occur in oval nucleated red cells in order to compare them with those in the discoidal non-nucleated red cells of man and of most mammals. The purpose of this paper, accordingly, is to give a description of the shape changes observed in the erythrocytes of camels, birds, reptiles, amphibians, and fishes.

In order to express the results in a small compass, I have condensed them into two tables and a series of observations which deal with points which require explanation or which ought to be put on record for the benefit of other investigators in the field.

(1) Material

As examples of the various vertebrate orders, I have used the camel (Came lus bactriens), the pigeon (Columba livia), the turtle (Pseudemis elegans), the frog (Rana esculenta), and the carp (Carassius auratus). The blood, drawn from a vein or from the heart, was received into heparin, and the measurements and other manipulations of the cells were made without delay.

(2) Measurements

The measurements of cell length and breadth, and of nuclear length and breadth, were made by the photographic method of Ponder & Millar (1924), with critical illumination and at a magnification of 480. No attempt was made to measure the thickness of the cells as seen on edge, for this is exceedingly difficult for technical reasons. Instead, the volume of the cells was found by dividing the percentage volume (high-speed hematocrit) by the red-cell count, and from the value for the volume, together with the figures for cell length 2a and the cell breadth 2b, the mean thickness can be calculated. The cell area is approximately 2πab. All measurements are given in μ.

A real difficulty arises in connexion with these measurements, for the red cells in birds, reptiles, amphibians, and fishes probably do not constitute an entirely homogeneous population. In birds, Kennedy & Climenko (1928) describe two types of cell, one with an oval nucleus and the other with a round nucleus, and I think that one can go further than this, and distinguish three types in fresh preparations: (i) oval cells with nuclei of varying degrees of eccentricity (the extent of which may be greatly exaggerated in fixed and stained films), (ii) oval cells with round nuclei, and (iii) round cells with round nuclei. The round cells with round nuclei are probably the younger cells, for they have an open nuclear structure and usually show diffuse basophilia; this is the view taken by Bizzofcero (1890) and by Kennedy & Climenko. In camels, Loo (1929) described a large round type of red cell in addition to the prevailing oval type, Cullen (1903) describes both oval and round cells in the blood of the skate and the dog fish, and in an early monograph (Ponder, 1924) I have referred to four types of erythrocyte in birds and to three types in reptiles, amphibians and fishes. It happens that the adult oval cell predominates (more than 90% of the cells seen), and So the measurements in Table 1 refer to this type. An investigation of the distribution of red-cell dimensions and an elaborate statistical treatment will be required before a complete description of red-cell shape in the lower vertebrates is available. Meantime we may remark upon the fact that young red cells (reticulocytes and erythrocyte precursors in the bone marrow) tend to be round, the distinctive shape characteristic of the animal (the biconcave disk in most mammals, the ovalocyte in some individuals, the oval cell in the camels, and the oval nucleated cell in the lower vertebrates) being fully exhibited in the adult red cell only.

(3) Slide and slip shape transformation

This shape transformation, due to the diffusion of alkali from glass surfaces and the transference of an ‘anti-sphering’ albumin from the cell surface to the glass by adsorption (Furchgott & Ponder, 1940), is peculiar to the discoidal mammalian red cell and the ovalocyte. Other types of cell show no shape change except that corresponding to slight crenation this consists of a mottling of the surface, not unlike that described by Millar (1925) in the mammahan ghost, together with wrinkles running out radially from the nucleus and terminating in fissures at the cell margins. A fine saw-toothed scalloping of the margins also occurs, but none of these changes are in any way comparable to the almost complete loss of shape observed in the crenated mammalian erythrocyte. Cells in saline seem to stick to the slide more readily than do cells in plasma.

(4) Effect of lecithin

The addition of an equal volume of a lecithin sol to a suspension of mammalian red cells in saline or in plasma causes a rapid transformation of disks to perfect spheres, irrespective of whether the cells are disks or ovalocytes (Ponder, 1939). No corresponding change occurs in the cells of camels or of pigeons, turtles, frogs, or fish, although a shape change which I call ‘lanceolation’ can sometimes be observed in the nucleated erythrocytes as a result of the addition of lecithin. When this change occurs, the cells lose their oval outline and become shaped like grains of wheat, with sharp ends and an almost diamond-shaped cross-section; sometimes this change occurs symmetrically, and sometimes one end of the cell remains oval while the other develops a lanceolate point. The extent of this lanceolation probably depends on the concentration of lecithin effective at the cell surface, for it is certainly quite variable throughout the preparation when a lecithin sol is run in under the cover-slip in the usual way and it is often entirely absent. If large amounts of lecithin are added, some of the cells may become prolytic forms, and ultimately haemolyse.

In the case of avian red cells at least, this lanceolation is increased by running pigeon plasma under the cover-slip, and for this reason I think that it bears little relation to the disk-sphere transformation of mammalian erythrocytes, which is reversed by plasma.

(5) Effects of rose Bengal

Washed mammalian disks and ovalocytes are promptly converted into spheres by rose bengal in concentrations 10−6M when illuminated, and the same shape change occurs rapidly in higher concentrations of the dye in the dark. The oval red cells of the camels and of the birds, reptiles, amphibians and fishes show no such immediate change, although they may show lanceolation. After some minutes (depending on the dye concentration, the intensity of illumination, etc.) the red cells of the mammals (including the ovalocytes and the cells of the Camelidae), and the cells of the pigeon also, become first less oval and then spherical, and show the characteristic homogeneous appearance and bright peripheral diffraction rings of the prolytic sphere. These shape changes occur in the cells one by one, without any apparent change in volume, and the process can be hastened by opening the iris diaphragm and flooding the cells with light. Each prolytic sphere ultimately haemolyses, losing its pigment and at the same time becoming an oval ghost.

The red cells of the turtle, the frog, and the carp, on the other hand, may first become lanceolate, but then turn into circular and flat prolytic disks or plates. There is great variation in this prolytic shape change ; some cells lose their pigment with scarcely any alteration in shape, others form oval disks, still others circular disks, and a few form what are apparently perfect prolytic spheres. The diversity is so great that one suspects that one is dealing with a mixed population in which there are difference of kind as well as of degree. These prolytic forms are flat even up to the moment of losing their pigment, and in this respect they differ from the prolytic form in birds, where the element of flatness is lost before lysis. As an extreme instance, some of the red cells of the frog are haemolysed by rose bengal without any shape change at all.

(6) Saponin and bile salts

After a few minutes the discoidal cells of mammals and the mammalian ovalocyte become prolytic spheres following the addition of saponin or bile salts. The red cell of the camel diminishes in its long axis, acquires a dull homogeneous appearance, and shows brilliant diffraction bands at its edges; this prolytic spheroid finally haemolyses, sometimes without a further change in shape, and sometimes after becoming a prolytic sphere. The pigeon red cell shows similar change, becoming either a sphere or an oval spheroidal body. The red cell of the turtle, the frog, and the carp, on the other hand, do not lose their element of flatness; as looked down upon, they remain ovals or become round, but even at the moment of haemolysis they are flat plates, with their c-axis apparently unchanged. Some of these shape changes have been described by Shattuck (1928), who has pointed out that sodium oleate breaks down both the cell membrane and the nuclear membrane, whereas saponin and sodium taurocholate break down the cell membrane only.

(7) Hypotonic plasma

All types of red cell swell in hypotonic plasma, behaving as more or less perfect osmometers. Oval cells become first lemon-shaped and later appear lemon-shaped or spherical. They finally haemolyse, leaving oval, flat, ghosts with a highly refractile nucleus and a very indistinct outline. The oval form of the ghost is assumed at the moment of haemolysis.

(8) Form of the ghost

Irrespective of the shape changes which precede lysis, the form of the haemolysed ghost is always substantially that of the red cell from which it is formed. This form is assumed at the moment of haemolysis, and constitutes very strong evidence for the existence of a pre-formed structure either at the surface or in the interior (see Ponder, 1942). The dimensions of the ghost are always smaller than those of the original cell. In the case of the pigeon, for example, the cell length is 15 μ. and the cell breadth is 8·6 μ; the ghost is 9·5 μ long and 6·3 μ broad. So far as can be gathered from photographic measurements which are not altogether satisfactory, the size of the nucleus in unchanged.

  1. The almost instantaneous, reversible, disk-sphere transformations (slide and coverslip, lecithin, and rose bengal) occur only in the biconcave red cells of mammals (including ovalocytes). They do not occur in the biconvex red cells of the Camelidae, nor in any type of nucleated erythrocyte.

  2. All types of red cell show prolytic changes (saponin, bile salts, high concentrations of rose bengal, hypotonicity). As we ascend from the fish to the mammal, the prolytic form shows an increasing departure from the original shape and an increasing approach to the form of a perfect sphere. Thus in fishes the prolytic form is an oval, or sometimes round, plate (at least two axes unequal), while in mammals it is a sphere (all three axes equal).

  3. The two foregoing conclusions mean that the structure responsible for the red-cell shape is more labile in the higher vertebrate forms, in which it seems to be an essentially surface structure, than in the lower vertebrate forms, in which it seems to be an organized cytoplasm in the usual sense of the term.

  4. The form of the ghost is always essentially that of the red cell from which it is derived, and this form is assumed at the moment of haemolysis. The sequence of shape changes suggests that the inherent structure of the cell (and of the ghost) is subjected to the action of obscure forces which develop during the haemolytic process, and which disappear when lysis takes place.

Bizzozero
(
1890
).
Arch. mikr. Anat
.
35
,
424
.
Cullen
(
1903
).
Johns Hopk. Heap. Bull
.
14
,
352
.
Furchgott
,
R. F.
&
Ponder
,
E.
(
1940
).
J. Exp. Biol
.
17
,
117
.
Kennedy
,
W. P.
&
Climenko
,
D. R.
(
1928
).
Quart. J. Exp. Physiol
.
18
,
43
.
Loo
,
C.-T.
(
1929
).
Chinese J. Physiol
.
3
,
323
.
Millar
,
W. G.
(
1925
).
Quart. J. Exp. Physiol
.
15
,
253
.
Ponder
,
E.
(
1924
).
The Erythrocyte and the Action of Simple Haemolysins
.
Edinburgh
:
Oliver and Boyd
.
Ponder
,
E.
. (
1939
).
J. Physiol
.
95
,
9
P.
Ponder
,
E.
(
1942
).
J. Exp. Biol
.
22
,
220
.
Ponder
,
E.
&
Millar
,
W. G.
(
1924
).
Quart. J. Exp. Physiol
.
14
,
67
.
Shattuck
,
G. E.
(
1928
).
J. Gen. Physiol
.
12
,
17
.