Organ Culture and Grafting of Choroid Plexus into the Ventricular CSF of Normal and Hydrocephalic HTx Rats

Abstract

Choroid plexus (CP) may aid brain development and repair by secreting growth factors and neurotrophins for CSF streaming to ventricular and subventricular zones. Disrupted ventricular/subventricular zone progenitors and stem cells lead to CNS maldevelopment. Exploring models, we organ cultured the CP and transplanted fresh CP into a lateral ventricle of postnatal hydrocephalic (hyHTx) and nonhydrocephalic (nHTx) rats. After 60 days in vitro, the cultured choroid ependyma formed spherical rings with beating cilia. Cultured CP expressed endocytotic caveolin 1 and apical aquaporin 1 and absorbed horseradish peroxidase from medium. Transthyretin secretory protein was secreted by organ-cultured CP into medium throughout 60 days in vitro. Fresh CP, surviving at 1 week after lateral ventricle implantation of nHTx or hyHTx did not block CSF flow. Avascular 1-week transplants in vivo expressed caveolin 1, aquaporin 1, and transthyretin, indicating that grafted CP may secrete trophic proteins but not CSF. Our findings encourage further exploration on CP organ culture and grafting for translational strategies. Because transplanted CP, though not producing CSF, may secrete beneficial molecules for developing brain injured by hydrocephalus, we propose that upon CP removal in hydrocephalus surgery, the fractionated tissue could be transplanted back (ventricular autograft).

INTRODUCTION

The multifunctional choroid plexus (CP) supports brain in diverse ways (1). CP appears in the seventh week of human gestation (2). Throughout life, the mammalian CP (mouse, rat, pig, and human) secretes/transports numerous hormones and growth factors into CSF (3–5). Here, the CP-derived trophins distribute by CSF flow to periventricular regions and subarachnoid space (6–8). As CNS develops perinatally (9), neurogenic regions in the dentate gyrus and subventricular zone (SVZ) require peptide signals to convert progenitors to neurons. From early development to adulthood (8), the choroidal protein secretion facilitates CSF-brain needs (10). Thus, brain development/maintenance depends on normal CSF composition and hydrodynamics (11).

CSF growth factors and neurotrophins promote proliferation, migration, and differentiation of neural stem/progenitor cells and neuroblasts. Lateral ventricle CP (mouse) secretes factors that augment proliferation of ventricular (VZ) and SVZ stem and progenitor cells (12). Ontogenetic transcriptome data (13) inform on rat choroidal transporters supplying trophins to CSF-brain (14). Disrupted VZ/ependyma in hydrocephalus upsets CNS ontogeny (15). Abnormal neurogenesis in VZ/SVZ decreases progenitor and stem cells (15–17). Distorted CSF flow/composition provides inadequate signals to germinal niches that disturb brain development (18).

The adult brain surrounded by CSF receives peptides/proteins from CP to modulate neuronal/glial metabolism (1, 19). Perinatal neurogenic niches (e.g., dentate gyrus and SVZ) contain stem cells/progenitors for generating neurons. In brain trauma, stroke, neurodegeneration, and infection, there is a greater need to repair/replace neurons. In severe central dyshomeostasis, the endogenous CP may not be able to supply sufficient trophic molecules to niches and injury sites for generating new brain cells. Such postulated deficiency of neurotrophic factors has led to the introduction of CP cells into human CSF-brain (20) by transplant or encapsulation in an attempt to alleviate neurodegeneration. Altogether, these considerations prompt translational investigation on how CP trophic power can be augmented when needed.

Neurosurgeons and neuroscientists proffer the therapeutic use of an exogenous “CSF cocktail” of neuropeptides or transplanted CP as in animal (pig) stroke models; these strategies may boost brain recovery (21, 22). This was prompted by findings that supplemental CSF growth factors/neurotrophins induce better recoveries following ischemia and other disorders (7, 23, 24). CP epithelial cell transplants into fourth ventricle CSF confer protection against ischemia in rats (25).

Huntington and Parkinson neurodegeneration models display less brain damage when encapsulated CP is implanted to counter chemical injury induction (26–28), most likely through secretion of brain derived neurotrophic factor (29), bFGF or TGF-β1 (30, 31). Both allo- and xenotransplants of CP attenuate pathologic sequelae postischemia and neural degeneration. CP epithelial cells transplanted into brain of APP/PS1 mice reduced β-amyloid (Aβ) and improved Alzheimer-like symptoms (28). Encapsulated choroid cells placed in Aβ-pretreated rat brains protected against hippocampal long-term memory dysfunction (32). Collectively, the data show benefits from supplemental factors.

Briefly, the reasoning is that grafted fresh CP (similar to organ-cultured CP) may secrete trophic proteins into hydrocephalic CSF. Such ventricular secretion into organisms with hydrocephalus would be an additional source of central neurotrophic factors; this rationale is drawn from experimental observations that supplementary growth factors to CSF augment brain recovery from injury (7). Grafted CP, by not forming CSF, might improve neuron microenvironment without increasing CSF volume. This could stimulate translational neurotrophic strategies to transplant the CP back into hydrocephalic patients whose CP was surgically removed to reduce CSF pressure.

We had 2 aims: to develop an organ culture of CP (explant) for maintaining CP ependyma intact for several weeks to enable morpho/functional studies of noninnervated, nonvascularized CP (to set the basis for future cultured CP grafting) and to transplant fresh CP tissue into the ventricles of newborn HTx rats, with and without fetal-onset hydrocephalus, for analyzing expression of transthyretin (TTR) and aquaporin 1 (AQP1) markers in the host and grafted CP.

MATERIALS AND METHODS

The HTx Hydrocephalus Rat

The hereditary hydrocephalus HTx rat was used for the explants and transplants. The HTx strain (from Dr Hazel Jones, University of Florida, Gainesville) has been long bred in the Animal Facility at the Instituto de Anatomía Histología y Patología, Facultad de Medicina, Universidad Austral de Chile. Housing and processing of the HTx animals were performed according to the regulations of the National Research Council of Chile. Nonhydrocephalic (nHTx) littermates develop normally. All protocols were approved by the Ethics Committee of Universidad Austral de Chile.

Postnatal (PN1) and PN7 hydrocephalic rats were distinguished from nHTx littermates by an overtly domed head and transillumination of the neonatal cranium. Hydrocephalic and nonhydrocephalic phenotypes were confirmed by microscopic observation of the ventricles in the brain slices collected for CP grafting, and in the tissue sections of the HTx brain hosting the grafts. nHTx littermates develop normally and thus serve as useful controls.

In contrast to several genetic hydrocephalus models for mice, there is just one rat model for hereditary hydrocephalus—the hyHTx. However, there is abundant information for normal CP-CSF transport/hydrodynamics in rats (but not mice). These factors, along with the wealth of information on hyHTx hydrocephalus physiopathology (33), contribute to evaluating various rat CP-CSF functions associated with our HTx grafting.

Organ Culture of HTx Choroid Plexus

The CP of both lateral ventricles of PN1 nHTx rats was excised under a microscope; the CP stalk was sectioned to obtain free-floating individual villi. Each villus was ∼50- to 100-µm wide and ∼500-µm long (Fig. 1B′). Villi were cultured for 2 months in DMEM/F-12 (Sigma D5648 and N6760). They were supplemented with 10% fetal bovine serum (Invitrogen) and 20 μg Ara-C (cytosine arabinoside, fibroblast inhibitor, Sigma C6645) at 37°C and 5% CO₂ (Fig. 1A, B, B′). Culture medium was renewed twice weekly. Explants were monitored by phase-contrast microscopy daily, from 1 to 22 days in vitro (DIV); thereafter, once a week. Attention was paid to cilia beating throughout the culturing; this complex ciliary function reflected viability of the choroid ependyma. A few CP explants were collected during the first, second, and third weeks in culture; most were procured after 60 DIV. Explants were fixed in Bouin’s fluid, embedded in paraffin, and serially sectioned. The conditioned medium was collected twice weekly for 7 consecutive weeks; the media were pooled and labeled as a 1-week sample. The 7 samples were immunoblotted (TTR).

FIGURE 1.

Explanted choroid plexus (CP), organ cultured for 2 months, remains morphologically and functionally differentiated. (A–C) CP tissues of the lateral ventricles (LV) of nonhydrocephalic postnatal 1 (PN1) and PN7 nonhydrocephalic (nHTx) rats were dissected out and organ cultured for 2 months. cc, corpus callosum (A) scanning electron microscopy (SEM). (B) Frontal section of the telencephalon of a PN7 nHTx immunostained for aquaporin 1 (AQP1) showing the selective immunoreaction of CP. (B′) Portion of a CP villus similar to those used for culture. c, core (interstitial). (C) In vivo phase contrast of explanted CP organ cultured for 60 days in vitro (60 DIV). (C′, C″) After 60 DIV, the ependymal cells of CP remain organized as a single layer forming ependymospheres and displayed cilia beating. Scale bars: A, 170 μm; B, 520 μm; B′, 120 μm; C, 40 μm; C′, 19 μm; C″, 13 μm.

Incubation of Organ-Cultured CP With Horseradish Peroxidase

After 60 DIV, CP explants were exposed for 1 minute to 1% horseradish peroxidase (HRP) (Grade VI, Sigma) diluted in phosphate buffered saline (PBS) pH 7.4. After PBS washing, they were fixed with Bouin’s, embedded in paraffin and processed for histology and immunocytochemistry with anti-HRP.

Transplants of Fresh HTx CP

Donor PN7 male nHTx rats were anesthetized and the lateral ventricle CPs collected in saline. Under a dissecting microscope, the CP stalk was sectioned for free-floating individual villi; each villus was ∼50- to 100-µm wide and ∼500-µm long. Simultaneously, PN7 male nHTx (n = 4) and PN7 male hyHTx (n = 4) rats were anesthetized with ketamine (40 mg/kg) plus acepromazine (100 mg/kg). Each recipient rat was grafted with 3–4 CP villi in 2 µL saline. The head of the operated rat was immobilized in a plastic cast for PN7 rats. A 21-Gauge needle connected through a cannula to a Hamilton syringe (containing fresh CP) was inserted into the left lateral ventricle, 0.5 mm posterior to bregma, 1.5 mm lateral to the sagittal suture, and 1.5 mm below the meninges. Five minutes after grafting, the cannula was slowly removed while injecting 2 µL saline to prevent the grafted CP from flowing back out of the ventricle. Animals recovered from surgery before being returned to the litter. Grafted rats were kept under constant temperature and photo regime (light: dark 12:12). After 1 week of grafting, the host brain was fixed by vascular Bouin’s and paraffin-embedded for serial sectioning.

Choice of Markers to Assess Key CP Functions

The hallmark protein to gauge choroid epithelial metabolism is TTR. Approximately, 40% of protein synthesis by CP is TTR, normally secreted into CSF. AQP1 at the apical membrane helps move water between CP and CSF (34). AQP4 is expressed by ependyma lining the ventricles (35). Immunohistochemically, AQP4 distinguishes the ependyma of the ventricular CSF-brain interface from CP ependyma (blood-CSF barrier). Caveolin 1 is a scaffolding protein in membrane microdomains (caveolae) modulated by signaling hormones/growth factors in endo- and transcytosis (36). HRP (45-kDa protein) marks intra- vs paracellular transport routes (37).

Immunohistochemical Procedures

Five-micrometer-thick sections of organ-cultured CP, and brain transplanted with CP, were processed with streptavidin/biotin (Vectastain Kit; Vector, Serva, Heidelberg, Germany) and diaminobenzidine (3,3-diaminobenzidine tetrahydrochloride; Sigma, St. Louis, MO). Primary antisera included: anti-TTR (rabbit) (Santa Cruz Biotech, sc-13098, 1:100); anti-AQP-1 (rabbit) (Abnova, pab27167, 1:500); and anti-HRP, polyclonal (rabbit) (Instituto de Histología y Patología, Universidad Austral de Chile), 1:1000. Antibodies were diluted in 0.1 M Tris, pH 7.8, 0.7% nongelling seaweed gelatin, lambda carrageenan, and 0.5% Triton X-100 (Sigma). Incubation was for 18 hours at room temperature. Primary antibody omission during incubation was the immunoreaction control.

For double immunofluorescence, sections were incubated overnight (16 hours) at room temperature with primary antibodies against: TTR (rabbits) (Santa Cruz Biotech, sc-13098), 1:100; caveolin 1 (rabbit) (Santa Cruz Biotech, sc894) 1:100; AQP1 (rabbit) (Abnova, pab27167), 1:500; AQP4 (rabbit) (Abnova, pab20767) 1:500; and HRP, polyclonal (rabbit) (Instituto de Histología y Patología, Universidad Austral de Chile), 1:1500. For the combinations TTR/caveolin 1 and AQP1/AQP4, the antibodies were used sequentially.

After incubating with primary and secondary antibodies, the sections were washed with Tris pH 7.8, incubated for 1 hour with Alexa-488-labeled antirabbit IgG and Alexa-594-labelled antirabbit IgG (1:500) (Invitrogen, Carlsbad, CA). All antibodies were diluted in Tris pH 7.8, containing 0.7% nongelling seaweed carrageenan (Sigma) and 0.5% Triton X-100 (Sigma). Slides were coverslipped (Vectashield, Dako, Barcelona, Spain) and inspected under epifluorescence for colocalization with multidimensional software (AxioVision Rel, version 4.6; Zeiss, Aalen, Germany).

Immunoblotting of Conditioned Medium Using Anti-TTR

Conditioned medium was collected weekly for 7 weeks. Additionally, an NH₄HCO₃ extract of fresh CP from PN7 nHTx was obtained. Protein content was determined by Bradford’s method. Samples of 20 μg protein of conditioned media and the culture medium were run in parallel in SDS-polyacrylamide gel electrophoresis. Blotting was done thrice. Immunoblotting of CP extract (60 μg protein) controlled for anti-TTR specificity for the 14- and 45-kDa forms.

Proteins were transferred to nitrocellulose membranes. To block nonspecific binding, blots were saturated with 5% nonfat milk in 0.1 M PBS containing 0.15 mM NaCl and 0.1% Tween-20 (Sigma, Madrid, Spain), for 2.5 hours. Anti-TTR developed in rabbits (Santacruz Biotech, sc-13098),1:500, was used as primary antibody; incubation was overnight. Goat antirabbit IgG labeled with HRP (Sigma), 1:50,000, was the secondary antibody, for 1 hour. Incubations were at room temperature and in darkness. Immunoreactive polypeptides were detected by enhanced chemiluminescence (SuperSignal West Pico Chemiluminescent Substrate, Thermo Scientific, Waltham, MA). Molecular weight standards were 10–250 kDa (Bio-Rad, Hercules, CA). Immunoblots were digitalized and linear densitograms obtained with UN-SCAN-IT software (Silk Scientific, Orem, UT).

Scanning Electron Microscopy

The brains of 2 PN7 nHTx rats (different from those used for immunohistochemistry) were quickly dissected out. Two coronal cuts were made through the frontal and occipital horns of the lateral ventricles. One percentage paraformaldehyde in 0.1 M cacodylate buffer, pH 7.4, was gently subperfused into the exposed brain cavities with a microliter syringe. After 20 minutes, ∼1-mm-thick coronal slices were cut and immersed in the fixative for 2 hours at room temperature. After buffer washing, they were fixed in 1% OsO₄ in 0.1 M cacodylate buffer for 2 hours at 4°C. Tissue slices were dehydrated in a graded series of ethanol, air-dried, and ion-coated with gold for scanning electron microscopy.

RESULTS

Organ Culture of CP

Morphology

During the first week in culture, the ependymal cells of CP (i.e., epithelium) and underlying blood capillaries had a normal spatial relationship. After 2–3 weeks, the choroid villi became ovoid/spherical and the capillaries attenuated. Choroid cells remained organized into a single layer (Fig. 2D). After 60 DIV, the vasculature was sparse. Choroid cells remained joined, forming spherical, and ring-shaped structures (Figs. 1C and 2E, I). Paraffin sections of 60 DIV explants revealed that choroid cells formed a continuous layer (∼15-µm height) surrounding a core with cell debris and a few eosinophilic cells, but no blood vessels (Fig. 2E, I, J). The apical plasma membrane appeared as a thick eosinophilic layer, ∼3 µm thick, most likely corresponding to the numerous microvilli, a distinct feature of the CP ependyma in the living animal (Fig. 2I′).

FIGURE 2.

Explanted choroid plexus (CP), organ cultured for 2 months, remains morphologically and functionally differentiated. (A) Frontal section of the telencephalon of a postnatal 7 (PN7) nonhydrocephalic rats (nHTx) immunostained for transthyretin (TTR), showing the selective immunoreaction of CP (arrow). (B) Detailed magnification of previous panel showing TTR throughout the cytoplasm of choroid ependyma. (B′) In the CP, aquaporin 1 (AQP1) is exclusively located at the apical plasma membrane domain (arrow). (C) Cross-section of CP villus processed for double immunofluorescence; AQP1 (light green) localizes in the apical membrane domain and TTR (red) as cytoplasmic granules. (D, D′) CP organ cultured for 19 days. Choroid cells remain organized into a single layer and express TTR. (E–G) Lateral ventricle CP of PN7 nHTx rats organ cultured for 2 months. Ependymal cells of CP explant remain organized as a single layer (E, arrow) and continue to express TTR in cytoplasm (F, arrow) and AQP1 at apical plasma membrane (G, arrow). (H, H′) CP explants secrete TTR into culture medium. Western blot of secreted proteins using anti-TTR (higher molecular weight; kD in top row, H) The conditioned medium was collected weekly for 7 weeks. A protein extract of CP and fresh culture medium (cm) without explants were used as control. ROD, relative optical density (I–J′) CP organ cultured for 2 months retains endocytic properties. Adjacent sections of CP were processed for H&E stain; arrow points to membrane exposed to cm (I, I′), and for horseradish peroxidase (HRP) immunocytochemistry (J, J′). Cultured CP exposed to HRP in the medium for 1 minute, concentrated the marker at the apical domain (full arrow) and in cytoplasmic structures (broken arrow). Scale bars: A, 500 μm; B, 25 μm; B′, 30 μm; C, 15 μm; D, 40 µm; D′, 60 µm; E, 30 μm; F, 30 μm; G, 25 μm; I, 20 μm; I′, 12 μm; J, 20 μm; J′, 12 μm.

Fresh CP strongly expressed TTR throughout the cytoplasm and AQP1 in the apical plasma membrane (Fig. 2A–C). Double immunofluorescence depicted TTR as cytoplasmic granules and AQP1 at the apical membrane (Fig. 2C). During early weeks in culture (Fig. 2D′) and after 60 DIV, choroid cells displayed TTR in the entire cell body and AQP1 in the apical membrane (Fig. 2F, G).

Physiology

Throughout the 60 DIV, the functional capacity of CP explants was evaluated by analyzing TTR secretion, HRP-absorptive capacity and cilia beating.

Secretory Capacity

Fifty-microliter samples of culture medium (control) and conditioned medium collected weekly during 7 weeks of organ culture were run in parallel for immunoblotting with anti-TTR. The 14 and 40 kDa TTR forms were detected in all samples (Fig. 2H); immunoreactive band intensity was determined by densitometry. As the medium contains fetal bovine serum, the anti-TTR used in the blots recognized both the plasma TTR and the CP TTR secreted into the culture medium. The density of the 14-kDa band of the culture medium was regarded as baseline and given a value of 1.00 optic density unit. The 14-kDa band density of all conditioned media was higher than the culture medium; the difference (58%–92%) was regarded as the TTR secreted by CP explants (Fig. 2H, H′). TTR secretion into conditioned medium was uniform for 7 weeks.

Absorptive Capacity

Sixty DIV CP cells were exposed for 1 minute to HRP added to the culture medium; the tracer strongly labeled the apical plasma membrane and numerous dots were distributed throughout the cytoplasm (possibly vesicles or phagosomes) (Fig. 2I–J′).

Cilia Beating

Explants were monitored by phase-contrast daily from 1 to 22 DIV; thereafter, once a week. As CP explants organized into ovoid/spherical structures lined by a single layer of ependyma ∼15-µm high (akin to ependymospheres, Fig. 1C, C′), the synchronized cilia beating became readily distinguishable. (See the recording at 60 DIV [Fig. 1C″ and the Supplementary Data Video, which shows in vivo phase contrast of explanted CP organ cultured for 60 DIV. The ependymal cells of CP remain organized forming ependymospheres displaying cilia beating [asterisk]. For orientation, see Fig. 1C, C″]).

Grafting of Fresh CP Tissue

nHTx Rats as Host

Fresh CP tissue from PN7 male nHTx rats was grafted into a lateral ventricle of PN7 male nHTx animals (Fig. 3A–D). In all rats, the grafted tissue was distinguishable from host rat CP after 1 week. The ependymal cells of grafted CP were cylindrical, displaying a nucleus with abundant euchromatin; they formed a continuous layer. Morphology was similar to the host CP (Fig. 3E, E″). A consistent feature of the grafted CP is a core of villi devoid of vasculature and containing eosinophilic cells (large nucleus), probably macrophages (Fig. 3E″). At variance, the intact choroid villi core of the host had blood vessels and a distinct perivascular space (Fig. 3E′). In all grafted rats, the transplanted CP tissue remained in the host lateral ventricle, close, or apparently attached to the external lateral wall (Figs. 3E, E″ and 4A, A″).

FIGURE 3.

Grafting of choroid plexus (CP) from a postnatal 7 (PN7) nonhydrocephalic (nHTx) rat (A–C) into a lateral ventricle of a PN7 nHTx rat (D, E–E″). (A) Scanning electron microscopy. (B) Head of a PN7 nHTx rat. (C) CP of lateral ventricle immunostained for transthyretin (TTR). (D) CP grafting into a host nHTx rat. (E) Magnification of panel D. The asterisk and broken arrow indicate site of entry of the cannula carrying the graft. After 1 week of transplantation (at PN14), the grafted CP was readily distinguishable from host CP. (E′–E″) The choroid cells of grafted CP display a morphology similar to host CP. At variance, the core of the choroid villi of the host and the grafted CP differed; the former was mostly occupied by blood vessels (E′, arrows), and the latter was devoid of vasculature, and contained eosinophilic cells with a large nucleus, probably corresponding to macrophages (E″, asterisk). Scale bars: A, 170 μm; C, 230 μm; D, 640 μm; E, 120 μm; E′, 14 μm; E″, 16 μm.

After 7 days, the grafted CP expressed AQP1 at the apical plasma membrane. AQP1 expression pattern in the host and grafted CP differed. In host CP, the apical pole of the ependymal cells with AQP1 channels was domed. At variance, the ependyma of grafted CP AQP1 appeared to form a smooth continuous layer (Fig. 4A, A″, B). The “broken-line” vs the “full-line” (graft) pattern of AQP1 distribution at the apical domain likely reflects the density/distribution of the choroid microvilli. In host CP of nHTx, TTR distributed throughout the ependyma (Fig. 4C). Although grafted CP from nHTx expressed TTR, the immunoreaction intensity was lower than host CP (Fig. 4C, D).

FIGURE 4.

Immunoreaction pattern for aquaporin 1 (AQP1) and transthyretin (TTR) allows delineation of grafted choroid plexus (CP) from host CP. (A–A″) Double immunofluorescence for AQP1 (red) and AQP4 (green). After 7 days, the grafted CP expressed AQP1 apically, similar to host CP. The pattern of AQP1 expression in the host and grafted CP differed. In the host CP, the AQP1 and apical cell pole had a dome shape (broken-line appearance) (A, A′, full arrow), whereas in the grafted CP, the AQP1 appeared as a smooth continuous layer (full-line appearance) (A, A″, full arrow). AQP4 was expressed by the multiciliated ependyma lining the lateral ventricle (A–A″). Here, in the ventricular wall, AQP4 was localized at the basolateral domain (A′, A″, broken arrow). (B) Immunoperoxidase staining for AQP1 of host CP; Nomarski optics. AQP1 is exclusively localized in the apical plasma membrane domain (arrows). Asterisk, cell body of choroid cells; bv, blood vessel of the core of the villus. (C, D) Although grafted CP expressed TTR, the immunoreaction intensity was significantly less than in host tissue. Scale bars: A, 80 μm; A′, 40 μm; A″, 40 μm; B, 14 μm; C, 15 μm; D, 15 μm.

hyHTx Rats as Host

CP tissue (fresh) from PN7 male nHTx rats was grafted into a lateral ventricle of hydrocephalic PN7 male hyHTx littermates. After 1 week, the transplanted CP displayed features different from host CP (Fig. 5A, A″). While the latter was vascularized, the former was not. In one operated rat, numerous ovoid cells with macrophage-resembling morphology were attached to the graft surface (Fig. 5A″). In another transplanted rat, the grafted CP was organized as a solid mass of ependymal cells with a core containing macrophage-like cells (Fig. 5C, C′). Throughout the serial sections, the grafted CP was always seen lying free in the dilated ventricle, in some cases, close to the medial wall and in another in the ventral horn (Fig. 5A, A″, C). Grafted CP was never seen in the Monroe foramen in the third ventricle. In the host and in grafted CP, TTR distributed throughout the cytoplasm of the choroid cells, although in the latter, the intensity of the immunoreaction was weaker (Figs. 5B, B″ and 6E, F). As in nHTx rats, the hyHTx rats had AQP1 distinctly expressed in CP of the lateral/third ventricles, whereas AQP4 was expressed in the multiciliated ependyma of the ventricles (Fig. 6A, B). In host CP, the CSF-proximate pole of the ependyma, containing apical AQP1 channels, was domed shape. This produced the “broken-line” pattern (Fig. 6C, inset). At variance, in the grafted CP, AQP1 continuously outlined the apical domain (“full-line” pattern) (Fig. 6D, inset). Caveolin 1 was expressed by host CP cells and blood vessels (Fig. 6E′). In the former, anticaveolin 1 reacted with numerous vesicles located in the supranuclear region of the cytoplasm (Fig. 6E′, inset). In the grafted CP, immunoreactive caveolin 1 appeared as vesicles and masses irregularly distributed in the choroid cells (Fig. 6F′, inset).

FIGURE 5.

Grafting of choroid plexus (CP) from a postnatal (PN7) nonhydrocephalic rat into a lateral ventricle of a PN7 hydrocephalic (hyHTx) rat. (A, B) H&E and immunoperoxidase staining for transthyretin (TTR) of a frontal section through the telencephalon of a PN14 hyHTx rat transplanted with CP at PN7. Notice the large dilatation of the lateral ventricles (LVs). The 2 boxes show the location of the host and grafted CPs. Inset: nontransplanted PN14 hyHTx showing ventriculomegaly similar to the transplanted hyHTx. Immunostaining for TTR. (A′–B″) After 1 week of transplantation, the host CP (A′, B′) was distinguishable from grafted CP (A″, B″). Choroid cells of the host CP (A′, inset) display morphology similar to the grafted CP (A″, inset). The core of the host CP was occupied by blood vessels (A′, arrows; inset, arrows); grafted CP was devoid of vasculature (A″, inset, asterisk), and macrophages were attached to the villus surface (broken arrows in A″ and inset). (B′, B″) Section adjacent to that shown in A′, A″, immunostained for TTR. Both the host and the transplanted CP displayed selective immunoreactivity; TTR distributed evenly in the cytoplasm of the choroid cells (insets). (C–D) hyHTx rat grafted with CP (the same experiment, different animals as that shown in A–B″). The grafted CP appears free in the ventral horn of the dilated LV (C). The grafted CP was disorganized, appearing as a mass with macrophages in the core (C′). The CP cells express TTR (D). Scale bars: A, B 830 μm; A′, 90 μm, inset 30 µm; A″, 90 μm, inset 30 μm; B′, 90 μm, inset 30 μm; B″, 90 μm, inset 30 μm; C, 80 μm; C′, 30 μm; D, 30 μm.

FIGURE 6.

In the grafted choroid plexus (CP) and the CP of the host hydrocephalic (hyHTx) rat, the immunoreaction patterns of aquaporin 1 (AQP1), transthyretin (TTR), and caveolin 1 are different. (A, B) Double immunofluorescence for AQP1 (green) and AQP4 (red). In hyHTx rats, AQP1 is expressed in the CP of third ventricle (IIIV, full arrow) and lateral ventricles (LV, full arrow), and AQP4 is expressed in the multiciliated ependyma (broken arrows). (C, D) Expression of AQP1 in the CP of the lateral ventricle of a host hyHTx rat (C) and in grafted CP (D). In host CP, the apical cell pole containing AQP1 channels has a dome shape; this produces a “broken-line” pattern (C, inset, arrow). In the grafted CP (D), AQP1 appears as a smooth continuous layer (full-line appearance) (D, inset, arrow). (E, F) Expression of TTR in CP of the LV of a host hyHTx rat (E) and in grafted CP (F). In the host CP, TTR appeared distributed throughout the cytoplasm of all choroid cells (E), whereas in the grafted CP, the intensity of the immunoreaction varied along the choroid ependyma (F). (E′, F′) Expression of caveolin 1 in the CP of a host hyHTx rat (E′) and in grafted CP (F′). In the host CP, caveolin 1 was strongly expressed by all choroid cells (E′) in numerous vesicles located in the supranuclear cytoplasm (E′, inset, arrow). It was also expressed in blood vessels (broken arrow). In grafted CP, caveolin 1 was heterogeneously expressed along the CP ependyma (F′); within choroid cells, it appeared as dots (inset, arrow) or irregular masses. Scale bars: A, 120 μm; B, 170 μm; C, 100 μm, inset 60 μm; D, 100 μm, inset 50 μm; E, 100 μm; E′, 100 μm, inset 10 µm; F, 100 μm; F′, 100 μm, inset 15 µm.

DISCUSSION

CP: A Promising Player in Innovative CSF-Hydrocephalus Research

Primary choroid ependyma adapts to microenvironments and can be manipulated into multiple configurations with distinct gene profiles (38). Ide et al pioneered transplanted choroid epithelial cells to facilitate axonal regeneration (39). There is burgeoning interest in using the neurotrophic potency of CP to reclaim lost neural functions; for example, Parkinson patients have been safely (although not yet efficaciously) implanted (striatum) with porcine CP embedded in immunoprotective capsules (20). Increasing attention to CP transplants identifies CSF as a valuable trophic system to expedite neurogenesis and brain healing in congenital hydrocephalus (40). CSF as a transplant site confers the advantage of being proximate to perinatal neurogenic targets (VZ).

We explored explanted CP (in vitro) and transplanted CP (in vivo) for retaining capacity to synthesize or release proteins over a long period. Successful pursuit now encourages future analysis of CP grafted into ventricular CSF, as cultured explant or fresh CP (supplemental). CP grafting opens new vistas of opportunities for CSF-CNS translational research.

In CP from nHTx rats, we explored properties of cultured explants and transplants of fresh CP into CSF, hydrocephalic animals, and nonhydrocephalic counterparts. Our long-term goal is to generate CP-CSF data in an ependyma-brain model for clinical translation to congenital hydrocephalus. CP, having an important role in CSF homeostasis, is a potential translational tool to restore ependymal/subventricular damage. Our short-term aim is to devise CP preparations for analyzing key aspects of cell and CSF transport biology. Such information may eventually help evaluate the benefits of cultured CP (possibly with modified RNA) placed into CSF of hydrocephalus hosts.

Organ Culture of CP: Evidence for HRP Absorption and TTR Secretion

Primary cultures of CP ependyma to examine secretion of proteins, absorption, transcytosis, and serotonin receptors have been widely explored (25, 41–46). By comparison, there are few reports dealing with organ culture of CP (47, 48).

Earlier CP explant investigations demonstrated secretion of epithelial aposomes/exosomes containing proteins fragmented from the apical membrane (49, 50); vesicle distribution via CSF to brain may occur. Recent explant studies emphasize: inflammation-sensitive choroid cells (51), knock-in proteins for circadian analyses (52), and utility of lipopolysaccharide-stimulated vesicle formation in CP to inform on neuroinflammation (53). We can now build upon the initial step (i.e., the culture phase) to an HTx hydrocephalus model that would permit the ability of the CP explants to rectify CSF system damage.

We found that organ-cultured CP from nHTx rats performed well for 2 months. Cultured CP displayed functions such as synchrony of cilia beating, absorption of marker proteins from culture medium, secretion of proteins, and expression of AQP1 at the apical plasma membrane. This is evidently the longest in vitro CP viability reported.

Ciliary Activity

Many ependymal cells in CP are endowed with 9 + 2 (microtubule configuration), with 20-µm-long kinocilia (54–56). A tuft of cilia emerges from the central region of each choroidal cell and projects to CSF. The number of choroidal cilia forming the tufts varies: 11–16 cilia/cell (monkey and rabbit), 15–18 in chick embryos (54), and 30–40 in the frog and salamander (55). Frog CP cilia beat ∼13 c/s. CP cilia stir CSF within 100 µm of the cell surface (55). Proper mixing of nascent and bulk CSF likely allows choroid cells to sense CSF (55, 56).

Cells extending a cilium, or a few primary cilia, into CSF are extant in CP (57, 58). In adult mice, CP ependymal cells have a small tuft of motile cilia or one/few primary cilia (57). CP cells with primary cilia are likely involved in osmosensation, chemosensation, transependymal fluid transport, and inflammatory phenomena (58). Considering the complex molecular structure of motile cilia and their multiple processes, our observed cilia beating by organ-cultured CP at 60 DIV probably reflects the explant’s viability and functionality.

Absorption Capacity

Patients with intraventricular hemorrhage accumulate hemosiderin in choroid ependyma, suggesting CP absorption (59). Strong evidence indicates that CP epithelial cells absorb cargo molecules from CSF and transport them along endocytosis and transcytosis pathways. Injection into CSF of tracer proteins, for example, HRP has clarified this absorptive property of CP (60). In the present study, organ-cultured CP for 2 months absorbed HRP from the culture medium.

Secretory Activity

Immunoreactive TTR in CP cultured for 2 months, and in the conditioned medium throughout the entire period, indicates that the cultured choroid cells synthesize and release TTR into the medium over the 2-month period. Lower TTR immunoreactivity of cultured CP compared with intact CP in vivo, and great amount of TTR in the conditioned medium, indicates that the CP may hypersecrete TTR in culture. Over the long culture period, secretory activity of CP is not controlled by neural or blood/CSF-borne signals. There is little information on the regulation of synthesis and release of CP TTR in untreated animals. Serotonin in CSF and sympathetic nerves in CP villi, while regulating CSF secretion would likely not affect TTR expression (61). Conserved protein reabsorption and protein secretion of cultured CP augur well for translational projects that graft CP into a ventricle to restore the disabled ependyma/brain interface.

Expression of Membrane Protein

AQP1 is needed to mediate water fluxes that support choroid cell volume regulation, which are important for choroid ependymal ion and microstructural (spatial) homeostasis (62). Interestingly, after 2 months in vitro, the CP continues to insert AQP1 into the apical membrane. Are there structural or non-CSF secretory functions of AQP1 in long-lasting culture? AQP1 expression/insertion may thus be a factor potentially affecting CP abilities in host CSF. Another prominent feature of explanted HTx CP was loss by 1 week in culture of blood vessels and extracellular matrix (stroma). This is expected given the blood flow deprivation in vitro.

Grafted CP Tissue Into CSF of Normal and Hydrocephalic Rats: Retained Morphology and Continued Expression of TTR, Caveolin 1, and AQP1

Neurotrophins/growth factor injections into experimental brain lesions improve function (63–65). Transplanting CP cells near lesions likely allows delivery of trophins to the injury site (22, 39, 66, 67). We noted transplant expression of TTR, but need to establish CP growth factor secretion into CSF. Significantly, several established CP and epithelial cell preparations secrete peptide/protein products (21, 26, 45, 68–70). Transplanted CPs, free or encapsulated, deliver biotherapeutic molecules to central target sites (39). Our transplants in HTx rats remained in the host ventricle, even during hydrocephalic enlargement. Importantly, we also ruled out CSF blockage.

Lateral ventricle CP transplants into HTx, like the organ-cultured CP in vitro, displayed vascular/interstitial regression. Vasculature diminution in transplants/explants begs the questions of conditions causing vascular atrophy. Reduced CP blood flow/CSF formation in clinical transplants would likely not further increase CSF volume in hydrocephalus. Thus, the impaired host brain may benefit from the choroidal graft protein secretions without a CSF volume-increment effect.

Elucidation is needed on how transplanted fresh CP obtains vitamins and other substrates. Distribution of plasma O₂ and glucose across the subventricular blood-brain barrier (capillaries) and the circumventricular blood vessels is likely; also, a supply of nutrients and growth factors from ventricular CSF to the apical membrane of transplanted CP probably occurs. Clearly, the transplant receives sufficient O₂, glucose, and other materials to maintain functions and epithelial structure.

Transplanted CP expresses caveolin 1, a membrane scaffolding protein. Caveolin 1 actions are integrated by hormones/growth factors that modulate CP function. Expression of caveolin 1 in CP transplants encourages analyses of how caveolin 1 and other membrane structural proteins impact the ventricular ependyma-brain viability of hosts with hydrocephalus.

New ways are sought to manage hydrocephalus rather than by diverting CSF. Neurospheres have potential for facilitating incapacitated ependymal/SVZ systems in hydrocephalus (71, 72). Various cells may aid CSF repair: CP ependyma, mesenchymal stem cells, subcommissural organ cells, and pluri-potential neural stem cells (73). RNA manipulation in cultured CP cells (for eventual ventricular transplanting) may be therapeutically helpful.

Challenges in Expanding Hydrocephalus Research Horizons

We demonstrated feasibility in organ culturing the HTx CP tissue and in transplanting fresh CP into lateral ventricles. Precise identification of macrophage and endothelial elements in CP grafts is needed. Certainly, the acquisition of CSF compositional data (hormones, growth factors, and neurotrophins) is key to translational advances. Another challenge is determining if intraventricular CP transplants improve brain/ependymal wall development in congenital hydrocephalus. Although exogenous stem cell supplements to CNS promote recovery, the mRNA transcript profile in CP may uniquely foster modulation of brain/ependyma (74).

The present findings prompt the notion that in hydrocephalic patients, the removal of CP tissue and its fractionation for transplant back into lateral ventricles, a sort of autograft, could be beneficial in neurosurgical treatment. The working model is that viable choroid cells (separated from matrix/blood flow) would not produce CSF (thereby not augmenting intracranial pressure) but would continue secreting beneficial trophic molecules into hydrocephalic CSF.

We are grateful for funding from the Hydrocephalus Association (JPM, CEJ, and EMR) and from Fondecyt 1000435, 1070241, and 1111018 (to EMR) that made these explant and transplant experiments possible.

The authors have no duality or conflicts of interest to declare.