The nervous system of amphioxus: structure, development, and evolutionary significance

Cross references:    Amphioxus PDFs       Children of the Amphioxus  

The nervous system of amphioxus: structure, development, and evolutionary significance 
 - Canadian Journal of Zoology 
The nervous system of amphioxus: structure, development, and evolutionary significance     
Note: 
The purple link is to the original article which is no longer available to the general public for free.  The green link is to the PDF which I was able to download at the local university library.  I made an initial attempt to alter the line breaks and fonts, but that did more harm than good, so here it is in its original form. 

The nervous system of amphioxus: structure,
development, and evolutionary significance
Helmut Wicht and Thurston C. Lacalli

Abstract:

    Amphioxus neuroanatomy is important not just in its
own right but also for the insights it provides regarding
the evolutionary origin and basic organization of the vertebrate
nervous system. This review summarizes the overall
layout of the central nervous system (CNS), peripheral nerves,
and nerve plexuses in amphioxus, and what is currently
known of their histology and cell types, with special attention
to new information on the anterior nerve cord.  
    The intercalated region (IR) is of special functional and
evolutionary interest. It extends caudally to the end of somite 4,
traditionally considered the limit of the brain-like region of the
amphioxus CNS, and is notable for the presence of a
number of migrated cell groups. Unlike most other
neurons in the cord, these migrated cells detach from the ventricular
lumen and move into the adjacent neuropile, much as developing
neurons do in vertebrates.    
    The larval nervous system is also considered, as there is a
wealth of new data on the organization and cell types of the
anterior nerve cord in young larvae, based on detailed
electron microscopical analyses and nerve tracing studies,
and an emerging consensus regarding how this region
relates to the vertebrate brain.  
    Much less is known about the intervening period of the life
history, i.e., the period between the young larva and the adult,
but a great deal of neural development must occur during
this time to generate a fully mature nervous system.  
    It is especially interesting that the vertebrate counterparts of at
least some postembryonic events of amphioxus neurogenesis
occur, in vertebrates, in the embryo. The implication is that the
whole of the postembryonic phase of neural development in
amphioxus needs to be considered when making phylogenetic
comparisons. Yet this is a period about which almost nothing is
known. Considering this, plus the number of new molecular and
 immunocytochemical techniques now available to researchers,
there is no shortage of worthwhile research topics using amphioxus,
of whatever stage, as a subject.  

Note:  Since this journal was published in Canada, there was an
obligatory translation of the Abstract into French, which I've deleted. 


Introduction 

    The common ancestor of present-day vertebrates and the
invertebrate cephalochordates has long been extinct, but if
one had to choose a living species likely to resemble it most
closely, amphioxus (Branchiostoma spp.; lancelets) would
be the organism of choice (Presley et al. 1996; Holland
2000). As such, amphioxus is of key importance to investigations
into vertebrate origins and characteristic features of
vertebrate organization. The nervous system is of special interest
because its basic plan is highly conserved among vertebrates,
yet almost nothing is known about how that plan
originated. One has little recourse except to conduct investigations
of amphioxus and other protochordates. This was the
rationale behind the anatomical studies of amphioxus carried
out in the late 19th century, principally on Branchiostoma
lanceolatum (Pallas, 1774), the European species, by leading
comparative zoologists of the day. There has been something
of a hiatus in the 20th century, in part because of concerns
that the apparent simplicity of amphioxus is due to its being
degenerate and derived, rather than primitive, but also because
of diminishing returns from studies using conventional
light microscopical methods. Much of the original literature
is in German, including major studies by Rohde (1887),
Retzius (1891), Dogiel (1903), and Franz (1923) and a comprehensive
review by Franz (1927).   Bone (1959, 1960a, 1961)
provides a comprehensive summary of previous work
for non-German speakers, along with much new information
from his own research. The most recent reviews on amphioxus
are by Ruppert (1997), for the general anatomy of the
animal as a whole, and Nieuwenhuys (1998), for the nervous
system.
     Why then review the nervous system yet again? The answer
has two parts. First, the nervous system of amphioxus
is surprisingly peculiar from a vertebrate perspective, so it is
not a trivial matter to get a good conceptual feel for its organization.
Just as one might have more than one tourist guide
when visiting an unfamiliar city for the first time (each preferably
with a good set of maps), there is a distinct benefit in
having more than one review to consult regarding amphioxus
neuroanatomy, preferably with good illustrations, as
each will inevitably have a slightly different perspective.
    Second, and more importantly, a whole new tool kit of molecular
and immunocytochemical techniques is now being
employed to reexamine the structure and development of
phylogenetically interesting animals, including amphioxus.
New studies of the amphioxus nervous system now appear
on a regular basis, and gene expression studies frequently
also focus on the nervous system. Given our very limited understanding
of amphioxus neural structure and function,
these new results are often difficult to interpret, typically
raising more questions than they answer. A reexamination of
what we do know, as of 2004, is therefore not out of place.
What is perhaps more important, however, is the way the
new results highlight major gaps in our knowledge — for
example, the postembryonic events that turn a larval nerve
cord into an adult one (see below). Thus, as well as a summary
of facts, we try here to provide insights into areas of
emerging interest and highlight issues where we think major
advances can be expected in the future. We hope, by this
means, to stimulate new thinking about the anatomy and
functional organization of the amphioxus nervous system,
and better inform research into the origin of vertebrates and
their nervous system.
     In the account that follows, the adult nervous system is
dealt with first, followed by a separate section on the larval
nervous system, specifically that of young larvae 12–14 days
old. Separate treatment is necessary because the nature of
the data is so different for the two stages, being largely conventional
anatomical descriptions based on light microscopy
and immunocytochemistry in the case of the adult, and
cellular-level details from electron microscopical (EM) reconstructions
in the case of the larva. Because our knowledge
of the intervening period of development is incomplete,
it is not always clear how structures and cell types in the
nervous system of young larvae relate to those in the adult,
and it is consequently impossible to make the two sections
as seamlessly complementary as one ideally might like.   

Anatomical overview   

General appearance   


     The central nervous system (CNS) of adult lancelets consists
of a tubular nerve cord, located directly above the notochord,
that extends most of the length of the body (Fig. 1).   






Fig. 1.
    A highly schematic lateral view of a lancelet (Brachiostoma sp.) with
63 myomeres (m1 to m63), showing the general organization of the central and
peripheral nervous systems in relation to peripheral structures. The first
four and last three members of the myomere series are highlighted
(light shading) to clarify the relationship between myomere number and
nerve number. Regions occupied by major blocks of non-myomeric muscle
are indicated by darker shading. Selected body regions are cut away,
as indicated by the jagged lines.  

Note:  An enlarged version of Fig. 1. in which the labels are more
easily read and in which the a
bbreviations and their corresponding
terms are listed
can be found at 
Amphioxus Nervous System .  



     Rostrally, the notochord extends forward of the nerve cord,
to almost the tip of the body, hence the name for this taxon:
Cephalochordata. The nerve cord itself begins a few hundred
micrometres caudal to the anteriormost coelomic cavity,
which supports the rostral fin, and just forward, by a few
dozen micrometres, of the first myomere. Caudally, the nerve
cord reaches to the tip of the caudal fin where, just above the
caudal end of the notochord, it forms a terminal enlargement
of variable form, the caudal ampulla (Retzius 1891; Franz
1923).
     Unlike the situation in vertebrates, the lancelet nerve cord
displays few externally visible landmarks except for the
above-mentioned ampulla and a transient anterior swelling
(the cerebral vesicle) in young larvae. There are thus no
neuromeres or other external indications of segmental arrangement
other than the serially repeated exit points for the
dorsal nerves. Internally, however, pronounced cytoarchitectural
differences can be identified along the rostrocaudal
axis, both in the adult and in larvae. These can be used to
define a series of major regions along the nerve cord, and
some of these can be further subdivided. Because of the absence
of external landmarks, the boundaries of these subdivisions
and their relation to the peripheral nervous structures
are best described with reference to the myomeres, of which,
on average, there are 63 pairs in B. lanceolatum (Poss and
Boschung 1996). This poses some problems if one is trying
to compare CNS architecture throughout development because
(i) though it is likely, it is not known for certain
whether the myomeres are permanently fixed in position in
relation to the nerve cord, and (ii) myomeres expand both
lengthwise and dorsoventrally during development, which,
combined with their chevron shape, means there is an increasing
zone of overlap between them that varies depending
on position along the dorsoventral axis. (iii) The left and
right rows of myomeres in the adult are not aligned; the first
larval somites, however, from which the myomeres develop,
are.   In the neurula stage, the left row of somites (as well as
the left row of dorsal nerves and neuromuscular contact
zones) starts to shift anteriorly (Conklin 1932). To make
things even more complicated, this “left forward shift” is
less pronounced in the region of the first five somites
(myomeres). Thus, in the adult, the plane defined by the
rostral tips of the first myomeres is almost perpendicular to
the long axis of the body, whereas the tips of increasingly
more posterior myomeres define increasingly more oblique
planes. Approximately at the level of the sixth pair of myomeres,
the “full” oblique angle of about 45° (i.e., the typical
“half-myomere staggering”) is reached and maintained
throughout the caudal part of the body. In the present review,
we shall use the transverse planes defined by the tips
of the left row of myomeres as landmarks. Past accounts,
however, have not always described the location of CNS
landmarks in a sufficiently precise way to overcome these
problems.
     The peripheral nervous system (PNS) consists of a set of
peripheral plexuses and a segmental series of intermyomeric
dorsal nerves (= dorsal roots or “true” nerves, as opposed
to the apparent ventral roots, which are in fact muscle
processes rather than nerves). The dorsal nerves are entirely
devoid of ganglia; in other words, amphioxus has no counterparts
of the vertebrate dorsal root ganglia. Peripheral
nerves in amphioxus instead consist solely of axons, which
derive from both central and peripheral neurons, and a few
glial-like support cells. The peripheral neurons reside in various
peripheral plexuses, which are especially well developed
around the gut, and in sense organs and the skin. The
nerves themselves issue from the dorsolateral margins of the
nerve cord, their proximal parts being located in the
myosepta, i.e., they pass between the myomeres. Owing to
the left-right asymmetry of the myomeres (the left row is
displaced a half-segment forward of the right one), the
nerves are staggered left to right.

The pattern of peripheral innervation for a typical nerve of the
trunk is shown in Fig. 2.   




Abbreviations and their corresponding terms:   

   
ap = atrial papillae   
    chd = notochord 
   
enc = encapsulated nerve endings   
   
gd = gut diverticulum   
   
gon = gonad   
    m = muscle 
   
mpf = metapleural folds   
   
ph = pharynx   
   
pm = pterygeal muscle   
    rdors = dorsal ramus of the dorsal nerve 
    rvent = ventral ramus of the dorsal nerve 
   
rventc = ventral ramus, ventral cutaneous branch   
   
rvisc = visceral ramus   
   
rvisca = visceral ramus, ascending branch   
   
rviscd = visceral ramus, descending branch   


Fig. 2.
    A highly schematic transverse section of a lancelet at the level of myomere 26,
at the junction of the pharynx and gut.  The notochord is lightly shaded; darker
shading indicates regions occupied by various peripheral plexuses. The nerve
cord and one dorsal nerve are shown in black; black dots indicate schematically
where nerve cells associated with the atrial plexus would lie.  



     Innervation patterns in the rostral and oral region are somewhat
different; Fig. 3 shows this in detail, while cytological
details and cell types as seen in typical transverse sections of
the nerve cord are shown in Figs. 5 and 6.   

Fig. 3





Abbreviations and their corresponding terms :   

   
cdQ = corpuscles of de Quatrefages   
   
chd = notochord   
   
CNS = central nervous system   
   
ibp = inner buccal plexus   
    n1, n2, n3… = dorsal nerve 1, 2, 3…
    obp = outer buccal plexus   
   
vp = velar plexus   


Fig. 3.    
    A ventrolateral view of the rostrum and oral region of a lancelet,
showing the first eight dorsal nerves (n1 to n8), but only
those on the left side. The buccal and velar plexuses supplying the
buccal chamber that extends from the base of the cirri to the velum
are shaded. Also shown are the first cell of Rohde (arrow),
Hatschek’s pit (white arrowhead), an anastomosis between nerves
1 and 2 (open arrowhead), and two of the neuromuscular contact
zones (black arrowheads), for myomeres 5 and 6 in this instance.
    Note that each contact zone consists of a small dorsal and a
larger ventral compartment.  



Anterior nerve cord
   


     The tip of the adult nerve cord rostral to myomere 1 is referred
to as the anterior vesicle (= cerebral vesicle (Willey
1894), Gehirn (Edinger 1906), archencephalon (von Kupffer
1906), and Stirnbläschen (Franz 1927)). The central canal is
expanded in this region and is roughly circular in cross section.
It opens on the left to the outside via Kölliker’s pit
(Kölliker 1843), a remnant of the anterior neuropore (Willey
1894). Two pairs of nerves emerge from the anterior vesicle
to supply the rostral region. By convention these are numbered
1 and 2 (e.g., Dogiel 1903; Kutchin 1913;
Nieuwenhuys 1998), but are referred to also as the rostral
and anterodorsal nerves, respectively (e.g., Lacalli 2002b).
They differ from remaining members of the dorsal nerve series
in several respects. First, because of the absence of
myomeres in the rostrum, both nerve pairs travel forward in
the connective tissue sheath surrounding the nerve cord and
notochord, whereas the rest of the dorsal nerves pass
through the myosepta. Second, the various branches of
nerves 1 and 2 carry small bulbous clusters of cells, the corpuscles
of de Quatrefages (Fig. 3), putative mechanosensory
organs consisting of primary sensory cells enclosed in a capsule
(Baatrup 1982) that occur nowhere else in the body.
     Nerve 1, the rostral nerve, is unusual also in that it enters the
nerve cord ventrally rather than dorsally. It has nevertheless
become customary to include it with the rest of the dorsal
nerves (e.g., Dogiel 1903; Kutchin 1913; Nieuwenhuys
1998). We follow this convention here, though it has the disadvantage
that nerve and myomere numbers then no longer
match; i.e., a given nerve n would emerge from between
myomeres n – 1 and n – 2 (see Fig. 1).
     The region of the adult nerve cord extending from the anterior
vesicle through the first four myomeres is recognized
as distinctive enough to require a special designation. It has
been referred to variously as a deuterencephalon (von
Kupfer 1906), hindbrain (Fritzsch 1996; Holland 1996), or
caudal brain region (Nieuwenhuys 1998), but was regarded
as a part of the spinal cord (Rückenmarksteil) by Franz
(1923). Most recently, Ekhart et al. (2003) have coined the
term intercalated region (IR) so as to avoid the implications
of homology associated with the older terms.  
    The IR as defined by Ekhart et al. extends from the
infundibular organ (Boeke 1902, 1908), which produces
Reissner’s fiber (Obermüller-Wilén 1976), to the first giant
cell of Rohde (Rohde 1887). It contains three subdivisions
(anterior, intermediate, and posterior), described below,
and is characterized by the presence of a number of cell
types and groupings not seen elsewhere in the spinal cord.
The most conspicuous of these are the large dorsal cells of
Joseph (Joseph 1904), which are putative photoreceptors
(see Figs. 6, 7).
    The IR gives rise to nerves 3–6, which, like all the dorsal nerves,
send branches to the corresponding component of the
subepidermal nerve plexus in the skin. These nerves also
connect with two peripheral nerve rings (or plexuses) associated
with the buccal region (Fig. 3), namely the buccal
plexus, which innervates the buccal cavity, cirri, and associated
muscles (the internal and external labial muscles, Franz
1927), and with the velar plexus, which innervates the tentacles
and sphincter muscle of the velum.
     An important non-neural structure located in this region is
Hatschek’s pit (Hatschek 1884), which lies ventral to the IR
at the junction of myomeres 3 and 4. This is actually a part
of the ciliated wheel organ of Müller (Müller 1844; Franz
1927), located in the roof of the buccal cavity (Figs. 1, 3,
7C). In some lancelet species the pit forms a direct contact
with the base of the nerve cord (see below).
     From both the adult anatomy and gene expression data it
is clear that the anterior vesicle and IR together form a CNS
region with a number of features in common with the vertebrate
brain: they form the anterior part of the neuraxis, their
nerves supply the region associated with the buccal cavity,
and cytoarchitectural specializations occur that are not present
in more caudal regions of the nerve cord. There is strong
evidence that these anterior regions also contain homologs
of neural centers and sense organs that occur in the brain of
craniates, as discussed more fully in the last section. It is
nevertheless difficult to determine the caudal extent of this
brain-like region with any precision and define it in a way
that applies equally to both larvae and adults. In adults, the
first Rohde cell is usually assumed to mark the posterior end
of the brain, but this is based more on convenience (the cell
is very easy to recognize) than on evidence, since there are
other internal landmarks and discontinuities that could be
used just as well. On the other hand, determining the exact
extent of the brain in craniates can be problematic, as the
boundary between spinal cord and brain, as well as the number
of cranial nerves, can vary from species to species
(Nieuwenhuys 1998).
     One way to define the brain is as that part of the CNS enclosed
in the cranium, but this is obviously not a useful criterion
when applied to acraniates.   If, instead, the vertebrate
brain is defined in terms of its peripheral nervous connections,
which supply the full extent of the pharyngeal region,
then the caudal end of the brain of lancelets would come to
lie somewhere in the middle of the body. In larval lancelets,
expression domains of Otx and various Hox genes indicate
that an equivalent of the vertebrate mid-hindbrain boundary
may be present somewhere in the front half of somite 2,
with a hindbrain-like region extending thereafter to about
somite 7 or 8 (Shimeld and Holland 2005). This would
move the junction between brain and spinal cord to a position
three or four myomeres caudal to the first Rohde cell.
The nerve cord in the neurula also extends from somite 1 to
somite 8, which has led to the suggestion that the entire
neurula is roughly equivalent in axial extent to the vertebrate
head (Gilland and Baker 1993). This would make the nerve
cord in the neurula coextensive with the vertebrate brain,
with most of its length being hindbrain-like in character.
While this may be true, the details of what it really means to
be “hindbrain-like” remain to be worked out (Jackman and
Kimmel 2002; Mazet and Shimeld 2002), so we must conclude
that the exact caudal boundary of the brain in lancelets
cannot currently be determined with any certainty.   


Spinal cord, dorsal nerves, and associated peripheral
structures
   

     The largest region of the CNS is commonly referred to as
the spinal cord, in recognition of its general similarity to that
structure in vertebrates. Using the first Rohde cell as the
landmark that divides it from the anterior CNS in the adult,
the spinal cord stretches from the fifth myomere to the last,
and thus gives rise to most of the dorsal nerves. Up to nerve
53, which innervates the anal sphincter muscle, the dorsal
nerves connect the spinal cord to the atrial nervous system,
the most extensive and complex component of the PNS. Figure
2 shows how the nerves and plexuses are arranged in the
transverse plane. Once a typical nerve leaves the nerve cord,
it enters the intermyotomal septum and passes laterally
along it, dividing into a dorsal and a ventral ramus as it
leaves. The rami travel under the skin lateral to the muscle
block (note that in vertebrates the corresponding nerve instead
lies medial to the trunk musculature), dorsally in the
case of the dorsal ramus and ventrally in the case of the ventral
ramus, to supply fibers to the subepidermal nerve plexus
underlying the skin of the dorsal fin (dorsal ramus) or the
subepidermal plexus beneath the skin on the flanks (ventral
ramus). At the ventral margin of the myomere, the ventral
ramus divides into a ventral cutaneous ramus, which supplies
the skin of the metapleural folds, and a visceral ramus,
which turns medially towards the wall of the atrial cavity,
where it again divides into an ascending and a descending
branch. The atrial epithelium and the epithelia of all organs 
that either flank it (e.g., gonads and pterygeal muscle) or are
located within it (pharynx and endostyle, the gut and its diverticulum)
are underlain by a massive system of nerve cells
and fibers collectively termed the atrial nervous system
(Bone 1961). The descending branch of the visceral ramus
contributes to the gonadal and pterygeal portions of that system
and, in addition, contributes motor fibers to the crossstriated
pterygeal (= transverse) muscle. The ascending
branch of the visceral ramus turns dorsally and gives off fibers
to the plexus covering the gonads and the lateral walls
of the atrium. It then turns medially and reaches the
denticulate ligament, which tethers the pharynx and the gut
to the roof of the atrium, and by this means it reaches the
nerve plexus of the pharynx, gill bars, endostyle, and gut
plus diverticulum.
     On the basis of its internal appearance, the adult spinal
cord can be subdivided into three major regions (this account
follows Franz 1923, 1927). The anterior part of the
cord, at the level of myomeres 5–11, is characterized by the
presence of large Rohde cells with descending axons and by
a high density of pigmented, photoreceptive organs of Hesse
(Hesse 1898). Peripherally, nerves connect this part of the
cord with the most caudal parts of the buccal nerve plexus,
via nerve 7, while the more posterior nerves in this region all
connect with the atrial nervous system (Bone 1961; see
Figs. 1, 2).
     The intermediate part of the spinal cord, from myomeres
12 to 38, lacks Rohde cells, and the density of Hesse organs
is reduced here. Peripherally, nerves from this region also
connect to the atrial nervous system. Its rostrocaudal extent
coincides roughly with that of the gonads, the caudal limit
being approximately at the level of the atriopore, whose
cross-striated sphincter is innervated by nerves 40 and 41
(Kutchin 1913). The transition between pharynx and gut is
located approximately at the level of myomere 26 or 27. The
atriocolemic funnels, enigmatic organs that may be sensory
in nature, are located just above this point. They are surrounded
by the cross-striated trapezius muscles (Franz 1927;
Bone 1961), which are innervated by the ascending visceral
branches of nerves 27–29.
     The posterior part of the spinal cord extends from the
atriopore, at the level of myomeres 38–40, to the last myomere,
typically number 63 in B. lanceolatum (Poss and
Boschung 1996). The number of Hesse organs increases again
in this region and the Rohde cells reappear. Nerves 51–53
supply the anal sphincter muscle (Kutchin 1913). Nerves
caudal to this point lack visceral rami (Franz 1927). The last
nerve (nerve 65 in an animal with 63 myomeres) leaves the
spinal cord at the caudal border of the last myomere and
innervates the skin adjacent to the tail fin.   


Terminal filament and caudal ampulla   

     Still more caudally, the spinal cord tapers into a thin
ependymal tubule, the terminal filament (filum terminale),
which connects the cord to a caudal ampulla just above the
posterior tip of the notochord. Reissner’s fiber, which is produced
in the infundibular organ, extends into that caudal
ampulla, where it is apparently phagocytosed by specialized
cells in the walls of the ampulla (Obermüller-Wilén and
Olsson 1974). There are numerous nerve fibers of unknown
origin along the ventral and lateral margins of the caudal
ampulla. These contain dense core vesicles, which implies
that this structure may function as an endocrine organ comparable
to the urophysis of anamniotic vertebrates (Ruiz and
Anadón 1991a).   


The adult PNS 
  

Peripheral plexuses
   

     The system of peripheral plexuses in amphioxus is unusual
in a number of respects. It supplies a fine meshwork of
free nerve endings that run through the entire epidermis
(Dogiel 1903; Kutchin 1913; Franz 1923, 1927; Welsch
1968b, see Fig. 2) such that every epidermal cell is apparently
in contact with them (Lele et al. 1958). The free nerve
endings probably arise from populations of intramedullary
sensory neurons, of which there are two types, according to
Bone (1960a, 1961): the dorsal bipolar or Retzius bipolar
cell and the dorsal root cell (see Fig. 4). The precise source
of the free nerve endings has not been determined with certainty,
however.
     The epithelia that line the buccal and atrial cavities, as
well as the organs embedded within them, are also supplied
with an extensive set of neural plexuses collectively known
as the atrial nervous system. Though the system is more or
less continuous, it is usually subdivided on the basis of the
organs it innervates; i.e., there are buccal, velar, gonadal, parietal,
pterygeal, pharyngeal, and endostylar subdivisions,
and so on (Bone 1961). While the various plexuses carry
motor fibers innervating muscles associated with the atrium,
they also contain the cell bodies and fibers of a large number
of peripheral neurons (Figs. 2, 4). It has been argued (e.g.,
Boeke 1935) that the plexuses are homologs of the autonomic
nervous system of craniates, but this is almost certainly
false. Enteric neurons in vertebrates develop from
migratory neural crest cells, a category of embryonic cells
that is entirely absent in amphioxus so far as can be determined
(Baker and Bronner-Fraser 1997). This implies by default
that neurons in the peripheral plexuses in amphioxus
arise locally, in situ, though Lacalli (2004) has pointed out
that the embryo is so small at the time that neuronal precursors
are probably first deployed that an origin close to the
neural plate, similar to that of placodes and the neural crest,
cannot be ruled out.    
    It is also true, however, that the neurotransmitters
identified to date in amphioxus peripheral neurons
differ from those released by autonomic neurons in
vertebrates. Specifically, neither acetylcholine (Flood 1974)
nor catecholamines (Moret et al. 2004) occur in peripheral
neurons, at least in the atrial nervous system.   FMRFamide,
however, does occur (Bone et al. 1996), and this is a transmitter
restricted mainly to the CNS in vertebrates. In addition,
many of the peripheral neurons in amphioxus send
axons into the nerve cord via the dorsal nerves (Holmes
1953; Bone 1961). This implies a sensory function, rather
than a motor one.
     Based on these unique features, Bone et al. (1996) concluded
that there were no obvious homologies between the
PNS of amphioxus and that of vertebrates. In general terms
this conclusion appears justified. Nevertheless, in rats, centripetal
projections are known from peptidergic peripheral
neurons (called rectospinal neurons) resident in the walls of
the anus (Dörffler-Melly and Neuhuber 1988). These are
evidently not simply displaced dorsal root ganglion cells, but
constitute instead a novel class of vertebrate enteric neurons
(Neuhuber et al. 1993). They could conceivably be a relic of
a primitive mode of visceral innervation related to that in
amphioxus. Conversely, they could be a derived feature of
no phylogenetic significance.
     The velar and buccal plexuses deserve special attention
owing to their curious asymmetry (Dogiel 1903; Kutchin
1913; Franz 1923, 1927). As shown in Fig. 3, there are two
buccal plexuses, an outer and an inner one, and a single velar
plexus (Bone 1961). The velar plexus derives developmentally
from the oral plexus that encircles the larval mouth
as it moves ventrally and caudally during metamorphosis
(Franz 1923; Kaji et al. 2001; the term “oral” should thus be
avoided when referring to neural structures associated with
the cirri because they are really preoral, buccal structures).
As with other parts of the atrial nervous system, the buccal
and velar plexuses combine sensory components (Bone
1961) with motor ones; the latter innervate the labial and velar
muscles.   
    The connection to the nerve cord via nerves 1–7
(1–8 according to Dogiel 1903; Kutchin 1913) is highly
asymmetrical. Nerves 3 and 4 on the left side are exceptional
in having contralateral branches that connect with the inner
buccal plexus on the right side. In addition, a subsidiary
branch from the contralateral branch of the left nerve 4 connects
to the right side of the velar plexus, while its left side
connects to nerve 5 by means of a caudal branch from that
nerve. This is all a consequence of the fact that the larval
mouth develops initially on the left side and is innervated
entirely by nerves emerging from the left side of the nerve
cord (Lacalli et al. 1999). The initial connections are then
retained during subsequent development, so the nerves are
dragged along as the mouth is repositioned.   


Peripheral sensory cells and organs
   

     Lancelets have an assortment of sensory cells and organs
located both inside the nerve cord and outside it. The former
are dealt with in relevant sections dealing with the adult and
larval CNS; they include the various photoreceptor systems,
which are all intramedullary, and Kölliker’s pit, for which an
olfactory function has been suggested. Hatschek’s pit, located
in the roof of the buccal cavity (Fig. 3), is not, strictly
speaking, part of the nervous system, but is considered here
because of its close association with it.   

Multicellular organs
   

     The peripheral tissues are well supplied with solitary sensory
cells and free nerve endings but harbor only a few
multicellular structures to which a sensory function can be
attributed. Four examples are considered here.    
    (1) The atriocolemic funnels, first described by Lankester (1875),
consist of paired conical recesses in the dorsal surface of the
atrial cavity that project anteriorly into the subchordal
coelom (Figs. 1, 2; see Willey 1894; Franz 1927). Both the
funnels and the surrounding striated trapezius muscle are
densely innervated by a branch from the ascending visceral
ramus of nerve 27 (Holmes 1953; Bone 1961) and, to a
lesser degree, by neighboring dorsal nerves (Franz 1927).
The fibers may be chiefly involved in the innervation of the
trapezius muscle, but the atriocoelomic funnels themselves
contain many uni- and multipolar neurons (Bone (1961)
distinguished three types) whose axons enter visceral rami;
from there they appear to travel to the nerve cord. No function
has yet been ascribed to the funnels, however.    
    (2) The atrial papillae of Müller (1844) were initially thought
to be excretory in nature (as renal papillae, Willey 1894). The
papillae are located in the floor of the atrial cavity (Fig. 2)
and are concentrated in the vicinity of the atriopore. They
consist of longitudinal strips of tall, densely packed cells,
many of which appear to be flagellated primary sensory
cells (Bone 1961; Bone et al. 1996; Ruppert 1997). Again,
their function is unknown.    
    (3) The encapsulated endings of
Fusari (1889) (see also Bone 1960b) are formed by free
nerve endings surrounded by clusters of cell nuclei. They are
located in the lateral walls of the metapleural folds (Figs. 2,
4) and may be mechanoreceptive.    
    (4) The corpuscles of de Quatrefages
(de Quatrefages 1845) are located in the connective
tissue of the rostrum along the branches of the first
and second nerves. They are typically located at branch
points, mainly at distal branches just proximal to the nerves’
entry into the subepidermal plexus (Dogiel 1903; Baatrup
1982; Fig. 3). The corpuscles consist of support cells and
peripheral neurons enclosed in a capsule of connective tissue
(Franz 1923). The neurons bear a pair of cilia that project
into a small central cavity and give rise to axons that enter
the branches of the adjacent nerves (Baatrup 1982). The
corpuscles are assumed to be mechanoreceptors sensitive
to the deformation of the rostral connective tissue.   

Solitary receptors   

     Solitary receptors are widely distributed over the entire
epidermis but are most common in the region of the rostrum,
buccal cirri, and tail (Dogiel 1903; Franz 1923; Bone 1960b;
Stokes and Holland 1995a; Holland and Yu 2002). They
form small clusters in some instances (Sinnesknospen, Franz
1923; Schulte and Riehl 1977), especially along the buccal
cirri. The most common receptor cell types are referred to
by convention as types I and II (Schulte and Riehl 1977;
Bone and Best 1978).  
    Type I cells are primary sensory neurons
with an apical circlet of microvilli, a single cilium, and
a basal neurite. There are several subtypes, but all are probably
mechanosensors (Baatrup 1981; Lacalli and Hou 1999).
Their axons project to the CNS via the dorsal nerves; once
there, they travel along the cord in two fiber tracts, dorsal
and subdorsal in the terminology of Holland and Yu (2002),
which may correspond to the somatosensory and viscerosensory
tracts of Bone (1960a; see Fig. 4). The central axons
of type I cells reach considerable lengths, so an axon entering
the CNS via the first nerve can typically project caudally
to mid-spinal levels, at least in larvae (Holland and Yu
2002).    
    Little is known about the neurotransmitters released
by peripheral neurons, but there is evidence that at least
some type I cells are GABAergic (Anadón et al. 1998).   
    Type II receptors (Fig. 4) are secondary sensory cells with
synaptic terminals borne on short basal processes, usually
three per cell (Stokes and Holland 1995a; Lacalli and Hou
1999). Apically, they have a modified nonmotile cilium surrounded
by a collar of branched microvilli. This extensive
elaboration of the apical surface suggests a chemoreceptive
function, but essentially nothing is known for certain about
chemoreception in amphioxus, either in terms of structures
or physiology (Lacalli 2004). The synaptic targets of type II
cells are not known, but most likely they are fibers belonging
to intramedullary Retzius bipolar cells (Holland and Yu
2002).
     Additional sensory cell types reported from larvae may
well be present in adults as well, but perhaps have simply
not yet been observed. Most notable are the (multi)ciliary
spines along the oral margin in larvae, each consisting of a
bundle of stiff cilia, one from each cell that contributes to
the spine (Lacalli et al. 1999). These cells are secondary
sense cells that synapse with local interneurons resident in
the oral plexus, whose axons then travel to the CNS. It
would be interesting to know whether this arrangement persists
to the adult. A second way of stiffening cilia is by altering
their internal support, and this is seen in ciliary spine
cells (Lacalli and Hou 1999). Ciliary spine cells are solitary
sensory cells present in small numbers on the larval rostrum,
and possibly elsewhere, in which the ciliary axoneme is replaced
with a lamellar matrix. Again, these cells are assumed
to be mechanoreceptors.
     As described above, there also are solitary sensory cells in
most parts of the atrial nervous system. Bone (1961) described
various types of multi- and unipolar cells that send
neurites into the CNS (see Fig. 4). It is not clear, however,
whether such cells are actually primary sensory cells or
whether they participate in complicated synaptic chains of
receptors, interneurons, and projection neurons, similar to
those described for the larval oral plexus.   

Hatschek’s pit   

     Hatschek’s pit (corresponding to the preoral pit of the
larva) is the central element of the wheel organ, a system of
ciliated ridges with an accessory feeding function, located in
the roof of the buccal cavity (Figs. 1, 3; see Ruppert 1997
for a detailed description). As is the case with many other
structures, Hatschek’s pit is asymmetrical: the bottom of the
pit comes to lie to the right of the notochord and points towards
the base of the CNS (Fig. 7C). In some species of
lancelets (Gorbman 1999; Gorbman et al. 1999) it may even
be in contact with the CNS, a situation strongly reminiscent
of the hypothalamus–pituitary relationship in vertebrates.
Consequently, Hatschek’s pit has been regarded as a good
candidate for homology with Rathke’s pouch or the adenohyophysis
of vertebrates (e.g., Tjoa and Welsch 1974;
Nozaki and Gorbman 1992). As discussed elsewhere in this
issue (Sherwood et al. 2005), this hypothesis does not receive
much support from an analysis of the secretions by
Hatschek’s pit (the evidence for the presence of typical
adenohypohyseal hormones is somewhat inconclusive); on
the other hand, pituitary-specific transcription factors are expressed
in Hatschek’s pit during development (Candiani and
Pestarino 1998).
     There are further complications, however. First, the zone
of contact between Hatschek’s pit and the CNS in amphioxus
is displaced about two myomeres caudal to where it
should be if it were an exact homolog of the vertebrate pituitary,
since — according to the gene expression pattens in
the larva — the amphioxus homolog of the forebrain would
be adjacent to myomere 1 (Shimeld and Holland 2005). The
junction of myomeres 3 and 4 lies, instead, in a region of the
CNS that expresses AmphiHox genes during early larval development.
Thus, this region is a more likely candidate for
homology with the vertebrate hindbrain.
     Second, the development of Hatschek’s pit differs from
that of the adenohypophysis of vertebrates, which is classically
supposed to be of ectodermal (placodal) origin.
Hatschek’s pit, on the other hand, arises from the left anterior
diverticulum of the endodermal embryonic foregut (see
Conklin 1932; Stach 2000). This diverticulum then opens to
the exterior by fusing with a preoral ectodermal invagination,
thus forming the larval preoral pit. This pit is initially
located at the level of somite 1, but there is no zone of contact
with the nerve cord at this stage. The preoral pit then
shifts caudally at metamorphosis, along with the whole assemblage
of oral and preoral structures, and finally develops into
Hatschek’s pit at the boundary between myomeres 3 and 4.   
     Thus, despite claims that the adenohypophysis also develops
from the endodermal foregut in certain craniates (i.e., in
myxinoids, Gorbman and Tamarin 1985), both the development
of Hatschek’s pit and its position in adult amphioxus
are sufficiently different from the development and position
of the adenohypophysis in typical vertebrates to warrant some
caution in interpreting the relationship between the two
structures. Nevertheless, it is noteworthy that the preoral pit
differentiates very early in amphioxus, so that it appears to
be functional by the time the larvae begin to feed. That function
is unknown, but feeding is the main larval activity besides
swimming at this early stage, which implies for the
preoral pit a role in either feeding or some related aspect of
metabolic processing. Jacobs and Gates (2003) have suggested
that the ancestral adenohypophysis may have been an
external sense organ that acted on the internal physiology of
the animal via some form of non-neural signaling. If the
preoral pit is indeed an adenohypophyseal homolog that acts
in this way, its involvement in feeding and metabolism in
amphioxus would provide a rationale for the central role of
its vertebrate counterpart in the control of metabolism and
growth.   


The adult CNS   

General histological appearance   

     Most parts of the nerve cord are roughly triangular in
transverse section, with curved sides and a concave base that
rests on the notochord (Figs. 4, 5). Only the anterior vesicle
and the caudal ampulla are more or less circular in cross section.
The shape of the central canal (= ventricle, ventricular
system) varies from region to region. However, with the exception
of the anterior vesicle and the caudal ampulla, it
generally has the form of a vertical slit expanded slightly at
the top and bottom. The ventral expansion, which runs caudally
from the level of the infundibular cells, houses
Reissner’s fiber. The dorsal expansion is variable, being pronounced
in some regions and absent in others, and may be
filled with fluid or with cellular processes. In the slit-like
part of the central canal (= intermediate zone in some accounts),
the opposing walls of the nerve cord are closely
apposed and the canal itself is almost obliterated. Translumenal
processes cross from both sides in this zone, and the
remaining open space contains cilia that arise from both
ependymal cells and neurons.   
    The cell bodies of most of the neurons are located close to
the central canal (Figs. 4–7), and most have an apical process
that connects the cell to the ventricular surface. In terms
of vertebrate neuroanatomy, the neurons thus form a dense
periventricular layer and most, if not all, of the neurons in
this layer are of the cerebrospinal fluid contacting type.    
    Populations of cells variously termed translumenal neurons
(Lacalli and Kelly 2003a) or commissural cells (e.g., Franz
1923; Bone 1960a) form a special subcategory of cerebrospinal
fluid contacting cells. Not only do they contact the
ventricular surface, but parts of their somata — apical processes
in some instances or the whole cell body in others —
lie transversely across the ventricle. Such cells occur in large
numbers in most regions of the CNS (Figs. 4, 5, 6D) and include
the largest cells in the cord, i.e., the cells of Rohde
(see below and Fig. 5), along with other large neurons.
Migrated neurons, i.e., those whose cell bodies are entirely
detached from the ventricular surface, are scarce and restricted
to the IR of the anterior nerve cord (see below).   


Fig. 4   







Abbreviations and their corresponding terms:       

   
DC = dorsal compartment   
   
DR dorsal root cell   
   
EC = Edinger cell   
   
enc = encapsulated nerve endings   
   
MC = mid-commissural cell   
   
Mg = Müller’s glia (Schwann cell analogues)   
   
mRoax = median descending Rohde axon   
   
ncmf = notochordal motor fibers/tract   
   
NS = nervous system   
   
RB = Retzius bipolar cell   
   
rg = radial glia   
   
Sg = Schneider’s glia   
   
SM1 = somatomotor cell type 1   
   
smf = somatomotor fibers/tract   
   
ssf somatosensory fibers/tract   
   
VC = ventral compartment   
   
VM1, VM2 = visceromotor cell type 1, 2   
   
vsf = viscerosensory fibers/tract   

Fig. 4.
    A highly schematic transverse section of amphioxus through the central nervous system and the top portion of the notochord, showing various sensory and motor cell types and the composition of the dorsal nerves. The grey spot in the ventral expansion of the central canal is Reissner’s fiber.

Fig. 5   






Abbreviations and their corresponding terms:       

   
AC = Anadón’s GABAergic cells   
   
antRo = anterior Rohde cell   
   
antRoax = anterior descending Rohde axons   
   
cJohn = Johnston cell   
    DCC = dorsal commissural cell   
    EC = Edinger cell   
    MC = mid-commissural cell   
    mRoax = median descending Rohde axon   
   
intf = internuncial fiber bundle   
   
oHe = organ of Hesse   
   
postRoax = posterior ascending Rohde axons   
   
SD = small dorsal cell   
   
VL = ventral longitudinal cell   

Fig. 5.
    A highly schematic transverse section of amphioxus through the anterior spinal cord, showing various types of interneurons and their characteristic arrangement. The grey spot in the ventral expansion of the central canal is Reissner’s fiber.



    The amphioxus CNS is not vascularized, and the axonal
and dendritic processes of its neurons are not myelinated or
otherwise enveloped by glial cells (Meves 1973). Nevertheless,
various types of glial cells do occur in the nerve cord
(Bone 1960a; see Fig. 4). Müller’s glia (Müller 1900) consists
of groups of very small cells with intensely staining nuclei
that are clustered in the vicinity of the dorsal nerves and
in their peripheral branches. Some workers (e.g., Johnston
1905) have interpreted these as dorsal root ganglion cells;
however, as noted by Franz (1927), they also occur between
the muscle tails, i.e., the false ventral roots (Lele et al.
1958), and they lack neurites. Bone (1960a) consequently
concluded that they were Schwann cell analogues. In some
parts of the peripheral nerves, as well as the muscle tails,
these cells form incomplete partial sheaths around axons and
muscle tails (Flood 1966, 1974; also see Fig. 8B in Schulte
and Riehl 1977). This may be a local phenomenon, however,
rather than being widespread throughout the PNS.
     A second type of glial cell with either short processes or
no processes is Schneider’s glia (Schneider 1879; Bone 1960a),
which lines the walls of the dorsal expansion of the central
canal. A third type of glia, the radial glia (= ependymal glia,
Bone 1960a), lines the ventricular walls. These glia send
long, fiber-filled processes through the white matter to the
connective tissue sheath surrounding the nerve cord, and
there they expand to form terminal endfeet. This forms essentially
an outer limiting membrane around the neural components
of the nerve cord and provides the mechanical
support necessary to maintain the nerve cord’s shape. Several
other glial cell types have been reported from the larval
nerve cord (Lacalli and Kelly 2002), including one, the axial
glia, that appears to have a transient function in axonal guidance.
These cells arise adjacent to the primary motoneurons,
which suggests they could be related to vertebrate
oligodendrocytes, which also develop from a restricted zone
adjacent to the region where the primary motoneurons arise.
We suggest, therefore, that the myelination function may
have evolved secondarily in a cell line that functioned first
in axonal guidance.   


Selected cell types   

Cells of Rohde
  

     The cells of Rohde (kollossale Ganglienzellen = Rohde
cells, not to be confused with the nucleus of Rohde, a ventral
cluster of neurons in the anterior IR) are the largest neurons
in the nerve cord. As described most thoroughly by
Rohde (1887), they are translumenal cells with a large soma
across the central canal and dendrites that ramify in the
white matter on both sides of the nerve cord. The anteriormost
Rohde cell marks the rostral boundary of the spinal
cord. Its cell body is always located in the vicinity of the left
sixth nerve (i.e., at the boundary of myomeres 4 and 5; see
Figs. 1, 3, 7). Its giant axon (kollossale Faser, Rohde 1887),
which arises from the left side of the cell, turns ventrally and
is the most conspicuous structure in the ventral midline of
the spinal cord (Fig. 6). It projects caudally to the level of
the last myomere (Franz 1923). Much of the input to the
dendrites of the first Rohde cell appears to be via gap junctions
(Ruiz and Anadón 1989), as only a few synapses to
them have yet been found. The giant axon displays numerous
en-passant synapses (Ruiz and Anadón 1989); processes
of somatomotor cells (see below) might be postsynaptic to
the giant axon (Castro et al. 2004).
     The remaining Rohde cells (Fig. 5) are much smaller than
the first and form separate anterior and posterior groups
(Fig. 1). The anterior group, excluding the first cell, consists
of roughly 15 cells in the region of myomeres 5–11. Their
cell bodies are in line with the transverse planes defined by
the staggered left and right dorsal nerves. Cell size decreases
progressively caudally, so that the most caudal of the anterior
Rohde cells that can be identified with any certainty lies
at the level of the right 13th nerve (Franz 1923). The cells
give rise to thick axons alternating to the left and right sides
in the anterior series of Rohde cells. The axons turn ventrally,
cross the midline, and then travel caudally in a lateral
fiber bundle (Fig. 5).
     Rohde cells reappear at the level of myomere 38, adjacent
to nerve 39. The posterior Rohde cells are, on average, much
smaller than those in the anterior group, and they are not
aligned with the dorsal nerves. There are between 14 (Rohde
1887) and 18 (Franz 1923) posterior Rohde cells, the most
caudal one being at the level of myomere 60. The posterior
cells have mainly ascending axons that travel in a ventrolateral
tract (Franz 1923, see Fig. 5).
     As in the case of the first and largest Rohde cell, almost
nothing is known about the synaptic relationships and the
neurochemistry of the rest of the Rohde cell series. It has,
however, recently been shown that all the Rohde cells, including
their axons, are immunopositive for progesterone
(Takeda et al. 2003). The significance of this finding is entirely
unclear. The most one can currently say is that the
Rohde cells, though conspicuous, remain enigmatic.   

Organs of Hesse   

     The organs of Hesse (= Hesse organs, dorsal ocelli) are
composite photoreceptors (Hesse 1898) located in the ventral
part of the periventricular grey (Figs. 5, 7). Each organ
consists of a single rhabdomeric photoreceptor cell whose
microvilli are enveloped by a cup-shaped pigment cell
(Eakin and Westfall 1962; Ruiz and Anadón 1991c; see
Fig. 5). The receptor cells are primary sensory cells, i.e.,
each has an axon (Franz 1923), and these project to the
ventrolateral part of the spinal cord (Guthrie 1975). The first
organ to form in development is unusual in consisting of
three cells — two photoreceptors and one pigment cell —
and the former in this case are known to innervate the dorsal
compartment (DC) motoneurons (Lacalli 2002a). It is not
known whether this is the case also for the rest of the series,
but it does suggest that Hesse organs have something to do
with controlling activities that depend on slow undulations
of the body, as opposed to the fast ones used for escape
swimming. Hesse organs may not express Pax6 during development
— certainly the first does not (Glardon et al.
1998) — but they do express S-antigen (arrestin), a protein
typical of the phototransduction cascade in many animals
(Mirshahi et al. 1985), and serotonin is present in the receptor
cells (Candiani et al. 2001).
     The first Hesse organ to develop is located adjacent to
myomere 5, but the anteriormost in the adult lies at the
boundary of myomeres 3 and 4; it is located in the posterior
part of the IR. Hesse organs become more numerous as one
moves caudally into the spinal cord, and they have a tendency
to cluster opposite the centers of adjacent myomeres
(Franz 1923) so as to form a segmental pattern. Moving further
along the spinal cord, they first decrease in number towards
the center of the body and then increase again more
caudally. The orientation of the pigment cups is not random.
Instead, groups of Hesse organs in certain regions of the
spinal cord point in particular directions. Franz (1923) undertook
the tedious work of documenting this; for details, refer
to his paper.

Motoneurons, with remarks on muscle innervation patterns   

     The innervation of the myomeric muscles in amphioxus is
highly unusual. The myomeres consist of cross-striated muscles
arranged in a series of chevron-shaped blocks along the
flank. The muscle cells each send long processes (= muscle
tails) towards the ventrolateral margins of the nerve cord.
These processes were formely interpreted, mistakenly, as
ventral nerve roots. Their true nature was first recognized by
Flood (1966): the muscle processes are the sites where the
synapses from motoneurons are received; the axons and synaptic
terminals remain entirely confined within the nerve
cord; and transmitter release occurs across the basal lamina
(Figs. 3, 4). The synaptic zones are serially repeated, one for
each myomere, and are staggered left to right in a manner
similar to that of the myomeres. Amphioxus thus lacks any
counterpart to the ventral nerve roots of vertebrates (Schneider
1879; Flood 1966, 1968).
     The synaptic zones in each segment consist of two distinct
domains, the ventral and dorsal synaptic compartments
(Figs. 3, 4). Both utilize acetylcholine as a transmitter (Flood
1974). The ventral synaptic compartments are where the
deep, anaerobic, fast muscle cells receive their innervation.
The presynaptic motoneurons involved belong to a class of
cells that Bone (1960a) called somatomotor (SM) cells; they
may therefore also be called ventral compartment motoneurons.
They are found in the ventral parts of the grey matter
and have a tendency to cluster opposite the synaptic
contact zones, and each has a broad apical process connecting
it to the ventricular cavity. Some have internal vacuoles,
and this character, together with size and positional differences,
has been used to define several subtypes (Bone 1960a;
only one such type, the SM1 cell, is shown in Fig. 4). The
axons of the SM cells project laterally into the bundle of
somatomotor fibers adjacent to the synaptic zone of the ventral
compartment.
     The dorsal compartment is where the superficial, aerobic,
slow muscle cells of the myomeres receive their innervation.
The DC motoneurons are known from larvae (Lacalli and
Kelly 1999; Lacalli 2002a) but have not yet been identified
with certainty in adults. From the larval data, however, it
seems that the whole of the DC innervation along the nerve
cord may derive from motoneurons located in the anterior
cord at the level of somites 2–6 (see below). This is approximately
equivalent to the zone fated to become the IR of the
anterior cord, which extends from myotomes 2 to 4. Tracing
experiments by Fritzsch (1996) and Ekhart et al. (2003) have
revealed ventrally located cells with descending projections
in the adult in this region, but it is not clear what types of
cells these are, or even whether they are motoneurons or
interneurons. It is also possible that local commissural cells
contribute to the presynapses of the dorsal compartments
(Fig. 4). The Edinger (EC) and mid-commissural (MC) cells
of Bone (1960a) send axons into the bundle of somatomotor
fibers adjacent to the synaptic zone of the dorsal compartments.
Bone (1960a) regarded the entire DC system as a somatic
sensory system; hence, he classified the EC and MC
cells as afferent cells. It is further possible that the adult dorsal
compartment is innervated by a subset of the somatic
motoneuron series, i.e., one or the other of the SM cell types
(Castro et al. 2004).
     Unlike the notochord of any other chordate, the notochord
of lancelets is itself a muscle, i.e., it consists mainly of specialized
striated muscle cells referred to as notochordal
lamellae (Welsch 1968c; Flood 1970). The notochordal muscle
cells contact the nerve cord by means of processes that
emerge from the cells dorsally, in small bundles (notochordal
horns, Flood 1970), and pierce the connective tissue
sheath separating the notochord and CNS (Fig. 4). There are
many thousands of these horns, roughly one every 50 ?m
(Flood 1970), arranged in two rows to the left and right of
the midline. Their contacts with the base of the CNS are
specialized as postsynaptic swellings that are apposed to
presynaptic terminals inside the CNS. Thus, there are serially
(but not segmentally) repeated neurochordal synaptic
contacts at the base of the nerve cord. These are thought to
be cholinergic (Flood 1970), but their source within the
nerve cord has not been identified. Lacalli (2004) has made
a tentative suggestion that they originate from sensory cells
located in the rostrum, but this remains to be proven.
     The remaining muscles of the body, besides the myomeric
muscles and the notochord, are innervated by a visceromotor
(VM) system of cells and fibers. The neurons are distinctive
in appearance and location (Bone 1960a), as they lie immediately
beneath the ventral expansion of the central canal
and dorsal to the medial axon of the giant Rohde cell
(Fig. 4). This places them ventral to the SM neurons (cf.
Figs. 4, 7), a situation that differs from that in vertebrates,
where the VM neurons are dorsal to SM neurons (e.g.,
Nieuwenhuys 1998). There are both large (VM1, one per
segment) and small (VM2, many per segment) VM cells in
adult amphioxus. They are multipolar, and axons from both
types leave the CNS in the dorsal nerves; similar cells are
found also in larvae (Lacalli and Kelly 2002). Candiani et al.
(2001) reported that the cell bodies of VM2 cells contain
serotonin, but VM axonal terminals are cholinergic (Flood
1974) and are found in various muscles (e.g., the pterygeal
muscle) associated with the atrium. Though referred to as
visceral muscles, most of them are cross-striated (Franz
1927; Ruppert 1997) rather than smooth. There is therefore
some question whether the VM system of lancelets is really
comparable to the visceral muscles of vertebrates. A better
comparison may be, instead, with the branchiomotor nerves
and muscles of vertebrates, as suggested by Fritzsch and
Northcutt (1993).   

Intramedullary sensory neurons
   

     As mentioned above, lancelets do not have dorsal root
ganglia. Instead, they have intramedullary sensory cells
comparable to the Rohon-Beard cells found in the larval
nerve cord of anamniotic vertebrates (Fritzsch and Northcutt
1993). Bone (1960a) distinguished two major classes, the
dorsal bipolar or Retzius bipolar (RB) cell and the dorsal
root (DR) cell. There are two subclasses of the former and
three of the latter. RB cells are situated in the most dorsal
part of the periventricular grey and have both ascending and
descending fibers. The fibers are a major component of the
longitudinal somatosensory tracts, and the peripheral RB
processes originate as branches from them and enter the dorsal
nerve. There are fewer DR cells and most of them have
translumenal processes (Fig. 4).
     As to function, the usual assumption is that the RB and
DR axons are the most probable sources of receptive sensory
fibers in the various peripheral plexuses and the skin.  
Presumably some branch and terminate as free endings, while
others are postsynaptic to secondary receptor cells, wherever
those occur. This is a logical supposition, and there is
circumstantial evidence on the innervation of the larval rostrum
that tends to support it (Lacalli 2002b, 2004). It has in practice,
however, been impossible to prove. The necessary tracing
studies have not yet been performed, for example, so we
have no detailed information on the peripheral connections
of either RB or DR cells, nor is it certain that their processes
actually travel any distance outside the cord. There is recent
immunocytochemical evidence that some DR cells contain
?-aminobutyric acid (GABA) (Anadón et al. 1998), but otherwise
nothing is known about their transmitters.   

Interneurons
   

     The middle (i.e., intermediate) zone of the periventricular
grey (Fig. 5) is occupied by a variety of cells that Bone
(1960a) regarded as interneurons. These include the Rohde
cells and, more dorsally, translumenal dorsal commissural
cells and small dorsal cells. The latter are related to the sensory
system in that they send small processes into somatosensory
tracts. Bone (1960a) thought that both types might
contribute axons to the dorsal nerves in some cases. The EC
and MC cells are found more ventrally. As mentioned above,
they may have something to do with the DC motor system.
The MC cells are translumenal neurons of an unusual type:
the cell body itself lies across the central canal, and short expansions
project directly into the canal itself (Figs. 4, 5).   
     GABA, neuropeptide Y, and several other neuropeptides
have been detected in various cells loosely classified as
interneurons (Uemura et al. 1994; Anadón et al. 1998; Castro
et al. 2003), but this clearly represents a very diverse assemblage
of disparate cell types about which very little is
known.
     Three other interneuronal cell types deserve mention.
     First, the cells of Johnston (Bone 1960a) occur segmentally
between the dorsal nerves, one on each side, and have a long
process that projects to the dorsal expansion of the central
canal and contacts the fluid it contains (Fig. 5). Their function
is entirely unknown.    
    Second, a novel class of interneurons, Anadón’s cells,
has been identified in the vicinity of the ventral expansion
of the central canal (Anadón et al. 1998; see Fig. 5).  
These are very small GABAergic cells interspersed between
the cell bodies of SM and VM neurons (cf. Figs. 4, 5).  
    Anadón et al. (1998) have suggested
that they might be comparable to the inhibitory Renshaw
cells of vertebrates. Notably, these small cerebrospinal fluid
contacting cells do not seem to be identical to a third group
of interneurons, the ventral longitudinal cells of Bone (Bone
1960a; Fig. 5). The cell bodies of this third group are also
found between those of the motor neurons (cf. Figs. 4, 5) but
are much larger than those of Anadón’s cells and are detached
from the ventricle. They have dorsal, ascending, and
descending processes, and Bone (1960a) thought that they
also coordinate the activity of the motor neurons, in particular
that of the VM cells.   

The anterior nerve cord in detail   

     The region anterior to the first cell of Rohde has attracted
less attention in cytological studies than the spinal cord,
though the two differ, as noted by Bone (1960a), in terms of
both their cell types and their general organization. What is
clearly lacking is a descriptive study comparable to Bone’s
comprehensive analysis of the spinal cord, cited so extensively
above. The closest is that of Ekhart et al. (2003), but
this is focused more on cell grouping, cytoarchitecture, and
general morphology, while providing much less information
on individual cell types.   

Fig. 6   





Abbreviations and their corresponding terms   
   
alm = anterolateral migrated cell group   
   
avm = anteroventral migrated cell group   
   
cJo = Joseph cells   
   
io infundibular organ   
   
Köp = Kölliker’s pit   
   
lac = large translumenal (= commissural) cell   
   
lc = lamellar cells   
   
m1, m2, m3… = myomere 1, 2, 3…   
   
n1, n2, n3… = dorsal nerve 1, 2, 3…   
   
nRo = nucleus of Rohde   
   
pdm = posterodorsal migrated cell group   
   
pig = pigment cells of frontal eye   
   
PMC = primary motor center   
   
Rf = Reissner’s fiber   
   
SM somatomotor cells   
   
VM = visceromotor cell   
   
vsc = ventral spindle-shaped cell group   

Fig. 6.
    A schematic lateral view of the rostral end of the nerve cord
of a lancelet, including the anterior vesicle plus the anterior and
intermediate parts of the intercalated region (A), with cresyl
violet-stained transverse sections (B–E) to show arrangement
of cells in the transverse plane at the levels indicated in A.
    The diagram shows, projected onto the midsagittal plane,
the various cells and (or) groups of cells that have been
identified either by cytoarchitectural criteria, immunocytochemistry,
or tracing experiments.
    Thin arrows indicate the general direction of axonal projections
for selected examples; their exact targets are not shown because
they are largely unknown.
   


Fig. 6a 
    This is the first myotome enlarged x 2 so that the " various cells and
(or) groups of cells that have been identified either by cytoarchitectural
criteria, immunocytochemistry, or tracing experiments
" are more easily seen. 


   



Anterior vesicle   

     The anterior vesicle (Fig. 6) corresponds to the anterior
part of the larval cerebral vesicle (= anterior CV; see below
and Fig. 8). The anterior vesicle has a central ventricular
space (= central canal) but, unlike other regions of the nerve
cord, this is relatively wide and lacks the translumenal cell
processes found elsewhere in the cord. Dorsally, on the lefthand
side, there is a remnant of the anterior neuropore in the
form of a ciliated pit. This is Kölliker’s pit (Kölliker 1843;
Franz 1923), which has been ascribed an olfactory function,
though without direct evidence. In fact, the neuropore remains
an open channel, even in the adult (Vallet et al. 1985).
However, it is so narrow and clogged with cilia that one can
question how effectively it acts as a connection between the
central canal and the outside. The outer pit lacks obvious receptor
cells, nor are there any nerve fibers in evidence to
connect it to the anterior vesicle (Edinger 1906; Tjoa and
Welsch 1974). This tends to reinforce the conclusion that although
Kolliker’s pit may contain some specialized cell
types, it is not a sense organ.
     The caudal end of the anterior vesicle is marked dorsally
by the beginning of the axial columns of Joseph cells, as
well as scattered lamellar cells, and ventrally by the infundibular
organ. The infundibular organ consists of several rows
of columnar cells that secrete Reissner’s fiber. It is though to
correspond to the flexural and subcommissural organs of
vertebrates (Obermüller-Wilén 1976; see Nieuwenhuys 1998
for details).
     The walls of the anterior vesicle are formed by closely
packed, ciliated epithelial cells, among which are the pigment
cells of the frontal eye. While this region has been examined
in larvae in great detail at the EM level (see below),
the only EM study of the adult anterior vesicle is that of
Meves (1973), which is much less complete. Meves observed
rows of densely stained cells (Fig. 6), immediately
ventral and caudal to the pigment cells of the frontal eye,
that may correspond to the receptors and neurons of the larval
frontal eye. More caudally, in the preinfundibular zone,
i.e., between the frontal eye and the infundibular organ, she
found two types of cells, the more densely stained of which
had basal processes. These may correspond to some of the
larval preinfundibular projection neurons described by
Lacalli and Kelly (2000), including the cells of the balance
organ. These authors found evidence for the latter structure
in newly metamorphosed juveniles, so it may persist to the
adult stage.
     Whether by light microscopy (e.g., Edinger 1906; Franz
1923; Ekhart et al. 2003) or EM (Meves 1973), it is difficult
to discern much about the neuronal and glial cells of the anterior
vesicle, since most cells are small and rather densely
stained and have few visible distinguishing features. Franz
(1923, 1927) therefore concluded that the entire anterior vesicle
consisted only of glial cells. However, GABAergic
(Anadón et al. 1998) and serotoninergic neurons (Moret et
al. 2004) have since been identified in this region in adult
specimens, and since the same region in the larva is well
supplied with neurons, it must continue to function in a neural
capacity throughout the animal’s life, though what it does
in the adult is not evident.
     The first nerve of the dorsal nerve series (= nerve 1,
rostral nerve) enters the anterior vesicle ventrally at its rostral
margin. The nerve’s fibers then spread out to form a thin veil
of white matter that covers the lateral and ventrolateral aspects
of the anterior vesicle. Peripherally, this first pair of
nerves traverses the top of the notochord, enclosed within
the connective tissue sheath that surrounds the latter. Near
the tip of the notochord, the nerves exit the sheath and ramify
underneath the skin, so the innervation is very much restricted
to just the tip of the rostrum. This is an important
point, because if there are any as yet undescribed sensory
cell types or subtypes that occur only at the tip of the rostrum,
it is very likely that they would enter the cord by
means of the first nerve, which is exceptional anyway for
not being, strictly speaking, a dorsal nerve, so its CNS targets
may also differ from those of the rest of the dorsal
nerve series. Among known fibers in the first nerve are axons
from peripheral primary sensory cells, mainly type I
mechanoreceptors, and centrally derived axons from RB
cells and possibly other CNS cells. 
    The second nerve of the dorsal series (= nerve 2,
anterodorsal nerve) enters the CNS dorsally, at the junction
between the anterior vesicle and the IR. Unlike the first
nerve, the second is not associated with the notochord, but
turns laterally and enters the subepidermal connective tissue.
The second nerve ramifies extensively in the more caudal
parts of the rostrum, and several anastomoses are formed
with the branches of the first nerve (Franz 1923, Fig. 3). The
second nerve probably carries fiber types similar to those of
the rostral nerves.
     The peripheral branches of both the first and the second
nerves bear numerous small swellings, the corpuscles of de
Quatrefages described above. The centripetal axons derived
from these enter the distal branches of both nerves 1 and 2,
but it is not known whether they actually enter the cord in
both nerves or only one. Some of the carbocyanine dye (DiI)
tracing data have been interpreted as indicating that the corpuscles
connect to the CNS mainly via nerve 2 (Fritzsch
1996; B. Fritzsch, personal communication), which is intriguing
but requires confirmation. The central targets of axons
from the corpuscles are unknown.   

The intercalated region (IR)

     The region between the caudal end of the anterior vesicle
and the first giant Rohde cell has sometimes been regarded
as a part of the spinal cord (e.g., Franz 1927), but it clearly
has unique features that justify its being regarded as a specialized
part of the CNS (see also Bone 1960a; Fritzsch 1996). 
The most prominent of such features are the dorsal
Joseph cells, present throughout most of this zone, and a
conspicuous ventral group of cells, the nucleus of Rohde, in
the posterior part of the IR. The IR is not uniform along its
length but instead can be subdivided into three parts on
cytoarchitectural grounds. The description of these that follows
is based mainly on Ekhart et al. (2003) and some more
recent unpublished findings (H. Wicht, personal observations).   

Anterior IR   

     The anterior part of the IR is located adjacent to the first
myomere (Fig. 6) and probably corresponds to the posterior
part of the larval cerebral vesicle. As in more posterior parts
of the CNS, the central part of the ventricle is slit-like and
traversed by cellular processes from translumenal neurons.
Ventrally, a small expansion of the ventricular cavity houses
Reissner’s fiber. Dorsally, there is another expansion partly
filled with processes belonging to lamellar cells (Meves
1973; Ekhart et al. 2003), a form of ciliary photoreceptor
(Ruiz and Anadón 1991b), though a compact lamellar body
like that seen in the larva is no longer present. The dorsal aspect
of the anterior part of the IR is capped by the
rhabdomeric Joseph cells (Welsch 1968a; Ruiz and Anadón
1991b, 1991c). These increase in number in a rostrocaudal
direction. As with the Hesse organs, Joseph cells do not express
Pax6 during development but do contain a rhodopsinlike
protein (Watanabe and Yoshida 1986), so they are almost
certainly photoreceptors. The ventral and lateral aspects
of the anterior IR are covered by a relatively thick
layer of white matter. 
    Neurons and glial cells are easily distinguished in this region.
The neuronal cell bodies are mostly located in the
periventricular grey close to the central ventricular slit.
Compared with neurons in more posterior regions of the
CNS, these cells are relatively small. In the ventral part of
the periventricular grey, however, just above the ventral expansion
of the central canal, there are individual large cells
that resemble the somatic motoneurons (SM cells) described
by Bone (1960a) from spinal cord. Based on the larval data
(e.g., Lacalli and Kelly 2003b), several classes of ventral
interneurons with descending axons would also be expected
to reside in this area. 
    Immunocytochemical studies by Holland and Holland
(1993), Uemura et al. (1994), Anadón et al. (1998), Castro et
al. (2003), and Moret et al. (2004), and tracing studies by
Fritzsch (1996) and Ekhart et al. (2003), all show that the
periventricular grey of the anterior IR contains distinctive
cell types and groupings besides those evident in standard
histological preparations. First, in the periventricular grey,
there are numerous translumenal cells whose axons project
to the spinal cord (Ekhart et al. 2003). Then, just ventral to
the Joseph cells and surrounding the dorsal expansion of the
central canal, there are bilateral, longitudinal bands of neurons
immunoreactive for urotensin and FMRFamide (Uemura
et al. 1994), GABA (Anadón et al. 1998), neuropeptide Y
(Castro et al. 2003), and catecholamines (the catecholaminergic
population I of Moret et al. (2004)). Axons from some
of these catecholamine-containing cells travel anteriorly into
the anterior vesicle, but fibers from most of the other cell
types are directed ventrally into the lateral and ventral
neuropile, where they form a dense commissure or plexus
beneath the central canal. 
    Unlike the situation elsewhere in the CNS, there are
groups of neurons in the anterior IR that detach from the
ependymal layer and migrate into the white matter. In standard
sections of the anterior IR they are seen in the ventral
midline, below the ventral expansion of the central canal
(the anteroventral migrated (avm) group, Fig. 6B), and bilaterally
in the dorsal and lateral parts of the white matter (the
anterolateral migrated (alm) group, Fig. 6B; see Ekhart et al.
2003 for details). There is no information on the nature of
the avm cells, but recently published data, as well as some
personal observations of immunocytochemical preparations
of B. lanceolatum, have yielded interesting details on the
alm cell group. Firstly, the more posterior alm cells (slightly
rostral to the junction of myomeres 1 and 2; black squares in
Fig. 6) seem to correspond to the anterolateral serotoninergic
cells of Holland and Holland (1993) that were also observed
by Moret et al. (2004). Slightly more anterior (black circles
in Fig. 6) is another group of immunocytochemically identifiable
cells within the alm group. This is the catecholaminergic
population II of Moret et al. (2004). There is some
uncertainty about the exact positions of these two cell
groups, however. Moret et al. (2004) place them adjacent to
the rostral half of the second myomere. In an independent
immunocytochemical study, H. Wicht (unpublished data) localized
them more anteriorly, adjacent to myomere 1 and
thus within the confines of the alm group (see Fig. 6). Wicht’s
study did confirm, however, that both the catecholaminergic
population II and the anterolateral sertoninergic neurons
have long descending projections to the spinal cord. In retrograde
tracing experiments, Fritzsch (1996) found pairs of labelled
cells in late larvae that may correspond to the
anterolateral serotoninergic cells, even though he did not
specify their exact position, but Ekhart et al. (2003), in a
similar study in adults, did not find such cells. Assuming the
latter result is a false negative, the cells and projections appear
to be real; it is only their exact axial position that is a
matter of some uncertainty.   

Intermediate IR   

     The intermediate IR (Fig. 6) is located adjacent to
myomere 2, i.e., between dorsal nerves 3 and 4. It is characterized
by having several layers of Joseph cells. The dorsal
expansion of the central canal is lacking in this region, and
the canal has instead the shape of an inverted keyhole. The
lateral and ventral migrated cell groups that occur in the anterior
part of the IR are also absent. Instead, there is another
such group (the posterodorsal migrated (pdm) group,
Fig. 6C) located just ventral to the Joseph cells. Cells in this
group probably correspond to the B cells described by Bone
(1959, 1960a), which represent the anteriormost cluster of
somatosensory RB cells found elsewhere along the nerve
cord. A group of small, ventral spindle-shaped cells (vsc in
Fig. 6) is found directly beneath the ventral expansion of the
central canal. A number of large cells, presumably
motoneurons of one kind or another, are found in the ventral
part of the periventricular grey, and some of these send descending
projections to the spinal cord (Ekhart et al. 2003).
     There also is a single large VM neuron in the ventral
midline just posterior to the ventral spindle-shaped cell
group (H. Wicht, unpublished data; see Fig. 6E). Bone
(1960a) claimed that VM neurons were present only in the
spinal cord, but there must be at least enough in the IR to
supply dorsal nerves 3–7, which innervate the labial muscles
and other buccal structures. In addition, there are some very
large translumenal neurons in this region (lac, Fig. 6D) at
approximately mid-level in the periventricular grey. They
probably correspond to the giant translumenal cells described
by Lacalli and Kelly (2003b) from late-stage larvae
and the largest of the FMRFamide-positive cells of Uemura
et al. (1994). Axons from at least some of the large
translumenal neurons project to the spinal cord (Ekhart et al.
2003).
     The periventricular grey of the intermediate IR also contains
neurons positive for GABA (Anadón et al. 1998) and
peptides (Uemura et al. 1994), but the pattern differs from
that further forward. Specifically, neurons of corresponding
type are shifted more ventrally compared with the anterior
part of the IR.
     The periventricular grey in this region also contains
translumenal neurons with descending projections (Ekhart et
al. 2003). These decrease in frequency towards the middle
and caudal parts of the intermediate IR. From studies on larvae
(see below), a major locomotory control center, the primary
motor center (PMC), is expected to reside somewhere
in this region, beginning roughly at the anterior tip of
myomere 2 (Fig. 6, shaded region). There is no obvious
adult counterpart to the PMC at this location, but even if the
larval neurons persist to the adult, there are relatively few of
them and they could easily be overlooked.   

Posterior IR   

     The posterior IR (Fig. 7) is located adjacent to myomeres
3 and 4, between dorsal nerves 4 and 6. Its most conspicuous
feature is the nucleus of Rohde (Ekhart et al. 2003, first
described by Rohde in 1887). This is an agglomeration of
relatively large cells with intensely staining cytoplasm
(Figs. 7B, 7D) that surround the ventral expansion of the
central canal. The rostrocaudal extent of Rohde’s nucleus
coincides with that of the columnar epithelium of the wheel
organ and Hatschek’s pit, located in the roof of the buccal
cavity. The tip of Hatschek’s pit extends to the side of the
notochord and projects towards the base of the CNS at the
junction of myomeres 3 and 4 (Fig. 7C). Above, we have
discussed the evidence for homology between Hatschek’s pit
and the vertebrate adenohypophysis. In light of this hypothesis,
it is of course tempting to speculate that Rohde’s nucleus
is the equivalent of the neurosecretory hypothalamic
cell groups of vertebrates; however, as in the case of the
adenohypophyis, the evidence is so far not very convincing.
The intensely staining cytoplasm and the large amounts of
Nissl substance within it actually point towards an intense
secretory activity; however, none of the neuropeptides typical
of vertebrate hypothalamic endocrine cells have been observed
within the confines of the nucleus so far. Similarly,
nothing is known about the axonal projections of these cells,
so it is not clear whether they are the source of the fibers
that, from circumstantial evidence, appear to innervate
Hatschek’s pit (Tjoa and Welsch 1974).   

Fig. 7.   



Abbreviations and their corresponding terms   

   
chd = notochord   
   
cJo = Joseph cells   
   
cRo = Rohde cell   
   
Hp = Hatschek’s pit   
   
lac = large translumenal (= commissural) cell   
    m1, m2, m3… = myomere 1, 2, 3…   
   
n1, n2, n3… = dorsal nerve 1, 2, 3…   
   
nRo = nucleus of Rohde   
   
oHe organ of Hesse   
   
pdm = posterodorsal migrated cell group   
    SM = somatomotor cells   
   
vsc = ventral spindle-shaped cell group   
See Fig. 6a for the "various cells and (or) groups of cells
that have been identified either by cytoarchitectural
criteria, immunocytochemistry, or tracing experiments
". 

Fig. 7.
    A schematic lateral view of the region just behind that shown
in Fig. 6, to show the posterior part of the intercalated region
(A) and cresyl violet-stained transverse sections (B–D) at the
levels indicated in A.  
    The black structures in the ventral part of the nerve cord are
the pigment cells belonging to the Hesse organs.   



     The posterior part of the IR is also where the rostralmost
Hesse organs occur, while dorsally the Joseph cells vanish,
roughly at the boundary between myomeres 3 and 4. The
dorsal part of the central canal displays some isolated,
bubble-shaped expansions, but these are filled with fluid
rather than cell processes, and their functional significance is
not clear.
     The posterior IR contains only a few migrated cells. The
column of pdm cells that originates farther forward (see
Fig. 7) continues through this region into the spinal cord.
However, roughly coincident with the last of the Joseph
cells, the pdm columns move medially so as to effectively
merge with the periventricular grey.
     The periventricular grey of the posterior IR, flanking the
slit-like part of the central canal, contains numerous translumenal
neurons, many of whose axons descend to the spinal
cord, including several especially large examples (Ekhart
et al. 2003). In the ventral part of the periventricular grey,
immediately dorsal to the nucleus of Rohde (Fig. 7), there
are a few large neurons of the motoneuron series, some of
which have descending projections (Ekhart et al. 2003).
Serotonin-containing neurons are absent in this region
(Moret et al. 2004), but a relatively large number of
GABAergic and peptidergic cells (Uemura et al. 1994;
Anadón et al. 1998; Castro et al. 2003) do occur. In addition,
there are four relatively large catecholaminergic cells (population
III of Moret et al. 2004) with translumenal processes
in the vicinity of the roots of the fifth dorsal nerves.   


The larval nervous system   

     Early development in amphioxus resembles that of a typical
marine invertebrate: the egg is small, ca. 120 mu in diameter,
and hatches after gastrulation as a free-swimming,
but nonfeeding, ciliated larva. This stage undergoes
neurulation and elongates to produce a more typical
chordate-type larva, roughly 1.3 mm in length, with a
notochord, somites, dorsal nerve cord, and one pharyngeal
slit, the first of the series. The larvae then feed and grow in
the plankton until, after about 30 days under optimal conditions,
they metamorphose to juveniles ca. 5 mm long. Excellent
surveys of the morphological changes that occur during
development have been published by Hirakow and Kajita
(1990, 1991, 1994) and Stokes and Holland (1995a). Most
of the recent larval research has been done on different species
than those used for classical anatomical studies of the
adult, notably Branchiostoma floridae (Hubbs, 1922) in
North America and Branchiostoma belcheri (Gray, 1847) in
Asia. The neuroanatomical differences between species appear
minimal, though the differences observed in the extent
of the Joseph cell columns (H. Uemura, personal communication)
indicate that some caution is required when comparing
data across species.
     Amphioxus larvae are well known for their asymmetry,
the mouth being initially on the left side and the pharyngeal
slits on the right. This arrangement necessitates a major
repositioning of structures at metamorphosis, especially of the
mouth, which shifts caudally over a distance of several somites,
and the pharyngeal slits. This is a dramatic process,
achieved largely through differential growth, and has attracted
a good deal of attention in the past. What is perhaps
less appreciated is the magnitude of the size increase that
occurs during the larval phase, for both the body as a whole
and the internal structures as well. The nerve cord, for example,
increases from an initial diameter of 15 mu to ca.
100 mu at metamorphosis. By then, there is substantially
more sensory input, so the dorsal roots are larger and the
cord has a greater diversity of neuronal cell types and much
more neuropile, implying greater integrative capacity.
    Correlated with this, the animal’s behavior becomes much
more complex. A great deal of neural development thus takes
place during the larval phase, and the nervous system consequently
looks quite different, even at a fairly gross anatomical
level, depending on the stage examined. Reconstructing
the developmental events responsible for generating a characteristic
adult neuroanatomy thus requires that a range of
stages be examined, spanning a period of weeks, but this has
seldom been achieved in practice. Our knowledge of neural
development is therefore somewhat fragmentary, and only
provisional conclusions can be drawn in many instances.
     The first detailed description of larval neuroanatomy was
that of Bone (1959), who made the prescient remark that larvae
are likely to be more revealing about phylogenetic issues
than the adult. Bone’s study, and much subsequent larval
work, concentrated on rather late stages, since these are
better subjects for most staining techniques and are easier to
handle than young larvae. The young stages are nevertheless
the most important ones in terms of understanding early patterns
of neural differentiation and tract formation and for
comparison with the rapidly increasing body of gene expression
data (Holland and Holland 1999; Shimeld and Holland
2005). Recent relevant work on young larvae includes studies
of peripheral innervation by means of whole mount
immunostaining and DiI tracing (Yasui et al. 1998; Kaji et
al. 2001; Holland and Yu 2002), as well as a detailed study
of the internal microanatomy of the anterior nerve cord at
the EM level, using serial sections and three-dimensional
reconstruction (e.g., Lacalli et al. 1994; Lacalli 1996, 2002a,
2002b; Lacalli and Kelly 1999, 2000, 2003a, 2003b). The
latter study, in particular, has revealed much new
information of interest, from both a neuroanatomical and an
evolutionary perspective, and is the main focus of the account
that follows.   


Anterior nerve cord

Fig. 8   






Fig. 8.
    (A) Side view of the head of an amphioxus larva showing
the position of the nc = nerve cord in relation to the notochord
(shaded). The head is highly asymmetric, with the mouth
(heavy outline) on the left side and the pharyngeal (gill) slits
(lighter outlines) on the right.  
    (B) Oblique dorsal view of the anterior nerve cord showing
its main landmarks and selected cells. Asterisks indicate
the LPN3s = third pair of large paired neurons, which are putative
locomotory pacemaker neurons. Landmarks include the fe = frontal
eye, io = infundibular organ, lb = lamellar body, and PMC =
primary motor center; the zone of neuromuscular  junctions is shaded.
     This view extends to just beyond the boundary between
somites 1 and 2, roughly the extent of the cerebral vesicle
as anatomically defined and coextensive with the zone of
Otx expression. The io marks the junction between the
anterior and posterior parts of the cerebral vesicle.
    The two regions differ in the shape of the central canal
and the direction in which most cilia project into it. 

    Other abbreviations and their corresponding terms:   

    aRB = anterior Retzius bipolar cells   
   
bo = balance organ   
   
n1, n2, n3… = dorsal nerve 1, 2, 3…   
   
pinfn = postinfundibular neuropile   

Modified from Lacalli (1996). 



     Several distinctive subdomains can be recognized in the
anterior nerve cord. Externally, a slightly bulbous anterior
zone, the cerebral vesicle (CV), is easily distinguished from
the rest of the cord. The transition occurs slightly forward of
the boundary between somites 1 and 2. Internally, however,
the main anatomical landmark is the transition in the shape
of the central canal of the cord that occurs at a level coincident
with the ventral cluster of infundibular cells (Lacalli et
al. 1994; Lacalli 1996; see Fig. 8). Forward of this point is
the anterior CV, essentially the larval equivalent of the adult
anterior vesicle. Here the central canal is cylindrical and
lacks a floor plate, and most cilia project forward, some
even escaping out the neuropore. The principal landmarks of
the anterior CV are the frontal eye, located at the anterior tip
of the cord, and a putative balance organ (Lacalli and Kelly
2000), positioned just in front of the infundibular cells. Most
of the neurons are ventral, columnar, and closely packed,
especially in the immediate preinfundibular region, and have
caudally projecting neurites.
     The posterior part of the CV (probably corresponding to
the anterior part of the IR of adults) begins at the infundibular
cells. The floor plate begins here, the central canal narrows
to a slit shaped like an inverted keyhole, so that only
the ventral portion remains open, and the cilia of the
ependyma and floor plate project backwards. The dorsal part
of the cord in this zone is occupied by an ovoid mass of ciliary
lamellae, which together constitute the lamellar body.
The corresponding ventral part of the cord is taken up first
by a postinfundibular (= tegmental) neuropile and, behind
this, the primary motor center (PMC), containing the
anteriormost motoneurons and sets of interneurons with caudal
projections that are involved in locomotory control
(Lacalli 1996; Lacalli and Kelly 2003b). The transition from
this region to a type of organization more typical of the rest
of the nerve cord, whatever that entails, appears to be a gradual
one, with no obvious landmarks (e.g., no Rohde cells) to
indicate where it occurs. However, the caudal limit of the
lamellar body, which extends almost to the boundary between
somites 1 and 2 in late-stage larvae, and the similar
extent of Otx expression suggest that there is something distinctive
about the nerve cord to about the level of the boundary
between somites 1 and 2 or slightly beyond.
     Despite the usefulness of the infundibular cells as anatomical
markers, there is no obvious transition in terms of
neuronal cell type at this point. Instead, cells of essentially
anterior character are found from the preinfundibular region
to the beginning of the PMC. “Anterior” here refers to cells
with irregular basal neurites that form repeated varicosities
containing mixed vesicle types and few, if any, synapses.
These are features that are generally associated with slow
transmission, often involving neuropeptides (Burns and Augustine
1995). Beginning in the PMC, most of the neurons
have well-defined axons and separate dendritic structures, either
arbors or spines (both occur), and synaptic junctions,
often with clear vesicles, predominate. This implies fast
transmission and aminergic or amino acid transmitters,
which is perhaps logical for neurons directly involved in the
locomotory control circuits.
     The very fact that distinctive regions can be recognized
within the anterior cord, however this is defined anatomically,
raises the question of how the subdivision of the anterior
cord is controlled at the molecular level. The question is
addressed in detail elsewhere in this symposium (Shimeld
and Holland 2005), but a few remarks are useful here. Genes
such as Otx, expressed throughout most of the CV, presumably
define the character of that zone by some means,
though how is still poorly understood. But this does not explain
how specific structures within each zone are specified;
for example, the frontal eye and infundibular cells. Signalling
from boundaries is one possibility, and the anterior
boundary of the neural plate (ANB) is a good candidate,
since it is now recognized as an important signalling center
in vertebrates (Grove 2002). A similar role for the ANB in
amphioxus would help explain the organization of the frontal
eye, in which cells of like type are precisely oriented in
the transverse plane, parallel to the ANB. The infundibular
region is a second possible candidate, not least because cells
of similar type, including subtypes among the various
preinfundibular, parainfundibular, and tegmental neurons, occur
at roughly equal distances in both directions from this site.
Any gene expressed early in a pattern centered in this region
is then of potential interest. FoxD is one such gene (Yu et al.
2002), and further research may turn up more, since we are
still at a very early stage in understanding how regional subdivision
of amphioxus CNS is controlled.   

Frontal eye

     The frontal eye consists of a pigment cup, oriented so it
opens dorsally, and four rows of neurons. The first two rows
consist of simple sensory neurons, 6 in the first row and 10
in the second, with cilia that project out the neuropore, and
basal axons. The two rows differ in terms of the extent and
type of their varicosities, but both project to the ventrolateral
tracts and continue through the anterior CV, but probably not
much further. Their close association with the pigment cup
indicates that these cells are probably photoreceptors, though
this has not been tested experimentally. Behind the putative
photoreceptors are the other two rows of neurons. The first
of these rows (row 3) consists of 6 cells with multiple processes,
typically short, by means of which the cells form
multiple points of contact with each other. In the fourth row,
only the two most medial neurons show a close association
with the frontal eye. They have basal neurites that communicate
synaptically via two routes (one anterior, one at the
level of the postinfundibular neuropile) with the dendrites of
the third pair of large paired neurons (LPN3 cells), which
are key components of the locomotory control center (see
below). This arrangement of photoreceptors and neurons has
been compared to the vertebrate retina by Lacalli (1996),
who suggested possible homology between cells in rows 3
and 4 and retinal amacrine and bipolar cells, respectively. 
    A further point of similarity between the frontal eye and the
paired eyes of vertebrates is that both develop at the anterior
margin of the neural plate in what is essentially a ventromedial
position. Also, projections in both cases are to regions
caudal to the infundibulum, to roughly midbrain level.
The argument for homology is thus reasonably strong.  
    However, while the vertebrate retina has a two-
dimensional array of photoreceptors, cells in the frontal eye
form strictly onedimensional files, and there is no evidence
that this is a secondarily degenerate condition, i.e., that
amphioxus ever had an image-forming eye.
     Behavioral experiments show that amphioxus larvae can
orient to light while suspended and feeding at the water
surface, probably by modulating ciliary beat on the body
surface (Stokes and Holland 1995b). The frontal eye is implicated
in this, though an appropriate neurociliary effector
pathway has yet to be demonstrated. As the larva grows, the
pigment spot enlarges somewhat, but the complement of
photoreceptors and neurons appears to change very little.
The function of the frontal eye in the adult is not known.
Further information on this and other amphioxus photoreceptor
systems can be found in Ruppert (1997) and Lacalli (2004). 


Infundibular region

     The infundibular cells are secretory cells rather than
neurons, but they lie within a ventral mass of about 80 closely
packed neurons that resemble primary sensory cells. There
are a variety of subtypes among these. The most distinctive
(14 cells) have expanded, club-shaped cilia, which suggests
they may function to detect displacement, i.e., as a balance
organ (Lacalli and Kelly 2000). Like axons of many of the
surrounding cells, axons from this putative balance organ
project to the postinfundibular neuropile and terminate in
large varicosities.  
    Among the other neuronal subtypes in this region are
three classes of preinfundibular projection neurons
(PPN1–3) with mixed clear and dense-core vesicles and
comparatively short axons (Lacalli and Kelly 2003b); four
cells, the PPN2s, have clear vesicles and axons that travel at
least to somite 7 and possibly farther (Lacalli 2002a). In
general, from the paucity of synaptic specializations within
the postinfundibular neuropile, it appears that paracrine release
is the predominant mode of transmission, suggesting
that this region is mainly a modulatory center.
     The lamellar body is the second major contributor to the
postinfundibular neuropile. Each of its cells has a single
large axon that travels down the side of the cord to the
neuropile, where a tangled mass of subsidiary branches is
formed (Lacalli et al. 1994). The lamellar body is generally
accepted as a homolog of the vertebrate pineal organ, which
suggests that either the cells themselves or their downstream
targets in the neuropile generate a circadian rhythm. There is
no direct experimental evidence for such rhythms in either
adults or larvae, but the larvae have diurnal patterns of vertical
migration in the plankton under some conditions (Wickstead and
Bone 1959), which implies the presence of a circadian clock.   


Primary motor center

     The PMC contains the anteriormost motoneurons in the
cord and a number of large premotor interneurons. These
cell types occur elsewhere in the cord, but not in such a
large cluster. The important cells, from an organizational
standpoint, are three pairs of large paired neurons (LPNs).
These are extensively innervated by sensory inputs, both directly
by primary sensory cells in the periphery and by synapses
from the anteriormost RB cells (= B cells of Bone
1959, aRB cells in Fig. 8). The third pair, the LPN3s, are the
most important and are cross-innervated in a bilaterally symmetrical
fashion, an indication that they may be mutually inhibitory
and hence capable of pacemaker function (Lacalli
1996; Lacalli and Kelly 2003b). Their output is to ventral
compartment (VC, or fast) motoneurons via synapses and to
DC (slow) motoneurons via an unusual class of intercellular
junctions (juxtareticular junctions; see Lacalli 2002a).
     The LPN3s are thus the best candidates for neurons exerting
a direct controlling influence over both fast and slow
swimming, which appear to have a similar neuromuscular
basis in amphioxus and vertebrates (Bone 1989). Fast or escape
swimming occurs in response to sensory inputs, which
are a massive and redundant input to the VC system. The
VC system also receives synaptic input from fibers in the
postinfundibular neuropile and may be subject to additional
paracrine input as well, via fibers passing through the
neuropile, all of which provides an opportunity to modulate
the response to sensory stimuli.    
    In contrast, the slow system, which drives vertical
migration, is almost devoid of synaptic input. Besides
its link via junctions to the LPN3s, this pathway
seems to be mainly under the control of the PPN2s
mentioned above, a class of preinfundibular projection neurons
that make repeated junctional contacts with the axons
of the DC motoneurons. What this means in functional terms
is not clear, but the circuitry (Fig. 9) suggests a switching
device of some kind. Perhaps the escape response is suppressed
during migration, which might itself be under circadian
control. To assess such proposals, however, much more
information is needed on the nature of the various types of
preinfundibular neurons than is currently available.   



Fig. 9.    




Fig. 9.
    Schematic diagram of the main locomotory control circuits
in the anterior cord in young amphioxus larvae. The LPN3s
(third pair of large paired neurons) are central control neurons
that probably act as pacemakers. They receive external sensory
inputs via several pathways and communicate with the two classes
of motoneurons (DCm and VCm) by synapses or junctions as
shown. There is extensive additional synaptic input (not shown)
to the VC (fast) system, but almost none to the DC (slow) system
except for junctions with a single class of preinfundibular
neurons (type 2 preinfundibular projection neurons, PPN2s) and,
more caudally (not shown), input from the dorsal ocelli. The
function of the ipsilateral projection neurons (IPNs) is not clear,
but they appear to provide some kind of link between the two
systems. The postinfundibular (= tegmental) neuropile is a paracrine
center, and the specific interactions among its components
are not clear from the morphological data. Modified from Lacalli
(2002a). 



     VC motoneurons in larvae resemble the somatic motoneurons
(SM cells) reported from the adult in overall morphology
(cf. Lacalli and Kelly 1999; Bone 1959, 1960a).  They are
distributed rather irregularly in the anterior cord,
with roughly equal numbers on each of the two sides, but
there is no sign of bilateral pairing. They receive synapses
on dendritic spines of varying length, located all along the
axon, which confirms the supposition that the thin collaterals
reported from adult motoneurons (Bone 1960a; Castro et al.
2004) are dendrites. It is useful to note that as the cord
grows, and its neuropile expands, early dendrites would have
to lengthen to maintain their original connections. Spines in
the adult cord will thus be longer than those in the larval
cord, and the longest spines are the earliest, and presumably
most important, functional connections. Since the longest
spines in the larva are postsynaptic to LPNs, this interpretation
supports the central role proposed for these cells in initiating
swimming.
     It is not known whether the larval motoneurons persist
through to the adult stage or whether the larval cells are replaced
at some point in development. Lacalli (2000) has argued
for the former, based on the measured lengths of the
motoneuron apices. These are axially elongated by an
amount that roughly matches the axial expansion of the somites
during development. 
    DC motoneurons differ from VC motoneurons in being
restricted to the anterior part of the cord, specifically somites
2–6. This restriction was first inferred from EM data, which
showed that while axons project both rostrally and caudally
from the last two members of the series, located in somites 4
and 5, none travel forward from more caudal segments
(Lacalli and Kelly 1999). Confirming this, the amphioxus
homolog of the estrogen-related receptor gene (ERR)
selectively marks the same cells, revealing six pairs in the
anterior somites and none more caudally (Bardet et al.
2005). Various molecular data support the idea of a segmental
or otherwise periodic repeat in the arrangement of cell
types in the cord at the level of somites 2–7 (Jackman and
Kimmel 2002; Mazet and Shimeld 2002), which is essentially
the amphioxus homolog of the hindbrain. The DC
motoneurons evidently form a compressed series, with more
than one pair per segment. The true nature of patterning in
this part of the anterior cord is still, therefore, not clear. It
may be that some cell types show a strictly repeating segmental
pattern, while others are more loosely controlled, or
there may be several quasi-segmental patterns superimposed
over one another. See Shimeld and Holland (2005) for further
discussion.
     In contrast to the detailed information now published on
the microanatomy of the anterior cord, nothing comparable
is yet available for more caudal regions. Swimming behavior
changes as the larva grows, from a phased side-to-side bending
of the whole body in very young stages to what looks
like a propagated wave of contractions (Stokes 1997). The
latter implies a locomotory signal propagated from segment
to segment, more like the situation in vertebrates. One interpretation
is that the pacemaker circuits identified in the anterior
cord of young larvae are involved in initiating locomotory
contractions, but these are probably propagated through the
more caudal segments by a series of local pacemakers.
Regardless of details, it seems clear that the control
circuits described from the anterior cord of young larvae
cannot account fully for the complexity and dynamics of behavior
in older larvae.   


Peripheral sensory cells and nerves 
  

     This section is brief, as a recent review by Lacalli (2004)
covers most aspects of the larval sensory system and includes
a summary of what is known of the early circuitry. As
in the adult, the surface epithelium in the larva is supplied
with sensory cells of various types. The first evident functional
response of the larva is to mechanical stimulation, and
this correlates with the early appearance of primary type I
sensory neurons in the rostrum and tail. Axons from these
enter the cord at each end and travel long distances within it
(Holland and Yu 2002), usually in the ventrolateral tracts,
where they make repeated synapses with ventral interneurons
involved in locomotory control (Lacalli 2002b, 2004).  
Those located at the tip of the rostrum enter the cord
via the paired rostral nerves, which are substantial (ca. 25–
30 fibers) at a time when the dorsal nerves consist of, at
most, a few fibers. As the larva grows, primary sensory neurons
differentiate over much of the body surface (Stokes and
Holland 1995a; Holland and Yu 2002), and the dorsal roots
become much larger as their fibers grow into the cord. In
contrast to the rostral and caudal fibers, those entering via
dorsal nerves pass into the expanding dorsal tract, which, by
the late larval phase, has subdivided along most of the length
of the cord into separate dorsal and subdorsal tracts that run
in parallel (Holland and Yu 2002).
     Other neuronal cell types identified in the epidermal tissues
are (i) structures in the rostrum at the neurula stage
identified as growth cones by Yasui et al. (1998) because of
their apparently transitory nature, but which may be (see
Lacalli 2003a) cell bodies of pioneering rostral neurons that
differentiate early; (ii) neurons associated with the various
peripheral plexuses that synapse peripherally, including two
types (intrinsic and extrinsic neurons) in the oral nerve
plexus (Lacalli et al. 1999); (iii) additional, more specialized,
type I sensory cells, including a variant with a modified,
spine-like cilium (Lacalli and Hou 1999); (iv) type II
sensory neurons, putative chemoreceptors with a collar of
branched microvilli and basal synapses to peripheral nerves,
which develop as the larva matures (Stokes and Holland
1995a; Lacalli and Hou 1999); and (v) ventral pit cells,
which lie in rows along the developing metapleural folds
(Stokes and Holland 1995a) and are present also in late larvae,
though the evidence that these are neurons is equivocal.
     Further research is likely to reveal additional types and define
subtypes among those already described. The above
conclusion regarding the adult — that the overall organization
of the peripheral system is complex and its function is
poorly understood — applies equally to the larva.   


Postembryonic growth: generating an adult nerve cord
from the larval one
   

     As the amphioxus larva grows, the various fiber tracts and
regions of neuropile increase in size, expanding the cord
both ventrally and laterally. The addition of new neurons,
however, occurs mainly through proliferation and differentiation
of the dorsal two thirds of the ventricular layer, a region
chiefly occupied by the intermediate zone (not to be
confused with other uses of this term in reference to subdivisions
of the nerve cord along its longitudinal axis), which, in
the transverse plane, is defined by the presence of populations
of various types of translumenal neurons (Fig. 10; see
Lacalli 2002b).

Fig. 10   



Fig. 10.
    The nerve cord in transverse section, at about the level of
somite 2, in early- and late-stage amphioxus larvae. The locomotory
control system in young larvae (left) depends on sets of motoneurons
(mn) and caudally projecting interneurons, all of which are ventral,
adjacent to the floor plate. Later development (right) mainly
involves proliferation and expansion of the intermediate and dorsal
parts of the cord, which results in increasing numbers of interneurons
(tn, since most have translumenal processes) and complex circuitry
as increasing numbers of peripheral sensory fibers enter the cord via
dorsal nerves (dn). Thus, there is a clear directional bias in
the way the cord differentiates: ventral first, dorsal later.
Modified from Lacalli (2004).   


    In some cases the translumenal processes are
large enough to bridge to the neuropile on the opposite side,
but more commonly they are small and only just cross the
central canal. The largest belong to the larval giant cells, of
which there are five forward of the first Rohde cell (Bone
1959); these probably correspond to the large translumenal
neurons described above from the adult IR. The larval cells
have ipsilateral axons but otherwise resemble Rohde cells,
which have medial or contralateral axons. Bone (1960a) suggests
an internuncial function for cells of this general type,
and tracings of some of the more abundant small
translumenal neurons in larvae tend to confirm this (Lacalli
and Kelly 2003a). The development of the intermediate
zone, in fact, parallels the increase in peripheral input via
dorsal nerve roots, which itself is a consequence of the increase
in the surface area of the body and the number of peripheral
sensory cells. Young larvae have very simple sensorimotor
circuits, but it is unlikely that these would cope
well with vastly increased input, presumably excitatory in
nature, if it is not filtered or modulated in some way. This is
the problem of gain control, which is common to developing
neural systems (e.g., see Priebe and Ferster 2002). The additional
level of processing inserted during larval development
between the sensory input and the motor output is evidently
a way of solving this problem.   
     Except for the addition of radial glial cells with fibers that
bridge to the top of the notochord (e.g., as in Fig. 4), the
ventral part of the ependymal region, where the central canal
remains open, changes very little throughout the larval phase
(Lacalli and Kelly 2002). The progressive increase in radial
glial cell numbers does appear to have a consequence, however,
because the apices of motoneurons in the juvenile and
the adult open into the central canal more dorsally than in
the early larva, i.e., slightly above the ventral expansion
rather than adjacent to the floor plate. Unless there is a
wholesale replacement of larval neurons by juvenile and
adult ones, which seems unlikely, this is best explained by a
passive displacement of early neurons up the sides of the
central canal as more ventrally positioned glial precursors
proliferate.
     In summary, the neurons in the amphioxus nerve cord appear
to differentiate along most of the cord’s length in a ventral
to dorsal sequence. This can be interpreted in various
ways. If ontogeny were, in fact, no more than a recapitulation
of phylogeny, it would mean that the ventral locomotory
circuits were evolutionarily older than the dorsal modulatory
ones. While it is certainly true that some dorsal neurons are
highly specialized, and may well be late-evolving cell types,
this can be true of all dorsal cells only if there was an ancestral
form that swam but lacked any way of modulating its locomotion
in response to peripheral sensory input. This is highly
unlikely. A better explanation is that the hatching
stage in amphioxus has been secondarily reduced in size, to
at least some degree, during evolution. As the hatching larva
evolved ways to utilize the resources of the egg most effectively,
the differentiation of the essential parts of the locomotory
circuits was accelerated at the expense of everything
else. In this interpretation, the early differentiation of the
ventral circuits is a clear indication of their crucial importance
to the hatching larva, whereas the dorsal modulatory
pathways are evidently less important, such that their development
can be delayed.
     There are a number of specialized cell groupings at the
anterior end of the adult nerve cord that are not present in
early larvae, including various types of migrated cells described
above from the adult IR. Judging from the time that
the anterolateral serotoninergic cells first appear (Holland
and Holland 1993), these cell groupings probably develop in
the late larval phase or during metamorphosis. Despite the
proliferative activity this entails, the anteriormost region
fails to thicken as much as the rest of the cord, so the CV
progressively disappears as an externally recognizable zone.
     Of the late-developing cell groups, the dorsal (population I)
dopamine-containing cells reported by Moret et al. (2004)
are especially noteworthy. These are as dorsal and anterior
as one can get in the nerve cord, which is precisely where a
telencephalic homolog would be predicted to form if
amphioxus had one. For this and other reasons, Lacalli (2004)
suggested that the population I cells may represent a primitive
version of the olfactory bulb. Regardless of whether this
is eventually confirmed, the key point is to recognize that
these cells, like the Joseph cells and other anterior migrated
cell clusters, are all late-developing centers that probably act
in a modulatory way on established circuits. It should be
possible, in principle, to determine what functions such cells
perform by correlating their time of appearance during development
with changes in behavior. This could prove a useful
experimental strategy in future.
     The existence of migrated cell groups of various types in
the anterior nerve cord raises an interpretive problem that
deserves some attention. Elsewhere in the cord, the majority
of neurons remain attached to the ventricular surface, so
their site of origin with reference to the dorsoventral axis of
the nerve cord is clear by inspection. Where this is not the
case, e.g., for the anterior migrated cell groups, the final position
of the cells with respect to the dorsoventral axis,
whether they reside close to the periventricular layer or deep
in the neuropile, may be secondarily altered. This can happen
in two ways: either the cells themselves migrate dorsoventrally
or the expanding ventricular layer leaves them behind.  
Thus, a migrated cell or group of such cells might be
quite dorsal in terms of its point of origin in the ventricular
layer, but could ultimately occupy a position well below the
dorsalmost cells of the mature nerve cord. The time element
in development therefore needs to be explicitly considered
in such cases. This is an important issue, particularly in
cases where one is trying to identify possible homologs of
vertebrate CNS neurons, as we do in the section below. For
example, in earlier attempts to deal with the different subcategories
of RB cells (e.g., Lacalli 1996), the term “tectal cell”
was applied to the anterior group (aRB cells in Fig. 8, equivalent
to the A cells of Bone 1959). The term was chosen in
part, and perhaps unfortunately, because it seemed that the
dorsalmost cells in a region judged to be midbrain-like on the
basis of molecular criteria could, in the course of evolution,
have generated major new dorsal structures, including the
optic tectum. However, cells in the dorsal part of the
amphioxus ventricular layer in young larvae have only just
begun to proliferate and generate neurons, so it is not obvious
whether the late progeny of these cells are more or less
dorsal in character than early ones. How this all relates to
the appearance of entirely new dorsal structures and brain
regions during the early stages of vertebrate evolution remains
to be determined.   


Comparison with vertebrates

     Amphioxus is now generally accepted as the best available
model for the immediate invertebrate ancestor of vertebrates
(Holland 2000).  As such, it is of key importance to
investigations into vertebrate origins and characteristic features
of vertebrate organization. The PNS, in particular, has
been the subject of much comparative analysis in the past.
Amphioxus differs from vertebrates in its reliance on peripherally
derived sensory neurons and extensive peripheral
plexuses, which are largely replaced in vertebrates by products
of the neural crest and placodes. The reason for this
transition and the relation (i.e., possible homology) between
the component cell types is currently a matter of some interest,
but many key questions remain unanswered (Lacalli
2004). At the anatomical level, it is the arrangement and
innervation patterns of the dorsal nerves that have received
the most attention, mainly because of the clues these provide
concerning the segmental structure of the vertebrate head.
The key issue here is the relation between the serially repeating
units in the hindbrain (i.e., rhombomeres), pharynx
(gills and gill arches), and paraxial mesoderm, and how
these all relate to the somite series. Do they, for example, reflect
a single underlying pattern of repeats, or are they independent
patterns secondarily superimposed on one another?
To date, the contribution of amphioxus to this debate has
proven less than enlightening (e.g., see Northcutt 2001), perhaps
because patterns of peripheral innervation are less conservative
than one would ideally like. In addition, however,
the peculiarities of larval growth and metamorphosis in
amphioxus, especially the caudal shift of the mouth and oral
apparatus, ensure that the spatial relation between the nerve
cord and peripheral structures in the head is not only different
than that in vertebrates, but also undergoes developmental
changes that have no vertebrate counterpart.
     More recently, the wealth of data on patterns of gene
expression during development has renewed efforts to identify
regional homologies between the vertebrate brain and the
anterior nerve cord of amphioxus. Here the molecular and
anatomical data are largely in agreement, and a comparatively
consistent story is emerging. The presence of a pineal
homolog (the lamellar body) and a comparable infundibular
region has long been accepted as evidence that the CV is basically
a primitive counterpart of the diencephalon (Olsson 1986).
     Patterns of Otx, FoxB, and Hox gene expression indicate
the presence of regions homologous with the forebrain +
midbrain to about the level of the boundary between somites
1 and 2 (Holland and Holland 1999; Shimeld and Holland
2005). This places the PMC, with its pacemaker neurons, at
roughly midbrain level. Then, somewhere adjacent to somite
2, a zone begins that is hindbrain-like in character, in which
gene expression occurs in a segmental or quasi-segmental
pattern of repeats (Jackman and Kimmel 2002; Mazet and
Shimeld 2002). The microanatomy shows, however, that except
for the lamellar body, it is only the ventral structures
that are represented in the larval brain of amphioxus
(Fig. 11). There are zones similar in organization to the ventral
diencephalon, extending from the preoptic area to the
hypothalamus and infundibulum, the floor of the midbrain,
roughly equivalent to the tegmentum, and the anterior end of
the reticulospinal system. Further, the arrangement of longitudinal
fiber tracts is similar in vertebrates and amphioxus
larvae (cf. Hjorth and Key 2002; Lacalli et al. 1994), and
both develop an early ventral connection between the two
sides of the cord in the immediate postinfundibular region,
namely the ventral commissure in vertebrates and the
postinfundibular neuropile in amphioxus larvae.   


Fig. 11.   




Abbreviations and their corresponding terms: 

   
CV = (larval) cerebral vesicle    
   
io = infundibular organ   
   
PMC = primary motor center   

Fig. 11.
    A proposal, consistent with both anatomical and molecular data,
for how the anterior nerve cord in amphioxus larvae maps to
the vertebrate brain. Most of the former finds its closest counterpart,
and probable homologs, in the ventral region (variously shaded)
of the latter, with the exception of the lamellar body, a pineal homolog.  
    Caudal to somite 2, the amphioxus nerve cord becomes more
hindbrain-like in character, according to the molecular data, but the
exact point of the transition is uncertain (indicated by the question
mark), as there is no obvious corresponding anatomical transition.
    Modified from Lacalli (1996). 


     In addition, what is so far inferred concerning the functions
performed in the anterior nerve cord in amphioxus larvae
also indicates similarities to the vertebrate brain. Basically,
a variety of modulatory inputs, including signals from sensory
neurons located in the hypothalamus or its amphioxus
equivalent, converge on a ventral locomotory control center
(tegmentum and reticulospinal system in vertebrates, PMC
in amphioxus), and this initiates a locomotory response. The
ancestral plan one infers from this would thus have a series
of basal centers and connecting tracts that activate swimming
under the control of external sensory inputs, as well as

internal homeostatic signals, and would incorporate a
switching center to coordinate escape swimming with other
basic activities, e.g., migratory swimming and feeding.
     In contrast, there is no evidence in amphioxus for homologs
of any of the dorsal brain structures involved in vertebrate
sensory processing except for the pineal organ. A second
possible exception is the olfactory bulb, which Lacalli
(2004) has argued may be represented in rudimentary form
in the advanced-stage larva of amphioxus (and presumably
also the adult), but this remains to be proven. The significance
of the Joseph cells, if any, in relation to dorsal visual
centers in vertebrates, remains unresolved. The various major
centers involved in processing inputs from the vertebrate
organs of special sense therefore appear to be absent in
amphioxus, again with the possible exception of the pineal
organ and olfactory bulb. This is presumably because the
sense organs themselves, and the corresponding CNS processing
centers, evolved as vertebrates evolved, after their
divergence from more basal chordate lineages.
     A more detailed look at the nerve cord reveals that the
arrangement of cell types across the dorsoventral axis has a
number of recognizably vertebrate-like features: there is a
floor plate, and motoneurons are ventral, while sensory
interneurons are dorsal. There are also enough similarities
between the RB cells in amphioxus and the Rohon-Beard
cells of vertebrates to suggest homology (Fritzsch and
Northcutt 1993). The anatomical evidence for a relation between
Rohde cells and the various types of reticulospinal giant
cells found in vertebrates is somewhat less convincing,
however. Dorsal expansion of the cord is a noticeable feature
of CNS development during the late larval phase; in
contrast, the ventral components of the cord probably
change very little. This reflects the need for a vastly increased
capacity for sensory integration as the larva grows
and approaches metamorphosis. Comparable events in vertebrate
development occur in the embryo, when proliferation
expands the cord dorsally and somatosensory reflex circuits
complete their differentiation. These changes correlate with
increased egg and embryo size and a prolonged period of
embryogenesis in vertebrates, which allows for a much
larger and better developed nervous system at hatching (see
Fig. 12), possibly in response to the increased predatory
pressures vertebrates (and their hatchlings) experienced during
their early evolution. This is an important point, because
it adds a developmental dimension to previous ideas about
CNS evolution. Though many of the events of neurogenesis
in vertebrate embryos have probably always been embryonic,
it is likely that there are others that evolved as additions
to an already functioning CNS in an active animal.
These would have been incorporated into the embryo only
secondarily, so their mechanism of formation and the way in
which the existing circuitry accommodates new inputs
should reflect this in some way.   

Fig. 12   


Abbreviations and their corresponding terms:   

   
comv = ventral commissure   
   
ep = epiphysis / pineal organ   
   
inf = infundibulum   
   
os = optic stalk   
   
r1, r2, r3…  = rhombomere 1, 2, 3…   
   
s1, s2, s3…  = somite 1, 2, 3…   
   
t = telencephalon   
    tc = optic tectum
   


Fig. 12.
    Differences in scale between the anterior nerve cord of amphioxus
at two stages (14-day-old larva, middle diagram, and newly
metamorphosed juvenile, bottom diagram) and an embryonic
vertebrate brain (zebrafish at 24 h, top diagram). For the zebrafish, cell
clusters involved in early tract formation are shown. Arrowheads
indicate the locations of the infundibular cells in amphioxus; the
small arrow points to the postinfundibular (tegmental) neuropile.
Based on a previous analysis of size differences; see Lacalli (2004). 
 
     By way of example, consider the development in amphioxus
of midbrain-level dopaminergic and serotoninergic neurons
reported by Moret et al. (2004). Neurons utilizing these
transmitters are present at a roughly comparable location in
vertebrates and contribute to several functionally important
modulatory systems. Why did dopaminergic neurons evolve
in this particular location, and what can one infer about their
original function? If they were not originally formed in the
embryo, at what point in the life history did they first appear,
and why? These questions are raised here not because
we have answers, but to illustrate the following general
point: when constructing evolutionary scenarios, one always
needs to be thinking in terms of a sequence of changing life
histories. From our analysis, this will be especially true in
the case of innovations in CNS organization and structures
that develop comparatively late in embryogenesis.
     In the more immediate term, future work on the amphioxus
nervous system might usefully be directed at refining
our understanding of the molecular differences between
neuronal subclasses, in terms of both gene expression patterns
and neurotransmitters. Work of this type is now in
progress in a number of laboratories, but much remains to be
done. The amphioxus CNS is well suited for such studies
because it is small and compact enough that a thorough inventory
of cell types is a feasible objective in the larva at
least, if not in the adult. A disadvantage is that many aspects
of amphioxus neural organization are so peculiar that the
conventional wisdom as to how a neural system ought to operate
is sometimes more a hindrance than a help. On the
evolutionary side, there are questions that can be addressed
concerning the cell types and circuits in the ventral part of
the anterior nerve cord, and here there may be direct
homologs in the ventral brainstem centers of vertebrates. In
contrast, amphioxus will likely provide very little in the way
of useful clues regarding the origin and basic cell types of
dorsal structures in the vertebrate brain, as these evidently
evolved largely after lancelets and vertebrates diverged. A final
point, especially relevant to the subject of this issue, is
the relation to hemichordates. Here, amphioxus offers a possible
bridge between the highly centralized system of vertebrates
and the diffuse one of hemichordates (Lacalli 2003b).
     One would hope to find intimations of the latter in
amphioxus sufficient to indicate whether the hemichordate
system is derived or whether it retains at least some important
ancestral features. 


References 

*
Anadón, R., Adrio, F., and Rodríguez-Moldes, I. 1998.
Distribution of GABA immunoreactivity in the central and peripheral nervous
system of amphioxus (Branchiostoma lanceolatum Pallas).
J. Comp. Neurol. 410: 293–307. 

Baatrup, E. 1981.
Primary sensory cells in the skin of amphioxus
(Branchiostoma lanceolatum (P)).
Acta Zool. (Stockh.) 62: 147– 157. 

Baatrup, E. 1982.
On the structure of the corpuscles of de Quatrefages
(Branchiostoma lanceolatum (P)).
Acta Zool. (Stockh.) 63: 39–44. 

Baker, C.V.H., and Bronner-Fraser, M. 1997.
The origins of the neural crest. Part II. An evolutionary perspective.
Mech. Dev. 69: 13–29. 

Bardet, P.L., Schubert, M., Horard, B., Holland, L.Z., Laudet, V.,
Holland, N.D., and Vanacker, J.-M. 2005.
Expression of estrogen-related receptors in amphioxus and zebrafish: implications
for the evolution of hindbrain segmentation at the invertebrate-to-vertebrate transition.
Evol. Dev. In press. 

Boeke, J. 1902.
Über das Homologon des Infundibularorganes bei Amphioxus lanceolatus.
Anat. Anz. 21: 411–414. 

Boeke, J. 1908.
Das Infundibularorgan im Gehirn des Amphioxus.
Anat. Anz. 32: 473–488. 

Boeke, J. 1935.
The autonomic (enteric) nervous system of Amphioxus lanceolatus.
Q. J. Microsc. Sci. 77: 623–658. 

Bone, Q. 1959.
The central nervous system in larval acraniates.
Q. J. Microsc. Sci. 100: 509–527. 

Bone, Q. 1960a.
The central nervous system in amphioxus.
J. Comp. Neurol. 115: 27–64. 

Bone, Q. 1960b.
A note on the innervation of the integument in amphioxus, and its bearing on the mechanism of cutaneous sensibility.
Q. J. Microsc. Sci. 101: 371–379. 

Bone, Q. 1961.
The organization of the atrial nervous system of amphioxus [Branchiostoma lanceolatum (Pallas)].
Philos. Trans. R. Soc. Lond. B Biol. Sci. No. 243: 241–269. 

Bone, Q. 1989.
Evolutionary patterns of axial muscle systems in some invertebrates and fish.
Am. Zool. 29: 5–18. 

Bone, Q., and Best, A.C.G. 1978.
Ciliated sensory cells in amphioxus.
J. Mar. Biol. Assoc. U.K. 58: 479–486. 

Bone, Q., Chubb, A.D., Pulsford, A., and Ryan, K.P. 1996.
FMRFamide immunoreactivity in the peripheral (atrial) nervous
system of amphioxus (Branchiostoma).
Isr. J. Zool. 42(Suppl.): 213–225. 

Burns, M.E., and Augustine, G.J. 1995.
Synaptic structure and function: dynamic organization yields architectural precision.
Cell, 83: 187–194. 

Candiani, S., and Pestarino, M. 1998.
Expression of the tissuespecific transcription factor Pit-1 in the lancelet, Branchiostoma
lanceolatum.
J. Comp. Neurol. 392: 343–351. 

Candiani, S., Augello., A., Oliveri, D., Passalacqua, M., Pennati,
R., de Bernardi, F., and Pestarino, M. 2001.
Immunocytochemical localization of serotonin in embryos, larvae and adults
of the lancelet, Branchiostoma floridae.
Histochem. J. 33: 413– 420. 

Castro, A., Manso, M.J., and Anadón, R. 2003.
Distribution of neuropeptide Y immunoreactivity in the central and peripheral
nervous systems of amphioxus (Branchiostoma lanceolatum Pallas).
J. Comp. Neurol. 461: 350–361. 

Castro, A., Becerra, M., Manso, M.J., and Anadón, R. 2004.
Somatomotor system of the adult amphioxus (Branchiostoma
lanceolatum) revealed by an anticalretinin antiserum: an
immunocytochemical study.
J. Comp. Neurol. 477: 161–171. 

Conklin, E.G. 1932.
The embryology of amphioxus.
J. Morphol. 54: 69–151. 

de Quatrefages, M.A. 1845.
Mémoire sur le systéme nerveux et sur l’histologie du Branchiostome ou amphioxus. Annls. Sci. Nat. 4: 197–248. 

Dogiel, A.S. 1903.
Das periphere Nervensystem des Amphioxus (Branchiostoma lanceolatum).
Anat. Hefte, 21[= Heft 66]: 145– 213. 

Dörffler-Melly, J., and Neuhuber, W.L. 1988.
Rectospinal neurons: evidence for a direct projection from the enteric to the central
nervous system in the rat.
Neurosci. Lett. 92: 121–125. 

Eakin, R.M., and Westfall, J.A. 1962.
Fine structure of photoreceptors in amphioxus.
J. Ultrastruct. Res. 6: 531–539. 

Edinger, L. 1906.
Einiges vom “Gehirn” des Amphioxus.
Anat. Anz. 28: 417–428. 

Ekhart, D., Korf., H.-W., and Wicht, H. 2003.
Cytoarchitecture, topography, and descending supraspinal projections in the anterior
nervous system of Branchiostoma lanceolatum.
J. Comp. Neurol. 466: 319–330. 

Flood, P.R. 1966.
A peculiar mode of muscular innervation in amphioxus. Light and electron microscopic studies of the socalled ventral roots.
J. Comp. Neurol. 126: 181–218. 

Flood, P.R. 1968.
Structure of the segmental trunk muscles in amphioxus. With notes on the course and “endings” of the socalled ventral root fibres.
Z. Zellforsch. Mikrosk. Anat. 84: 389– 416. 

Flood, P.R. 1970.
The connection between spinal cord and notochord in amphioxus (Branchiostoma lanceolatum).
Z. Zellforsch. Mikrosk. Anat. 103: 115–128. 

Flood, P.R. 1974.
Histochemistry of cholinesterase in amphioxus (Branchiostoma lanceolatum, Pallas).
J. Comp. Neurol. 157: 407–438. 

Franz, V. 1923.
Haut, Sinnesorgane und Nervensystem der Akranier. Jena.
Z. Naturw. 59: 402–526. 

Franz, V. 1927.
Morphologie der Akranier. Ergebn.
Anat. Entwgesch. 27: 465–692. 

Fritzsch, B. 1996.
Similarities and differences in lancelet and craniate nervous systems.
Isr. J. Zool. 42(Suppl.): 147–160. 

Fritzsch, B., and Northcutt, R.G. 1993.
Cranial and spinal nerve organization in amphioxus and lampreys: evidence for an ancestral craniate pattern.
Acta Anat. 148: 96–109. 

Fusari, R. 1889.
Beitrag zum Studium des peripherischen Nervensystems von Amphioxus lanceolatus. Int. Mschr. Anat. Physiol.
6: 120–140. 

Gilland, E., and Baker, R. 1993.
Conservation of neuroepithelial and mesodermal segments in the embryonic vertebrate head.
Acta Anat. 148: 110–123. 

Glardon, S., Holland, L.Z., Gehring, W.J., and Holland, N.D. 1998.
Isolation and developmental expression of the amphioxus Pax-6
gene (AmphiPax-6): insights into eye and photoreceptor evolution.
Development, 125: 2701–2710. 

Gorbman, A. 1999.
Brain – Hatschek’s pit relationships in amphioxus species.
Acta Zool. 80: 301–305. 

Gorbman, A., and Tamarin, A. 1985.
Early development of oral, olfactory and adenohypophyseal structures of agnathans and its evolutionary implications. In Evolutionary biology of primitive
fishes. Edited by R.E. Foreman, A. Gorbman, J.M. Dodd, and R. Olsson.
Plenum Press, New York. pp. 165–186. 

Gorbman, A., Nozaki., M., and Kubokawa, K. 1999.
A brain – Hatschek’s pit connection in amphioxus.
Gen. Comp. Endocrinol. 113: 251–254. 

Grove, E.A. 2002.
The telencephalon saved by TLC.
 Neuron, 35: 215–217. 

Guthrie, D.M. 1975.
The physiology and structure of the nervous system of amphioxus (the lancelet), Branchiostoma lanceolatum Pallas.
In Protochordates. Edited by E.J.W. Barrington and R.P.S. Jefferies.
Symp. Zool. Soc. Lond. No. 36: 43–80. 

Hatschek, B. 1884.
Mittheilungen über Amphioxus.
Zool. Anz. 7: 517–520. 

Hesse, R. 1898.
Untersuchungen über die Organe der Lichtempfindung
bei niederen Tieren. IV. Die Sehorgane des Amphioxus.
Z. Wiss. Zool. 63: 456–464. 

Hirakow, R., and Kajita, N. 1990.
An electron microscopic study of the development of amphioxus, Branchiostoma belcheri: cleavage.
J. Morphol. 203: 331–344. 

Hirakow, R., and Kajita, N. 1991.
Electron microscopic study of the development of amphioxus, Branchiostoma belcheri: gastrulation.
J. Morphol. 207: 37–52. 

Hirakow, R., and Kajita, N. 1994.
Electron microscopic study of the development of amphioxus, Branchiostoma belcheri: the neurula and larva.
Acta Anat. Nippon. 69: 1–13. 

Hjorth, J., and Key, B. 2002.
Development of axon pathways in the zebrafish central nervous system.
Int. J. Dev. Biol. 46: 609–619. 

Holland, L.Z., and Holland, N.D. 1999.
Chordate origins of the vertebrate central nervous system.
Curr. Opin. Neurobiol. 9: 596–602. 

Holland, N.D., and Holland, L.Z. 1993.
Serotonin-containing cells in the nervous system and other tissues during ontogeny of a
lancelet, Branchiostoma floridae.
Acta Zool. 74: 195–204. 

Holland, N.D., and Yu, J.-K. 2002.
Epidermal receptor development and sensory axon pathways in vitally stained amphioxus.
Acta Zool. 83: 309–321. 

Holland, P.W.H. 1996.
Molecular biology of lancelets: insights into development and evolution.
Isr. J. Zool. 42(Suppl.): 247–272. 

Holland, P.W.H. 2000.
Embryonic development of heads, skeletons and amphioxus:
Edwin S. Goodrich revisited.
Int. J. Dev. Biol. 44: 29–34. 

Holmes, W. 1953.
The atrial nervous system of amphioxus (Branchiostoma).
Q. J. Microsc. Sci. 94: 523–535. 

Jackman, W.R., and Kimmel, C.B. 2002.
Coincident iterated gene expression in the amphioxus neural tube.
Evol. Dev. 4: 366–374. 

Jacobs, D.K., and Gates, R.D. 2003.
Developmental genes and the reconstruction of metazoan evolution — implications of evolutionary loss, limitations on inference of ancestry, and type 2 errors.
Integr. Comp. Biol. 43: 11–18. 

Johnston, J.B. 1905.
The cranial and spinal ganglia and the viscero-motor roots in amphioxus.
Biol. Bull. (Woods Hole), 9: 112–127. 

Joseph, H. 1904.
Über eigentümliche Zellstrukturen im Zentralnervensystem von Amphioxus.
Anat. Anz. 25(Suppl.): 16–26. 

Kaji, T., Aizawa, S., Uemura, M., and Yasui, K. 2001.
Establishment of left-right asymmetric innervation in the lancelet oral region.
J. Comp. Neurol. 435: 394–405. 

Kölliker, A. 1843.
Über das Geruchsorgan von Amphioxus. (Müllers)
Arch. Anat. Physiol. (Berlin), 1843: 33–55. 

Kutchin, H.L. 1913.
Studies on the peripheral nervous system of amphioxus.
Proc. Am. Acad. Arts Sci. 49: 569–624. 

Lacalli, T.C. 1996.
Frontal eye circuitry, rostral sensory pathways and brain organization in amphioxus larvae: evidence from 3D reconstructions.
Philos. Trans. R. Soc. Lond. B Biol. Sci. No. 351: 243–263. 

Lacalli, T.C. 2000.
Cell morphology in amphioxus nerve cord may reflect the time course of cell differentiation.
Int. J. Dev. Biol. 44: 903–906. 

Lacalli, T.C. 2002a.
The dorsal compartment locomotory control system in amphioxus larvae.
J. Morphol. 252: 227–237. 

Lacalli, T.C. 2002b.
Sensory pathways in amphioxus larvae. I. Constituent fibres of the rostral and anterodorsal nerves, their targets and evolutionary significance.
Acta Zool. 83: 149–166. 

Lacalli, T.C. 2003a.
Ventral neurons in the anterior nerve cord of amphioxus larvae. II. Further data on the pacemaker circuit.
J. Morphol. 257: 212–218. 

Lacalli, T.C. 2003b.
Body plans and simple brains.
Nature (Lond.), 424: 263–264. 

Lacalli, T.C. 2004.
Sensory systems in amphioxus: a window on the ancestral chordate condition.
Brain Behav. Evol. 64: 148– 162. 

Lacalli, T.C., and Hou, S. 1999.
A re-examination of the epithelial sensory cells of amphioxus.
 Acta Zool. 80: 125–134. 

Lacalli, T.C., and Kelly, S.J. 1999.
Somatic motoneurons in the anterior nerve cord of amphioxus larvae: cell types, cell position and innervation patterns.
Acta Zool. 80: 113–124. 

Lacalli, T.C., and Kelly, S.J. 2000.
The infundibular balance organ and related aspects of cerebral vesicle organization. Acta Zool. 81: 37–47. 

Lacalli, T.C., and Kelly, S.J. 2002.
Floor plate, glia and other support cells in the anterior nerve cord of amphioxus larvae. Acta Zool. 83: 87–98. 

Lacalli, T.C., and Kelly, S.J. 2003a.
Sensory pathways in amphioxus larvae. II. Dorsal tracts and translumenal cells.
Acta Zool. 84: 1–13. 

Lacalli, T.C., and Kelly, S.J. 2003b.
Ventral neurons in the anterior nerve cord of amphioxus larvae. I. An inventory of cell types and synaptic patterns.
J. Morphol. 257: 190–211. 

Lacalli, T.C., Holland, N.D., and West, J.E. 1994.
Landmarks in the anterior central nervous system of amphioxus larvae.
Philos. Trans. R. Soc. Lond. B Biol. Sci. No. 344: 165–185. 

Lacalli, T.C., Gilmour, T.H.J., and Kelly, S.J. 1999.
The oral nerve plexus in amphioxus larvae: function, cell types and phylogenetic
sugnificance.
Proc. R. Soc. Lond. B Biol. Sci. 266: 1461– 1470. 

Lankester, E.R. 1875.
On some new points in the structure of amphioxus and their bearing on the morphology of Vertebrata.
Q. J. Microsc. Sci. 15: 257–267. 

Lele, P.P., Palmer, E., and Weddell, G. 1958.
Observations on the innervation of the integument of amphioxus, Branchiostoma
lanceolatum.
Q. J. Microsc. Sci. 99: 421–440. 

Mazet, F., and Shimeld, S.M. 2002.
The evolution of chordate neural segmentation.
Dev. Biol. 251: 258–270. 

Meves, A. 1973.
Elektronenmikroskopische Untersuchungen über die Zytoarchitektur des Gehirns von Branchiostoma lanceolatum.
Z. Zellforsch. Mikrosk. Anat. 139: 511–532. 

Mirshahi, M., Bouchieux, C., Collenot, G., Thillaye, B., and Faure,
J.-P. 1985.
Retinal S-antigen epitopes in vertebrate and invertebrate photoreceptors.
Investig. Ophthalmol. Vis. Sci. 26: 1016– 1021. 

Moret, F., Guilland, J.-C., Coudouel, S., Rochette, L., and Vernier, P. 2004.
Distribution of tyrosine hydroxylase, dopamine and serotonin
in the central nervous system of amphioxus: implications
for the evolution of catecholamine systems in vertebrates.
J. Comp. Neurol. 468: 135–150. 

Müller, E. 1900.
Studien über Neuroglia.
Arch. Mikrosk. Anat. Entwmech. 60: 11–62. 

Müller, J. 1844.
Über den Bau und die Lebenserscheinungen des Branchiostoma lubricum Costa, Amphioxus lanceolatus Yarrell.
Abh. K. Preuss. Akad. Wiss. 1844: 79–116. 

Neuhuber, W.L., Appelt, M., Polak, J.M., Baier-Kustermann, W.,
Abelli, L., and Ferri, G.-L. 1993.
Rectospinal neurons: cell bodies, pathways, immunocytochemistry and ultrastructure. Neuroscience, 56: 367–378. 

Nieuwenhuys, R. 1998.
Amphioxus. In The central nervous system of vertebrates.
Vol. 1. Edited by R. Nieuwenhuys, H.J. ten Donkelaar, and C. Nicholson.
Springer-Verlag, Berlin. pp. 365–396. 

Northcutt, R.G. 2001.
Lancelet lessons: evaluating a phylogenetic model.
J. Comp. Neurol. 435: 391–393. 

Nozaki, M., and Gorbman, A. 1992.
The question of functional homology of Hatschek’s pit of amphioxus (Branchiostoma
belcheri) and the vertebrate adenohypophysis.
Zool. Sci. (Tokyo), 9: 387–395. 

Obermüller-Wilén, H. 1976.
The infundibular organ of Branchiostoma lanceolatum.
Acta Zool. 57: 211–216. 

Obermüller-Wilén, H., and Olsson, R. 1974.
The Reissner’s fiber termination in some lower chordates.
Acta Zool. 55: 71–79. 

Olsson, R. 1986.
Basic design of the chordate brain.
In Proceedings of the 2nd International Conference on Indo-Pacific Fishes:
Indo-Pacific fish biology, Tokyo, Japan, 29 July – 3 August
1985. Edited by T. Uyeno, R. Arai, T. Taniuchi, and K. Matsuura.
Ichthyological Society of Japan, Tokyo. pp. 86–93. 

Poss, S.G., and Boschung, H.T. 1996.
Lancelets (Cephalochordata: Branchiostomatidae): How many species are valid?
Isr. J. Zool. 42(Suppl.): 13–66. 

Presley, R., Horder, T.J., and Slipka, J. 1996.
Lancelet development as evidence of ancestral chordate structure.
Isr. J. Zool. 42(Suppl.): 97–116. 

Priebe, N.J., and Ferster, D. 2002.
A new mechanism for neuronal gain control (or how the gain in brains has mainly been explained). Neuron, 35: 602–604. 

Retzius, G. 1891.
Zur Kenntniss des Centralnervensystems von Amphioxus lanceolatus.
Biol. Unters. 2: 29–42. 

Rohde, E. 1887.
Histologische Untersuchungen über das Nervensystem von Amphioxus lanceolatus. Zool. Beitr. 2(1): 169–211. 

Ruiz, M.S., and Anadón, R. 1989.
Some observations on the fine structure of the Rohde cells of the spinal cord of the amphioxus, Branchiostoma lanceolatum (Cephalochordata).
J. Hirnforsch. 30: 671–677. 

Ruiz, M.S., and Anadón, R. 1991a.
Ultrastructural study of the filum terminale and caudal ampulla of the spinal cord of
amphioxus (Branchiostoma lanceolatum Pallas).
Acta Zool. 72: 63–71. 

Ruiz, M.S., and Anadón, R. 1991b.
The fine structure of lamellate cells in the brain of amphioxus (Branchiostoma lanceolatum, Cephalochordata).
Cell Tissue Res. 263: 597–600. 

Ruiz, M.S., and Anadón, R. 1991c.
Some considerations on the fine structure of rhabdomeric photoreceptors in the amphioxus, Branchiostoma lanceolatum (Cephalochordata).
J. Hirnforsch. 32: 159–164. 

Ruppert, E.E. 1997.
Cephalochordata (Acrania). In Microscopic anatomy of invertebrates.
Vol. 15. Edited by F.W. Harrison and E.E. Ruppert.
Wiley, New York. pp. 349–504. 

Schneider, A. 1879.
Beiträge zur vergleichenden Anatomie und Entwicklungsgeschichte der Wirbelthiere. Verlag von Georg Reimer, Berlin. 

Schulte, E., and Riehl, R. 1977.
Elektronenmikroskopische Untersuchungen an den Oralcirren und der Haut von Branchiostoma lanceolatum.
Helgol. Wiss. Meeresunters. 29: 337–357. 

Sherwood, N.M., Adams, B.A., and Tello, J.A. 2005.
Endocrinology of protochordates.
Can. J. Zool. 83: 225–255. 

Shimeld, S.M., and Holland, N.D. 2005.
Amphioxus molecular biology: insights into vertebrate evolution and developmental
mechanisms.
Can. J. Zool. 83: 90–100. 

Stach, T. 2000.
Microscopic anatomy of developmental stages of Branchiostoma lanceolatum (Cephalochordata, Chordata).
Bonn. Zool. Monogr. 47. 

Stokes, M.D. 1997.
Larval locomotion of the lancelet Branchiostoma floridae.
J. Exp. Biol. 200: 1661–1680. 

Stokes, M.D., and Holland, N.D. 1995a.
Embryos and larvae of a lancelet, Branchiostoma floridae, from hatching to metamorphosis: growth in the laboratory and external morphology.
Acta Zool. 76: 105–120. 

Stokes, M.D., and Holland, N.D. 1995b.
Ciliary hovering in lancelets.
Biol. Bull. (Woods Hole), 188: 231–233. 

Takeda, N., Kubokawa, K., and Matsumoto, G. 2003.
Immunoreactivity for progesterone in the giant Rohde cells of the
amphioxus, Branchiostoma belcheri.
Gen. Comp. Endocrinol. 132: 379–383. 

Tjoa, L.T., and Welsch, U. 1974.
Electron microscopical observations on Kölliker’s and Hatschek’s pit and on the wheel organ in the head region of amphioxus (Branchiostoma lanceolatum).
Cell Tissue Res. 153: 175–187. 

Uemura, H., Tezuka., Y., Hasegawa, C., and Kobayashi, H. 1994.
Immunohistochemical investigation of neuropeptides in the central
nervous system of the amphioxus, Branchiostoma belcheri.
Cell Tissue Res. 277: 279–287. 

Vallet, P.G., Ody, M.G., and Huggel, H. 1985.
Étude ultrastructurale du neuropore d’amphioxus adulte (Branchiostoma
lanceolatum Pallas).
Rev. Suisse Zool. 92: 845–849. 

von Kupffer, C. 1906.
Die Morphogenie des Centralnervensystems.
In Handbuch der vergleichenden und experimentellen
Entwickelungslehre der Wirbeltiere. Edited by O. Hertwig.
Band 2, Teil 3. Verlag von Gustav Fischer, Jena. pp. 1–272. 

Watanabe, T., and Yoshida, M. 1986.
Morphological and histochemical studies on Joseph cells of amphioxus, Branchiostoma
belcheri Gray.
Exp. Biol. 46: 67–73. 

Welsch, U. 1968a.
Die Feinstruktur der Josephschen Zellen im Gehirn von Amphioxus.
Z. Zellforsch. Mikrosk. Anat. 86: 252–261. 

Welsch, U. 1968b.
Beobachtungen über die Feinstruktur der Haut und des äu?eren Atrialepithels von Branchiostoma lanceolatum Pall.
Z. Zellforsch. Mikrosk. Anat. 88: 565–575. 

Welsch, U. 1968c.
Über den Feinbau der Chorda dorsalis von Branchiostoma lanceolatum.
Z. Zellforsch. Mikrosk. Anat. 87: 69–81. 

Wickstead, J.H., and Bone, Q. 1959.
Ecology of acraniate larvae.
Nature (Lond.), 184: 1849–1851. 

Willey, A. 1894.
Amphioxus and the ancestry of the vertebrates.
MacMillan, New York. 

Yasui, K., Tabata, S., Ueki, T., Uemura, M., and Zhang, S.-C. 1998.
Early development of the peripheral nervous system in a lancelet species.
J. Comp. Neurol. 393: 415–425. 

Yu, J.K., Holland, N.D., and Holland, L.Z. 2002.
Am amphioxus winged helix/forkhead gene, AmphiFoxD: insights into vertebrate
neural crest evolution.
Dev. Dyn. 225: 289–297. 


Appendix A  

List of abbreviations used in figures. 

AC Anadón’s GABAergic cells
acf atriocoelomic funnels
alm anterolateral migrated cell group
ansm anal sphincter muscle
antRo anterior Rohde cell
antRoax anterior descending Rohde axons
ap atrial papillae
aRB anterior Retzius bipolar cells
atsm atriopore sphincter muscle
avm anteroventral migrated cell group
bc buccal cavity
bo balance organ
cdQ corpuscles of de Quatrefages
chd notochord
cJo Joseph cells
cJohn Johnston cell
CNS central nervous system
comv ventral commissure
cRo Rohde cell
CV (larval) cerebral vesicle
DC dorsal compartment
DCC dorsal commissural cell
DCm (larval) dorsal compartment motoneuron
DR dorsal root cell
e endostyle
EC Edinger cell
elm external labial muscle
enc encapsulated nerve endings
ep epiphysis / pineal organ

fe frontal eye
gd gut diverticulum
gon gonad
Hp Hatschek’s pit
ibp inner buccal plexus
ilm inner labial muscle
inf infundibulum
intf internuncial fiber bundle
io infundibular organ
Köp Kölliker’s pit
lac large translumenal (= commissural) cell
lb lamellar body
lc lamellar cells   

LPN3s  third pair of large paired neurons
m1, m2, m3… myomere 1, 2, 3…
MC mid-commissural cell
Mg Müller’s glia (Schwann cell analogues)
mpf metapleural folds
mRoax median descending Rohde axon
n1, n2, n3… dorsal nerve 1, 2, 3…

nc nerve cord
ncmf notochordal motor fibers/tract
nRo nucleus of Rohde
NS nervous system
obp outer buccal plexus
oHe organ of Hesse
os optic stalk
pdm posterodorsal migrated cell group
ph pharynx
pig pigment cells of frontal eye
pinfn postinfundibular neuropile
pm pterygeal muscle
PMC primary motor center
postRoax posterior ascending Rohde axons
r1, r2, r3… rhombomere 1, 2, 3…
RB Retzius bipolar cell
rdors dorsal ramus
Rf Reissner’s fiber
rg radial glia
rvent ventral ramus
rventc ventral ramus, ventral cutaneous branch
rvisc visceral ramus
rvisca visceral ramus, ascending branch
rviscd visceral ramus, descending branch
s1, s2, s3… somite 1, 2, 3…
SD small dorsal cell
Sg Schneider’s glia
SM somatomotor cells
SM1 somatomotor cell type 1
smf somatomotor fibers/tract
ssf somatosensory fibers/tract
t telencephalon
tc optic tectum
tm trapezius muscle of atriocoelomic funnel
VC ventral compartment
vcav ventricular cavity
VCm (larval) ventral compartment motoneuron
VL ventral longitudinal cell
VM visceromotor cell
VM1, VM2 visceromotor cell type 1, 2
vp velar plexus
vsc ventral spindle-shaped cell group
vsf viscerosensory fibers/tract
vsm velar sphincter muscle
© 2005 NRC Canada




CotA AmphPDFs - NervSysStrDevEv
140706 - 1118    




My comments

This 29 page paper raised a number of issues I'd like to come back to.  Below is a brief, initial survey: 

CotA AmphPDFs - NervSysStrDevEv
140702 - 2032    

@

The most recent reviews on amphioxus
are by Ruppert (1997), for the general anatomy of the
animal as a whole, and Nieuwenhuys (1998), for the nervous
system.


corpuscles
of de Quatrefages (Fig. 3), putative mechanosensory
organs consisting of primary sensory cells enclosed in a capsule
(Baatrup 1982)


     The epithelia that line the buccal and atrial cavities, as
well as the organs embedded within them, are also supplied
with an extensive set of neural plexuses collectively known
as the atrial nervous system. Though the system is more or
less continuous, it is usually subdivided on the basis of the
organs it innervates; i.e., there are buccal, velar, gonadal, parietal,
pterygeal, pharyngeal, and endostylar subdivisions,
and so on (Bone 1961).


    The connection to the nerve cord via nerves 1–7
(1–8 according to Dogiel 1903; Kutchin 1913) is highly
asymmetrical. Nerves 3 and 4 on the left side are exceptional
in having contralateral branches that connect with the inner
buccal plexus on the right side. In addition, a subsidiary
branch from the contralateral branch of the left nerve 4 connects
to the right side of the velar plexus, while its left side
connects to nerve 5 by means of a caudal branch from that
nerve. This is all a consequence of the fact that the larval
mouth develops initially on the left side and is innervated
entirely by nerves emerging from the left side of the nerve
cord (Lacalli et al. 1999). The initial connections are then
retained during subsequent development, so the nerves are
dragged along as the mouth is repositioned.   



     Solitary receptors are widely distributed over the entire
epidermis but are most common in the region of the rostrum,
buccal cirri, and tail (Dogiel 1903; Franz 1923; Bone 1960b;
Stokes and Holland 1995a; Holland and Yu 2002). They
form small clusters in some instances (Sinnesknospen, Franz
1923; Schulte and Riehl 1977), especially along the buccal
cirri. The most common receptor cell types are referred to
by convention as types I and II (Schulte and Riehl 1977;
Bone and Best 1978).  
    Type I cells are primary sensory neurons
with an apical circlet of microvilli, a single cilium, and
a basal neurite. There are several subtypes, but all are probably
mechanosensors (Baatrup 1981; Lacalli and Hou 1999).
Their axons project to the CNS via the dorsal nerves; once
there, they travel along the cord in two fiber tracts, dorsal
and subdorsal in the terminology of Holland and Yu (2002),
which may correspond to the somatosensory and viscerosensory
tracts of Bone (1960a; see Fig. 4). The central axons
of type I cells reach considerable lengths, so an axon entering
the CNS via the first nerve can typically project caudally
to mid-spinal levels, at least in larvae (Holland and Yu
2002).    
    Little is known about the neurotransmitters released
by peripheral neurons, but there is evidence that at least
some type I cells are GABAergic (Anadón et al. 1998).   
    Type II receptors (Fig. 4) are secondary sensory cells with
synaptic terminals borne on short basal processes, usually
three per cell (Stokes and Holland 1995a; Lacalli and Hou
1999). Apically, they have a modified nonmotile cilium surrounded
by a collar of branched microvilli. This extensive
elaboration of the apical surface suggests a chemoreceptive
function, but essentially nothing is known for certain about
chemoreception in amphioxus, either in terms of structures
or physiology (Lacalli 2004). 

    refs: 
Anadón, R., Adrio, F., and Rodríguez-Moldes, I. 1998.
Distribution of GABA immunoreactivity in the central and peripheral nervous
system of amphioxus (Branchiostoma lanceolatum Pallas).
J. Comp. Neurol. 410: 293–307.   


Castro, A., Manso, M.J., and Anadón, R. 2003.
Distribution of neuropeptide Y immunoreactivity in the central and peripheral
nervous systems of amphioxus (Branchiostoma lanceolatum Pallas).
J. Comp. Neurol. 461: 350–361. 

Castro, A., Becerra, M., Manso, M.J., and Anadón, R. 2004.
Somatomotor system of the adult amphioxus (Branchiostoma
lanceolatum) revealed by an anticalretinin antiserum: an
immunocytochemical study.
J. Comp. Neurol. 477: 161–171. 


    The synaptic zones in each segment consist of two distinct
domains, the ventral and dorsal synaptic compartments
(Figs. 3, 4). Both utilize acetylcholine as a transmitter (Flood
1974). The ventral synaptic compartments are where the
deep, anaerobic, fast muscle cells receive their innervation.
The presynaptic motoneurons involved belong to a class of
cells that Bone (1960a) called somatomotor (SM) cells; they
may therefore also be called ventral compartment motoneurons.
They are found in the ventral parts of the grey matter
and have a tendency to cluster opposite the synaptic
contact zones, and each has a broad apical process connecting
it to the ventricular cavity. Some have internal vacuoles,
and this character, together with size and positional differences,
has been used to define several subtypes (Bone 1960a;
only one such type, the SM1 cell, is shown in Fig. 4). The
axons of the SM cells project laterally into the bundle of
somatomotor fibers adjacent to the synaptic zone of the ventral
compartment.
     The dorsal compartment is where the superficial, aerobic,
slow muscle cells of the myomeres receive their innervation.
The DC motoneurons are known from larvae (Lacalli and
Kelly 1999; Lacalli 2002a) but have not yet been identified
with certainty in adults. From the larval data, however, it
seems that the whole of the DC innervation along the nerve
cord may derive from motoneurons located in the anterior
cord at the level of somites 2–6 (see below). This is approximately
equivalent to the zone fated to become the IR of the
anterior cord, which extends from myotomes 2 to 4.


Thus, there are serially
(but not segmentally) repeated neurochordal synaptic
contacts at the base of the nerve cord. These are thought to
be cholinergic (Flood 1970), but their source within the
nerve cord has not been identified.


    GABA, neuropeptide Y, and several other neuropeptides
have been detected in various cells loosely classified as
interneurons (Uemura et al. 1994; Anadón et al. 1998; Castro
et al. 2003)


    Second, a novel class of interneurons, Anadón’s cells,
has been identified in the vicinity of the ventral expansion
of the central canal (Anadón et al. 1998; see Fig. 5).  
These are very small GABAergic cells interspersed between
the cell bodies of SM and VM neurons (cf. Figs. 4, 5).  
    Anadón et al. (1998) have suggested
that they might be comparable to the inhibitory Renshaw
cells of vertebrates.


     Whether by light microscopy (e.g., Edinger 1906; Franz
1923; Ekhart et al. 2003) or EM (Meves 1973), it is difficult
to discern much about the neuronal and glial cells of the anterior
vesicle, since most cells are small and rather densely
stained and have few visible distinguishing features. Franz
(1923, 1927) therefore concluded that the entire anterior vesicle
consisted only of glial cells. However, GABAergic
(Anadón et al. 1998) and serotoninergic neurons (Moret et
al. 2004) have since been identified in this region in adult
specimens


Then, just ventral to
the Joseph cells and surrounding the dorsal expansion of the
central canal, there are bilateral, longitudinal bands of neurons
immunoreactive for urotensin and FMRFamide (Uemura
et al. 1994), GABA (Anadón et al. 1998), neuropeptide Y
(Castro et al. 2003), and catecholamines (the catecholaminergic
population I of Moret et al. (2004)).


Firstly, the more posterior alm cells (slightly
rostral to the junction of myomeres 1 and 2; black squares in
Fig. 6) seem to correspond to the anterolateral serotoninergic
cells of Holland and Holland (1993) that were also observed
by Moret et al. (2004). Slightly more anterior (black circles
in Fig. 6) is another group of immunocytochemically identifiable
cells within the alm group. This is the catecholaminergic
population II of Moret et al. (2004). There is some
uncertainty about the exact positions of these two cell
groups, however. Moret et al. (2004) place them adjacent to
the rostral half of the second myomere. In an independent
immunocytochemical study, H. Wicht (unpublished data) localized
them more anteriorly, adjacent to myomere 1 and
thus within the confines of the alm group (see Fig. 6). Wicht’s
study did confirm, however, that both the catecholaminergic
population II and the anterolateral sertoninergic neurons
have long descending projections to the spinal cord. In retrograde
tracing experiments, Fritzsch (1996) found pairs of labelled
cells in late larvae that may correspond to the
anterolateral serotoninergic cells, even though he did not
specify their exact position, but Ekhart et al. (2003), in a
similar study in adults, did not find such cells. Assuming the
latter result is a false negative, the cells and projections appear
to be real; it is only their exact axial position that is a
matter of some uncertainty.   


Serotonin-containing neurons are absent in this region
(Moret et al. 2004), but a relatively large number of
GABAergic and peptidergic cells (Uemura et al. 1994;
Anadón et al. 1998; Castro et al. 2003) do occur. In addition,
there are four relatively large catecholaminergic cells (population
III of Moret et al. 2004) with translumenal processes
in the vicinity of the roots of the fifth dorsal nerves.   


     Despite the usefulness of the infundibular cells as anatomical
markers, there is no obvious transition in terms of
neuronal cell type at this point. Instead, cells of essentially
anterior character are found from the preinfundibular region
to the beginning of the PMC. “Anterior” here refers to cells
with irregular basal neurites that form repeated varicosities
containing mixed vesicle types and few, if any, synapses.
These are features that are generally associated with slow
transmission, often involving neuropeptides (Burns and Augustine
1995). Beginning in the PMC, most of the neurons
have well-defined axons and separate dendritic structures, either
arbors or spines (both occur), and synaptic junctions,
often with clear vesicles, predominate. This implies fast
transmission and aminergic or amino acid transmitters,
which is perhaps logical for neurons directly involved in the
locomotory control circuits.


     The LPN3s are thus the best candidates for neurons exerting
a direct controlling influence over both fast and slow
swimming, which appear to have a similar neuromuscular
basis in amphioxus and vertebrates (Bone 1989). Fast or escape
swimming occurs in response to sensory inputs, which
are a massive and redundant input to the VC system. The
VC system also receives synaptic input from fibers in the
postinfundibular neuropile and may be subject to additional
paracrine input as well, via fibers passing through the
neuropile, all of which provides an opportunity to modulate
the response to sensory stimuli.    
    In contrast, the slow system, which drives vertical
migration, is almost devoid of synaptic input. Besides
its link via junctions to the LPN3s, this pathway
seems to be mainly under the control of the PPN2s
mentioned above, a class of preinfundibular projection neurons
that make repeated junctional contacts with the axons
of the DC motoneurons.


     There are a number of specialized cell groupings at the
anterior end of the adult nerve cord that are not present in
early larvae, including various types of migrated cells described
above from the adult IR. Judging from the time that
the anterolateral serotoninergic cells first appear (Holland
and Holland 1993), these cell groupings probably develop in
the late larval phase or during metamorphosis. Despite the
proliferative activity this entails, the anteriormost region
fails to thicken as much as the rest of the cord, so the CV
progressively disappears as an externally recognizable zone.
     Of the late-developing cell groups, the dorsal (population I)
dopamine-containing cells reported by Moret et al. (2004)
are especially noteworthy. These are as dorsal and anterior
as one can get in the nerve cord, which is precisely where a
telencephalic homolog would be predicted to form if
amphioxus had one. For this and other reasons, Lacalli (2004)
suggested that the population I cells may represent a primitive
version of the olfactory bulb.











    








Comments