Human Fetal Brain Tissue Research Articles

Development of human fetal ventral mesencephalic grafts in rats with 6-OHDA lesions of the nigrostriatal pathway

Neuronal transplantation is an approach that can be exploited to study the development of the human central nervous system as well as being used in attempts to restore neurological function. In the present study, we have examined cellular events that appear to precede the development of dopamine nerve fiber extension by neurons from the human fetal ventral mesencephalon. These cellular events were examined using neuronal cell suspensions from human fetal ventral mesencephalic tissue (gestational ages 7–10 weeks) transplanted into the striatum of unilaterally lesioned 6-hydroxydopamine (6-OHDA) rats. Animals were sacrificed for immunohistochemistry 9–10 weeks after the transplantation prior to the manifestation of behavioral recovery. Histological analysis revealed tyrosine hydroxylase (TH) immunoreactive neurons in the grafts. The majority of these neurons had very short TH positive processes (60–70 um), indicating that the maturation of grafted dopaminergic neurons was still incomplete. Immunostaining for the human specific intermediate neurofilament (hNF, clone: BF-10) showed dense neuronal fibers in the grafts. These fibers extended deeper into the host brain than the TH positive neuronal processes. The whole striatum, particularly the medial part of the striatum, exhibited long NF positive processes. Glial fibrillary acidic protein (GFAP) immunohistochemistry revealed fine astrocytic processes inside the grafts, which were clearly different from host reactive glial cells surrounding the grafts. These graft-derived glial processes tended to extend into the host brain deeper than the TH positive neuronal processes from the grafts. These early histological findings of the grafted human fetal ventral mesencephalon suggest that the graft-derived NF positive neuronal processes, as well as the glial processes, radiate from the grafted tissue and extend into the host brain prior to the extension of TH positive processes. These results further suggest that human-to-rat xenografts can be used to study the neural development of human fetal brain tissue.

Recovery of cellular activity in cryopreserved human foetal brain tissue

Since long-term cryopreservation can cause losses in neural tissue viability and function a prerequisite would be the ability to monitor and promote functional recovery in donor tissue intended for neural transplantation. Rapid assessment of cryopreserved tissue's functional status prior to grafting is presently difficult in a clinical setting. A convenient indicator of functional status may be the level of DNA synthesis activity taking place in the tissue. Using immunocytochemical detection of incorporated bromodeox-yuridine we have quantified and compared DNA synthesis activity (expressed as proliferative capacity (PC)) in human foetal mesencephalic, striatal, cortical and cerebellar tissue before and after a 275-376 day storage in liquid nitrogen. There was a post-storage reduction in viability of 48-73% and in PC of 26-59%; the higher the PC before storage the greater the reduction after. Incubation of cryopreserved tissue with fetal calf serum resulted in 2-4-fold higher PC levels than serum-untreated controls and reached 80% of fresh tissue levels in mesencephalic cells after 3-4 h incubation. Assuming that quantification of proliferative activity is a practical indicator of the tissue's functional status, these findings suggest that treatment of the tissue with serum can largely restore the lost function caused by cryopreservation.

Recovery of cellular activity in cryopreserved human foetal brain tissue.

Since long-term cryopreservation can cause losses in neural tissue viability and function a prerequisite would be the ability to monitor and promote functional recovery in donor tissue intended for neural transplantation. Rapid assessment of cryopreserved tissue's functional status prior to grafting is presently difficult in a clinical setting. A convenient indicator of functional status may be the level of DNA synthesis activity taking place in the tissue. Using immunocytochemical detection of incorporated bromodeox-yuridine we have quantified and compared DNA synthesis activity (expressed as proliferative capacity (PC)) in human foetal mesencephalic, striatal, cortical and cerebellar tissue before and after a 275-376 day storage in liquid nitrogen. There was a post-storage reduction in viability of 48-73% and in PC of 26-59%; the higher the PC before storage the greater the reduction after. Incubation of cryopreserved tissue with fetal calf serum resulted in 2-4-fold higher PC levels than serum-untreated controls and reached 80% of fresh tissue levels in mesencephalic cells after 3-4 h incubation. Assuming that quantification of proliferative activity is a practical indicator of the tissue's functional status, these findings suggest that treatment of the tissue with serum can largely restore the lost function caused by cryopreservation.

Human GAP-43 Gene Expression: Multiple Start Sites for Initiation of Transcription in Differentiating Human Neuroblastoma Cells

Phorbol ester treatment of human SH-SY5Y neuroblastoma cells, which leads to mature neuron-like cells with a sympathetic phenotype, induces outgrowth of neurites which are terminated by growth cones. The neurite extension is parallelled by an increased expression of the growth-associated protein, GAP-43. At the mRNA level, two GAP-43 mRNA species of 1.4 and 1.6 kb, respectively, were detected in SH-SY5Y cells. Both the low- and high-molecular-weight GAP-43 transcripts cosedimented with a polysomal fraction, indicating translation of both types of transcripts. To structurally characterize these GAP-43 mRNAs, several cDNA clones were isolated. The only difference identified corresponded to various size extensions in the 5′-untranslated region. A human genomic DNA fragment extending 1145 bp 5′ of the GAP-43 translation start site, including a putative promoter region of the GAP-43 gene, was also characterized. Comparison of human and rat GAP-43 genomic sequences revealed an 85% identity between the first 900 bp 5′ of translation start site. By RNase protection analysis, two clusters of putative transcription start sites, located approximately 200 bp apart, were identified. DNaseI footprinting analyses, using nuclear extracts prepared from untreated and TPA-treated SH-SY5Y cells, revealed specific footprints primarily detected in extracts prepared from differentiating cells. These clustered at positions immediately 5′ of the two putative transcription start site regions. GAP-43 mRNA expression was finally studied using a probe which specifically recognizes the high-molecular-weight GAP-43 transcripts. Five tested human neuroblastoma cell lines and human fetal brain tissue expressed these transcripts. Furthermore, during differentiation of SH-SY5Y and LA-N-5 cells, both sizes of GAP-43 transcripts were transiently induced with the larger slightly preceeding the smaller mRNA species.

Susceptibility of human foetal brain tissue to cool- and freeze-storage

Human second trimester foetal brain tissue was stored for a period of 1–6 weeks under various conditions in an attempt to evaluate factors influencing its susceptibility (cell loss) and survivability. Post-storage viability of mesencephalon, striatum, cerebellum and occipital cortex was assessed by a protocol combining vital staining with cell density counts so that tissue viability and cell loss could be evaluated simultaneously; tissue survivability was evaluated by cell culture. A significant amount of cell loss occurred after 24 h storage at room temperature, after one week at 4°C and by two weeks at −20°C in all structures; storage at −196°C resulted in 17–21% cell loss at the end of a 6 week period. At −20°C the cryoprotective effect of 20% FCS was equivalent to that of 15% FCS + 7% DMSO combined, suggesting potential use of serum in replacement of chemical additives. The procedure for removal of DMSO was critical to cell viability and survivability: single step dilution led to 27–39% greater cell loss than slow, multi-step dilutions. In comparison to fresh, non-stored tissue, immunocytochemical characterization of in vitro propagated stored tissue revealed no changes in the populations of major constituent cell types including neurones, dopaminergic neurones, glial and fibroblast cells. These results provide information on possible conditions under which transplant tissue can be satisfactorily stored depending on the prevailing requirements.

Cryopreservation and culture of the human fetal brain tissues.

Human embryos after 3-4.5 months of gestation were obtained with abortion. The brain tissue of the bodies was scissored up to obtain 1-3 mm3 pieces, and 7% dimethyl sulfoxide (DMSO), as a cryoprotectant, was added, and then stored at -70 degrees C for 1-30 days or at -196 degrees C for 1-84 days. The survival rate of stored cells was 64%-88%. During 6 days of storage with neuron culture medium, the survival rate of cells at 4 degrees C is over 50% each day, but, as time goes on, the count of the cells is getting less and less. The cells washed out DMSO after cryopreservation and the planting fresh cells can adhere to the wall of the culture bottle, grow, display various forms of neurons and gliacytes. From the above findings, it was suggested that: 1) The fetal human brain tissue, handled properly, can endure cryopreservation with 7% DMSO as a cryoprotective agent; 2) The storage time was related insignificantly to the survival rate of the tissues stored; 3) It is available for a short preservation at 4 degrees C; and 4) It is possible to set up a bank of fetal human brain tissue.

Effect of 3'-azido-3'-deoxythymidine on human immunodeficiency virus type 1 replication in human fetal brain macrophages.

We investigated whether cells derived from the fetal central nervous system can support productive infection by a human immunodeficiency virus type 1 (HIV-1) isolate termed UHC-1, produced by a cellular clone derived from HIV-1 strain HIV-IIIB chronically infected U-937 promonocytic cells, and what the effect of nucleoside analogs might be on viral replication in this system. Fractionation of human fetal brain tissue into two different populations, enriched for either astrocytes or macrophages, showed that only the latter were able to support productive UHC-1 replication and generation of detectable progeny virus. Pretreatment of fetal brain macrophages with either of two nucleoside analogs, 3'-azido-3'-deoxythymidine (AZT) or the (-) enantiomer of 2'-deoxy-3'-thiacytidine, efficiently blocked production of progeny virus. Generation of unintegrated proviral DNA and HIV-1 transcripts were inhibited by pretreatment of fetal brain macrophages with 1 microM AZT. Administration of AZT at 24 h postinfection led to a slight reduction in viral transcript levels and viral progeny production by day 15 postinfection; however, brain macrophages under these conditions did not contain detectable amounts of unintegrated viral DNA. These results suggest that AZT may interfere with the accumulation of unintegrated HIV-1 DNA in brain macrophages. This is the first demonstration that nucleoside analogs are able to block HIV-1 replication in primary cultures of brain cells.

Cloning and characterization of a fourth human somatostatin receptor.

We have isolated a gene coding for a fourth human somatostatin (somatotropin release-inhibiting factor) receptor. This additional somatostatin receptor (hSSTR4) is specifically expressed in human fetal and adult brain and lung tissue. The deduced amino acid sequence of the receptor displays both sequence and structural homology to three cloned somatostatin receptors as well as to other members of the family of GTP-binding-protein-coupled seven-helix transmembrane-spanning receptors. Pharmacological characterization of the expressed receptor reveals specific, high-affinity binding of somatostatin 14 and somatostatin 28. Surprisingly, several well-characterized synthetic somatostatin analogs fail to exhibit high-affinity binding to hSSTR4, indicating the existence of pharmacologically different receptor subtypes. Our data suggest that the diverse biological effects exerted by somatostatin are mediated by a family of receptors with discrete patterns of expression and different pharmacological properties.

Human fetal myelinated organotypic cultures

We have previously reported the establishment of organotypic cultures derived from human fetal brain tissue. Although these cultures permit the testing of multiple hypotheses about normal human neurodevelopment and neuropathologic conditions, they have the limitation of not being myelinated and therefore preclude the study of questions related to myelinogenesis and diseases of myelin. In the current communication, we describe recent developments that allow us to overcome this limitation and permit the establishment of a myelinated organotypic culture model. Sections of dorsal column dissected from the lumbar spinal cord of human fetuses ranging in age 21–23 weeks of gestation were placed in culture. The explants were maintained for up to 12 weeks during which time they were characterized and shown to express a number of CNS cell-type-specific markers including glial fibrillary acidic protein (astrocytes), nerve growth factor receptor and neurofilament protein (neurons), CD68 (microglia), and myelin basic protein, HNK-1 and galactocerebroside (oligodendrocytes). In addition, lectin histochemistry using Ricinus communis agglutinin-1 detected microglia and endothelial cells. Upon explantation, abundant myelin was seen by electron microscopy in the cultures. Although during the culture period there was degradation of myelin, there was also evidence of maintenance of intact myelin sheaths around small caliber axons and de novo myelin synthesis. This model system may permit the further use of human organotypic cultures to investigate issues related to neurodevelopment and to pathologic conditions including those relevant to dysmyelination and demyelination.

Initial studies of embryonic transplants of human hippocampus and cerebral cortex derived from schizophrenic women

Human fetal brain tissue was obtained from first-trimester elective abortions of two women who also has schizophrenia. Portions of the embryonic hippocampus or cerebral cortex were transplanted into the anterior eye chamber of immunologically compromised athymic nude rats. In this environment, embryonic brain tissue derived from normal women generally continues organotypic growth and development for many months. Although initial survival after transplantation was normal, the tissue derived from schizophrenic women manifested less robust growth. However, cells in the transplants showed typical neuronal differentiation, with development of different neuronal types, such as pyramidal cells, granule cells, and gamma-aminobutyric acid (GABA)-containing interneurons. Rhythmic electrical activity was also observed, indicative of some local synaptic organization. The presence of messenger RNA (mRNA) for brain-derived neuronotropic factor (BDNF) was observed using in situ hybridization. The reason for the decreased rate of growth of these transplants remains unknown and the significance of the finding cannot be assessed from only two fetuses. However, these preliminary findings suggest that fetal transplants may be a useful model system for the detection of developmental pathogenic processes in the expression and transmission of schizophrenia.

Successful xenografts of second trimester human fetal brain and retinal tissue in the anterior chamber of the eye of adult immunosuppressed rats.

Successful xenografting of first trimester human fetal CNS tissue and retina has been reported in the literature. We wished to test the feasibility ofusing the anterior chamber ofthe rat eye to support the development of more mature human fetal xenografts. Here we report on the successful outcome of human brain and retinal transplants. Adult host rats immunosuppressed with cyclosporin A accepted these xenografts and supported their further development. Periodic examination of the host eyes using a direct ophthalmoscope or an ophthalmic slit lamp permitted direct visual monitoring of the health and growth of the transplants. Histologically it was possible to identify neuronal, macroglial, and microglial (macrophage) cell types within the grafts. Mitotic activity and histogenetic differentiation took place. Blood vessels filled with hematic cells were commonly present within the grafts. The walls of these vessels prevented the leakageofhorseradish peroxidase, suggesting the presence of a functional brain-blood barrier in the graft. These results indicate that it is possible to use a small animal model to study normal and pathological phenomena oniate fetal human neural tissues. Our group has already taken advantage of the model to achieve HIV infectivity offetal human brain outside the human body.

Persistent human immunodeficiency virus type 1 infection in human fetal glial cells reactivated by T-cell factor(s) or by the cytokines tumor necrosis factor alpha and interleukin-1 beta

Human immunodeficiency virus type 1 (HIV-1) infection of the brain has been associated with a severe dementing illness in children and adults. However, HIV-1 antigens are most frequently found in macrophages and microglial cells. To determine the extent of susceptibility of neuroglial cells to infection, the HIV-1 genome was introduced into cells cultured from human fetal brain tissue. Astroglial cells rapidly transcribed the viral genome producing high levels of p24 protein and infectious virions which peaked two to three days posttransfection. Thereafter HIV-1 genome expression progressively diminished and a persistent phase of infection developed during which neither virus nor viral proteins could be demonstrated by immunodetection methods. Cocultivation with CD4+ T cells at any time during the persistent infection resulted in resumption of p24 synthesis and virus multiplication. The release of persistence did not require direct cell-cell contact between the glial and T cells, since separation of the two cell types across a permeable membrane resulted in a delayed but similar resumption of p24 synthesis and virus multiplication. The persistently infected glial cells could also be stimulated to produce viral p24 protein if either tumor necrosis factor alpha or interleukin-1 beta was added to the medium without T cells present. These results suggest that astrocytes may serve as an undetected reservoir for HIV-1 and disseminate the virus to other susceptible cells in the brain upon triggering by some cellular or biochemical signal.

Interaction between human brain tumour biopsies and fetal rat brain tissue in vitro

The invasiveness of human intracranial tumours was studied in an organ culture system. Biopsies from six glioblastomas, four astrocytomas, two mixed gliomas, one ependymoma, four meningiomas and two carcinoma metastases were cut into fragments of 0.5 mm diameter, and placed in agar overlay tissue culture. The tumour specimens formed spheroids which were co-cultured with cell aggregates or fragments from fetal rat brain for up to 10 days in vitro. The invasiveness of the glioblastoma spheroids was characterised by a gradual destruction of normal brain tissue by tumour cells, followed by replacement of normal tissue by these cells. Co-cultures from two glioblastomas showed lesions in the normal brain tissue in areas removed from the tumour cells. Tumour spheroids from four glioblastomas totally destroyed the normal brain tissue without any change in the original tumour spheroid configuration. The low-grade gliomas were less invasive than the glioblastomas. The meningiomas and the metastases were non-invasive. This organ culture assay appeared to reflect the in situ invasive behaviour of the brain tumours examined. It is suggested that it may be used for evaluating the aggressiveness of individual brain tumours with the specific aim of correlating clinical data with the biological character of the tumour.

Cryopreservation of human brain tissue

Tissues from products of conception were examined to determine the feasibility of obtaining viable neural tissue after suction abortion at 9–12 weeks of gestation. The ventral mesencephalon, a prototype region whose maturation can be monitored and which is a potential tissue for transplantation, was identified in 32 of 120 cases. The tissue was then screened for the presence of infectious agents, while being held at −196°C in cryopreservative solutions. There of 32 specimens were found to be contaminated by normal vaginal bacteria; all other viral, fungal, and mycoplasma testing was negative. Thawed brain fragments retained high viability after storage in liquid nitrogen and when grown in vitro exhibited neuronal morphology, tyrosine hydroxylase immunoreactivity, and dopamine production. We have demonstrated that human fetal brain tissue can be cryopreserved in a manner which not only retains viability but allows normal phenotypic differentiation after thawing.

Immunosuppression and transforming growth factor-beta in glioblastoma. Preferential production of transforming growth factor-beta 2.

Transforming growth factor (TGF)-beta 1 is a polypeptide that is assumed to play a fundamental role in the growth of both normal and neoplastic cells. TGF-beta 2 is a closely related polypeptide, originally described as glioblastoma cell-derived T cell suppressor factor (G-TsF) due to its immunosuppressive activity. Expression of the genes for TGF-beta 1 and G-TsF/TGF-beta 2 was examined in tumor cells and was found to be different in several cell lines and tissues that were tested. Whereas two glioblastoma cell lines expressed both TGF-beta 1 and G-TsF/TGF-beta 2 mRNA, one melanoma and neuroblastoma cell lines showed only TGF-beta 1 mRNA which in the case of the neuroblastoma required cycloheximide treatment for its detection. The coordinate expression of the genes for TGF-beta 1 and G-TsF/TGF-beta 2 in glioblastoma was not paralleled by secretion of both polypeptides as only G-TsF/TGF-beta 2 but not TGF-beta 1 was identified in supernatants of glioblastoma cells. These data provide evidence for a post-transcriptional level of regulation for production of the two forms of TGF-beta. As mRNA for G-TsF/TGF-beta 2 was also identified in fresh surgically removed human glioblastoma tissue, G-TsF/TGF-beta 2 may also be secreted within the tumor in vivo. Unlike glioblastoma, human fetal brain tissues or adult brain specimens studied did not express detectable levels of TGF-beta mRNA. Impaired cell-mediated immunity is an established finding in patients with glioblastoma. Secretion of G-TsF/TGF-beta 2 by tumor cells in vivo may contribute to decreased immune surveillance for tumor development, as well as neovascularization of the tumor tissue.

Monoclonal Antibodies M340 and UJ181.4 Recognize Antigens Associated with Primitive Neuroectodermal Tumours/Tissues

Monoclonal antibodies UJ181.4 and M340 were raised from separate fusions using either human foetal brain or medulloblastoma tissue as immunogens. The antibodies are both of IgG isotype and bind to their target antigens with affinities in excess of 10(-9) M. Both monoclonal antibodies have been shown to bind to primitive neuroectodermal tumours and human foetal brain. Expression of the two antigens is developmentally regulated as no binding is detected on adult brain as determined by a variety of indirect binding assays. The two monoclonal antibodies can be clearly distinguished by their migration in a pH gradient (isoelectric focussing gels). UJ181.4 and M340 clearly recognize two different epitopes, as it is not possible to block the binding of one antibody with an excess of the other. In addition, careful study of the binding profiles of the two reagents suggest that they recognize two distinct antigens, although these have not been biochemically characterized. The reagents have proved particularly useful in both the differential diagnosis of the small round cell tumours of childhood and anaplastic brain tumours. In addition, they have formed part of a panel of reagents used for the immunomagnetic depletion of neuroblasts from bone marrow and have been used for targeting radionuclides to tumour cells in the intrathecal compartment.

The Ethics of Fetal Tissue Transplants

The Ethics of Fetal Tissue Transplants Heretofore, damage to the brain or spinal cord, whether from injury or disease, has been considered irreversible. There is no satisfactory cure for Parkinson's or Alzheimer's disease, for paraplegia resulting from spinal cord injury, for certain forms of epilepsy, or for a host of other less common but equally grave neurologic disorders. This situation may soon change. Over the last ten years, animal experiments have demonstrated that transplantation of appropriate fetal brain tissue to an impaired adult brain can in certain cases restore normal function. Human fetal brain tissue recently was used in transplants for patients with Parkinson's disease in Mexico and Sweden, and similar procedures are planned or underway in China, Canada and elsewhere. Future uses of fetal tissue may not be restricted to neurosurgery but might include transplanting of fetal pancreas to treat diabetes mellitus or fetal liver for certain blood and metabolic disorders. The prospect of therapeutic use of human fetal tissue has aroused strong emotions. Objections have been raised on several grounds: * The requisite killing and dissection of the fetus both are an abuse to the developing human being and brutalize those who perform them. Further, the therapeutic use of fetal tissue will encourage abortion, may motivate conception with the express intent to abort, and might even lead to the sale of fetuses and fetal material. * Collection of fetal tissue in a medically useful form exposes the mother to unnecessary risk, while transplantation of such tissue exposes the recipient to unacceptable risk. * The medical use of human fetal tissue has unacceptable social consequences. * The legal safeguards necessary to overcome these objections are either undesirable or unenforceable. Such objections raise important issues for the most immediate practical application of fetal tissue, the transplantation of human fetal dopamine-secreting neurons to the brains of patients with medically unresponsive Parkinson's disease. Scientific Grounds Parkinson's disease is a devastating neurologic disorder that occurs when neurons degenerate in the region of the midbrain called the substantia nigra. Normally, fibers from these cells secrete the chemical dopamine in forebrain regions important for regulating movement. In the absence of normal dopamine secretion, the patient suffers from a variety of impairments including rigidity, difficulty initiating movements, and tremor. The underlying causes of the disease are unknown and there is currently no cure. Existing drug treatments inevitably lead to unacceptable side effects, and as the disease itself progresses the patient becomes unable to carry out the essential activities of daily life. Animal experiments over the last decade have demonstrated that fetal dopaminergic neurons can be grafted to the brain, restoring normal movement to rats and monkeys with experimentally induced Parkinson's disease. Similar grafts of tissue from adult adrenal glands, which also produce dopamine have been shown to have beneficial effects in animal experiments. Such a procedure has been performed in at least one hundred patients with Parkinson's disease in Mexico, China, Sweden, and the United States. While it would be premature to draw definitive conclusions about the therapeutic value of this procedure, there appear to be no significant adverse effects. Nonetheless, based on animal experiments, it is likely that grafts of fetal human dopamine-secreting neurons will be substantially more effective than adrenal grafts. Procedures for transplanting human fetal neurons are under development in a number of laboratories, and at least five operations have been performed in Mexico City, Stockholm, and Birmingham, England. [1] Numerous factors influence the success of neuronal transplantation to the brain. …

Cloning and expression of human apolipoprotein D cDNA.

The amino acid sequence of human apolipoprotein D, a component of high density lipoprotein, has been obtained from the cloned cDNA sequence. The 169-amino acid protein has no marked similarity to other apolipoprotein sequences, but has a high degree of homology to plasma retinol-binding protein and other members of the alpha 2u-globulin protein superfamily. Apolipoprotein D mRNA has been detected in human liver, intestine, pancreas, kidney, placenta, adrenal, spleen, and fetal brain tissue. Tissue culture cells transfected with the cloned cDNA secrete material that reacts with anti-apoD antibodies.

The clinical significance of antibody against oncofetal antigen-I (anti-OFA-I) in patients with gynecologic tumors.

AbstractOncofetal antigen‐I (OFA‐I) is a membrane antigen of human cancer cells that cross‐reacts with human fetal brain tissues. The antibody profile against this antigen was analyzed in relation to the course of disease.Fifteen out of 22 patients had persistently stable antibody levels and antibody titer in three of them increased by more than 4‐fold. These patients did not show either recurrence or death, but had a good prognosis. In four patients, antibody levels dropped by more than 4‐fold and was followed by clinical recurrence and death.These studies show that persistently stable antibody titers are predictive of non‐recurrence, and decreasing titers predictive of recurrence.

Characterization of somatomedins from human fetal brain: identification of a variant form of insulin-like growth factor I.

A fetal form of somatomedin (insulin-like growth factor) that crossreacts in the fetal brain radioreceptor assay has been proposed to exist in humans. Using this assay to monitor activity during purification, we have isolated a variant form of insulin-like growth factor I (IGF-I) from human fetal brain tissue. The variant IGF-I showed potent crossreaction in the fetal brain radioreceptor assay and stimulated DNA synthesis in fetal brain cells in vitro. Structural analysis revealed the variant IGF-I to have a truncated NH2-terminal region compared to IGF-I isolated from serum. An additional peptide, which displayed less potent crossreaction in the fetal brain radioreceptor assay, was also isolated from the human fetal brain. Partial amino acid sequence analysis revealed identity to insulin-like growth factor II.