The Fine Structure of the Vascular System of Amphioxus: Implications in tl!_e Development of Lymphatics and Fenestrated Blood Capillaries *

Abstract

The blood and lymph vessels of a few mammals have been quite extensively studied
by electron microscopy (3, 4, 5, 15, 22, 23, 2.J). By contrast, in lower animals even the
blood vessels have been relatively neglected. to say nothing of the lymphatics. The few
studies which have been made indicate that in reptiles (2), amphibians {31) and teleosts
(15) the blood vessels are similar to those in the mammals. In the elasmobranchs, however,
the interendothclial junctions appear less firmly closed, the basement membranes
are more tenuous and the venous vessels are intermittently attached to the connective
tissue by fine fibrils ( 11). These three features are strongly reminiscent of mammalian
lymphatics and arc probably associated with the low blood pressure of these fish, whid1
is sometimes "negative" in the venous vessels (13, 30).
It is of great interest that even in the fairly prim itive elasmobranchs, those which lack
true lymphatics (32) still have fenestrated capillaries in some organs {11 ). There is
evidence that the fenestrae allow the entry of large molecules and fluid into the venous
limbs of capillaries (7, 8, 11 ), both in these anima ls and in the higher vertebrates, where
they are very common in some regions. They may well supplement the lymphatic system,
especially in relatively motionless regions, or where the lymphatics are infrequent. It
appears, however, that the hagfishcs lack fenestrae, but have some open junctions in their
encocrine capillaries (9a). Thus their blood capillaries have some features in common
with the lympha tics of higher vertebrates. ( either open junctions nor fenestrae seem
to be present in the cerebral capill a ries of the myxines (27), but we have no information
about the rest of their vessels.)
The invertebrates have had even fewer studies of their vasculature. Only the cephalopocis
( 1) and earth-worms {17, 19. 20) have had any detailed description. The crustaceans
have been briefly mentioned (19) and the leech's neural ·endothelium • has been described
( 12~, but it is evident that this is very unusual in site, structure and function. In their
major vessels the cephalopods have per icytes with myofibrils, thick basement membranes
and endothelium which is nearly continuous, but with a few open junctions. In the more
peripheral vessels, the endoth elial cells gradually come to lie further and further apart,
until there are quite wide gaps 1- 10!! between them. However, the basement membranes
are always present, as is a complete investment of pericytcs - which come to lack the
myofibrils. The pericytes have closed junctions which, though they are not "tight june- 

tions" {16), contain dense material whid1 may present a considerable barrier to the
passage of large molecules. The higher blood pressures and plasma protein concentra tions
in the cephalopods have obviousl y caused developments analogous to those in the
higher vertebrates, but differing in d etail. The situation in the more primitive earthworm
is similar, but the endothelium is discontinuous even in the major vessels ( 17, 19,
20). (A point of nomenclature should be noted .h5:re: ·workers on the earthworm have
called the pericytes "endothelium", or "myoendotheliu~ and considered the endothelium
to be "amoebocytes lying on the basement membrane", which they considered
lay on the " lumenal side of the endothelium."' It was pointed out by Barber and Gra:ialdei
(1), however, that the true amoebocytes arc morphologica lly, and presumably functionally,
distinct from the true endothelium since they contain many granules, more
mitodlondria, etc. Hence they considered that there was true endothelium - though
often very discontinuous- inside the basement membrane, as in the mammals, and that
it was pericytes which contained the myofi brils.)
In order to help bridge the gap between the invertebrates a nd the vertebrates, it was
decided to study amphioxus, one of the most primi tive of chordates. Its low blood pressure
and plasma protein levels. and generally primitive development might be expected
to be associated with vascu lar structural and functional peculiarities. These would not
only be interesting in themselves, but, by their contrasts, might help to clarify what is
found in the higher vertebrates. In particula r, the way in which large molecules enter
these vessels from the ti ssues would be of significance both for the study of the lymphatic
system, and of the fenestrated blood capillaries.
The vascular structure of amphioxus detectable with the light microscope has been
well reviewed by Kampmeier (21), whose description has been used as the basis for this
paper. Briefly, there is a contractile ventral aorta, which pumps blood through the giU
arches and nephridiae, which then fl ows into the pair of dorsal aortae. These merge on
the stomach and intestine and supply the intestinal sinusoids, which run forwards on the
"liver ". Some of these converge into tbe contractile subintestinal vein, which supplies
the "liver" sinusoids, which in turn flow jnto lhe hepatic vein, and thence into the ventral
aor tae. In addition to this branchioenteric circulation. the dorsal aortae supplies sinusoids
to the segments of the body wall, which flow into the "cardinal" veins. (I shall use
"cen tra l" or "major" vessel to include the aortae and the main veins; "peripheral"
vessels refers to the rest of the vessels, which are basically sinusoids, i.e. they have wide,
irregular, often flat tened lumens, with - electron microscopicall y - gaps between the
endothelial cells.) Kamfnneier mentions a number of problems whid1 have been answered
in the present work, viz. the nature of the sinusoids, whether only the aortae have endothelial
linings, and the nature of the "lymph·· spaces. The way in which large molecules
and pa rticles enter the s inusoids of the gut and the vessels of the body wall has also
been studied.