Stability and interactions of recombinant human nerve growth factor in different biological matrices: in vitro and in vivo studies.

The purpose of this investigation was to characterize the stability, activity, and interactions of recombinant human nerve growth factor (rhNGF) in various biological matrices in vitro and in vivo. rhNGF (10 microg/ml) remained stable in human plasma for up to 4 days at 37 degrees C. There was a decrease in the recovery of rhNGF after incubation at lower concentrations (20 ng/ml) and for longer time periods (3 and 5 days at 37 degrees C). Size exclusion HPLC analysis indicated that rhNGF forms high molecular weight (HMW) complexes after long incubation periods. We confirmed that alpha(2)-macroglobulin (alpha(2)M) is the major plasma component that binds to rhNGF. Furthermore, this interaction was considerably increased by treatment of plasma with primary amines such as CH(3)NH(2). Changes in the pH environment did not affect the interaction of rhNGF with alpha(2)M. We also determined that the binding of rhNGF to CH(3)NH(2)-treated pure alpha(2)M or alpha(2)M present in human plasma substantially diminished its immunoreactivity and bioactivity detection. The interaction of rhNGF with activated alpha(2)M was reversed and inhibited by coincubation with dimethyl sulfoxide. Released rhNGF under these conditions was fully bioactive. (125)I-rhNGF also binds to alpha(2)M by forming similar (125)I-rhNGF/HMW complexes in plasma after i.v. administration in rats and mice. Sixty minutes after dosing in rats, most of the labeled material was in the form of a (125)I-rhNGF/HMW complex. These studies have provided a better understanding of the nature of the interactions of rhNGF with plasma components as well as methods to enhance, reverse, and inhibit these interactions.

Tissue disposition and pharmacokinetics of recombinant human nerve growth factor after acute and chronic subcutaneous administration in monkeys.

In this study, we have characterized the metabolism, tissue disposition, excretion routes, and plasma pharmacokinetics of recombinant human nerve growth factor after single and multiple s.c. administration in male cynomolgus monkeys. Unlabeled nerve growth factor (NGF; 2 mg/kg) was administered three times a week for 4 weeks and a full pharmacokinetic profile was obtained for doses 1 and 12. For the tissue distribution studies, 0.8 microg/kg of trace (125)I-labeled recombinant human nerve growth factor was dosed. Histological analysis of emulsion-microautoradiography indicated that specific (125)I-NGF labeling was confined to sections of nerves most frequently localized adjacent to large vessels in sections of kidney, spleen, liver, and salivary gland. A small percentage of large neurons within the sympathetic ganglia were intensely labeled, as well as large neurons within the dorsal root ganglia. We found an increased disposition of (125)I-NGF in parts of the peripheral nervous system (including sympathetic ganglia) from 8 to 24 h postdose. In contrast, radioactivity in most non-neuronal tissues declined. This suggests specific uptake in these target tissues known to express specific receptors for NGF. We also identified changes in pharmacokinetic parameters after single versus chronic s. c. administration. These studies demonstrated that s.c. administration of NGF at 0.8 microg/kg doses in monkeys is capable of accessing and localizing in the target tissues.

Heregulins and the ErbB-2/3/4 receptors in gliomas.

The activation of autocrine loops involving proto-oncogene related receptor tyrosine kinases has led to the analysis of a large number of growth factor systems in human glioma specimens and cell lines. The ErbB-2 system, also called HER-2 or neu, is analogous to the epidermal growth factor receptor system (EGF-R, ErbB-1). Neuregulins consist of a large family of proteins arising from alternative mRNA splicing of a single gene located at 8p22-p11. Activation of ErbB-2 by neuregulins occurs in heterodimeric complexes with ErbB-3 and ErbB-4. A panel of human glioma cell lines, which had previously been analyzed for ErbB-2 expression, was examined for ErbB-3 and ErbB-4 expression. Coordinate expression of ErbB-2, -3 or -4 was not observed in these cell lines. Despite the presence of a complete system capable of signaling in about half the cell lines, no constitutive activation of ErbB-2, -3 or -4 was observed, and autophosphorylation of ErbB-2 in response to heregulin was observed only in one cell line from the panel, NCE-G84. Moreover, the addition of recombinant heregulin or antibodies capable of disrupting ErbB-2/ErbB-3 complexes had no effect on cell proliferation. We conclude that the role of neuregulins and its receptors in the control of glioma cell proliferation may be limited or may be context dependent on in situ conditions which are lost in vitro. Alternatively, neuregulins may be involved in cell differentiation or survival in the central nervous system. Data supporting these conclusions are described in more detail herein.

Regulation of neurotrophin receptor expression during embryonic and postnatal development.

Members of the NGF family of proteins act as neurotrophic agents for defined populations of peripheral and central neurons during embryonic and postnatal development. We have studied the presence of receptors for brain-derived neurotrophic factor (BDNF) and neurotrophin-3 and -4/5 (NT-3, NT-4/5) by cross-linking radioiodinated neurotrophins to specific cell surface receptors. We have identified neurotrophin receptors representing full-length TrkB and TrkC and their truncated forms (lacking a functional cytoplasmic kinase domain) in neuronal as well as in non-neuronal tissues. During chicken embryonic and early postnatal brain development, expression of full-length TrkB and TrkC proteins preceded the onset of the truncated forms of these receptors. A similar pattern was also observed in mouse embryonic and early postnatal brain. The relative levels of neurotrophin receptors in the basal forebrain and in the hippocampus did not change significantly with age in mice. High levels of receptors for the three neurotrophins were detected in the nigrostriatal system. Full-length TrkB and TrkC receptors were found in chicken and rat embryonic ventral spinal cord, as well as on purified motoneurons. Again, truncated TrkB appeared significantly later than the full-length form on spinal motoneurons. In chicken embryonic retina and optic tectum we detected full-length TrkB and TrkC; however, the optic tectum also expressed large amounts of the truncated form of TrkB. TrkC but not TrkB was detected in chicken embryonic skeletal muscle, suggesting that NT-3 may have a novel function in this tissue. The presence of neurotrophin receptors in a wide variety of embryonic and postnatal tissues underlines the significant role of BDNF, NT-3, and NT-4/5 in embryonic and postnatal development. The regulation of the ratio of full-length versus truncated neurotrophin receptors may play an important role in the development, maturation, and maintenance of various neuronal networks.

Characterization of neurotrophin receptors by affinity crosslinking.

Neurotrophic factors regulate the developmental survival and differentiation of specific neuronal populations. Brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3) are members of the nerve growth factor (NGF) protein family, also known as the neurotrophins. Insights into the different roles of neurotrophins can be gained by studying the expression of their functional receptors. Here we report the development of procedures for their radiolabeling and efficient crosslinking to specific cell-surface receptors. BDNF and NT-3 receptors in cell lines and tissue preparations expressing receptors for the 2 neurotrophins have been identified using this affinity crosslinking procedure. Like NGF, BDNF and NT-3 crosslinked to the low affinity NGF receptor (p75NGFR) on PC12 cells. BDNF and NT-3 also crosslinked to cells expressing p145trkB protein, producing an approximately 160 kD neurotrophin-receptor complex. Crosslinking of the 2 neurotrophins in vivo to specific trk family members in many areas of the central nervous system also produced a 160 kD receptor complex. However, in all brain regions a complex of approx. 100 kD could also be identified, all or most of which represents crosslinking to a truncated form of trkB. The broad distribution of BDNF and NT-3 receptors throughout the CNS suggests that neurotrophins may have yet unrecognized functions on specific neuronal populations. BDNF and NT-3 receptors were also found in brain areas in which the neurotrophins themselves are also synthesized, suggesting that beyond long-range trophic effects, these proteins may also act as autocrine or short-range paracrine regulators.

Neurotrophin gene expression by cell lines derived from human gliomas.

The expression of neurotrophin (NGF, BDNF, and NT-3) mRNAs in 24 cell lines derived from human malignant gliomas was studied by Northern analysis. Widespread expression of neurotrophin genes was found with BDNF being the most abundantly expressed. Nearly all cell lines expressed BDNF, and about two-thirds of the cell lines expressed NGF and NT-3. Half of the cell lines analyzed expressed all three neurotrophins. Secretion of NGF into the medium of several cell lines could be detected by ELISA and a PC12 neurite outgrowth assay. Immuno- and bioactive NGF was isolated from conditioned medium of one cell line. No evidence of expression of the neurotrophin receptors trk and trkB by Northern analysis was found. Receptor crosslinking with radiolabeled cognate ligands failed to detect functional receptors in all but one cell line. In this cell line a receptor complex for BDNF was found that corresponded to truncated trkB receptors that lack the signal transducing tyrosine kinase domain. Neurotrophins did not stimulate mitosis of the glioma cultures. The findings suggest that production of neurotrophins by glioma cells is a general phenomenon, although neurotrophins made by gliomas lacking their receptors may not play an autocrine but rather a paracrine role.

Activity and biospecificity of proteolyzed forms and dimeric combinations of recombinant human and murine nerve growth factor.

Purified recombinant human nerve growth factor (rhNGF) and submaxillary gland-derived murine NGF (muNGF) were characterized by amino acid composition, polyacrylamide gel electrophoresis (PAGE), reversed-phase HPLC (RP-HPLC), and high-performance ion-exchange chromatography (HPIEC). Limited tryptic digest of the N and C termini of the 120-residue form of rhNGF produced a species of 109 residues (10-118). The previously observed natural murine analogue of this variant, muNGF lacking the first eight N-terminal amino acids, was also isolated as a homodimer. Both species were purified using HPIEC and characterized by amino acid analysis, N-terminal sequence, PAGE, and RP-HPLC analysis. Each of the four homodimeric species was evaluated in some or all of the following biological assays for NGF: chick dorsal root and sympathetic ganglion assays and rat pheochromocytoma-12 cell line neurite extension assay. The 118-residue homodimeric versions of both rhNGF and muNGF displayed equivalent bioactivity, whereas the N terminal-modified molecules presented activity reduced by 50- to 100-fold. Utilizing HPIEC, we have examined the ability of the monomeric forms of any two of the homogeneous dimeric species of rhNGF to recombine. We have shown that not only can all of the previously described species form dimers by recombination, but an interspecies dimer can be created between muNGF and rhNGF.

Isolation and characterization of two peptides with prolactin release-inhibiting activity from porcine hypothalami.

Two peptides with in vitro prolactin release-inhibiting activity were purified from stalk median eminence (SME) fragments of 20,000 pig hypothalami. Monolayer cultures of rat anterior pituitary cells were incubated with aliquots of chromatographic fractions and the inhibition of release of prolactin in vitro was measured by RIA in order to monitor the purification. The hypothalamic tissue extract was separated into 11 fractions by high-performance aqueous size-exclusion chromatography with one fraction showing a 4-fold increase in prolactin release-inhibiting factor (PIF) activity. This material was further purified by semipreparative reversed-phase (RP) HPLC. This process resulted in the separation of two distinct fractions that showed high PIF activity. These were further purified by semipreparative and analytical RP-HPLC to apparent homogeneity as judged by the UV absorbance profiles. Neither of the two peptides showed cross-reactivity with gonadotropin releasing hormone-associated peptide or with somatostatin-14 antibodies. Protein sequence analysis revealed that one of the PIF peptides was Trp-Cys-Leu-Glu-Ser-Ser-Gln-Cys-Gln-Asp-Leu-Ser-Thr-Glu-Ser-Asn-Leu-Leu- Ala-Cys - Ile-Arg-Ala-Cys-Lys-Pro, identical to residues 27-52 of the N-terminal region of the proopiomelanocortin (POMC) precursor (corresponding to amino acids 1-26 of the 16-kDa fragment). The sequence of the other PIF was Ala-Ser-Asp-Arg-Ser-Asn-Ala-Thr-Leu-Leu-Asp-Gly-Pro-Ser-Gly-Ala-Leu-Leu- Leu-Arg - Leu-Val-Gln-Leu-Ala-Gly-Ala-Pro-Glu-Pro-Ala-Glu-Pro-Ala-Gln-Pro-Gly-Val- Tyr, representing residues 109-147 of the vasopressin-neurophysin precursor. Synthetic peptides corresponding to the N-terminal region of POMC had significant PIF activity in vitro.

Secretion of activin by interstitial cells in the testis.

Activin, a dimer formed by the beta subunits of inhibin, has an effect that is opposite to that of inhibin in a number of biological systems. Which cell types secrete activin in vivo is not known. TM3 cells, a Leydig-derived cell line, contained messenger RNAs that hybridized with human beta A and beta B complementary DNA probes and were similar in size to the porcine messenger RNA for the beta subunits of inhibin. No hybridization to the inhibin alpha subunit was detectable in the TM3 cells. Conditioned medium from TM3 cells and from primary cultures of rat and porcine interstitial cells stimulated the release of follicle-stimulating hormone in a pituitary cell culture assay. It is likely that, in the testis, the Leydig cells secrete activin and the Sertoli cells produce inhibin, or a combination of both.

Recombinant expression and characterization of human activin A.

Activin, which stimulates the secretion of FSH from anterior pituitary cells, is a dimer of the beta-subunits of inhibin. Two species of activin (A and AB) have been purified from ovarian follicular fluid and characterized. We have been able to biosynthetically produce recombinant human activin A by constructing stable cell lines expressing the mRNA for the beta A subunit of human inhibin. These cell lines secreted a 24 kilodalton beta A dimer which stimulated FSH secretion in cultured pituitary cells. The ability of this protein to stimulate FSH secretion was sensitive to reduction of disulfide bonds, exhibited a slow onset of action, and was blocked by actinomycin D. In addition, recombinant activin A stimulated hemoglobin accumulation in K562 cells. These data show that recombinant activin A has the biochemical properties and biological activities that have been ascribed to native activin A. In addition, these results provide an independent confirmation, and thereby a final proof, of the structure and function of activin.

Activin stimulates secretion of follicle-stimulating hormone from pituitary cells desensitized to gonadotropin-releasing hormone.

The secretion of follicle-stimulating hormone (FSH) by pituitary cells is stimulated by activin and gonadotropin-releasing hormone, GnRH. To examine the possible interrelationships between the intracellular actions of these secretagogues, responsiveness to activin was tested following pretreatment with 0, 0.1, or 10 nM GnRH. In cells pretreated with 0 or 0.1 nM GnRH, FSH secretion was increased approximately 2-fold during a subsequent challenge with either activin or GnRH. In contrast, in cells pretreated with 10 nM GnRH, FSH secretion became unresponsive to GnRH but could still be stimulated 2-fold by activin. These results demonstrate that activin is able to stimulate FSH secretion in cells that have undergone desensitization to GnRH.

Photoaffinity labeling of pituitary GnRH receptors: significance of the position of photolabel on the ligand.

Photoreactive derivatives of GnRH and its analogues were prepared by incorporation of the 2-nitro-4(5)-azidophenylsulfenyl [2,4(5)-NAPS] group into amino acid residues at positions 1, 3, 6, or 8 of the decapeptide sequence. The modification of Trp3 by the 2,4-NAPS group led to a complete loss of the luteinizing hormone (LH) releasing as well as LH-release-inhibiting activity of the peptide. The [D-Lys(2,4-NAPS)]6 analogue was a very potent agonist that, after covalent attachment by photoaffinity labeling, caused prolonged LH secretion at a submaximal rate. [Orn(2,4-NAPS)]8-GnRH, a full agonist with a relative potency of 7% of GnRH, after photoaffinity labeling caused prolonged maximal LH release from cultured pituitary cells. In contrast, [Orn(2,5-NAPS)]8-GnRH, although being equipotent with the 2,4-NAPS isomer in terms of LH releasing ability, was unable to cause prolonged LH release after photoaffinity labeling. Thus, [Orn(2,4-NAPS)]8-GnRH is a very effective photolabeling ligand of the functionally significant pituitary GnRH receptor. Based on this compound, a pituitary peptidase resistant derivative, D-Phe6,[Orn(2,4-NAPS)]8-GnRH-(1-9)-ethylamide, was synthesized. This derivative showed high-affinity binding to pituitary membranes with a Kd comparable to those of other GnRH analogues. A radioiodinated form of this peptide was used for pituitary GnRH-receptor labeling. This derivative labeled 59- and 57-kDa proteins in rat and 58- and 56-kDa proteins in bovine pituitary membrane preparations, respectively. This peptide also labeled pituitary GnRH receptors in the solubilized state and therefore appears to be a suitable ligand for the isolation and further characterization of the receptor.

Gonadal development and gametogenesis in the hypogonadal mouse are restored by gene transfer.

These results describe controlled regulation of a mammalian neural gene in transgenic mice. Analysis of truncated GnRH-GAP genes in transgenic mice will enable us to define the DNA sequences responsible for this control. Furthermore, by separate mutation of the GnRH and GAP coding sequences we will be able to determine the relative importance of these two peptides in the development and maintenance of reproductive function.

The hypogonadal mouse: reproductive functions restored by gene therapy.

The hypogonadal (hpg) mouse lacks a complete gonadotropin-releasing hormone (GnRH) gene and consequently cannot reproduce. Introduction of an intact GnRH gene into the genome of these mutant mice resulted in complete reversal of the hypogonadal phenotype. Transgenic hpg/hpg homozygotes of both sexes were capable of mating and producing offspring. Pituitary and serum concentrations of luteinizing hormone, follicle-stimulating hormone, and prolactin were restored to those of normal animals. Immunocytochemistry and in situ hybridization showed that GnRH expression was restored in the appropriate hypothalamic neurons of the transgenic hpg animals, an indication of neural-specific expression of the introduced gene.

A prolactin-inhibiting factor within the precursor for human gonadotropin-releasing hormone.

The cloned complementary DNA sequence encoding the human gonadotropin-releasing hormone (GnRH) precursor protein was used to construct an expression vector for the bacterial synthesis of the 56-amino acid GnRH-associated peptide (GAP). GAP was found to be a potent inhibitor of prolactin secretion and to stimulate the release of gonadotropins in rat pituitary cell cultures. Active immunization with peptides corresponding to GAP sequences led to greatly increased prolactin secretion in rabbits.