Protection against ischemic brain injury by protein therapeutics

Abstract

Preventing massive cell death is an important therapeutic strategy for various injuries and disorders. Protein therapeutics have the advantage of delivering proteins in a short period. We have engineered the antiapoptotic bcl-x gene to generate the super antiapoptotic factor, FNK, with a more powerful cytoprotective activity. In this study, we fused the protein transduction domain (PTD) of the HIV/Tat protein to FNK and used the construct in an animal model of ischemic brain injury. When added into culture media of human neuroblastoma cells and rat neocortical neurons, PTD-FNK rapidly transduced into cells and localized to mitochondria within 1 h. It protected the neuroblastomas and neurons against staurosporine-induced apoptosis and glutamate-induced excitotoxicity, respectively. The cytoprotective activity of PTD-FNK was found at concentrations as low as 0.3 pM. Additionally, PTD-FNK affected the cytosolic movement of calcium ions, which may relate to its neuroprotective action. Immunohistochemical analysis revealed that myc-tagged PTD-FNK (PTD-myc-FNK) injected i.p. into mice can have access into brain neurons. When injected i.p. into gerbils, PTD-FNK prevented delayed neuronal death in the hippocampus caused by transient global ischemia. These results suggest that PTD-FNK has a potential for clinical utility as a protein therapeutic strategy to prevent cell death in the brain.

Proteins, rather than genes, can be directly and readily introduced into cells when fused with the protein transduction domain (PTD) of HIV/Tat protein (1). In fact, PTD-fused proteins can be successfully delivered to several tissues, including the brain when injected into mice systemically (2). Although this technology was originally described in 1994 (3) and many proteins fused with the PTD have been introduced into cells in vitro and in vivo (2, 4), few animal models of disease or injury have successfully precluded necrosis as well as apoptosis, by treating with transducing full-length therapeutic proteins. Demonstrating the in vivo efficacy of proteins with strong therapeutic activity would be important for practical applications.

Because antiapoptotic factors prevent presumed necrosis as well as apoptosis (5), preventing massive cell death by antiapoptotic factors could be an efficacious strategy for the treatment of various disorders and injuries. Antiapoptotic members of the Bcl-2 family, Bcl-2 and Bcl-x_L, are good candidates for protein therapeutics. They interact with the proapoptotic Bcl-2 family members to neutralize apoptotic activity and are localized mainly in mitochondria to regulate the release of cytochrome c (6, 7). Virus-mediated transfer of the bcl-2 gene into the brain has been shown to inhibit neuronal cell death induced by stroke, ischemia, and 6-hydroxydopamine (8–12). Delivery of the bcl-2 and bcl-x genes by particle gun also prolonged neuronal survival (13). However, these applications have difficulties expressing sufficient amounts of the active proteins in a short time to prevent cell death in acute cases. In addition, gene therapy approaches cannot exclude the possibility of hazardous insertions of transgenes.

Because antiapoptotic activity of Bcl-2 and Bcl-x_L is suppressed by phosphorylation, mutants of Bcl-2 and Bcl-x_L that lack their phosphorylation sites have been generated to prevent a loss of function (14, 15). We have constructed a powerful artificial cytoprotective protein, FNK (originally designated Bcl-xFNK in ref. 16) from Bcl-x_L (17) by site-directed mutagenesis, based on the high-resolution crystal structure of the rat Bcl-x_L (18). FNK has three amino acid substitutions, Tyr-22 to Phe (F), Gln-26 to Asn (N), and Arg-165 to Lys (K), in which three hydrogen bonds stabilizing the central α5-α6 helices (the putative pore-forming domain) are abolished. Compared with Bcl-x_L, FNK protects cultured cells more potently from cell death induced by oxidative stress (hydrogen peroxide and paraquat), a calcium ionophore (A23187), growth factor withdraw (serum and IL-3), anti-Fas, cell cycle inhibitors (TN-16, camptothecin, hydroxyurea, and trichostatin A), a protein kinase inhibitor (staurosporine, STS), and heat treatment (16). FNK is the first and sole mutant with a gain-of-function phenotype among the mammalian antiapoptotic factors. To test the efficacy of FNK in preventing neuronal cell death induced by injuries in vivo, we generated PTD-FNK fusion protein by fusing FNK with the PTD of HIV/Tat. Here, we show that PTD-FNK protects cultured neuronal cells against neuroexcitoxicity as well as apoptosis and that treatment of gerbils with PTD-FNK successfully reduces ischemic injury of hippocampal CA1 neurons.

Materials and Methods

Construction and Preparation of PTD-FNK, PTD-Bcl-x_L, and PTD-myc-FNK.

An oligonucleotide encoding MGYGRKKRRQRRRG (the TAT protein transduction domain of 11 aa is underlined) (1) was ligated to the 5′ end of FNK or Bcl-x_L coding sequence by PCR to construct PTD-FNK or PTD-Bcl-x_L, respectively. An oligonucleotide encoding GEQKLISEEDLG (myc TAG sequence is underlined) was inserted between the PTD and FNK sequences of PTD-FNK by PCR to obtain PTD-myc-FNK. The ligated DNA fragment was inserted between the NcoI and HindIII sites of the pPROEX1 expression vector (Invitrogen). The coding region of FNK was also cloned into the vector. The integrity of the constructs was confirmed by DNA sequencing. The resultant plasmids were introduced into Escherichia coli DH5αMCR (Invitrogen), followed by induction of expression with isopropyl 1-thio-β-d-galactoside. The induced proteins were recovered as inclusion bodies. To prepare PTD-FNK, PTD-Bcl-x_L, and FNK for in vitro experiments, the inclusion bodies were solubilized with 7 M urea, 1 mM DTT, 50 mM Tris⋅HCl (pH 8.0), and 150 mM NaCl, and finally dialyzed against saline containing 33% PBS. After a brief centrifugation, the soluble fraction was added into the media for neuroblastomas or primary neocortical neurons. Protein purity and concentration were confirmed by SDS/PAGE. To prepare PTD-FNK and PTD-myc-FNK for in vivo injections, the inclusion bodies were solubilized in a buffer containing 7 M urea, 2% SDS, 1 mM DTT, 62.5 mM Tris⋅HCl (pH 6.8), and 150 mM NaCl. The PTD-FNK extracted was then subjected to SDS/PAGE to remove endotoxin and contaminated proteins. The gel was briefly immersed in 1 M KCl, which makes an insoluble complex with free SDS, and a band corresponding to PTD-FNK (transparent because of reduced free SDS) was cut out. PTD-FNK was electrophoretically extracted from the gel slice in extraction buffer (25 mM Tris/0.2 M glycine/0.1% SDS) and was injected into animals. The extraction buffer was used as a control (vehicle). The concentration of extracted PTD-FNK ranged from 1 to 2 mg/ml. Protein concentration was determined by Coomassie brilliant blue staining after SDS/PAGE, followed by comparison with BSA standard.

Cell Culture.

Neuroblastoma SH-SY5Y cells were cultured in DMEM (Invitrogen) with 15% FBS and 20 units/ml penicillin–20 μg/ml streptomycin. Primary cultures of neocortical neurons were prepared from 18-day rat embryos as described (19, 20) with slight modifications. In brief, neocortical tissues were cleaned of meninges, minced, and treated with 0.25% trypsin and 0.01% DNase I. After mechanical dissociation by pipetting, cells were resuspended in MEM (Invitrogen) supplemented with 25 mM Hepes buffer, 10 μg/ml insulin, 5.5 μg/ml transferrin, 100 μM putrescine, 6.7 ng/ml sodium selenite, and 50 units/ml penicillin–100 μg/ml streptomycin, and then plated onto poly-l-lysine-coated plates at a density of 1 × 10⁵ cells/cm². Neurons were used at days 6–8. Neuronal identity was confirmed by immunostaining with anti-MAP2 (a marker of neurons; Sigma), anti-glial fibrillary acidic protein (a marker of astrocyte; Immunon, Pittsburgh), and anti-Nestin (a marker of neural stem cells; PharMingen). Preparations containing >95% neurons were used for experiments.

Immunostaining.

Cell cultures were prepared on 4-well plastic dishes (SonicSeal Slide, Nalge Nunc). The cells were rinsed with PBS and fixed in 4% paraformaldehyde in PBS for 30 min. After washing with PBS, the cells were incubated for 30 min in 0.2% Triton X-100, 30 min in a blocking buffer (3% BSA and 3% goat serum in PBS), and overnight at 4°C in a blocking buffer containing primary antibody. After another wash with PBS, cells were incubated in a blocking buffer containing BODIPY FL goat anti-mouse IgG (1:500, Molecular Probes) for 1 h and imaged with a confocal scanning microscope Fluoroview FV/300 (Olympus, Tokyo) using excitation and emission filters of 488 and 510 nm, respectively. Monoclonal anti-Bcl-x (no. 35-32) antibody was prepared as described (21). Because this antibody recognizes its N-terminal region, the amino acid alterations in FNK would not affect the affinity to the antibody. When cells were prestained with 100 nM MitoTracker red, its fluorescence was imaged with excitation and emission filters of 543 and 565 nm, respectively.

STS-Inducing Apoptosis of SH-SY5Y.

After incubation with 800 nM STS for 1 day, 5 μM propidium iodide (PI, for detecting dead nuclei) and 5 μM Hoechst 33342 (for detecting living nuclei) were added to each well containing SH-SY5Y that had been pretreated with a given concentration of PTD-FNK or vehicle. Cells stained with either dye were counted per well. Viability was calculated as follows: 100 × (1 − PI-stained cells/total cells).

Exposure of Neurons to Glutamate.

Cells were preincubated with PTD-FNK at a given concentration or vehicle for 2 h. For the determination of cell survival, neurons were washed once with Hanks' balanced salt solution (HBSS) with 2 mM CaCl₂ and exposed to 1 mM l-glutamate in HBSS containing 2 mM CaCl₂ at 37°C for 1 h. Then, HBSS was replaced with the medium and incubated for 1 day. Viable neurons with a triangular soma and neurites were enumerated under a phase-contrast microscope (×200) in five fields per well for four independent wells.

For the determination of a transient elevation in intracellular free Ca²⁺ ([Ca²⁺]_i) the fluorescence Ca²⁺ indicator Fluo-3 AM (Molecular Probes) was added to the medium at a final concentration of 5 μM and incubated for 30 min. Then, neurons were washed once with HBSS containing 2 mM CaCl₂ and exposed to 1 mM l-glutamate for a period. Fluorescence was imaged by confocal scanning microscopy using excitation and emission filters of 488 and 510 nm, respectively. Images obtained every 5 s were analyzed by using the nih image program. Analysis of data were performed as described (22, 23).

Immunohistochemical Staining of Brain Sections.

Male C57BL/6N mice (4 weeks old; Seac Yoshitomi, Fukuoka, Japan) were used. To examine the distribution of PTD-myc-FNK (50 mg/kg) delivered the brain at 10 h after its i.p. injection, mice were killed with an overdose of pentobarbital and perfused transcardially with cold heparinized physiological saline followed by 4% paraformaldehyde in PBS. Brains were removed, fixed by immersion into 4% paraformaldehyde in PBS overnight at 4°C, dehydrated, and embedded in paraffin. Paraffin sections (4 μm) were prepared for immunohistochemical staining. To detect PTD-myc-FNK in situ, rabbit anti-Myc Tag polyclonal antibody (1:50, 4°C overnight, Upstate Biotechnology, Lake Placid, NY) was used coupled with a DAKO Envision⁺ system.

Cerebral Ischemia.

Male Mongolian gerbils (Hoshino Laboratory Animals, Saitama, Japan), weighing 60–80 g, were injected i.p. with vehicle or PTD-FNK (5 mg/kg) 3 h before ischemic insults. Ischemia was induced for 5 min by occluding the common carotid arteries bilaterally with aneurysm clips under halothane anesthesia (1.5%) in N₂O/O₂ (70:30) (24, 25). Rectal and temporal muscle temperatures were maintained close to 37°C during ischemia and the first 6 h of reperfusion. Seven days later, the brains were transcardially perfused with 4% formaldehyde in 0.1 M phosphate buffer (pH 7.4). Then, 5-μm-thick paraffin cross sections containing hippocampal tissue were stained with hematoxylin–eosin. Intact neurons with distinct nuclei were counted per 1-mm CA1 length on each side, and the average was expressed as neuronal cell density (cells per mm).

Animal protocols were approved by the Animal Care and Use Committee of Nippon Medical School.

Results and Discussion

Transduction of PTD-FNK into Neuroblastoma Cells and Primary Cultured Neocortical Neurons.

To examine the ability and time course of the PTD-FNK protein to transfer into cells, 30 nM PTD-FNK was added to culture medium of the neuroblastoma cell line, SH-SY5Y, and incubated for the indicated periods (Fig. 1 A and C). The cells were cotreated with 100 nM MitoTracker red, a fluorescent indicator of mitochondria, for 10 min just before stopping the incubation with PTD-FNK and fixed with paraformaldehyde. The cells were stained with anti-Bcl-x mAb and the fluorescence was detected by confocal scanning fluorescent microscopy. Green fluorescence indicating Bcl-x derivatives was faint at time 0, but increased over time, achieving a steady-state level in ≈1 h. The PTD-FNK protein was observed to transduce into virtually all of the cells. In addition, the majority of green fluorescence colocalized with the red fluorescence of MitoTracker red as revealed by the merged images. The decay of the PTD-FNK protein was examined, revealing that the half-span of the incorporated protein is ≈2 h (Fig. 1D).

Fig 1.

(A) Transduction of PTD-FNK into neuroblastoma SH-SY5Y cells. At the time indicated after the addition of 30 nM PTD-FNK, cells were washed once with PBS and fixed. Ten minutes before fixation, 100 nM MitoTracker red was added as the mitochondria indicator. PTD-FNK was immunodetected with anti-Bcl-x antibody (green) and colocalized with mitochondria (red). (Scale bar: 20 μm.) (B) Transduction of PTD-FNK into primary cultured neocortical neurons. Thirty minutes after adding PTD-FNK at the indicated concentrations, cells were washed once with PBS and fixed. Ten minutes before fixation, 100 nM MitoTracker red was added. Immunoreactivity of PTD-FNK (green) localized with mitochondria (red). (Scale bar: 20 μm.) (C) Quantification of the incorporated PTD-FNK protein in SH-SY5Y and primary neocortical neurons. The fluorescent intensity (arbitrary unit) of each cell in A (Left) and B (Right) were estimated by using the nih image program to obtain averages with SD (vertical bars). *, P < 0.003, compared with time 0 h (Left) and with no PTD-FNK (Right) by one-way ANOVA. (D) Decay of PTD-FNK incorporated into SH-SY5Y cells. Cells were incubated with 30 nM PTD-FNK for 1 h. After being washed once with DMEM, the cells were incubated in the medium with FBS for indicated periods, followed by fixation for immunostaining with anti-Bcl-x. Fluorescence of the cells was imaged with a confocal scanning microscope. The fluorescent images (four fields of view, each field containing five cells) were quantified with nih image. The intensity of cells without PTD-FNK was used as a background to be subtracted from those of cells with PTD-FNK. The average at 0 h is taken as 100%. Vertical bar, SD.

To study the concentration of intracellular protein delivery, neocortical neurons were prepared from rat embryos and treated with 30 or 300 nM PTD-FNK for 30 min. Endogenous Bcl-x_L protein gave punctate fluorescent signal and appeared to colocalize with mitochondria. In neurons incubated with PTD-FNK, the intense green fluorescent signals increased depending on PTD-FNK concentration and colocalized with the red fluorescence from MitoTracker red (Fig. 1 B and C Right). Taken together, these observations indicate that PTD-FNK is rapidly transduced into neuroblastomas or primary cultured neurons and localizes to mitochondria.

Protection of Neuronal Cell Death by PTD-FNK Transduction in Vitro.

To examine the cytoprotective activity of PTD-FNK after transduction into cells, we performed two in vitro cell-death protection experiments on (i) STS-induced apoptosis of SH-SY5Y cells (26) and (ii) glutamate excitotoxicity of primary neocortical neurons (27).

As expected, STS induced apoptosis in SH-SY5Y cells as judged by nuclear fragmentation (Fig. 2 C–E). The PTD-FNK protein markedly inhibited cell death in a PTD-FNK concentration-manner (Fig. 2A). More importantly, PTD-FNK was more potent in inhibiting cell death compared with its WT version PTD-Bcl-x_L and FNK itself (Fig. 2B).

Fig 2.

(A) Dose-dependent cytoprotective effect of PTD-FNK on STS-induced apoptosis of neuroblastoma SH-SY5Y cells. One hour after PTD-FNK addition at the concentrations indicated, STS was added. The cells were incubated for another day. Surviving cells were counted by staining with PI and Hoechest 33342. The mean of three to four independent experiments is presented with SD (vertical bars). *, P < 0.01; **, P < 0.001, compared with no PTD-FNK by one-way ANOVA. (B) Comparison between PTD-FNK and PTD-Bcl-x_L in the cytoprotective activity against STS. One hour after the addition of 30 pM of each protein (PTD-FNK, PTD-Bcl-x_L) or buffer (Ctl), STS was added at the concentrations indicated. Cells were incubated for 1 more day and stained with PI (red) and Hoechest 33342. The mean of three to four independent experiments is presented with SD (vertical bars). *, P < 0.05; **, P < 0.0001, compared with buffer only by one-way ANOVA. (C–E) Representative pictures of apoptotic nuclei (indicated by arrows) with (D and E) or without (C) 800 nM of STS (×200). One hour before the addition of STS, 300 pM PTD-FNK was added (E). The cells were incubated for 1 day and stained with PI (red) and Hoechst 33342 (blue).

To induce excitotoxicity, neocortical neurons that had been pretreated with 30 nM PTD-FNK for 2 h were incubated with HBSS containing 2 mM CaCl₂ and 1 mM glutamate for 1 h. Neurons were washed with HBSS in the absence of Ca²⁺ and glutamate and placed in complete medium for 24 h, and then viable cells were counted under a microscope. PTD-FNK markedly reduced glutamate-induced toxicity (Fig. 3 A and C). The cytoprotective effect of PTD-FNK was maximum at 3 pM to 3 nM. Interestingly, PTD-FNK significantly protected neurons from glutamate excitotoxicity at concentrations as low as 0.3 pM. Next, we compared the cytoprotective effect of PTD-FNK with that of PTD-Bcl-x_L or the FNK protein without PTD at 30 pM (Fig. 3B). PTD-FNK was twice as effective as PTD-Bcl-x_L in protecting neurons. Glutamate excitotoxicity is widely considered to induce necrotic cell death (28). Thus, PTD-FNK appeared to inhibit necrosis in addition to apoptosis. The latter observation is not surprising because FNK gene expression prevents against cell death induced by a calcium ionophore (16).

Fig 3.

(A) Dose-dependent neuroprotective effect of PTD-FNK on excitotoxic cell death in primary cultured neocortical neurons induced to glutamate. Two hours after addition of PTD-FNK at the concentrations indicated, neocortical neurons were exposed to 1 mM glutamate for 1 h. The cells were washed and incubated in medium for 1 day. The number of viable neurons was counted under a phase-contrast microscope in five fields of view (FOV) per well for four independent wells, and the average is presented with SD (vertical bars). *, P < 0.05; **, P < 0.001, compared with 0 pM by one-way ANOVA. (B) Comparison between PTD-FNK and PTD-Bcl-x_L in the neuroprotective activity against glutamate. Neocortical neurons were incubated with 30 pM of each protein (PTD-FNK, PTD-Bcl-x_L, and FNK) or buffer only (Ctl) for 2 h. Washed cells were exposed to glutamate as described in A. After 1 day, the number of viable neurons was counted under a phase-contrast microscope in five FOV per well for four independent wells. The average is presented with SD (vertical bars). *, P < 0.05, compared with no protein (Ctl) by one-way ANOVA. (C) Representative images of viable neurons observed in A by immunostaining with anti-MAP2 antibody (×100). (D) Delayed [Ca²⁺]_i deregulation in neurons exposed to glutamate. Representative tracings of changes in [Ca²⁺]_i are shown. Neurons were pretreated with or without PTD-FNK (3 nM) for 2 h in medium. Washed cells were placed in HBSS containing 2 mM Ca²⁺ and Fluo-3 AM. Twenty individual cells were assessed by confocal scanning fluorescent microscopy. Many neurons without PTD-FNK showed a secondary delayed increase in [Ca²⁺]_i indicated by * (Left), after the initial response induced by the addition of glutamate (arrows). In most neurons treated with PTD-FNK, the secondary increase did not occur (Right).

We also examined how PTD-FNK protects neurons against the excitotoxicity. Cultured neocortical neurons treated with or without PTD-FNK were placed in HBSS containing 2 mM CaCl₂ and Fluo-3 AM to measure [Ca²⁺]_i. Fluorescence signal intensity was monitored from 20 neurons for 2 h by confocal scanning fluorescent microscopy. [Ca²⁺]_i increased rapidly after the addition of glutamate and gradually to reach a plateau. The second increase of [Ca²⁺]_i clearly occurred ≈50 min later in 55–60% (11–12/20) of the neurons monitored (Fig. 3D Left). This finding agrees with previous reports (22) and the second increase could be responsible for neuronal cell death (23). In contrast, the second increase in [Ca²⁺]_i was observed significantly in only 20% (4/20) of the neurons that had been pretreated with 3 nM PTD-FNK during the measurement time (P = 0.02 by the Mann–Whitney U test). Thus, these experiments suggest that PTD-FNK affects the movement of calcium ions, leading to the protection against excitotoxicity, directly or indirectly.

Cerebral Ischemic Damage Reduced by PTD-FNK Transduction in Vivo.

To test whether PTD-FNK can be delivered into neurons in the brain, myc-tagged PTD-FNK (PTD-myc-FNK) was first injected i.p. into a mouse. Immunohistochemical staining using anti-myc-tag antibody revealed strong immunoreactivity in the microvessels as well as the cytoplasm of cortical neurons, compared with those of a control mouse injected with vehicle (Fig. 4 A and B), indicating that PTD-FNK can indeed cross the blood–brain barrier and get into the neuronal cells. Transient global ischemia causes the slow progressive degeneration of hippocampal CA1 neurons in gerbils. Glutamate release and apoptotic regulation have been shown to contribute to delayed-neuronal cell death (29–34). We chose ischemic brain injury as a model system to test the cytoprotective effect of PTD-FNK in vivo.

Fig 4.

(A and B) Transduction of myc-tagged PTD-FNK proteins into neurons in the brain. PTD-myc-FNK (50 mg/kg) (A) or vehicle (B) was injected i.p. into mice. After 10 h brains were perfused transcardially with 4% paraformaldehyde. Paraffin sections (4 μm) were incubated with a rabbit anti-Myc Tag polyclonal antibody. The antibody complex was detected with a DAKO Envision⁺ system. Cortices are shown. (Scale bars: A and B, 50 μm.) (C–H) PTD-FNK reduces cerebral ischemic damage. Gerbils injected with vehicle (n = 8) or PTD-FNK (5 mg/kg; n = 5) were subjected to ischemic insults for 5 min. After 7 days, animals were perfused transcardially with 4% formaldehyde and 5-μm thick cross sections containing hippocampal tissue were stained with hematoxylin–eosin. (C) Neuronal cell density in the hippocampal CA1 region of gerbils injected with vehicle (Ctl) or PTD-FNK. The number of surviving CA1 neurons is shown as intact neurons per 1-mm CA1 length (neuronal cell density; see Materials and Methods). The horizontal bar represents the average of neuronal cell density in gerbils injected with PTD-FNK. The average in gerbils without ischemic insult (untreated; n = 3) is 196 cells per mm. The histopathological damage (percentage dead neurons) was statistically assessed by the Mann–Whitney U test. (D–K) The hippocampus stained with hematoxylin–eosin. The hippocampus of a gerbil injected with vehicle (D and E) or PTD-FNK (F–I). The hippocampus of a gerbil without ischemic insult is also shown (J and K). The boxes in D, F, H, and J have been enlarged in E, G, I, and K, respectively. (Scale bars: D, F, H, and J, 250 μm; E, G, I, and K, 50 μm.) The number of surviving CA1 neurons in F and H was 35 and 174 cells per mm, respectively, and is shown in C.

We asked whether in vivo injection of PTD-FNK could protect CA1 neurons against ischemic damage. In control gerbils injected with vehicle, fewer than 1% of CA1 neurons survived the ischemic insult (Fig. 4 C–E). In contrast, injection of PTD-FNK apparently inhibited the delayed neuronal cell death (Fig. 4 C and F–I). The magnitude of the protective effect of PTD-FNK ranged from 16% to 89% survival of CA1 neurons (Fig. 4C), which is statistically significant. A neuronal survival of 89% in ischemic insult suggests that PTD-FNK is as effective as ischemic preconditioning of 2 min (24) and mild hypothermia (35) and much more effective than virus-mediated introduction of the bcl-2 gene (12). As shown in this study, the neuroblastomas and primary neurons transduced with a low amount of PTD-FNK (pM order) exhibited resistance to STS-inducing apoptosis and glutamate excitotoxicity in vitro, respectively. Thus, it is a plausible explanation that PTD-FNK delivered to CA1 neurons protects cells from death caused by the collapse of Ca²⁺ homeostasis and/or oxidative stress caused by ischemic insult as suggested by our in vitro experiments (16). Additionally, cells of the other types, such as endothelial and glial cells, transduced with PTD-FNK may also contribute to the amelioration of the ischemic brain injury.

The induction of hypothermia is known to inhibit neuronal death after ischemia (35). Because the gerbils were kept at 37°C for 6 h after the ischemic episode, it is unlikely that the anti-ischemic efficacy of PTD-FNK is mediated through its hypothermic properties. Thus, a greater synergistic effect may be expected on combining PTD-FNK injection with hypothermic treatment.

In conclusion, PTD-FNK injected into rodents can be successfully delivered to the brain and prevents the slow progressive death of neurons. Proteins delivered into cells by this strategy would remain in cells only transiently. In addition, transgenic mice expressing FNK have shown no particular phenotype, indicating lack of gross adverse effects (unpublished work). Thus, the toxicity of PTD-FNK would not be critical. PTD-FNK was delivered in a neuroprotective paradigm before the brain insult in this study. Because neurons die in a week after the ischemic insult, this strategy may be useful as the neurorestorative treatment. The therapeutic time frame and effective administration protocols remain to be determined. Thus, time course and dose effects need to be investigated in the near future for details. In addition, the careful comparison of PTD-FNK with PTD-Bcl-x_L or FNK itself may be required.

Cao et al. (36) have recently reported that PTD-Bcl-x_L inhibited STS-induced apoptosis in primary cultured neocortical neurons and that PTD-Bcl-x_L delivered to the brain when i.p. injected and inhibited neuronal cell death induced by transient focal ischemia. Because pancreatic islets transduced with PTD-Bcl-x_L fusion protein in vitro were more resistant to apoptosis and were transplanted without functional failure (37), PTD-FNK is also expected to be more useful for the transplantation of organs or tissues. Our results strongly suggest that PTD-FNK is a potent protein with various therapeutic applications.

Abbreviations

• [Ca²⁺]_i, intracellular free Ca²⁺

• PTD, protein transduction domain

• PI, propidium iodide

• STS, staurosporine