Prior to tissue acquisition by the authors, informed consent for the donation of tissues was obtained under the auspices of the protocol for the National Human Neural Stem Cell Resource. This protocol is approved by Institutional Review Board (IRB) of Children's Hospital of Orange County, and follows a protocol approved by the UC Davis School of Medicine IRB. All tissues were acquired in compliance with NIH and institutional guidelines. Two patients were used for the present study: a patient with fragile X syndrome and a patient with no neurogenetic disease.

The autopsy followed standard procedures as described 20. The periventricular zone in the area of the head of the caudate nucleus regions was identified and dissected from the appropriate brain sections. Brain region specimens were then placed in separate Petri dishes and rinsed three times with DGA (see below). Tissues were minced with sterile scalpel blades, triturated in DGF (see below) containing 10% DMSO, taken to -80°C overnight in controlled-rate freezing containers, and then transferred to liquid nitrogen Dewars for long term storage.

Portions of fresh brain, heart, and testicular tissue were received, and fixed in 10% formalin for 10 days prior to sampling and processing for paraffin sections in standard fashion. All tissue sections were stained with hematoxylin and eosin, with cardiac valves and aorta additionally stained for elastin and mucin. All sections were examined by standard light microscopy.

The base medium was a high glucose 1:1 DMEM:F12 (Irvine Scientific). The basal medium (DGA) used for all other media was the base medium containing glutamine, penicillin, streptomycin, gentamicin, ciprofloxacin, and amphotericin as previously described 20. Medium (DGF) used for all washes consisted of DGA containing 20% fetal bovine serum.

All procedures were performed as previously described 20, with modifications. Tissues were quickly thawed, diluted by drop-wise addition and agitated in 10 volumes of DGF, then further dissociated by trituration and three washes in DGF with centrifugation – no enzymatic digestion was used. Whole tissue homogenates were plated directly on fibronectin-coated tissue culture plates (6-well, tissue culture treated, Falcon) in primary growth medium (PGM) composed of DGF containing 10% BIT 9500 (Stem Cell Technologies), 40 ng/mL basic fibroblast growth factor (FGF-2; InVitrogen), 20 ng/mL epidermal growth factor (EGF; InVitrogen), and 20 ng/mL platelet-derived growth factor-AB (PDGF-AB; Peprotech). Plates had been previously incubated with 200 uL/cm² of fibronectin (5 ug/mL; Sigma) overnight at 37°C, the fibronectin solution aspirated, and the plates allowed to air dry before the introduction of tissue homogenates.

Approximately 300 mg fresh tissue was subjected to mincing and trituration, and the resulting crude tissue homogenate plated into six wells of a fibronectin-coated six-well tissue culture plastic plate (≈60 cm² total surface area).

After plating, 50% of the medium was replaced, 3 times weekly. Non-adherent cells and debris from the removed supernatant were pelleted by centrifugation and re-introduced into the cultures together with the fresh medium. After 7 days in culture, plates were agitated by sharp rapping with a marking pen and 100% of the culture medium and non-adherent material was removed. Fifty percent of the volume removed was replaced with fresh medium, while the removed medium was centrifuged to pellet cell debris and non-adherent cells and to recover conditioned medium as supernatant. Fifty percent, by volume, of the conditioned medium was then returned to the original plates. The pellet, containing the non-adherent fraction, was resuspended in 50% conditioned medium and 50% fresh medium, by volume, and then transferred to a fresh fibronectin-coated 6-well plate. After one week, the procedure was repeated, except that the non-adherent fraction was discarded. In this way, an additional population of cells was recovered from the non-adherent fraction. All the cells were eventually combined to form a single population of cultured cells.

At near confluence, cultures were passaged by lifting with a solution of Cell Dissociation Buffer (GIBCO) supplemented with trypsin. The cells were washed twice with DGF and plated in 1:1 (conditioned:fresh) medium into a fibronectin-coated T75 flask. Thereafter, and at approximately one week intervals, cells were lifted and similarly plated into twice the surface area from which they were removed. After the cells had reached a confluent surface area of 600 cm², the medium from one T75 flask was exchanged with GM (PGM without serum), and these cells were cultured for two weeks before immunocytochemical analysis, or differentiation and immunocytochemical analysis, as previously described 20.

Immunocytochemistry was performed as previously described 20. Primary antibodies and dilutions were used as follows: nestin (1:100; mouse; Chemicon), type III β-tubulin (Tuj20; 1:100; mouse; Chemicon), MAP2ab (1:250; mouse; Sigma), GFAP (1:500, guinea pig; Advance Immuno), CD133-APC (1:100; mouse; Miltenyi), NCAM (1:100, rabbit, Chemicon), fusin (1:100, mouse, Chemicon), and FMRP (1:100; mouse; Chemicon). Coverslips were mounted with Prolong^® Antifade Kit (Molecular Probes, Eugene, OR). Some cells were stained with 4',6-diamidino-2-phenylindole (DAPI, Sigma) before being rinsed and mounted. Pictures were imaged on an Olympus IX70 Microscope and digitally photographed via a Microfire digital camera (Optronics, Goleta, CA) using Image Pro Plus 4.5 with AFA plugin 4.5 software.

JS was a 25-year-old man with fragile X syndrome. His history included motor and language delays in childhood, with walking at 20 months, phrases at three years, and sentences at 6 years of age. He was diagnosed with fragile X syndrome at 11 years of age, by cytogenetic testing. His behavior included hyperactivity, anxiety, shyness, poor eye contact, hand flapping, finger biting, and perseverative speech. At age 24, cognitive testing with the WAIS III demonstrated a verbal IQ of 58, performance IQ of 51, and full scale IQ of 51. He did not have autism; childhood autism rating scale (CARS) score was 28.5 (below autism range).

Previous medical history included severe mitral valve prolapse; echocardiography revealed moderate thickening and redundancy of both mitral leaflets, with central mitral regurgitation (grade 2 to 3+ by Doppler), mild tricuspid regurgitation, and moderate left ventricular and mild left atrial enlargement. Upon physical examination at age 24, JS had a long face, prominent ear pinna, high arched palate, and macroorchidism with testicular volume of 60 ml bilaterally. Blood pressure was 142/74, and a grade III/IV systolic and diastolic murmur with click was heard on examination. He died unexpectedly at age 25, presumably due to a cardiac arrhythmia secondary to mitral valve prolapse.

A complete autopsy was performed 16 hours after death. Abnormal findings included increased brain weight (1600 gm; normal, 1440 ± 20 g), an enlarged heart, (400 gm; normal, 349 ± 40 g), with features of mitral valve prolapse (myxomatous thickening of both leaflets; "hooding" of the posterior leaflet), and increased testicular weight of 73 g (normal average, 25 g). Our laboratory did not receive the intact brain, but portions of it, therefore detailed gross and histological evaluation was not possible. In particular, neither the hippocampus nor the amygdala was available for histopathological examination. Evaluation of available brain sections showed diffuse, acute early ischemic damage corresponding to the manner of death; no intraneuronal inclusions were seen, nor was there spongiosis of white matter in the cerebrum, cerebellum, or middle cerebellar peduncles. Cerebellar folia showed moderate patchy absence of Purkinje cells; because of the manner of death, loss due to ischemic damage cannot be ruled out. However, the majority of Purkinje cells present appeared histologically normal.

Minced brain tissue, derived from the periventricular zone in the area of the head of the caudate nucleus and maintained under proliferation conditions in serum-containing medium with growth factors, yielded viable cells that formed an adherent monolayer on fibronectin-coated plates. When lifted and plated without serum or fibronectin substrate, the cells grew in suspended clusters/spheres (Figure 1), similar to previously described neurospheres 20. The morphology within the adherent population was variable and included small, rounded profiles, medium-sized bipolar and spindle-shaped profiles, and larger cells with polygonal and multipolar morphologies. Once a robust primary culture had been established in the six-well plates (approximately one month after plating of tissue homogenates), the cultures were passaged into T75 flasks approximately once per week until 600 cm² of confluent adherent cells had been produced. Cells were then further expanded under serum-free conditions for immunocytochemical analysis, or under growth-factor-free conditions for biochemical analysis.

Although the methodology used was similar to that previously reported 20, four conditions employed in the current work are noteworthy. (i) The cells were grown from cryopreserved, rather than fresh, tissue. This modification allows pathologists with no local access to a stem cell culture laboratory to preserve tissues for later stem cell harvest at a remote collaborating laboratory. (ii) No enzymatic digestion, only trituration, was used to generate the crude tissue homogenates. Preliminary studies showed that enzymatic digestion, typically used with fresh tissue, adversely affected our ability to harvest living cells from cryopreserved tissues. (iii) Serum (20%) was maintained in the culture until it was expanded to 600 cm² of confluent adherent cells, after which the cells were cultured in serum-free medium. Preliminary studies indicated that, unlike cells harvested from fresh tissues, cells harvested from cryopreserved tissues required application of serum for a longer time in culture to sustain a sufficient rate of proliferation. (iv) The non-adherent fraction was transferred to new fibronectin-coated plates after one week. Preliminary studies showed that the non-adherent fraction from cryopreserved tissues retained a significant population of viable cells for a longer period of time than that from fresh tissue. These plates were cultured for an additional week before the remaining non-adherent fraction was finally discarded. Cells grown from both sets of plates were combined for expansion.

Immunocytochemical analysis of fragile X neural progenitor cells grown under expansion conditions demonstrated the expression of a range of developmental and mature neural markers (Figure 2). Many of the current results are similar to previous findings with control human neural progenitor cell cultures 20. In particular, the distribution of the multipotential neural progenitor lineage marker, CD133, the neuroepithelial marker nestin, and the neural cell adhesion molecule, NCAM, are found to be widespread in these cultures, consistent with earlier observations of immunocytochemistry and flow cytometry in control human neural progenitor cells 202324. The CXCL12 (SDF-1) cytokine receptor, CXCR4 (fusin, CD184), and the glial neurofilament marker, glial fibrillary acidic protein (GFAP), are also widely expressed in the fragile X neural progenitor cell cultures, in agreement with a previous report 20. The expression of β-III-tubulin, a mature neuronal marker, is restricted to subpopulations of cells, again consistent with control human neural progenitor cells 20. Finally, the primitive neuroepithelial (intermediate filament) markers, nestin and CD133, seen in the proliferating neural progenitor cells (Figure 2), disappear under differentiation conditions (data not shown). Importantly, FMRP staining is markedly reduced in the fragile X neural progenitor cells (Figure 3) compared to control cultures.

Southern Blot analysis of DNA isolated from brain tissue (frontal cortex) showed the presence of a mosaic pattern. Specifically, full mutation alleles were present in 82% of the cells (435, 528, 652, 727, 847 CGG repeats) with the remaining 18% harboring a premutation allele. Sizing of the CGG repeat number by PCR analysis demonstrated the presence of a premutation allele of 90 CGG repeats and an allele with the deletion of the CGG element and the flanking region, the latter of which was not detected by Southern Blot analysis. Sequence analysis of this allele indicated the presence of a deletion extending from nucleotide 13742 to 13915 of the FMR1 gene (GenBank accession number L29074). Southern blot analysis of DNA isolated from neural progenitor cells revealed the presence only of hypermethylated full mutation alleles (536 and 591 CGG repeats). No premutation alleles were detected by PCR analysis.

The brain FMR1 mRNA level (frontal cortex) relative to the reference gene (glucuronidase) was low (0.08 ± 0.009, relative value). In agreement with the observed lack of FMRP expression, no FMR1 message was detected after 40 cycles of PCR in total RNA isolated from the neural progenitor cells.