Nerve Growth Factor Receptor-Mediated Gene Transfer

ARTICLE

doi:10.1016/j.ymthe.2004.11.005

Nerve Growth Factor Receptor-Mediated Gene Transfer Nan Ma,1,2,* Shan Shan Wu,1,2,* Yue Xia Ma,1,2 Xu Wang,1,2 Jieming Zeng,1,2 Guping Tong,1,2 Yan Huang,3 and Shu Wang1,2,4,y 1

Institute of Bioengineering and Nanotechnology, Singapore 117602 2 Institute of Materials Research & Engineering, Singapore 3 Ngee Ann Polytechnic, Singapore 4 Department of Biological Sciences, National University of Singapore, Singapore *

These authors have contributed equally to the work.

y

To whom correspondence and reprint requests should be addressed at the Institute of Bioengineering and Nanotechnology, 3 Research Link, Singapore 117602. Fax: (65)-6872-0785. E-mail: [email protected].

One obstacle to effective gene therapies for neurological disorders lies in the cell-type diversity of the nervous system, making it difficult to direct gene delivery vectors to specific types of cells. To meet this challenge, we have developed a recombinant peptide-based gene delivery vector that targets nerve growth factor (NGF) receptors. The peptide comprises a cell-targeting domain derived from the NGF hairpin motif containing loops 1 and 2 linked to a DNA-binding domain composed of SPKR repeats. In PC12 cells, it activated the high-affinity NGF receptor, TrkA, and displayed NGF-like bioactivity by promoting neurite outgrowth and cell survival after serum deprivation. When combined with a low molecular weight of polyethylenimine (PEI), the peptide condensed plasmid DNA into nanoparticles that efficiently transferred exogenous genes into PC12 cells, enhancing reporter gene expression 5600-fold over peptide-free DNA/PEI complexes. Co-incubation with free NGF inhibited this effect. Furthermore, the peptide enhanced gene expression in NGF-receptor-rich rat primary cortex neurons but not glial cells. An in vivo experiment targeting TrkA-expressing dorsal root ganglia demonstrated that the peptidecontaining complexes were 9- to 14-fold more efficient in transfection than controls. These properties make the chimeric peptide a promising gene delivery vector for targeting specific subtypes of neurons. Key Words: nerve growth factor, receptor, gene delivery, PC12 cells, neurons, dorsal root ganglia

INTRODUCTION Targeted gene delivery to selected cell types is a crucial aspect of enhancing the specificity of transgene expression, for the purposes of gene therapy as well as functional genomics research. The transfection efficiency of a given dose of genetic material would be improved by augmenting entry into the desired cells and reducing wasteful uptake by nontarget cells. For therapeutic applications, targeted gene delivery would ensure therapeutic efficacy in the cells of interest while limiting side effects, including immune, inflammatory, and cytotoxic responses, caused by the expression of exogenous genes in nontarget cells. One approach to targeted gene delivery is to use ligand-associated delivery vectors to take advantage of the natural process of receptor-mediated endocytosis. Many ligand – receptor systems have been investigated to date for targeted gene delivery [1 – 3]. Little has been established, however, on how to use such a system to target selected subtypes of neurons.

270

Neurons differ in their chemical signals and receptors, among which are nerve growth factor (NGF) and its receptors. NGF is an important molecule capable of promoting the development, growth, and survival of its responsive neurons, such as basal forebrain cholinergic neurons and sympathetic sensory neurons [4]. NGF is synthesized and secreted from neuronal target cells as noncovalently bound homodimers and functions through retrograde signaling after binding to its nerve-terminal-localized receptors and being internalized in endosomes [5]. NGF interacts selectively with one of the receptor protein tyrosine kinases, TrkA [6]. Internalized TrkA after NGF stimulation could be identified in endosomes and transported to neuron cell bodies to activate its signaling targets [7,8]. NGF also binds with relatively low affinity to a 75-kDa transmembrane glycoprotein, p75NTR [9], a receptor with two main physiological functions: acting as a coreceptor to modulate Trk signaling and initiating autono-

MOLECULAR THERAPY Vol. 9, No. 2, February 2004 Copyright D The American Society of Gene Therapy 1525-0016/$30.00

ARTICLE

doi:10.1016/j.ymthe.2004.11.005

mous signaling cascades that result in either the induction of apoptosis or the promotion of survival [9,10]. Since receptor-mediated endocytosis of NGF is a rapid and efficient process [5], NGF may serve as a candidate targeting ligand to direct gene vectors to TrkA- and/or p75NTR-positive cells. Extensive studies on the structural determinants of NGF in interacting with its receptors have identified NGF loop 1 as a key active site for p75NTR receptorbinding [11,12] and the N-terminal region, loop 2, and loop 4 as important structures for binding to TrkA [13,14]. Based on these studies, small bioactive peptide mimetics using sequences from the NGF loop regions have been developed [15 – 18]. In the current study we developed a targeted gene delivery vector by combining the NGF-receptor-binding activities of NGF loops 1 and 2 with the DNA-binding function of SPKR repeats in a single chimeric peptide. We demonstrated that the peptide enhanced gene expression mediated by DNA/poly-

mer complexes specifically in cells expressing NGF receptors.

RESULTS Chimeric Peptide SPKR4NL1-2 and Its Biochemical and Biological Effects SPKR4NL1-2 is a chimeric peptide containing a DNAbinding domain, SPKRSPKRSPKRSPKR, and a targeting domain incorporating the NGF hairpin motif of loops 1 and 2. The two parts are linked by an a-helical linker, TYLSEDELKAAEAAFKRHNPT. The SPKR4 sequence, derived from the histone H1 DNA-binding domain, and the a-helical linker were first used together by Fortunati et al. [19] in a DNA delivery vector. The linker is flanked by flexible glycine residues and serves to allow independent action of the DNA-binding and targeting domains. The targeting domain comprises a cysteine residue followed by ami-

FIG. 1. SPKR4NL1-2 activates TrkA and ERK. PC12 cells preincubated in RPMI 1640 medium containing 0.5% FBS and 0.25% horse serum for 2 days were treated for 20 min with NGF or peptides diluted in serumfree RPMI 1640. The cell lysates were analyzed by immunoblotting using primary antibodies specific to either phospho-TrkA or phosphorylated ERK 1 and 2. (A) PhosphoTrkA Western blot. The cells were treated with NGF (20 ng/ml), SPKR4NL1-2 (8 AM), SPKR4NL1-2/DNA complexes (8 AM, N/P ratio of 5), or (SPKR)4 (8 AM). (B) PhosphoERK Western blot. The cells were treated with NGF (20 ng/ml), various concentrations of SPKR4NL1-2 from 1 to 8 AM, or 8 AM (SPKR) 4 . (C) A TrkA inhibitor blocks SPKR4NL1-2-induced ERK activation. The PC12 cells were preincubated with 0, 10, 20, 50, and 100 nM K-252a (TrkA tyrosine kinase inhibitor) for 10 min before treatment with NGF (20 ng/ml) or SPKR4NL1-2 (8 AM). Molecular weights of protein standards are shown on the left.

MOLECULAR THERAPY Vol. 9, No. 2, February 2004 Copyright D The American Society of Gene Therapy

271

ARTICLE

no acids 17 – 67 of human NGF, such that a disulfide bridge can form between this cysteine residue and C58 in NGF. Based on the crystal structure of NGF [14], we judged that such a disulfide bond would help the

doi:10.1016/j.ymthe.2004.11.005

targeting domain assume the native conformation of loops 1 and 2. The chimeric peptide was expressed with a His10-tag in the Escherichia coli strain BL21(DE3). The eluates

FIG. 2. SPKR4NL1-2 has NGF-like bioactivity. (A and B) SPKR4NL1-2 promotes neurite outgrowth. PC12 cells were treated with 8 AM (A) (SPKR)4 or (B) SPKR4NL1-2 for 3 days. (C) SPKR4NL1-2 promotes survival of PC12 cells deprived of serum for 3 days. Different concentrations of SPKR4NL1-2, from 0 to 16 AM, were added at the time of serum withdrawal and 10 ng/ml NGF was used as a positive control. Cell survival was estimated by an MTT assay and expressed as a percentage of maximal NGF-promoted cell survival.

272


doi:10.1016/j.ymthe.2004.11.005

from nickel-chelate affinity chromatography of E. coli lysates predominantly contained a protein similar in size to the expected molecular weight of SPKR4NL1-2 at 12.9 kDa. The produced chimeric peptide was able to activate TrkA and the ERK 1 and 2 proteins in its downstream signaling pathways (Fig. 1). We observed TrkA phosphorylation by Western blot analysis of the cell lysate of TrkA-positive PC12 cells treated with SPKR4NL1-2, whereas this response was not elicited by similar treatment with a control peptide, (SPKR)4, that lacked the targeting domain (Fig. 1A). We also observed enhanced phosphorylation of ERK 1 and 2 after treatment with SPKR4NL1-2 at concentrations ranging from 1 to 8 AM for 15 min, compared to that after treatment with 8 AM (SPKR)4 control peptide (Fig. 1B). To dem-

ARTICLE

onstrate that the ERK activation occurred specifically through TrkA phosphorylation, we pretreated cells with K-252a, a TrkA tyrosine kinase inhibitor, to see if ERK phosphorylation would be reduced. Fig. 1C shows that ERK phosphorylation induced by 8 AM SPKR4NL1-2 was reduced and eventually eliminated by increasing concentrations of K-252a. In response to NGF, PC12 cells stop proliferating and differentiate into sympathetic neuron-like cells. Within 2 – 3 days of NGF treatment, cell morphology changes and neurites can be seen projecting from the cells. We tested SPKR4NL1-2 to see if it could also promote neurite outgrowth in PC12 cells. We treated the cells with 8 AM (SPKR)4 or SPKR4NL1-2 for 3 days. Compared to (SPKR)4, SPKR4NL1-2 displayed NGF-like FIG. 3. SPKR4NL1-2 binds to and condenses plasmid DNA. (A) Electrophoretic mobility of plasmid DNA through a 1% agarose gel was reduced by SPKR4NL1-2 binding. Various amounts of peptide were mixed with 0.1 Ag of DNA in a volume of 20 Al for 30 min before electrophoresis. (B) The fluorescence of ethidium bromide intercalated in DNA was reduced by the addition of SPKR4NL1-2 that displaced ethidium bromide. The indicated amounts of peptide were added to 0.8 Ag DNA premixed with ethidium bromide. (C and D) Atomic force microscopy images of supercoiled plasmid DNA and SPKR4NL1-2/DNA/PEI600 complexes, respectively. The N/P ratios of peptide/DNA and PEI600/DNA were 2/1 and 10/1, respectively. Both images were collected as 4-Am2 fields and the scale bar represents 0.5 Am.


273

ARTICLE

bioactivity by promoting neurite outgrowth (Figs. 2A and 2B). We also investigated whether SPKR4NL1-2 shared the ability of NGF to promote survival of PC12 cells deprived of serum for 3 days. The addition of SPKR4NL1-2 (8 AM) to the serum-free medium maintained PC12 viability at 80% of the maximal survival rate promoted by 10 ng/ml NGF (Fig. 2C). The survivalpromoting effect exhibited a dose – response trend over the range of 2 – 8 AM polypeptide. SPKR4NL1-2 Binds to Plasmid DNA and Enhances Gene Delivery to PC12 Cells We assessed the DNA-binding activity of SPKR4NL1-2 first in a DNA retardation assay. When mixed with 0.1 Ag of plasmid DNA, SPKR4NL1-2 dose-dependently reduced the mobility of DNA through an agarose gel under electrophoresis (Fig. 3A), indicating that the peptide had bound to the DNA and reduced its charge/mass ratio. A significant reduction in DNA migration occurred when we used 1.25 Ag of peptide (Fig. 3A). We also characterized the ability of SPKR4NL1-2 to bind to and condense DNA by

doi:10.1016/j.ymthe.2004.11.005

measuring the quenching of fluorescence when intercalated ethidium bromide was displaced from DNA by the peptide. The fluorescence dropped sharply when 8 Ag of peptide was added to 0.8 Ag of DNA, indicating the amount of SPKR4NL1-2 needed for DNA condensation (Fig. 3B). This result was in agreement with that from the DNA retardation assay. Atomic force microscopy revealed, however, that SPKR4NL1-2 only partially condenses plasmid DNA at a nitrogen/phosphate (N/P) ratio of 2.5 (result not shown). The N/P ratio is defined as the number of protonatable nitrogen atoms per DNA phosphate group and is a frequently used measure of the charge balance in polycation or peptide/DNA complexes. For the purposes of calculating the N/P ratio for the peptides used in the study, we take the number of basic amino acid residues in the DNAbinding domain to be the number of ‘‘N’’ moieties per peptide molecule, that is, a total of 8 from the 4 lysine and 4 arginine residues in (SPKR)4. With the addition of low-molecular-weight polyethylenimine, PEI600, at an N/P ratio of 10, the loosely looped structure of free

FIG. 4. SPKR4NL1-2 enhances polycation-mediated gene transfection of PC12 cells. To form complexes used for each well of a 48-well plate, 0.5 Ag of pCAGluc plasmid was first mixed with varying amounts of the peptide and incubated for 30 min, after which polycation was added and the mixture incubated for a further 30 min. Luciferase activity is given in relative light units (RLU)/mg total protein. (A) PEI600 at an N/P ratio of 10. (B) Poly-L-lysine (PLL) at an N/P ratio of 10.

274


doi:10.1016/j.ymthe.2004.11.005

ARTICLE

plasmid DNA molecules was no longer seen and compact nanoparticles in the size range of 100 to 200 nm were formed (Figs. 3C and 3D). Encouraged by the results from these DNA binding assays and the observation that SPKR4NL1-2 could activate TrkA even after binding to DNA (Fig. 1A), we went further to test whether SPKR4NL1-2 could improve gene delivery efficiency using NGF-receptor-positive PC12 cells. Complexes made from a luciferase-encoding reporter plasmid, pCAGluc, and SPKR4NL1-2 alone had relatively weak transfection efficiencies, but the addition of polycationic polymer to fully condense the DNA created an effective gene delivery system. We chose PEI600 and poly-L-lysine (PLL) for their DNA-condensing abilities and low background transfection efficiencies. In 48-well plates, we transfected PC12 cells with triple complexes containing 0.5 Ag of pCAGluc, either PEI600 or PLL at an N/P ratio of 10, and various amounts of SPKR4NL1-2. To prepare the complexes, we added the peptide to the DNA in OptiMEM medium and incubated them for 30 min at room temperature, after which we added the polycation and incubated the mixture for a further 30 min before use. We also added chloroquine, an endosomolytic agent, with the PLLcontaining complexes. Fig. 4A shows that SPKR4NL1-2 enhanced PEI600-mediated gene transfer in a dose-dependent manner. Transgene expression using the highest amount of SPKR4NL1-2 tested was 5600-fold higher than that achieved with PEI600 alone. Similarly, with PLL-mediated gene delivery, transfection efficiency with the optimal amount of SPKR4NL1-2 was 1000 times higher than without SPKR4NL1-2 (Fig. 4B). Specificity of SPKR4NL1-2-Mediated Gene Delivery To examine the specificity of the SPKR4NL1-2 peptide in mediating gene delivery, we first compared the transfection efficiencies of the peptide and its control, (SPKR)4, in PC12 cells using a reporter gene encoding an enhanced green fluorescent protein (EGFP). When we applied (SPKR) 4 /DNA/PEI600 triplexes, only a few EGFP-positive cells could be seen under a fluorescence microscope (result not shown). The number of EGFPpositive cells increased dramatically when we used SPKR4NL1-2/DNA/PEI600 triplexes, with approximately 30% of the PC12 cells exhibiting bright green fluorescence (result not shown). The result indicates that the targeting domain designed to recognize NGF receptors was essential to the enhanced gene expression mediated by SPKR4NL1-2. To demonstrate further that SPKR4NL1-2-mediated gene delivery depends on the peptide binding to NGF receptors, we studied the effect of adding exogenous NGF (200 ng/ml) during transfection with SPKR4NL1-2/ pCAGluc/PEI600 complexes in PC12 cells. Fig. 5 shows that the presence of NGF reduced the efficiency of intracellular accumulation of labeled plasmid DNA


FIG. 5. NGF inhibits (A) DNA accumulation and (B) gene expression mediated by SPKR4NL1-2 in PC12 cells. To form complexes, fluorescein-labeled luciferase plasmid or pCAGluc plasmid was first mixed with either SPKR4NL1-2 or (SPKR)4 at an N/P ratio of 2.5 and incubated for 30 min, after which PEI600 (N/P ratio of 10) was added and the mixture incubated for a further 30 min before transfection. NGF (200 ng/ml) was added to some wells during transfection. Flow cytometric analysis was carried out after transfection to investigate intracellular accumulation of DNA/SPKR4NL1-2 complexes. To investigate gene expression, luciferase activities were measured 24 h later and are shown in RLU/mg protein. *P < 0.05 compared to the cells treated with the same SPKR4NL1-2/DNA/PEI600 complexes but without NGF.

and the luciferase gene expression by 82 and 97%, respectively. When SPKR 4 NL1-2 was replaced with (SPKR)4, the transfection efficiency was very low and NGF treatment had no significant effects. The action of NGF as a competitive inhibitor of SPKR4NL1-2/pCAGluc/PEI600 complexes indicates that SPKR4NL1-2-mediated gene delivery was targeted to NGF receptors. We subsequently tested cell-type specificity of SPKR4NL1-2-enhanced PEI600-mediated gene delivery on primary cultures of NGF-receptor-expressing cortical neurons and NGF-receptor-poor glial cells isolated from 20-day-old embryonic rats. We first assayed cells for the

275

ARTICLE

expression of their high-affinity NGF receptor TrkA by flow cytometric analysis using an antibody against TrkA. Approximately 48% of rat primary cortical neurons were TrkA positive, while almost no primary cortical glial cells were positively stained (Fig. 6A). We then transfected cortical neurons and glial cells in 48-well plates with SPKR4NL1-2/pCAGluc/PEI600 or SPKR4/pCAGluc/ PEI600 complexes containing 0.25 Ag/well pCAGluc, peptide at an N/P ratio of 2.5, and PEI600 at an N/P ratio of 5. Fig. 6B shows that gene delivery by PEI600 alone was negligible in both neuronal and glial cells. When the control peptide, (SPKR)4, was added to the complexes, the additional DNA-condensing ability enhanced transgene expression in both types of cells, although the impact was higher in glia than in neurons. Triple complexes containing SPKR4NL1-2 were nine

doi:10.1016/j.ymthe.2004.11.005

times more effective in neurons than (SPKR)4 triple complexes, whereas in glial cells, there was essentially no difference in performance between SPKR4NL1-2 and (SPKR)4. We compared SPKR4NL1-2 with commonly used nonviral gene carriers PEI25kDa and Lipofectamine2000 in terms of gene delivery efficiency in primary cortical neurons and glial cells (Fig. 7). With optimal carrier/ DNA ratios, Lipofectamine2000 gave the highest transfection efficiency in both types of cells among three tested gene carriers. Although SPKR4NL1-2 was less efficient in glial cells compared with PEI25kDa, it mediated the transgene expression in neurons comparable to that of PEI25kDa. Notably, SPKR4NL1-2 was the only one among the three carriers giving a much higher level of gene expression in neurons compared to glial cells (Fig.

FIG. 6. SPKR4NL1-2 mediates gene delivery to primary neurons and glial cells. (A) Flow cytometric analysis of the percentage of cells expressing TrkA receptors. Primary rat cortical neurons or glial cells were incubated with rabbit anti-TrkA, followed by staining with sheep anti-rabbit IgG-FITC. (B) Luciferase gene expression. In a 48-well plate, primary rat cortical neurons or glial cells were transfected with SPKR4NL1-2/DNA/PEI600, (SPKR)4/DNA/PEI600, or DNA/PEI600 complexes. The complexes were prepared with 0.25 Ag of pCAGluc plasmid mixed with peptides, if present, and PEI600 at N/P ratios of 2.5 and 5, respectively. Luciferase activities were measured 24 h later and are shown in RLU/mg protein. **P < 0.01 compared to the neurons treated with DNA complexed with (SPKR)4 and PEI600.

276


ARTICLE

doi:10.1016/j.ymthe.2004.11.005

SPKR4NL1-2-Mediated in Vivo Gene Delivery to Dorsal Root Ganglia (DRG) To evaluate the efficacy of SPKR4NL1-2-mediated transfection in vivo, we injected gene delivery complexes intrathecally into the lumbar regions of the rat spinal cord. The lumbar spinal cord is in general an NGFreceptor-poor region. A large portion of DRG neurons, however, expresses TrkA receptors [20]. As DRGs lie at the distal end of the dorsal root of the spinal cord inside the apex of the dural sleeve [21], intrathecally applied peptide/DNA complexes in the cerebrospinal fluid would be able to penetrate into the DRG. We injected complexes containing 4 Ag of pCAGluc per rat. We dissected the lumbar spinal cord and the lumbar DRG 3 days later and analyzed them for luciferase activity. While luciferase activity was barely detectable in the spinal cord, DRG transgene expression produced by SPKR4NL1-2/pCAGluc/ PEI600 complexes was 9 to 14 times that mediated by the pCAGluc/PEI600, (SPKR)4/pCAGluc/PEI600, or pCAGluc/ PEI25kDa controls (Fig. 8A). To evaluate in vivo cell-type specificity of SPKR4NL1-2mediated transfection, we carried out immunohistochemical double labeling of DRG sections using antibodies against luciferase to visualize transfected cells and antibodies against the neuron-specific nuclear protein (NeuN) to visualize neurons. The results demonstrated that almost all of the transfected cells in DRG were neurons (Fig. 8B).

DISCUSSION

FIG. 7. Comparison of various gene carriers for transfer efficiency in primary neurons and glial cells. In a 48-well plate, primary rat cortical neurons or glial cells were transfected with SPKR4NL1-2/DNA/PEI600, DNA/PEI25kDa, or DNA/Lipofectamine complexes. The complexes were prepared with 0.25 Ag of pCAGluc plasmid mixed with peptides and PEI600 as described for Fig. 6, PEI25kDa at an N/P ratio of 10, or Lipofectamine2000 with a w/w ratio of 1:2. Luciferase activities were measured 24 h later and are shown in RLU/mg protein in (A) and the percentage of neuronal readings from glial RLU in (B).

7A). The level in neurons mediated by this peptide was 587% of that in glial cells (Fig. 7B). The corresponding percentages produced by Lipofectamine2000 and PEI25kDa were 23 and 20%, respectively (Fig. 7B). The difference suggested that SPKR4NL1-2 preferentially functioned in neurons over glial cells.


To develop a gene delivery system capable of targeting specific subtypes of neurons in complex organs like the brain and the spinal cord, the present study has designed, developed, and tested a chimeric peptide, SPKR4NL1-2, that can bind to DNA, direct DNA complexes to cells equipped with NGF receptors, and mediate intracellular gene delivery. Several lines of evidence implicate the involvement of NGF receptors in gene delivery mediated by this peptide. First, it possessed NGF-like bioactivity in NGF-receptor-positive PC12 cells, even after binding to DNA, and its activity in signal transduction was blocked by the TrkA inhibitor K252a. Second, the peptide dose-dependently mediated gene delivery to cells expressing NGF receptors, but had no enhancement effect on the cells lacking these receptors. Third, only chimeric polypeptides containing both a receptor-targeting domain and a DNAbinding domain were functioning in mediating gene expression; the targeting domain contains NGF loops 1 and 2, two domains known to be important in interacting with TrkA and p75NTR. Fourth, gene transfer to NGF-receptorexpressing cells was inhibited by co-incubation with an excess amount of NGF. Fifth, the gene delivery vector was able to target a region expressing NGF receptors (the DRG) even in a complex in vivo environment, a feature making it promising for clinical application.

277

ARTICLE

doi:10.1016/j.ymthe.2004.11.005

FIG. 8. SPKR4NL1-2 mediates in vivo gene delivery to dorsal root ganglia (DRG). Complexes formed from 4 Ag of pCAGluc, SPKR4NL1-2 at an N/P ratio of 2.5, and PEI600 at an N/P ratio of 10 were injected intrathecally into the lumbar spinal cord in rats. Complexes formed with pCAGluc and PEI600, PEI600 plus SPKR4, or PEI25kDa were used as controls. DRG and the lumbar spinal cords were collected 3 days after injection. (A) Luciferase gene expression. Results are expressed in RLU/mg protein. **P < 0.01 compared to the rats treated with DNA complexed with (SPKR)4/PEI600. (B) Confocal images of luciferase expression in neurons in DRG. Frozen sections of DRG collected from animals injected with SPKR4NL1-2/DNA/PEI600 complexes were used for double immunostaining against luciferase protein to show transfected cells and against neuron-specific nuclear protein (NeuN) to show neurons.

Small peptide mimetics of NGF using sequences from its hairpin loop regions have been developed to activate NGF-receptor-related signal transduction and promote NGF-like neurotrophic effects [12,15 – 18,22]. These peptide mimetics need to be structural analogs, not just sequence analogs, of NGF loops to be biochemically and biologically active, and hence, cyclization is a common feature. Differing from the short, cyclized loops used in the previous studies, an NGF hairpin motif, a supersecondary structural element built up from two adjacent h strands that are connected by a loop region, was utilized in the present study. The A and B strands of NGF are connected together at one end by NGF loops 1 and 2. These elements of the hairpin motif, comprising amino acids 17 – 67 of NGF, have been used as the

278

targeting ligand of our chimeric peptide-based gene delivery vector. By using a hairpin motif instead of a short, artificially cyclized loop, we expect the native loop conformation to be well stabilized by hydrogen-bonding between h strands, while the flexibility required for induced-fit ligand recognition by NGF receptors will be better conserved. NGF, like other members of the neurotrophin family, is a noncovalently bound homodimer that stimulates dimerization of its receptors, thus triggering a signal cascade and inducing receptor-mediated endocytosis [23 – 25]. In designing short peptide mimetics of neurotrophins, much work has been put into emulating this effect using cyclic dimers or bicyclic compounds [12,15 – 18]. In contrast, the peptide SPKR4NL1-2 used in this


doi:10.1016/j.ymthe.2004.11.005

study is monomeric and monovalent but exhibits detectable NGF-like bioactivity. During its design, we anticipated that after the polypeptide was combined with DNA, nanoparticle formation would result in many monovalent ligands being tethered close together on the particle surface, such that neighboring pairs of ligands would be able to facilitate receptor dimerization. Theoretically, SPKR4NL1-2/pCAGluc complexes formed at an N/P ratio of 1 would contain on the order of 1000 polypeptide molecules available for binding to each plasmid DNA molecule; at the higher N/P ratios used, the number becomes even larger. However, this does not explain why free SPKR4NL1-2 peptides were able to activate TrkA and promote neurite outgrowth and neuron survival. We propose that the positively charged DNA-binding domain allows the peptide to bind loosely to the weakly negatively charged cell surface, creating a high local concentration of the ligand that offers the opportunity for receptor dimerization. It has also been shown recently that genuine monomeric and monovalent ligands of TrkA can be partial agonists, functioning through an allosteric effect on the receptor [22]. Further structural analysis of receptor – ligand complexes is necessary to explore these hypotheses. Targeted gene delivery in the CNS is currently achieved simply through direct stereotactic injection of naked DNA or gene vectors into well-defined anatomical locations, transducing various types of cells around the injection site. Retrograde axonal transport of viral or nonviral gene vectors offers another way to target neurons in different regions by choosing an appropriate injection site either in the periphery or in more accessible regions of the CNS. This method, however, transduces all projection neurons and does not distinguish between subtypes of neurons. When used in combination with the above injection procedures, the NGF-receptor-targeting peptide reported here might be able to direct gene vectors specifically to selected neurons in a particular region. This approach can potentially be used advantageously in gene therapy to treat disorders in which neurons expressing NGF receptors are affected. In Alzheimer’s disease, the neurons with the most prominent pathological changes are basal forebrain cholinergic neurons, which express TrkA [26,27]. A gene delivery system that targets TrkA may therefore help to transfer therapeutic genes into the neurons to augment cholinergic functions.

MATERIALS AND METHODS Design, expression, and purification of peptides. By a PCR-mediated gene assembly method [28], a DNA fragment encoding (SPKR)4 and the ahelical linker, TYLSEDELKAAEAAFKRHNPT, was first constructed and cloned into the pET-16b expression vector (Novagen, Madison, WI, USA) between the NdeI and the BamHI site, downstream of a His10 tag, to create pET16-SPKR4linker. The DNA sequence encoding full-length NGF was first cloned into a TA vector by PCR amplification of human brain cDNA (Clontech, Palo Alto, CA, USA). The DNA sequence


ARTICLE

corresponding to amino acids 17 – 67 of NGF was amplified with the forward primer 5V-CTGGATCCTGCAGTGTCAGCGTGTGG-3V and the reverse primer 5V-ATGGATCCTCACCCGCTGTCAACG-3V. These primers introduced BamHI restriction sites at both ends to enable the digested fragment to be ligated in-frame with the a-helical linker sequence. In addition, the forward primer introduced an additional Cys codon. The construct, designated pET16-SPKR4NL1-2, was checked for correct insert orientation and sequenced by conventional methods. The peptide produced will comprise the following sequence: MGHHHHHHHH HHSSGHIEGR HMSPKRSPKR SPKRSPKRGG TYLSEDELKA AEAAFKRHNP TGSCSVSVWV GDKTTATDIK GKEVMVLGEV NINNSVFKQY FFETKCRDPN PVDSG. For peptide expression, the E. coli strain BL21(DE3) bearing pET16SPKR4NL1-2 was vigorously shaken at 37jC in LB medium containing 100 Ag/ml ampicillin until the optical density reached 0.7. Expression of the recombinant protein was then induced at 30jC by the addition of 1 mM isopropyl-h-D-thiogalactopyranoside (Bio-Rad, Hercules, CA, USA). The bacteria were harvested by centrifugation 90 min later and frozen. The cell pellet was resuspended in cold lysis buffer (20 mM Tris – HCl, pH 7.9, 0.5 M NaCl, 20 mM imidazole) supplemented with EDTA-free protease inhibitor (Calbiochem, San Diego, CA, USA) and 0.1 mg/ml lysozyme (Sigma – Aldrich, St. Louis, MO, USA). The cell lysate was sonicated until no longer viscous and cleared by centrifugation. The supernatant was ¨ KTAexplorer purified by nickel-chelate affinity chromatography on an A FPLC (Amersham Biosciences, Uppsala, Sweden). A 1-ml HisTrap column (Amersham Biosciences) equilibrated with lysis buffer was used to purify lysate from 1.6 L of culture; washing and eluting were performed by increasing the imidazole concentration. The eluted fractions were analyzed by SDS gel electrophoresis and Western blotting using antibodies against His-tag (Qiagen, Hilden, Germany). Approximately 4 to 6 mg of protein was expressed per liter of the culture, giving a typical yield of 1 mg per liter of culture after purification. The fractions containing the purified protein were pooled and dialyzed against distilled water at 4jC. The control peptide (SPKR)4, SPKRSPKRSPKRSPKR, was chemically synthesized and purified by Alpha Diagnostic (San Antonio, TX, USA). Reporter plasmid and preparation of DNA complexes. The firefly luciferase reporter plasmid used was pCAGluc (kindly donated by Yoshiharu Matsuura, National Institute of Infectious Diseases, Tokyo, Japan), with a composite promoter CAG consisting of the cytomegalovirus immediateearly enhancer, chicken h-actin promoter, and rabbit h-globin polyadenylation signal. The enhanced green fluorescent protein reporter plasmid, pEGFPC1, was obtained from Clontech. Plasmid DNA was amplified in E. coli and purified with the HiSpeed Plasmid Kit (Qiagen). The quantity and quality of the purified plasmid DNA were assessed by its optical density at 260 and 280 nm and by electrophoresis in 1% agarose gel. The purified plasmid DNA was resuspended in TE buffer and kept in aliquots at a concentration of 1 mg/ml. These plasmids were complexed with the above peptides and either polyethylenimine (PEI600, MW 600; Sigma – Aldrich) or low-molecularweight PLL (MW 1000; Sigma – Aldrich) for gene transfection. To form triplexes, equal volumes of plasmid in 5% glucose and polypeptides in OptiMEM (Invitrogen, Carlsbad, CA, USA) were first mixed and incubated for 30 min at room temperature, after which PEI600 or PLL in OptiMEM was added and the mixture incubated for a further 30 min before use. The required amount of PEI600 or PLL was calculated based on the nitrogen/ phosphate ratio used and by taking into account that 1 Ag of DNA contains 3 nmol of phosphate. DNA binding assay. A DNA retardation assay was carried out to assess qualitatively the ability of SPKR4NL1-2 to bind to DNA and thus to retard its migration (by decreasing its charge/mass ratio) through an agarose gel under electrophoresis. Various amounts of the peptide were added to 0.1 Ag of plasmid DNA made up to 20 Al with 5% glucose. The mixtures were vortexed and incubated at room temperature for 30 min before being subjected to electrophoresis in a 1.0% Tris – borate – EDTA agarose gel stained with ethidium bromide for visualization under ultraviolet light.

279

ARTICLE

The quenching of the fluorescence of ethidium bromide intercalated in DNA was also used to measure the DNA binding and condensing ability of SPKR4NL1-2. Ethidium bromide (153 Al of a 0.01% solution) was added to 96 Ag of plasmid DNA and diluted with water to a final volume of 6 ml. Various amounts of the peptide dissolved in a volume of 50 Al were placed in wells of a 96-well microplate, to which 50 Al of the DNA – ethidium bromide solution was added. After 5 min of mixing, 100 Al of water was added. The fluorescence reading was taken using wavelengths of 485 nm for excitation and 593 nm for emission on a SPECTRAFluor Plus microplate reader (TECAN, Maennedorf, Switzerland). The topology of DNA samples, SPKR4NL1-2/DNA complexes, and SPKR4NL1-2/DNA/PEI600 complexes was examined by atomic force microscopy (AFM). Naked plasmid DNA was diluted with Hepes – Mg buffer (10 mM MgCl2, 40 mM Hepes, pH 7.6) to a final concentration of 30 ng/ml. After incubation for 5 – 10 min at room temperature, 2 Al of the solution was deposited onto the center of a freshly cleaved mica surface. After being left to stand for 10 min, the sample was rinsed with 2 – 5 ml of deionized water. The sample was then blown dry with a gentle stream of nitrogen. SPKR4NL1-2/DNA complexes and SPKR4NL12/DNA/PEI600 complexes were prepared at a peptide/DNA N/P ratio of 2.5/1 and diluted with deionized water. After incubation for 30 min at room temperature, 3 Al of each aqueous complex solution (with final DNA amounts of 10 ng in each sample) was deposited onto mica, rinsed, and dried as described above. AFM (Multi Mode, Nanoscope III; Digital Instruments, Santa Barbara, CA, USA) was used to obtain images in air at room temperature in tapping mode using Nanoprobe silicon tips (Digital Instruments, Woodbury, NY, USA) with a spring constant of 34 N/m and a resonance frequency of about 280 kHz. Scanning was performed at a scan speed of 0.5 – 2 Hz. The 256 256-pixel images were captured with scan size 1 – 2 Am. The tip loading force was minimized to avoid structural changes to the samples. Cell cultures. Rat PC12 pheochromocytoma cells obtained from American Type Culture Collection (Manassas, VA, USA) were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS; Sigma – Aldrich) and 5% horse serum (Sigma – Aldrich) in a humidified incubator with 5% CO2 at 37jC. PC12 cells are known to express TrkA and p75NTR receptors. Primary cultures of neurons were established from the cortices of embryonic Wistar rats at gestational day 20. Cortices freed of meninges were dissected and individual cells were dispersed mechanically by trituration of minced tissue in 3 ml DMEM supplemented with 2% FBS. The suspension was allowed to settle in a centrifuge tube. The cells in the supernatant were collected by centrifugation and resuspended gently in DMEM with 10% FBS. Viability of the cells was assessed prior to plating using trypan blue. Cells were plated onto microplates precoated with poly-L-lysine/laminin at a density of 7.5 105 viable cells/cm2. After a 2h incubation period to allow neurons to attach, the medium and unattached cells were removed and replaced by serum-free DMEM/F12 medium with 1% N2 supplement (Invitrogen). The cell were then incubated at 37jC in 5% CO2 in a humidified incubator. After 2 – 5 days of culture, the neuron cells were used for transfection experiments. Primary cultures of glial cells were also established from the cortices of 20-day-old embryonic Wistar rats. Individual cells were collected as described above. The cells were plated at a density of 4 105 viable cells/cm2 on poly-L-lysine/laminin-coated dishes and grown to confluence for 7 days in DMEM/F12 medium supplemented with 10% FBS. The day before transfection, the cells were detached and plated into coated microplates at a density of 2.5 104 cells per well. Biochemical and biological assays. To detect the activation of TrkA and ERK, PC12 cells were seeded into six-well plates at a density of 2 106 cells per well and cultured in RPMI 1640 containing 0.5% FBS and 0.25% horse serum for 2 days. The low serum concentrations were necessary to reduce the basal levels of phosphorylation. The cells were incubated in fresh RPMI 1640 with 0.5% FBS and 0.25% horse serum 2 h before treatment. The peptides or NGF-positive controls (Invitrogen) were then added in serum-free medium to stimulate the PC12 cells for 15 to 20 min. The cells were then washed and sonicated in SDS – PAGE loading buffer.

280

doi:10.1016/j.ymthe.2004.11.005

The cell lysates were electrophoresed on 12.5% SDS gels and transferred to nitrocellulose membranes. Membranes were incubated with phosphoTrkA (Tyr490) antibody or phospho-p44/p42 MAPK (Thr202/Tyr204) monoclonal antibody obtained from Cell Signaling Technology (Beverly, MA, USA). After the application of appropriate horseradish peroxidaseconjugated secondary antibodies, colorimetric detection was performed using the Amplified Opti-4CN Substrate Kit from Bio-Rad. In some experiments, PC12 cells were preincubated, before addition of polypeptides or NGF, with various concentrations of K-252a (Calbiochem), a TrkA tyrosine kinase inhibitor. In the PC12 neurite outgrowth assay, SPKR4NL1-2 and it control peptide, (SPKR)4, were examined for their ability to induce neurite outgrowth in PC12 cells. Peptides in a range of concentrations were added to PC12 cells cultured in serum-enriched RPMI 1640 medium. The cells were incubated for 3 days before being photographed. For the PC12 survival assay, the cells were seeded in collagen-coated 96-well plates in serum-free RPMI 1640 containing 10 ng/ml NGF and cultured for 6 days. The medium was then changed to serum-free RPMI 1640 with peptides at the indicated concentrations. NGF at 10 ng/ml was found to be an optimal concentration for PC12 survival (data not shown), and this was used as the positive control. After 4 days of incubation, cell viability was estimated with the MTT assay. Cell survival was expressed as a percentage of the viability in the NGF-treated controls. Flow cytometry. For flow-cytometric analysis of TrkA receptor expression, primary cortical neurons or glial cells were dislodged from 60-mm culture dishes with 1 mM EDTA in Hanks’ buffered saline, fixed by 4% paraformaldehyde in PBS, blocked with 1% BSA in PBS, and stained with rabbit anti-TrkA antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 1 h. After being washed with PBS, cells were stained with sheep anti-rabbit IgG-FITC (Sigma – Aldrich) for 30 min. Cells were washed and resuspended in PBS. Flow cytometry was processed with a Coulter flow cytometer at the Clinical Research Center of the National University of Singapore. For flow-cytometric analysis of intracellular accumulation of DNA/ SPKR4NL1-2 complexes, PC12 cells were plated in six-well plates at a cell density of 2 106 cell per well and incubated overnight. The culture medium was replaced with 400 Al OptiMEM, and an aliquot of 75 Al of complexes containing 1.5 Ag of fluorescein-labeled luciferase plasmid (Gene Therapy Systems, San Diego, CA, USA) was added to each well and incubated with the cells for 3 h at 37jC. After the incubation, the cells were washed with PBS three times, fixed with 4% paraformaldehyde at room temperature for 30 min, and then analyzed. Gene delivery in vitro. Gene delivery complexes were tested for transfection efficiency in vitro in TrkA- and p75NTR-expressing PC12 cells and primary cortex neurons, as well as NGF-receptor-negative primary cortex glial cells. For PC12 cell transfection, the cells were grown to 50 – 70% confluency in microplate wells. The gene delivery complexes were prepared according to the specifications presented for each experiment. Transfection of PC12 cells and glial cells was accomplished by replacing the serum-enriched medium with half the usual volume of serum-free medium, OptiMEM (Invitrogen), containing the gene delivery complexes. After 4 h of incubation, the medium was changed back to the usual serum-enriched medium. Transfection in primary cultures of cortical neurons was done in a similar fashion. However, since the serum-free DMEM-F12/N2 medium was used for culturing the neurons, the complexes were simply added directly to an amount of DMEM-F12/N2 to make up half the usual volume. Four hours later, an equal volume of DMEM-F12/N2 was added. One to two days after transfection, the cell culture medium was removed and the cells were rinsed in PBS. Enough Reporter Lysis Buffer (Promega, Madison, WI, USA) to cover the bottom surface was added to each well (50 Al/well for 48-well plates). After one freeze – thaw cycle, the lysates were tested for luciferase activity using the Luciferase Assay System (Promega) and a single-tube luminometer (Berthold Lumat LB 9507; Bad Wildbad, Germany). The total protein concentration of each lysate was determined using the DC Protein Assay (Bio-Rad). The results were expressed in relative light units (RLU) per milligram of total protein.


ARTICLE

doi:10.1016/j.ymthe.2004.11.005

Gene delivery in vivo. Adult male Wistar rats (8 weeks of age, 180 – 200 g) were used for the in vivo study. The injection procedure was performed under anesthesia by peritoneal injection of sodium pentobarbital (60 mg/ kg). After the skin around L4 – L5 was exposed, the intralumbar injection was accomplished via a 100-Al syringe connected to a 26-gauge needle. A slight movement of the tail usually accompanied the proper injection into the subarachnoid space. Twenty microliters of triple complexes containing 4 Ag DNA, peptide at an N/P ratio of 2.5, and PEI600 at an N/P ratio of 10 was injected into each rat. After the triple complexes were slowly injected over 2 – 5 min, the syringe was left in place for 5 min to limit diffusion of the complexes from the injection site due to backflow pressure. The skin was closed with surgical clips after the injection and the animals were kept warm until recovered. After 3 days, the animals were perfused transcardially with 0.1 M PBS (200 ml/rat) under anesthesia. The lumbar spinal cord and dorsal root ganglia were collected and homogenized with 400 and 100 Al of Reporter Lysis Buffer (Promega), respectively. After sonication for 10 s on ice, the lysates were centrifuged at 14,000g for 5 min at 4jC. The luciferase activity of 20 Al of supernatant was measured as described for the in vitro experiments. The total protein concentration of each lysate was determined using the DC Protein Assay (Bio-Rad), and the results were expressed in RLU per milligram of total protein. Immunohistochemical staining. Following deep anesthesia, rats were perfused 3 days after injection with Ringer’s solution followed by 2% paraformaldehyde in 0.1 M PBS, pH 7.4. After perfusion, dorsal root ganglia were removed and postfixed in the same fixative for 2 – 4 h before they were transferred into 0.1 M PBS containing 15% sucrose and kept overnight at 4jC. Frozen sections were cut at 30-Am thickness mounted on coated slides. The sections were washed for 20 min in 0.1 M PBS at pH 7.4 containing 0.2% Triton X-100, blocked with 5% normal goat serum in PBS for 1 h, and then incubated overnight with polyclonal antiluciferase (Promega; dilution 1:150) and a monoclonal antibody against NeuN (Chemicon International, Temecula, CA, USA; dilution 1:500). Sections were washed in 0.1 M PBS and further incubated with antirabbit IgG TRITC conjugate (Sigma – Aldrich; dilution 1:100) and antimouse IgG FITC conjugate (Sigma – Aldrich; dilution 1:100) for 1 h. Control sections were incubated without primary antibodies. Sections were examined with a Carl Zeiss LSM410 confocal laser scanning microscope. Each section was initially scanned with a 488-nm laser line and an emission filter BP 510-525 for the detection of FITC fluorescein and then with a 543-nm laser line and an emission filter LP 570, for the detection of TRITC fluorescein.

3. 4. 5. 6. 7. 8.

9. 10.

11. 12.

13. 14.

15.

16.

17.

18.

19. 20.

21. 22.

ACKNOWLEDGMENTS This work was funded by the Agency for Science, Technology, and Research (A* STAR), Singapore. The authors thank Mr. S.J. Gao (Institute of Bioengineering and Nanotechnology, Singapore) for his technical assistance and other members in the Wang lab for their helpful discussion and support.

23. 24. 25.

RECEIVED FOR PUBLICATION JULY 3, 2003; ACCEPTED NOVEMBER 9, 2003.

REFERENCES 1. Wu, G. Y., and Wu, C. H. (1987). Receptor-mediated in vitro gene transformation by a soluble DNA carrier system. J. Biol. Chem. 262: 4429 – 4432. 2. Varga, C. M., Wickham, T. J., and Lauffenburger, D. A. (2002). Receptor-mediated


26. 27. 28.

targeting of gene delivery vectors: insights from molecular mechanisms for improved vehicle design. Biotechnol. Bioeng. 70: 593 – 605. Merdan, T., Kopecek, J., and Kissel, T. (2002). Prospects for cationic polymers in gene and oligonucleotide therapy against cancer. Adv. Drug Delivery Rev. 54: 715 – 758. Thorne, R. G. , and Frey, W. H. II (2001). Delivery of neurotrophic factors to the central nervous system: pharmacokinetic considerations. Clin. Pharmacokinet. 40: 907 – 946. Neet, K. E., and Campenot, R. B. (2001). Receptor binding, internalization, and retrograde transport of neurotrophic factors. Cell. Mol. Life Sci. 58: 1021 – 1035. Patapoutian, A., and Reichardt, L. F. (2001). Trk receptors: mediators of neurotrophin action. Curr. Opin. Neurobiol. 11: 272 – 280. Grimes, M. L., et al. (1996). Endocytosis of activated TrkA: evidence that nerve growth factor induces formation of signaling endosomes. J. Neurosci. 16: 7950 – 7964. Watson, F. L., Heerssen, H. M., Bhattacharyya, A., Klesse, L., Lin, M. Z., and Segal, R. A. (2001). Neurotrophins use the Erk5 pathway to mediate a retrograde survival response. Nat. Neurosci. 4: 981 – 988. Roux, P. P., and Barker, P. A. (2002). Neurotrophin signaling through the p75 neurotrophin receptor. Prog. Neurobiol. 67: 203 – 233. Khursigara, G., Bertin, J., Yano, H., Moffett, H., DiStefano, P. S., and Chao, M. V. (2001). A prosurvival function for the p75 receptor death domain mediated via the caspase recruitment domain receptor-interacting protein 2. J. Neurosci. 21: 5854 – 5863. Longo, F. M., Vu, T. K., and Mobley, W. C. (1990). The in vitro biological effect of nerve growth factor is inhibited by synthetic peptides. Cell Regul. 1: 189 – 195. LeSauteur, L., Wei, L., Gibbs, B., and Saragovi, H. U. (1995). Small peptide mimics of nerve growth factor bind TrkA receptors and affect biological responses. J. Biol. Chem. 270: 6564 – 6569. Ibanez, C. F. (1995). Neurotrophic factors: from structure – function studies to designing effective therapeutics. TIBTECH 13: 217 – 227. Wiesmann, C., Ultsch, M. H., Bass, S. H., and de Vos, A. M. (1999). Crystal structure of nerve growth factor in complex with the ligand-binding domain of the TrkA receptor. Nature 401: 184 – 188. Longo, F. M., Manthorpe, M., Xie, Y. M., and Varon, S. (1997). Synthetic NGF peptide derivatives prevent neuronal death via a p75 receptor-dependent mechanism. J. Neurosci. 48: 1 – 17. Beglova, N., LeSauteur, L., Ekiel, I., Saragovi, H. U., and Gehring, K. (1998). Solution structure and internal motion of a bioactive peptide derived from nerve growth factor. J. Biol. Chem. 273: 23652 – 23658. Beglova, N., Maliartchouk, S., Ekiel, I., Zaccaro, M. C., Saragovi, H. U., and Gehring, K. (2000). Design and solution structure of functional peptide mimetics of nerve growth factor. J. Med. Chem. 43: 3530 – 3540. Xie, Y., Tisi, M. A., Yeo, T. T., and Longo, F. M. (2000). Nerve growth factor (NGF) loop 4 dimeric mimetics activate ERK and AKT and promote NGF-like neurotrophic effects. J. Biol. Chem. 275: 29868 – 29874. Fortunati, E., et al. (2000). A multi-domain protein for beta1 integrin-targeted DNA delivery. Gene Ther. 7: 1505 – 1515. Ruit, K. G., Osborne, P. A., Schmidt, R. E., Johnson, E. M. , and Snider, W. D. Jr. (1990). Nerve growth factor regulates sympathetic ganglion cell morphology and survival in the adult mouse. Neuroscience 10: 2412 – 2419. Cohen, M. S., Wall, E. J., Brown, R. A., Rydevik, B., and Garfin, S. R. (1990). Cauda equina anatomy. II. Extrathecal nerve roots and dorsal root ganglia. Spine 15: 1248 – 1251. Maliartchouk, S., Debeir, T., Beglovai, N., Cuello, A. C., Gehringi, K., and Saragovi, H. U. (2000). Genuine monovalent ligands of TrkA nerve growth factor receptors reveal a novel pharmacological mechanism of action. J. Biol. Chem. 275: 9946 – 9956. Ceresa, B. P., and Schmid, S. L. (2000). Regulation of signal transduction by endocytosis. Curr. Opin. Cell Biol. 12: 204 – 210. Whitmarsh, A. J., and Davis, R. J. (2001). Signal transduction by target-derived neurotrophins. Nat. Neurosci. 4: 963 – 964. Strous, G. J., and Gent, J. (2002). Dimerization, ubiquitylation and endocytosis go together in growth hormone receptor function. FEBS Lett. 529: 102 – 109. Winkler, J., Thal, L. J., and Gage, F. H. (1998). Cholinergic strategies for Alzheimer’s disease. J. Mol. Med. 76: 555 – 567. Siegel, G. J., and Chauhan, N. B. (2000). Neurotrophic factors in Alzheimer’s and Parkinson’s disease brain. Brain Res. Brain Res. Rev. 33: 199 – 227. Jayaraman, K., and Puccini, C. J. (1992). A PCR-mediated gene synthesis strategy involving the assembly of oligonucleotides representing only one of the strands. BioTechniques 12: 392 – 398.

281