Gene therapy for preventing neuronal death using hepatocyte growth factor: in vivo gene transfer of HGF to subarachnoid space prevents delayed neuronal death in gerbil hippocampal CA1 neurons

Article metrics


To develop a novel strategy to prevent delayed neuronal death (DND) following transient occlusion of arteries, the gene of hepatocyte growth factor (HGF), a novel neurotrophic factor, was transfected into the subarachnoid space of gerbils after transient forebrain ischemia. Importantly, transfection of HGF gene into the subarachnoid space prevented DND, accompanied by a significant increase in HGF in the cerebrospinal fluid. Prevention of DND by HGF is due to the inhibition of apoptosis through the blockade of bax translocation from the cytoplasm to the nucleus. HGF gene transfer into the subarachnoid space may provide a new therapeutic strategy for cerebrovascular disease.


Neurons are highly sensitive to hypoxia-ischemia. Given this susceptibility and their postmitotic nature, the development of effective protective therapeutic strategies is essential. Pyramidal neurons in the CA1 subfield of the hippocampus are known to be the most vulnerable to cerebral ischemia.1,2,3 Following transient occlusion of the bilateral common carotid arteries in the gerbil, delayed neuronal death begins in CA1 pyramidal neurons a few days after recirculation, during which time no energy crisis or morphological change is observed. Therefore, to prevent delayed neuronal death in CA1, pyramidal neurons might be of therapeutic value. This therapeutic strategy might be useful for the treatment of other neuronal disease, such as motor neuron degenerative disease and cerebral ischemic disease. From this viewpoint, interest has centered on various cytokines, especially hepatocyte growth factor (HGF), which is a well-known potent pleiotrophic cytokine that exhibits mitogenic, motogenic, and morphogenic activities in a variety of cells.4,5,6 Here, we demonstrated that gene transfer of HGF into the subarachnoid space could cause beneficial effects on delayed neuronal death in gerbils, resulting in a treatment for cerebral ischemic disease induced by occlusion of cerebral arteries.

Both HGF and the c-Met/HGF receptor of membrane-spanning tyrosine kinase are expressed in various regions of the brain.7 It is reported that sympathetic neurons of the superior cervical ganglion coexpress bioactive HGF and its receptor, Met tyrosine kinase, both in vivo and in vitro,8 and that permanent middle cerebral artery occlusion induces HGF and its activator in rat brain.9 Exogenous HGF selectively promotes the growth of cultured sympathetic neurons; the magnitude of this growth effect is similar to that observed with exogenous nerve growth factor (NGF).10 In addition, recent studies suggest that HGF prevents apoptosis in cerebellar granular neurons via the phosphatidylinositol-3 (PI3)-kinase/Akt pathway.11 It is also reported that HGF cooperates with ciliary neurotrophic factor (CNTF) in promoting the survival and growth of parasympathetic and proprioceptive neurons, and that within the same neurons, the effects of HGF on survival and growth are selectively dependent on which other signaling pathways are concurrently activated.12 Given these unique properties, HGF is considered to be a novel neuroprotective agent. Indeed, administration of recombinant HGF has been reported to protect hippocampal neuron from ischemia-induced delayed neuronal death.13 In addition, a recent study provides a strong rationale for the potential clinical use of HGF for the treatment of motor neuron degenerative disease.14 Based upon these studies, we examined the possibility that gene transfer of HGF into the subarachnoid space could cause beneficial effects on delayed neuronal death in gerbils in this study.


In vivo transfection of β-galactosidase gene into brain using HVJ–liposome delivery system

To develop an efficient gene transfer method into the CNS, we chose the direct infusion of HVJ–liposome complex into the cisterna magna. In our preliminary experiments using β-galactosidase, β-galactosidase expression was observed in both sides of the brain following injection into the cisterna magna. In this study, we also confirmed this observation. As shown in Figure 1, positive staining of β-galactosidase was clearly observed in the cerebral cortical cells (Figure 1a), hippocampus (Figure 1b), and choroid plexus (Figure 1c) at 7 days after transfection. Expression of β-galactosidase could be detected 6 h after transfection in some populations of cerebral cortical cells (data not shown). In addition, no positive staining for β-galactosidase could be detected in brain transfected with control vector or untransfected brain. Considering the treatment of cerebral ischemic disease in the clinical setting, it seems best to employ infusion into the subarachnoid space rather than injection into the lateral ventricle using a stereotactic frame.

Figure 1

Representative microphotographs of brain stained with β-galactosidase at 7 days after injection of HVJ–liposomes containing β-galactosidase gene. (a) Low magnification (× 40). (b) CA1 subfield of hippocampus (× 400). (c) Choroid plexus in third ventricle (× 400).

To demonstrate the successful gene transfer of HGF gene into the central nervous system (CNS), we first measured the protein expression of HGF in the cerebrospinal fluid (CSF) by ELISA. Initially, we measured human HGF in the CSF of control gerbils before and 7 days after transient occlusion of the bilateral carotid arteries. Expectedly, human HGF protein could not be detected using specific anti-human antibody before transfection, since anti-human HGF antibody could not react with gerbil HGF. Similarly, human HGF could not be detected in the CSF of gerbils transfected with control vector. Then, we measured human HGF protein concentration in the CSF of gerbils transfected with human HGF gene. At 7 days after transfection, human HGF was readily detected (control vector, not detected; HGF vector: 2.16 ± 0.03 ng/ml).

Prevention of delayed neural death by transfection of human HGF gene

Next, we examined the effects of over-expression of HGF in delayed neuronal death induced by occlusion and reperfusion of the carotid arteries in gerbils. In sham-operated gerbils, the CA1 pyramidal cell layer was clearly visible (Figure 2a). However, almost all of the CA1 pyramidal cells progressed to death within 7 days after transient forebrain ischemia (Figure 2b). In contrast, transfection of HGF gene resulted in the prevention of neuronal death, since many normal pyramidal cells could be observed at 4 days after transient forebrain ischemia in gerbils transfected with HGF gene as compared with gerbils transfected with control vector (Figure 2c). Even at 7 days after transient forebrain ischemia, surviving neurons were clearly observed despite the presence of some atrophied neurons (Figure 2d). These morphological features were confirmed by quantitative analysis to measure the number of intact neurons per 1 mm in the bilateral CA1 subfields of the hippocampus at 4 and 7 days after reperfusion (Figure 3). At 4 days after reperfusion, transfection of HGF gene attenuated the neuronal death, whereas in gerbils transfected with control vector demonstrated a significant increase in the dead neurons as compared with sham-operated non-ischemic animals (P < 0.01), as shown in Figure 3a. At 7 days after reperfusion, in gerbils transfected with control vector, only a few neurons survived after transient forebrain ischemia as compared with sham-operated non-ischemic animals (P < 0.01, Figure 3b). In contrast, approximately 70% of the neurons survived in gerbils transfected with human HGF vector as compared with sham-operated non-ischemic gerbils (P < 0.01, Figure 3b). There was no significant difference in the number of surviving neurons between untransfected ischemic animals and gerbils transfected with control vector.

Figure 2

Representative photographs of pyramidal cells in hippocampal CA1 subfield in ischemic gerbil forebrain after transfection of HGF gene (× 400). These sections were stained with hematoxylin and eosin. Arrows demonstrated the atrophied neurons. (a) Sham-operated gerbils; (b) gerbils transfected with control vector, at 7 days after transfection; (c) gerbils transfected with human HGF vector, at 4 days after transfection; (d) gerbils transfected with human HGF vector, at 7 days after transfection.

Figure 3

Effect of HGF gene transfer on survival of hippocampal pyramidal neurons are (a) and 7 (b) days after reperfusion. (a) n = 8 for each group; sham, sham-operated gerbils; control, gerbils transfected with control vector at 4 days following transient forebrain ischemia; HGF gene, gerbils transfected with human HGF vector at 4 days following transient forebrain ischemia; *P < 0.01 versus control. (b) n = 8 for each group; sham, sham-operated gerbils; control, gerbils transfected with control vector at 7 days following transient forebrain ischemia; HGF gene, gerbils transfected with human HGF vector at 7 days following transient forebrain ischemia; rHGF, gerbils treated with human rHGF at 7 days following transient forebrain ischemia; *P < 0.01.

Moreover, we compared the protective effects of HGF gene transfer on delayed neuronal death with a single administration of recombinant HGF. As shown in Figure 3b, in animals infused with 30 μg recombinant HGF, approximately 20% of the neurons survived as compared with sham-operated non-ischemic gerbils, whereas a significant protective effect on delayed neuronal death was observed in gerbils treated with recombinant HGF (P < 0.01). However, the protective effect on neuronal death was more potent with HGF gene transfer as compared with a single administration of recombinant protein (P < 0.01). Probably, recombinant HGF injected into the CSF would rapidly diminish. Although a previous report revealed that continuous intrastriatal administration of HGF for 7 days during reperfusion prevented delayed neuronal death,13 continuous expression of HGF by HGF gene transfer might be ideal for human clinical trials.

Molecular mechanisms of prevention of delayed neuronal death by HGF

To confirm the protective effects of HGF, we performed immunohistochemical staining against c-met, the specific receptor of HGF, at 7 days after transfection. Expression of c-met was detected in CA1 of non-ischemic gerbils (Figure 4a). In the brain of gerbils transfected with control vector, c-met expression was not changed as compared with brain before ischemia (Figure 4b). Of particular importance, a marked increase in c-met expression was detected in gerbils transfected with human HGF gene as compared with those transfected with control vector (Figure 4c). The present observation that HGF up-regulated c-met expression is consistent with our previous report.7 Increased expression of c-met might contribute to enhancement of the inhibitory effects of HGF on neuronal death.

Figure 4

Representative photographs of immunohistochemical staining for c-met/HGF receptor in hippocampal CA1 subfield (× 400). (a) Sham-operated gerbils; (b) gerbils transfected with control vector, at 7 days after transfection. (c) gerbils transfected with human HGF vector, at 7 days after transfection.

Finally, we explored the molecular mechanisms by which HGF prevented delayed neuronal death. To examine whether prevention of enhanced apoptosis in the CA1 subfield of gerbils transfected with HGF gene contributed to a decrease of lesion size, DNA fragmentation analysis using terminal deoxynucleotidyl transferase-mediated dUTP-biotin in situ nick end labeling (TUNEL) was performed on coronal sections at the hippocampal level that were harvested 7 days after transient forebrain ischemia (Figure 5). In the hippocampal level section of gerbils transfected with HGF gene, a marked decrease in TUNEL-positive CA1 cells was observed as compared with gerbils transfected with control vector (Figure 5b). Thus, the reduction in delayed neuronal death after forebrain ischemia by over-expression of HGF might be due to the prevention of apoptosis. Moreover, we focused on the mechanisms of inhibition of apoptosis by HGF. Interestingly, cytoplasmic granular immunostaining for bax protein was observed in sham-operated animals (Figure 6a). This staining pattern was almost identical to the distribution of each neuron in the CA1 section. However, at 4 days following forebrain ischemia, the subcellular distribution of bax protein differed markedly. The present data clearly demonstrated that bax protein was translocated into the nucleus from the cytoplasm following forebrain ischemia (Figure 6b). In contrast, bax expression was still observed in the cytoplasm in CA1 neurons of gerbils transfected with HGF vector at 4 days following forebrain ischemia (Figure 6c). As discussed later, translocation of bax from the cytoplasm to the nucleus is believed to promote apoptosis in many cells.25,26,27 The preventive role of HGF in ischemic neuronal injury of CA1 neurons might be due to the blockade of translocation of bax by HGF.

Figure 5

Representative photographs of TUNEL-positive staining in CA1 hippocampal neurons at 7 days following transient forebrain ischemia (× 400). TUNEL-positive cells were counter-stained with hematoxylin. (a) Sham-operated gerbils; (b) gerbils transfected with control vector, at 7 days after transfection; (c) gerbils transfected with human HGF vector, at 7 days after transfection.

Figure 6

Representative photographs of immunohistochemical staining for bax protein in hippocampal CA1 subfield at 4 days following transient forebrain ischemia (× 400). (a) Sham-operated gerbils; (b) gerbils transfected with control vector, at 4 days after transfection; (c) gerbils transfected with human HGF vector, at 4 days after transfection.


Disruption of blood flow to the brain initiates a cascade of events that produces neuronal death and leads to neurological dysfunction. A number of extracellular mediators are implicated in ischemic cell death, including excitatory amino acids, free radicals, and calcium. Recently, many studies have examined the intracellular events that accompany ischemic cell death, to determine whether neurons are passively killed or actively participate in self-destruction. These studies indicate that some neurons contribute to their own demise through a series of biochemical events similar to programmed cell death. Therefore, to prevent brain injury, numerous studies have focused on the development of neuroprotective agents that effectively prevent delayed neuronal death following transient forebrain ischemia.15,16,17,18,19,20,21,22,23 Recently, HGF has been the center of interest in neuroprotective substances, since HGF is both a chemoattractant and a survival factor for embryonic motor neurons.8 In addition, sensory and sympathetic neurons and their precursors respond to HGF with increased differentiation, survival and axonal outgrowth.8 The broad spectrum of HGF activities and its observed synergy with other neurotrophic factors suggest that the major role of HGF is to potentiate the response of developing neurons to specific signals.

Nevertheless, the clinical utility of neuroprotective agents such as HGF is quite limited due to the presence of the blood–brain barrier, which makes the central nervous system relatively inaccessible to circulating proteins and peptides. Since the molecular size of numerous neuroprotective agents (drugs and cytokines), including HGF, is too large to penetrate the blood–brain barrier, these agents seem to be ineffective without the direct and continuous injection into the ventricle, striatum or cerebral cortex by a surgical technique.12,13,14 From the standpoint of clinical use, it is clear that these methods are less useful, because they entail surgical insult and prolonged endurance for the patient. Indeed, previous studies employed the infusion of recombinant protein continuously into the brain or the subarachnoid space, whose manipulation is rather harmful in clinical situations. This procedure is necessary because of the rapid disappearance of recombinant neurotrophic growth factors. For example, Miyazawa et al13 reported that continuous post-ischemic intrastriatal administration of human recombinant HGF (10 or 30 μg) for 7 days potently prevented the delayed death of hippocampal neurons in the Mongolian gerbil. To explore the clinical application, it is necessary to overcome the issue of inaccessibility of the brain.

One method to overcome these limitations is to utilize a drug delivery system into the central nervous system. We focused especially on gene transfer into the subarachnoid space, since intrathecal injection into the cisterna magna with a needle involves no systemic anesthesia, no burr hole and no pain. Thus, this procedure seems to be less invasive and more practical than transfection methods such as injection into the lateral ventricle with a stereotactic procedure. The feasibility of this less invasive transfection method is supported by the observations that (1) gene transfer of β-galactosidase into the subarachnoid space resulted in positive staining in the cerebral cortical cells, hippocampus and choroid plexus; (2) it was possible to detect human HGF in CSF; (3) transfection of HGF gene was sufficient to decrease delayed neuronal death; and (4) prevention of neuronal death by HGF gene transfer was more effective than recombinant HGF. Our procedure has another advantage, that is, to use HGF, a secreted protein, for the treatment of ischemic damaged brain. Together with the fact that HGF can be secreted from cells due to the presence of signal peptides, the secretion of HGF into the CSF from transfected cells in the brain have beneficial effects on brain injury. It is not necessary for cells transfected with HGF gene to be directly the ischemic damaged neurons.28,29 To consider the utility, gene therapy rather than recombinant therapy may be superior due to the short half-life of rHGF. For example, the alpha-phase of intravenous administration of rHGF was 3.2 min, while the beta-phase was 26.5 min.30 Similarly, Liu KX et al31 reported 4 min via intravenous administration. Even intramuscular or subcutaneous injection of rHGF was less than 2.7 h.31 Thus, the continuous expression of transgene over a long time may be ideal to demonstrate therapeutic effects. It may be also be preferable to deliver a lower dose over a period of several days or more from an actively expressed transgene in the brain, rather than a single or multiple bolus doses of recombinant protein, to avoid side-effects.

It is unclear how HGF prevented delayed neuronal death, although previous reports have mentioned the neurotrophic action of HGF.12 Recent studies suggest that HGF prevents apoptosis in cerebellar granule neurons via the PI3-kinase/Akt pathway.10 Thus, we measured TUNEL-positive neuronal cells in the present study. Expectedly, TUNEL-positive cells were markedly decreased by over-expression of HGF. Moreover, we found that HGF inhibited translocation of a pro-apoptotic molecule, bax, from the cytoplasm to the nucleus. Recent studies have documented that the subcellular localization of bax determines the fate of cells, as bax protein moves to the mitochondria and other membrane sites, and triggers a catastrophic change of mitochondrial function after delivery of death signals to cells.30,31,32 Probably, one of the potential mechanisms of how HGF works as an anti-apoptotic molecule is to inhibit the translocation of bax, although further studies are necessary.

Overall, the present study demonstrated that gene transfer of HGF, a novel neurotrophic and anti-apoptotic factor, into the subarachnoid space has a profound neuroprotective effect against post-ischemic delayed neuronal death in the hippocampus. Therefore, it is reasonable to apply gene transfer rather than recombinant therapy to obtain sustained expression and secretion of neurotrophic factors such as HGF in the ischemic brain. Continuous development of systems involving vectors, promoters or alternative routes of administration may help to achieve human gene therapy for cerebrovascular disease in the future.

Materials and methods

Construction of plasmids

To produce the HGF expression vector, human HGF cDNA (2.2 kb) was inserted into a simple eukaryotic expression plasmid (pcDNA3) that utilizes the cytomegalovirus (CMV) promoter/enhancer.24 This promoter/ enhancer has been used to express reporter genes in a variety of cell types and can be considered to be constitutive. Downstream from the HGF cDNA is the SV40 polyadenylation sequence. The vector used as a control was pcDNA3 expression vector plasmid, which does not contain HGF cDNA. We obtained β-galactosidase gene expression vector from a commercially available source (Promega, Madison, WI, USA). The β-galactosidase expression vector is driven by SV 40 promoter.

Ligation of bilateral common carotid arteries

Male Mongolian gerbils (50–70 g; Charles River Japan, Atsugi, Japan) were anesthetized with an initial concentration of 3% halothane, and then maintained with 1.5% halothane in a mixture of 20% O2/80% N2O under a face mask, and breathed spontaneously throughout the surgical procedure.13,33,34 Through a midline cervical incision, the bilateral carotid arteries were exposed and occluded with clips (Sugita aneurysm clips; Mizuho Ikakogyo, Tokyo, Japan) to induce 5 min forebrain ischemia. The body (rectal) temperature was monitored and maintained at 37°C throughout the experiments using a feedback-controlled heating pad. After surgery, the animals were transfected with HGF gene or control gene by the HVJ–liposome method. After gene transfection, animals were placed on a heating pad till anesthesia was terminated. Sham-operated animals were treated in the same manner except for common carotid artery occlusion. The animals were divided into four experimental groups as follows: sham-operated group (no ischemia), vehicle-treated group (5 min of ischemia), post-HGF gene group (HGF gene transfection immediately after 5 min of ischemia), post-rHGF group (injection of recombinant HGF immediately after 5 min of ischemia). Histological evaluation was performed at 4 and 7 days after ischemia. The experimental procedure was approved by the Committee for Animal Experimentation of Osaka University, Graduate School of Medicine, and met the guidelines of the Japanese Association for Laboratory Animal Science.

Preparation of HVJ–liposome complex

The procedures used for the preparation of HVJ–liposomes have been described previously.32,35,36,37,38,39 In brief, phosphatidylserine, phosphatidylcholine and cholesterol were mixed in a weight ratio of 1:4.8:2. The lipid mixture (10 mg) was deposited on the sides of a flask by removal of tetrahydrofuran in a rotary evaporator. Dried lipid was hydrated in 200 μl balanced salt solution (BSS; 137 μM NaCl, 5.4 μM KCl, 10 μM Tris-HCl, pH 7.6) with plasmid (200 μg). Empty liposome which did not contain plasmid (BSS 200 μl) were used as control. Liposomes were prepared by shaking and sonication. Purified HVJ (Z strain) was inactivated by UV irradiation (110 erg/mm2/s) for 3 min just before use. The liposome suspension (0.5 ml, containing 10 mg lipid) was mixed with HVJ (10 000) hemagglutinating units in a total volume of 4 ml BSS). The mixture was incubated at 4°C for 10 min and then for 60 min with gentle shaking at 37°C. Free HVJ was removed from the HVJ–liposomes by sucrose density gradient centrifugation. The top layer of the sucrose gradient was collected for use. This preparation method has been optimized to achieve maximal transfection efficiency.

In vivo gene transfer to subarachnoid space

In this study, we employed gene transfer into the cisterna magna using infusion of HVJ–liposome complex as in vivo gene transfer into the CNS. For infusion into the subarachnoid space, the head of each animal was fixed in a prone position and the atlantooccipital membrane was exposed through an occipitocervical midline incision. A stainless steel cannula (27 gauge; Becton Dickinson, Franklin Lakes, NJ, USA) was introduced into the cisterna magna. After withdrawal of 50 μl CSF for confirmation of cannula position and to avoid increased intracranial pressure, HVJ–liposome solution (100 μl: 10 μg/ml) was carefully injected over 1 min into the cisterna magna (subarachnoid space). Thereafter, the animals were placed head down for 30 min. The sterile procedure was completed by the delivery of a prophylactic dose of antibiotic (30 000 U penicillin G i.m.). No behavioral change such as convulsion or abnormal movement of extremities was observed in any animal undergoing injections.


CSF (100 μl) from gerbils 7 days after transient forebrain ischemia was used for the experiments. The concentration of HGF in the CSF was determined by enzyme-immunoassay using anti-human HGF antibody (Institute of Immunology, Tokyo, Japan), as described previously.33 Briefly, rabbit anti-human HGF IgG was coated on 96-well plates (Corning, NY, USA) at 4°C for 15 h. After blocking with 3% bovine serum albumin in phosphate-buffered saline (PBS), conditioned medium was added to each well. The preparation was incubated for 2 h at 25°C. Wells were washed three times with PBS containing 0.025% Tween 20 (PBS-Tween). Biotinylated rabbit anti-human HGF IgG was added and the preparation was incubated for 2 h at 25°C. After washing with PBS-Tween, wells were incubated with horseradish peroxidase-conjugated streptavidin–biotin complex in PBS-Tween. The enzyme reaction was initiated by adding substrate solution composed of 2.5 mg/ml o-phenylenediamine, 100 mM sodium phosphate, 50 mM citric acid, and 0.015% H2O2. The enzyme reaction was halted by adding 1 M H2SO4, and absorbance at 490 nm was measured. The antibody against human HGF reacts with only human HGF, and not with rat and gerbil HGF.33

Histopathological examination

For X-gal staining, after perfusion-fixation in 3% paraformaldehyde: 20% sucrose solution for 1 day, 25-μm thick frozen sections in the coronal plane were taken at 100-μm intervals. Sections were stained with X-gal to identify infected neurons expressing β-galactosidase.

Immunohistochemistry and morphology

At each point tested, animals were anesthetized with sodium pentobarbital (10 mg/kg, i.p.) and perfused transcardially with 0.9% heparinized saline, followed by 70% ethanol. The brains were then removed and preserved in a fixative overnight and processed for paraffin embedding. Coronal sections (5-μm) were cut at the level of the dorsal hippocampus for conventional hematoxylin and cosin (H&E) staining, and immunohistochemical staining. Sections for immunohistochemical staining were deparaffinized and then hydrated by transferring them through the following solutions: xylene bath twice for 5 min, 100% ethanol for 5 min twice, and then 90% ethanol, 80% ethanol, 70% ethanol, and PBS, for 3 min each.

Anti-bax antibody (p-19) and anti-c-met antibody (m-Met: SP-260) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-bax antibody is an affinity-purified rabbit polyclonal antibody raised against a peptide corresponding to amino acids 43–61, mapping within an amino terminal domain of bax protein of mouse origin. Anti-bax antibody reacts with bax protein of mouse, rat and human origin and is non-cross-reactive with bcl-2 and bcl-x proteins. Anti-c-met antibody (SP260) is provided as goat affinity-purified polyclonal antibody raised against a peptide mapping at the carboxy terminus of c-met p140 of mouse origin. Anti-mouse-met reacts with c-Met of mouse, rat and to a lesser degree, human origin. The deparaffinized sections were heated and boiled three times for 5 min by microwave in 10 mM citrate buffer, pH 6.0. To diminish nonspecific staining, each section was treated with methanol containing 3% hydrogen peroxide for 5 min. Anti-bax or anti-c-met antibodies used at dilutions of 1:500 and 1:250, respectively, in 0.05 M Tris-buffered saline, pH 7.6 (TBS) were added to the slides, which were incubated overnight in 4°C. Expression of bax protein was demonstrated by the labeled streptavidin biotin (LSAB) method using LSAB kit (Dako, Carpinteria, CA, USA) containing blocking reagent, biotinylated link antibody and peroxidase-labeled streptavidin reagents. The peroxidase-binding sites were detected by staining with 3,3-diaminobenzidine in TBS. Finally, counterstaining was performed with Mayer's hematoxylin. Slides were photographed with a light microscope (Olympus, Tokyo, Japan; model VANOX-S) equipped with a 35 mm camera (Olympus model C-35AD-4) with Ektar 100 film (Eastman Kodak, Rochester, NY, USA), and processed by the standard procedure for light-microscopic observations.

In situ end labeling of fragmented DNA

Sections were deparaffinized, rehydrated, treated with proteinase K (20 μg/ml), and incubated with PBS containing 0.3% hydrogen peroxide to reduce endogenous peroxidase activity. Nick end-labeling of apoptotic cells was measured using ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit (Intergen Inc., USA).25 It is based on the preferential binding of digoxigenin-dUTP by therminal deoxynucleotidyl transferase (TdT) to 3′-OH-end-labeling (TUNEL) detected by an antidigoxigenin antibody conjugated with peroxidase, a reporter enzyme that catalytically generates a brown-colored product from the chromogenic substrate diaminobenzidine (Dako). Counterstaining was performed by immersing the slides in 0.5% methyl green in 0.1 M sodium acetate solution (pH 4.0) for 5 min at room temperature.


Human recombinant HGF was purified from the culture medium of Chinese hamster ovary cells or C-127 cells, and transfected with expression plasmid containing human HGF cDNA.4

Statistical analysis

All values are expressed as mean ± s.e.m. Analysis of variance with subsequent Duncan's test was used to determine the significance of differences in multiple comparisons. Differences with P values less than 0.05 were considered significant.


  1. 1

    Ito U, Spatz MJ, Walker JT, Klatzo I . Experimental cerebral ischemia in mongolian gerbils Acta Neuropathol 1975 32: 209–223

  2. 2

    Kirino T . Delayed neuronal death in the gerbil hippocampus following ischemia Brain Res 1982 239: 57–59

  3. 3

    Kirino T, Sano K . Selective vulnerability in the gerbil hippocampus following transient ischemia Acta Neuropathol 1984 62: 201–208

  4. 4

    Nakamura T et al. Molecular cloning and expression of human hepatocyte growth factor Nature 1989 342: 440–443

  5. 5

    Zarnegar R, Michalopoulos GK . The many faces of hepatocyte growth factor: from hepatopoiesis to hematopoiesis J Cell Biol 1995 129: 1177–1180

  6. 6

    Matsumoto K, Nakamura T . Emerging multipoint aspects of hepatocyte growth factor J Biochem 1996 119: 591–600

  7. 7

    Honda S et al. Localization and functional coupling of HGF and c-Met/HGF receptor in rat brain: implication as neurotrophic factor Brain Res Mol Brain Res 1995 32: 197–210

  8. 8

    Maina F, Klein R . Hepatocyte growth factor, a versatile signal for developing neurons Nat Neurosci 1999 2: 213–217

  9. 9

    Hayashi T, Abe K, Sakurai M, Itoyama Y . Inductions of hepatocyto growth factor and its activator in rat brain with permanent middle cerebral artery occlusion Brain Res 1998 799: 311–316

  10. 10

    Yang XM et al. Autocrine hepatocyte growth factor provides a local mechanism for promoting axonal growth J Neurosci 1998 18: 8369–8381

  11. 11

    Zhang L, Himi T, Morita I, Murota S . Hepatocyte growth factor protects cultured rat cerebellar granule neurons from apoptosis via the phosphatidylinositol-3 kinasc/Akt pathway J Neurosci Res 2000 59: 489–496

  12. 12

    Davey F, Hilton M, Davies AM . Cooperation between HGF and CNTF in promoting the survival and growth of sensory and parasympathetic neurons Mol Cell Neurosci 2000 15: 79–87

  13. 13

    Miyazawa T et al. Protection of hippocampal neurons from ischemia-induced delayed neuronal death by hepatocyte growth factor: a novel neurotrophic factor J Cereb Blood Flow Metab 1998 18: 345–348

  14. 14

    Okura Y et al. Analysis of neurotrophic effects of hepatocyte growth factor in the adult hypoglossal nerve axotomy model Eur J Neurosci 1999 11: 4139–4144

  15. 15

    Yoshidomi M, Hayashi T, Abe K, Kogure K . Effects of a new calcium channel blocker, KB-2796, on protein synthesis of the CA1 pyramidal cell and delayed neuronal death following transient forebrain ischemia J Neurochem 1989 53: 1589–1594

  16. 16

    Kohno K et al. Intraventricular administration of nitric oxide synthase inhibitors prevents delayed neuronal death in gerbil hippocampal CA1 neurons Neurosci Let 1995 199: 65–68

  17. 17

    Xu DG et al. Elevation of neuronal expression of NAIP reduces ischemic damage in the rat hippocampus Nat Med 1997 3: 997–1004

  18. 18

    Yang GY, Zhao YJ, Davidson BL, Betz AL . Overexpression of interleukin-1 receptor antagonist in the mouse brain reduces ischemic brain injury Brain Res 1997 751: 181–188

  19. 19

    Nakajima Y, Fujimiya M, Maeda T, Mori A . Morphological investigation of the neuroprotective effects of graded hypothermia after diverse periods of global cerebral ischemia in gerbils Brain Res 1997 765: 113–121

  20. 20

    Cuevas P, Carceller F, Munoz-Willery I, Gimenez-Gallego G . Intravenous fibroblast growth factor penetrates the blood–brain barrier and protects hippocampal neurons against ischemia-reperfusion injury Surg Neurol 1998 49: 77–84

  21. 21

    Hayashi T, Abe K, Itoyama Y . Reduction of ischemic damage by application of vascular endothelial growth factor in rat brain after transient ischemia J Cereb Blood Flow Metab 1998 18: 887–895

  22. 22

    Yang GY et al. Attenuation of temporary focal cerebral ischemic injury in the mouse following transfection with interleukin-1 receptor antagonist Brain Res Mol Brain Res 1999 72: 129–137

  23. 23

    Colbourne F, Li H, Buchan AM . Continuing postischemic neuronal death in CA1: influence of ischemia duration and cytoprotective doses of NBQX and SNX-111 in rats Stroke 1999 30: 562–668

  24. 24

    Aoki M et al. Angiogenesis induced by hepatocyte growth factor in non-infarcted myocardium and infarcted myocardium: up-regulation of essential transcription factor for angiogenesis, ets Gene Therapy 2000 7: 417–427

  25. 25

    Gavrieli Y, Sherman Y, Ben-Sasson SA . Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation J Cell Biol 1992 119: 493–501

  26. 26

    Lopez E, Pozas E, Rivera R, Ferrer I . Bcl-2, Bax and Bcl-x expression following kainic acid administration at convulsant doses in the rat Neuroscience 1999 91: 1461–1470

  27. 27

    Nishita M et al. Nuclear translocation and increased expression of Bax and disturbance in cell cycle progression without prominent apoptosis induced by hyperthermia Exp Cell Res 1998 244: 357–366

  28. 28

    Lawrence MS et al. Herpes simplex viral vectors expressing Bcl-2 are neuroprotective when delivered after a stroke J Cereb Blood Flow Metab 1997 17: 740–744

  29. 29

    Antonawich FJ, Federoff HJ, Davis JN . Bcl-2 transduction, using a herpes simplex virus amplicon, protects hippocampal neurons from transient global ischemia Exp Neurol 1999 156: 130–137

  30. 30

    Oematsu Y et al. Effective administration route for the deleted form of hepatocyte growth factor To exert its pharmacological effects J Pharm Sci 1999 88: 131–135

  31. 31

    Flu KX et al. Importance of the liver in plasma clearance of hepatocyte growth factors in rats Am J Physiol 1992 263: G642–G649

  32. 32

    Nishii T et al. Angiotensinogen gene activating elements regulate blood pressure in the brain Circ Res 1999 85: 257–263

  33. 33

    Yamada A et al. Rapid and sensitive enzyme-linked immunosorbent assay for measurement of HGF in rat and human tissues Biomed Res 1995 16: 105–114

  34. 34

    Krajewski S et al. Release of caspase-9 from mitochondria during neuronal apoptosis and cerebral ischemia Proc Natl Acad Sci USA 1999 96: 5752–5757

  35. 35

    Ishimaru H, Takahashi A, Ikarashi Y, Maruyama Y . NGF delays rather than prevents the cholinergic terminal damage and delayed neuronal death in the hippocampus after ischemia Brain Res 1998 789: 194–200

  36. 36

    Suzuki S et al. Cerebral neurons express interleukin-6 after transient forebrain ischemia in gerbils Neurosci Lett 1999 262: 117–120

  37. 37

    Kaneda Y, Iwai K, Uchida T . Increased expression of DNA cointroduced with nuclear protein in adult rat liver Science 1989 243: 375–378

  38. 38

    Kato K et al. Gene transfer and the expression of a foreign gene in vivo in post-mitotic neurons of the adult rat brain using the hemagglutinating virus of the Japan–liposome method Brain Res Mol Brain Res 1994 25: 359–363

  39. 39

    Yamada K et al. Efficient oligonucleotide delivery using the HVJ–liposome method in the central nervous system Am J Physiol 1996 271: R1212–R1220

Download references


We wish to thank Rie Kosai and Keiko Yamaguchi for their excellent technical assistance. This work was partially supported by grants from the Japan Health Sciences Foundation, and the Japan Cardiovascular Research Foundation, a Japan Heart Foundation Research Grant, a Grant-in-Aid from the Japan Society for the Promotion of Science, the Ministry of Education, Science, Sports and Culture, and Ministry of Public Health.

Author information

Correspondence to R Morishita.

Rights and permissions

Reprints and Permissions

About this article


  • HGF
  • angiogenesis
  • stroke
  • gene therapy
  • delayed neuronal death

Further reading