Introduction
The purposes of this review are to summarize newer information
about the structure and function of the adult choroid plexus (CP), focusing
on issues with clinical relevance, and to correct several misconceptions
in the literature. Particularly, we stress the predominant role of
the CP in secreting ventricular cerebrospinal fluid (CSF) that maintains
ICP and bulk flow down the neuraxis into the subarachnoid space. Our
emphasis will be on humans greater than age two since the function
of the CP in younger children awaits elucidation (Bateman and Brown,
2012). In this section, we underscore key points expanded upon in the
body of the text.
The living anatomy of the CP is very revealing. As aptly remarked by
Wei Zheng, “In the human brain at autopsy duringwhich the CSF is usually
completely drained, the choroid plexus can be seen to extend along
the floor of the lateral ventricles, hang down from the roof of the third
ventricle, and overly the roof of the fourth ventricle. The size of the tissue
in one ventricle when in a dry, condensed state appears to be nothingmore
than that of an index finger. These autopsy impressions,which
have over the years become the popular perception among many neuropathologists,
can be dreadfully misleading and may lead to misjudgments
of the function of this tissue. Recent advances in technology
have enabled amicro-video probe to be inserted directly into the lateral
ventricles. The vivid images illustrate that the live choroid plexus pervades
the ventricles, stretching and pulsating in concert with the heart
pulse. To better understand this, it may be helpful to make an analogy
between the choroid plexus and a fishnet. The fishnet occupies barely
one corner of the fishing boat; yet upon spreading in the water, it extends
to cover a large area. Likewise, in the life situation where the
brain ventricles are full of CSF, the choroid plexus expands to fill nearly
all the cerebral ventricles. Unlike the fishnet, however, the plexus tissue
possesseswell-developed brush-type borders, i.e., microvilli, on the apical
epithelial surface. These brush borders further protrude into the CSF
and increase the choroidal epithelial surface area, enabling rapid and efficient
delivery of the CSF as well as other material included with the
CSF secretion” (Zheng and Chodobski, 2005).
Notwithstanding recent definitive data, reference is stillmade (Tang
et al., 2014) to older erroneous histological data claiming that the CP
apical surface area is small and therefore not physiologically important
for mediating brain-wide molecular distribution (Pardridge, 2011).
However, in fact, as discussed below CP surface area is substantial,
~25 to 50% the size of the inner capillary surface area of the brain. This
provides an enormous surface area for performing amyriad of choroidal
transport functions.Moreover, the large surface area of brain capillaries
and CP epithelial cells also allowswater to diffuse freely and bidirectionally
between blood, brain and CSF. Sweet et al. (1950) first showed this
phenomenon in humans. Subsequently Oldendorf demonstrated in rats
that greater than 50% of plasma water exchanges with the water of
brain and CSF in a single pass of blood through the brain (Oldendorf,
1970). The exact mechanism of this huge bidirectional flux of water
through the brain capillaries and CP is unknown since in adults there
are no obvious aquaporins in brain capillary endothelium or on the
basal side of the choroidal epithelium (Speake et al., 2003). We also
know that secretion of CSF is volumetrically b1/100 that of CNS water
diffusional exchange (Bateman and Brown, 2012; Brinker et al., 2014).
The preponderance of CSF is formed by CP and consists of salts, micronutrient
vitamins, secreted proteins, hormones and water pulled along
passively to maintain osmotic equilibrium, as described below.
One group states that CP does not produce CSF (Oreskovic and
Klarica, 2010), a result inconsistent with over 60 years of data obtained
for animals and humans. Among the multiple experiments countering
this view, we cite three recent examples. First, in accordance with the
most powerful of the causality arguments of John Stuart Mill, the
method of commitment variation, often termed dose–response causality
in pharmacology, one would predict that more CSF would be produced
in CP hyperplasia (Smith et al., 2007) if the CP produces CSF. As
part of a definitive review of CP hyperplasia cases, Hallaert and colleagues
present as an example (Hallaert et al., 2012), a very wellstudied
case of a 3-year-old child with communicating hydrocephalus
in which the lateral CP, but not the third and fourth ventricular CPs,
were enlarged on MRI due to a genetic defect; with the external drainage
systemat the level of the foramen of Monroe, this 10 kg child copiously
produced 2 l of CSF per day! On removal of the two enlarged
lateral plexuses, the histological appearance of the tissue was normal
and the child's CSF overproduction was prevented. There is no doubt
that her CP secreted CSF in this case. Similarly, inmany other cases of increased
ICP (e.g., hydrocephalus) due to relative or absolute overproduction
of CSF, as discussed below, removal of the lateral CP tissues is
beneficial by substantially lowering the ICP.
Secondly, for decadeswe have also known that the CP containsmultiple
transporters capable of transferring Na+, Cl−, and HCO3
− from
blood into CSF across the CP — a typical secretory epithelium. Recently,
in one example of a sophisticated study in mice (Kao et al., 2011), the
investigators knocked out the CP transporter sla4a5 that transports
NaHCO3 into CSF. Sla4a5 exists only on the apical surface of CP, apparently
nowhere else in the CNS. These sla4a5 knockout (KO) mice had
a ventricular volume and CSF pressure only ~25% of normal. Moreover,
their CSF had a lower HCO3
− concentration, as expected. This study
clearly showed that the CP, with the aid of sla4a5 and other ion transporters,
produces HCO3
−-containing CSF. And thirdly, the CP epithelial
monolayer in an in vitro trans-well actively transports not only vitamin
C (Angelowet al., 2003) but also “CSF” fromthe basal to the apical (ventricular)
side against a considerable hydrostatic pressure gradient
(Hakvoort et al., 1998). Such fluid pumping against pressure clearly
mimics the in vivo situation of elevated ICP in hydrocephalus.
Another set of concepts needing revision is that drugs and nutrients
pass into and out of CSF mainly by passive forces largely dependent on
molecularweight, ionic charge and lipid solubility; and do not penetrate
deeply into the substance of the brain from CSF (Nagaraja et al., 2005;
Pardridge, 2011, 2012). These concepts need qualification in order to
promote precise understanding and sound therapeutic regimens, as
discussed below.
Anatomy of choroid plexus: structural–functional relationships
Choroid plexus ultrastructure is similar in the lateral, third and
fourth ventricles. There is a vascular network surrounded by a single
layer of roughly cuboidal epithelial cells that are joined by tight junctions
and make up the blood–CSF barrier (Fig. 1) (Smith et al., 2004).
The vascular endothelial cells differ from those forming the blood–
brain barrier in being fenestrated (Wolburg and Paulus, 2010). Evidence
indicates that endothelial fenestrations, in the kidney and CP, facilitate
the movement of fluid out of the capillaries (Ballermann and Stan,
2007). Between the endothelium and the epithelium is an area of stroma.
The choroidal epithelial cells, linked by tight junctions at the apical,
CSF-facing pole o