Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Paraneoplastic neurologic syndromes are disorders of the nervous system that occur in association with cancer but are not a direct result of either primary or secondary lesions within the brain itself (1). Paraneoplastic encephalomyelitis (PEM) and/or paraneoplastic sensory neuronopathy (PSN) is one such disorder sometimes associated with small-cell lung cancer (SCLC) (2). Pathologically, the disease is characterized by neuronal loss, gliosis, and inflammatory infiltrates of the neuraxis, indicative of an immunologic pathogenesis. HuD antigen is a human neuronal RNA-binding protein and a homologue of the drosophila protein embryonic lethal abnormal visual system (Elav) (3). Patients with PEM/PSN and SCLC develop an intense immune response against HuD antigen or its homologous Hu antigens, HuC and Hel-N1, which are expressed in the nuclei of tumors and neurons of the peripheral and central nervous systems (2). This immune response is characterized by the presence of anti-Hu antibodies in serum and cerebrospinal fluid and deposits of anti-Hu immunoglobulin (Ig)G within the nervous system and tumor itself (2, 9, 10). The mechanism by which this immunoresponse to Hu antigens occurs in PEM/PSN is largely unknown. However, since HuD is the only Hu antigen expressed in SCLC tumors, it would appear to play a central role in triggering an anti-Hu immune response in SCLC (5, 11, 12). More importantly, the presence of anti-HuD antibodies in sera of patients with SCLC is directly related to the limitation of tumor spread. Approximately 15% of patients who have SCLC without neurologic dysfunction have anti-Hu antibodies in their sera. Most of these patients' tumors are confined to the thorax at diagnosis (13). In contrast, more than 50% of patients who have SCLC without anti-Hu antibodies have metastatic disease at the time of diagnosis (2). Clinical reports of spontaneous SCLC tumor remission in patients with PEM/ PSN and anti-Hu antibodies (14) suggest that the activation of anti-HuD immunity leads to a more favorable prognosis.

The effect of DNA vaccine against cancer and various infections (such as influenza) has recently been investigated. In these studies, it has been demonstrated that direct injection of a plasmid coding for carcinoembryonic antigen (15) or viral proteins (16) produces protective antibodies and elicits a cell-mediated immune response. Compared with orthodox vaccines consisting of tumor proteins or viral components, DNA vaccination allows evaluation of host immunity against transgene-encoding protein without any process of protein purification.

The function and distribution of HuD in the murine nervous system are similar to that in humans. We have therefore hypothesized that DNA vaccination with HuD- expressing plasmid will induce antitumor immunity against HuD-expressing tumors in a murine model. Because there is no available murine SCLC cell line, we have used a murine adenocarcinoma cell line with stable expression of mouse HuD antigen as an SCLC tumor model in this study.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of the Plasmid

The full-length sequence (1,158 base pairs [bp]) of mouse HuD antigen (mHuD) was amplified by polymerase chain reaction using complentary DNA (cDNA) obtained from the brains of BALB/c mice (17). Unique SmaI restriction sites were added to both ends of the mHuD sequence by using the following primers: 5'-TCCCCCGGGTCAAAGATGGAGTGGAATGGCTTGAAG-3' and 5'-TCCCCCGGGTCAGGATTTGTGGGCTTTGTTGGTTTT-3'. After treatment with SmaI, the mHuD sequence was subcloned into the multiple cloning site of the mammalian cell expression plasmid pCI-neo Vector (Promega, Madison, WI) driven by the cytomegalovirus (CMV) major immediate/ early promoter/enhancer. It generated a plasmid with a right-oriented mHuD sequence, designated pCMV.mHuD(+), and a plasmid with the mHuD sequence in the opposite orientation, pCMV.mHuD(), which was used as a negative control plasmid. Preparation of the plasmids was performed using a plasmid purification kit (Qiagen Plasmid Mega kit; Qiagen, Inc., Valencia, CA). The nucleotide sequence of mHuD in pCMV.mHuD was determined by dye terminator cycle sequencing and was identical with the alternatively spliced short form of mHuD. A 42-bp fragment was deleted from mHuD at the same site, as reported in rat and human HuD sequences (2, 18).

Mice and Cell Lines

Female 5-wk-old BALB/c mice (H-2^d), purchased from Charles River Japan (Yokohama, Japan), were used throughout this study. Mouse fibroblasts NIH/3T3 (ATCC CRL-1658) were grown in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine, 50 U/ml penicillin, and 50 µg/ml streptomycin (Life Technologies, Inc., Rockville, MD), and 10% heat-inactivated calf serum at 37°C in 5% CO₂ (19). Mouse adenocarcinoma Colon 26 cells (H-2^d) were grown in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 50 U/ml penicillin G, and 50 µg/ml streptomycin at 37°C in 5% CO₂. Because SCLC cell lines from BALB/c mice were not available, mHuD stable transfectants (Colon 26/HuD) were prepared by electroporation (290 V, 625 mF, 7.70 ms) of 10 µg of pCMV.mHuD(+) into parent Colon 26 cells, and selected in medium containing 800 µg/ml of G418 sulfate (Geneticin; Life Technologies). Several lots of transfectants (Colon 26/HuD cells) were grown separately in two subculture generations and frozen. The same passages of Colon 26/ HuD and Colon 26 cells were used in the experiments.

Northern Blot Analysis

To evaluate mouse HuD antigen expression in vitro, we transfected 10 µg of either the (+) or () isomer of plasmid pCMV.mHuD into NIH/3T3 cells by electroporation (290 V, 625 mF, 6.86 ms). After 72 h incubation, total RNA was extracted as previously described (20). Ten micrograms of total RNA was electrophoresed in 0.8% agarose gel containing 1.6 M formaldehyde, 1× 50 mM 3-(N-morpholino)propanesulfonic acid, and 10 mM ethylenediaminetetraacetic acid (EDTA), and transferred to a Hibond N nylon membrane (Amersham, Arlington Heights, IL). The membrane was hybridized with [³²P]-labeled HuD cDNA probe, washed at high stringency, and exposed to X-ray film for 3 h at room temperature. After probing with mHuD cDNA, the membrane was stripped and probed with [³²P]-labeled -actin as a control (20).

Western Blot Analysis

NIH/3T3 cells transfected with 10 µg of HuD plasmid by electroporation were harvested after 72 h incubation. Mouse HuD stable transfectant (Colon 26/HuD) or parent Colon 26 cells were also harvested by treatment with trypsin-EDTA and washed with phosphate-buffered saline (PBS). These harvested cells were lysed with a solution (150 µl per 10-cm culture dish) containing 1 mM (2S,3S)-trans-epoxysuccinyl-L - leucylamido - 3 - methylbutane (Sigma, St. Louis, MO), 1 mM phenylmethylsulfonyl fluoride (Sigma), 0.5% Nonidet P-40, 100 mM NaCl, and 50 mM phosphate buffer (pH 7.4). After centrifugation at 18,000 × g for 5 min at 4°C, the solubilized protein was subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 7% polyacrylamide gel and transferred onto a polyvinylidene difluoride (PVDF) membrane (Immobilon-PSQ; Millipore, Bedford, MA). The blots were blocked in BlockAce (Dai-nippon Pharmaceutial Co., Ltd., Tokyo, Japan) containing 0.1 mg/ml bovine serum albumin (Fraction V; Sigma). The initial antibody was obtained from diluted serum (1:5,000) of a patient with PEM diagnosed at the Department of Neurology, Juntendo University Hospital (Tokyo, Japan). Immunohistochemical staining demonstrated that the patient's serum reacted with mouse cerebral tissue (21). In addition, Western blot analysis demonstrated that the patient's serum reacted with a 41-kD HuD band in the cell lysate of BALB/c mouse brain (see Figure 1b). After incubation for 16 h, the blots were probed with peroxidase-conjugated goat antihuman IgA + IgG + IgM (H+L) (1:5,000; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 1 h at 25°C. The blots were developed with the ECL Western blotting system (Amersham International PLC, Buckinghamshire, UK).

View larger version (22K): [in this window] [in a new window]   Figure 1.   (a) Expression of mHuD mRNA transcript in NIH/3T3 cells transfected with mHuD-expressing plasmid. Northern blot analyses of transfected NIH/3T3 cells hybridized with mouse HuD cDNA probe (lanes 1-3) or -actin cDNA probe as a control (lanes 4-6). Lane 1, NIH/3T3 cells transfected with plasmid pCMV.mHuD(+); lane 2, NIH/3T3 cells transfected with negative control plasmid pCMV.mHuD(); lane 3, sham-treated cells; lanes 4-6, parallel to lanes 1-3, respectively. Molecular markers are present at the right side of lane 3. (b) Expression of mHuD protein in mouse cells. Western blot analyses of mouse brain lysates (lanes 1 and 2), lysates of NIH/3T3 cells (lanes 3-5), and lysates of Colon 26/HuD (lane 6) and Colon 26 (lane 7). The patient's serum was incubated as a first antibody in all lanes, except for lane 2. In lane 2, diluted (1:500) serum from a normal volunteer was used. Lane 3, NIH/3T3 cells transfected with pCMV.mHuD(+); lane 4, NIH/3T3 cells transfected with pCMV.mHuD(); lane 5, NIH/3T3 cells without transfection; lane 6, HuD stable transfectant (Colon 26/HuD) cells; lane 7, parent Colon 26 cells. Molecular marker is to the right of lane 7. Arrow indicates the position of a 41-kD band from mHuD.

MTT Assay

In order to evaluate the in vitro growth of Colon 26/HuD mHuD transfectants, Colon 26/HuD (lot.1) or parent Colon 26 cells were harvested, washed, and seeded at a density of 5 × 10⁴ cells/well (2.5 × 10⁴ cells/cm²) in 500-µl aliquots of RPMI-1640-10% FCS using 24-well culture plates (Day 0). In this assay, Geneticin was omitted from Colon 26/HuD. The culture medium was changed every 3 d for both cells. At 24 h (Day 1), on Days 2, 4, or 7 after seeding, 100 µl of sterile 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; 5 mg/ml) was added and incubated for 3.5 h at 37°C (22, 23). After rinsing with PBS, the resulting blue formazane product was solubilized with 400 µl of dimethyl sulfoxide. The optimal density was determined in triplicate against a reagent blank (no cell) at a test wavelength of 570 nm and a reference of 630 nm. The value reflected the viable cell population in each well.

Immunization of Animals

Plasmid pCMV.mHuD(+) or (), 75 µg, in 0.1 ml of sterile PBS was injected intramuscularly into 5-wk-old female BALB/c mice. Plasmid was injected three times at 2-wk intervals, the initial injection at Day 0, then Days 14 and 28 alternatively into either side of the quadriceps muscle.

In Vivo Tumor Models

To evaluate the feasibility of DNA vaccination against HuD antigen to suppress in vivo tumor growth, two protocols were used.

Protocol 1. Subcutaneous tumors were established by injection of 1 × 10⁶ Colon 26/HuD (lot.2) cells or Colon 26 cells into the right or left flank of BALB/c mice vaccinated with either pCMV.mHuD(+) or pCMV.mHuD() at Day 36 after immunization. Each group consisted of four mice. This protocol was performed to evaluate the growth of Colon 26/HuD tumor, compared with the Colon 26 tumor control in the same animal. After shaving, tumor size was measured using calipers and calculated as a product of width and length. Final measurement of size was taken at the time of death.

Protocol 2. To examine the growth of Colon 26/HuD tumor without the influence of Colon 26 cells, Colon 26/ HuD cells alone were injected into the mice vaccinated with pCMV.mHuD(+) or pCMV.mHuD() or into sham-treated mice. Sham-treated control mice received only PBS three times at 2-wk intervals. Each group consisted of five mice. At Day 36 after vaccination, 1 × 10⁶ of Colon 26/HuD cells (lot.3) were injected into both flanks of all mice. Tumor size was measured until 19 d after inoculation of tumor cells. Tumor size was calculated as a product of the mean width and length from both sides.

Assay of Thymidine Uptake in Spleen Cells

The spleen was excised 19 d after inoculation of tumor cells from DNA-vaccinated or sham-treated animals. Each group consisted of four mice. Single-cell suspensions of spleen cells were depleted of erythrocytes by centrifugation onto Ficoll-Conray solution M-SMF (JIMRO, Takasaki, Japan). Spleen cells were rinsed and resuspended in RPMI-1640 medium, supplemented with 10% FCS, 2 mM L-glutamine, 0.1 mM nonessential amino acids, 10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid, and 50 µM 2-mercaptoethanol (24). Irradiated (8,000 rad) Colon 26/HuD cells were used as target cells. A mixture of 1 × 10⁵ spleen cells and 1 × 10⁴ irradiated target cells were cultured with 5 U/ml of human recombinant interleukin-2 in round-bottomed 96-well plates for 24 h. Tritiated thymidine (20 Ci/mmol; NEM Life Science Products, Inc., Boston, MA) was added at 1 µCi/well and incubated for a further 24 h. The intact cells were harvested using a Micro 96 Harvester 11057 (Skatron Instruments, Lier, Norway), and radioactivity from thymidine uptake for DNA synthesis was measured by a liquid scintillation counter 1450 MicroBeta (Wallac, Turku, Finland) in triplicate. Radioactivity was compared between the groups of animals vaccinated with pCMV.mHuD(+) and () after subtraction of the value of sham-treated mice.

Antimouse HuD Antibody Production

The cell lysate of Colon 26/HuD cells was applied to 7% SDS-PAGE and transferred to a PVDF membrane (Immobilon-PSQ). The membrane strip was blotted for 2 h at 24°C with 1:500 diluted serum obtained from DNA-vaccinated animals at Day 36 after vaccination. The patient's serum was used as a positive control, as described previously. As a second antibody, 1:5,000 diluted horseradish peroxidase-labeled goat antimouse Ig (IgA + IgG + IgM [H+L]) (Southern Biotechnology Associates, Inc., Birmingham, AL) was used for the sera of DNA-vaccinated animals, or 1:5,000 diluted peroxidase-conjugated goat antihuman IgA + IgG + IgM (H+L) (Jackson ImmunoResearch Laboratories, Inc.) for the patient's serum. The blots were incubated for 1 h at 37°C, and developed with the ECL Western blotting system (Amersham).

Data Analysis

Data are presented as means ± standard deviation. Comparison of tumor size in each group was made by one-way analysis of variance. A probability of < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In Vitro Expression of Mouse HuD Antigen

Northern blot analysis revealed 3.7-kb messenger RNA (mRNA) transcripts of mouse HuD antigen in NIH/3T3 cells transfected with plasmid pCMV.mHuD(+) but not in NIH/3T3 cells transfected with plasmid pCMV.mHuD() or NIH/3T3 without transfection (Figure 1a). NIH/3T3 cells transfected with pCMV.mHuD() showed no band in the Northern blot hybridized with a [³²P]-labeled double-stranded HuD cDNA. This would represent the instability of RNA transcripts after transient transfection in these cells. It has been reported that the serum of the patient with PEM/PSN reacted with Hu antigens between 35 and 41 kD (25). Western blot analysis demonstrated that our patient's serum reacted with a 41-kD band in the cell lysate of mouse brain (Figure 1b, lane 1). Further, lower and higher molecular-weight bands were also detected. However, these bands may be nonspecific because they were also detected in serum obtained from a normal volunteer (Figure 1b, lane 2). Lysate of transient transfectant (NIH/3T3) or stable transfectant (Colon 26/HuD) with pCMV.mHuD(+) reacted with the same sized band when incubated with serum from the patient (Figure 1b, lanes 3 and 6). In contrast, no 41-kD band was detected in the lysate of NIH/3T3 cells transfected with pCMV.mHuD(), sham treatment, or parent Colon 26 cells (Figure 1b, lanes 4, 5, and 7). Thus, HuD-expressing plasmid can induce transient expression of HuD mRNA transcripts and protein in NIH/3T3. Moreover, stably transfected Colon 26 cells could produce mouse HuD antigen.

Suppression of Tumor Growth by DNA Vaccination

Mice were divided into two groups that received either plasmid pCMV.mHuD(+) or (). No death was observed during the vaccination. Body weight gain was comparable in each group. Subcutaneous inoculation of either Colon 26/HuD or Colon 26 cells in each side was performed at Day 36 after vaccination. In Protocol 1, subcutaneous tumor was first recognized 7 d after inoculation of cells. The growth of Colon 26/HuD tumor in pCMV.mHuD(+)- treated groups was slightly suppressed compared with that of the pCMV.mHuD()-treated group 11 d after the inoculation of tumor cells (Figure 2). At Day 19, the size of Colon 26/HuD tumor in pCMV.mHuD(+)-treated animals (86.6 ± 29.9 mm²) was significantly less than that in pCMV.mHuD()-treated animals (195.3 ± 48.1 mm², P < 0.05). On the other hand, tumor size of Colon 26 cells in pCMV.mHuD(+)-vaccinated animals (318.2 ± 78.0 mm²) was similar to that of pCMV.mHuD()-vaccinated mice (320.1 ± 87.9 mm², P > 0.9). In vivo growth of Colon 26/ HuD tumor was less than that of Colon 26 tumor even in the mice who received negative control plasmid (Figure 2). This might reflect the slower growth of Colon 26/HuD cells compared with that of Colon 26 cells, because MTT assay revealed that in vitro growth of Colon 26/HuD was slightly less than that of parent Colon 26 cells (Figure 3).

View larger version (23K): [in this window] [in a new window]   Figure 2.   Suppression of the growth of HuD-expressing cell tumors (Colon 26/HuD) in mice vaccinated with HuD-expressing plasmid pCMV.mHuD(+). Tumors were established by the injection of 1 × 106 Colon 26/HuD or Colon 26 cells subcutaneously into the right or left flank of previously vaccinated BALB/c mice (Protocol 1 in MATERIALS AND METHODS). Size of Colon 26/ HuD and Colon 26 tumors were calculated at each time point.

View larger version (13K): [in this window] [in a new window]   Figure 3.   In vitro growth of mouse HuD-expressing cells (Colon 26/HuD) and Colon 26 cells analyzed by MTT assay. Colon 26/ HuD or Colon 26 cells were seeded at a density of 5 × 104 cells/ well in 24-well culture plates (Day 0). Optical density (OD) at 570 nm subtracted from the value at 630 nm reflects the number of viable cells at each time point. Growth of Colon 26/HuD cells is slower than that of Colon 26 cells.

In Protocol 2, injection of a different lot of Colon 26/HuD cells was performed in DNA-vaccinated animals. At Day 36 after vaccination, 1 × 10⁶ Colon 26/HuD cells were injected into both flanks of all animals. Measurement of tumor size was continued until 19 d after inoculation of tumor cells (Figure 4). Mean tumor size at Day 19 in pCMV.mHuD(+)- vaccinated mice (107.7 ± 12.8 mm²) was significantly smaller than that in pCMV.mHuD()-vaccinated mice (156.6 ± 22.8 mm², P < 0.05) or in sham-treated animals (145.6 ± 10.2 mm², P < 0.05). Tumor size in the animals vaccinated with pCMV.mHuD() was similar to that of sham-treated animals (P > 0.3). In Protocol 2, tumor suppression was also observed by using a different lot of Colon 26/HuD cells without Colon 26 control cells. Thus, consistent antitumor activity of DNA vaccination was shown in two protocols against different lots (subpopulations) of tumor cells.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Suppression of the growth of HuD-expressing cell tumors (Colon 26/HuD) in mice vaccinated with mouse HuD- expressing plasmid. Colon 26/HuD cells, 1 × 106, were injected into both flanks of animals who received DNA vaccine (Protocol 2). Tumor size was measured on 19 consecutive days after inoculation and is represented as a mean value of both sides.

Histologic examination of brain tissue obtained from the animals vaccinated by pCMV.mHuD(+) with or without inoculation of Colon 26/HuD tumor cells revealed no obvious inflammation in the cerebrum and cerebellum (data not shown).

Thymidine Uptake of Spleen Cells

Spleen cells were obtained from mice vaccinated with pCMV.mHuD(+) or () or sham-treated mice, and de novo DNA synthesis was evaluated by incubation with Colon 26/HuD cells in the presence of tritiated thymidine. The uptake of radioactive thymidine by spleen cells obtained from the animals vaccinated with pCMV.mHuD(+) (7,212 ± 502 cpm) was significantly increased compared with that obtained from animals vaccinated with pCMV.mHuD() (574 ± 2,649 cpm, P < 0.05).

Antimouse HuD Antibody Production

Sera were collected from the animals that received either plasmid pCMV.mHuD(+) or () or from sham-treated mice to detect the production of anti-HuD antibody. Tumor cells were not inoculated into these mice. Each vaccinated group consisted of four animals, and individual sera were subjected to Western blot analysis. In the group of animals vaccinated with pCMV.mHuD(+), a 41-kD band was detected in all examined sera, although no similar band was seen in animals that received pCMV.mHuD() or in sham-treated animals (Figure 5). These results indicate that HuD-expressing DNA vaccine elicited anti-HuD antibody production in these mice.

View larger version (43K): [in this window] [in a new window]   Figure 5.   Presence of anti-HuD antibody in individual sera obtained from mice immunized with HuD-expressing plasmid pCMV.mHuD(+). Lysate of HuD stable transfectants (Colon 26/ HuD) was electrophoresed in 7% SDS-PAGE and transferred to a PVDF membrane. Western blot analyses were performed with diluted (1:500) serum obtained from pCMV.mHuD(+)-vaccinated mice as a first antibody. Lane 1, incubation with serum of the patient with PEM as a positive control. Arrow indicates a 41-kD band corresponding to mHuD; lanes 2-6, incubation with individual sera obtained from pCMV.mHuD(+)-vaccinated mice. Lane 7, incubation with serum obtained from a pCMV.mHuD()- vaccinated mouse; lane 8, incubation with serum obtained from a sham-treated mouse.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Although the present study shows that vaccination with HuD-expressing plasmid suppressed HuD-expressing tumor growth, the immune mechanism for the tumor suppression remains obscure. Cytotoxic T-cell killing of tumor plays a role in suppressing the tumor growth. However, antibody-dependent cell-mediated cytotoxicity is also an attractive explanation for tumor suppression from the clinical observation. The humoral immunity is related to antitumor immunity, from the observation that the presence of anti-Hu antibodies at the time of SCLC diagnosis is a strong and independent predictor of complete response to treatment and prolonged survival (26).

Since Liu and colleagues published the sequence of human HuD antigen (in 1995), alternatively spliced isoforms of HuD have been reported among various species (27). Liu and colleagues reported two alternative spliced forms of human HuD, namely HuDpro and HuDmex. HuDpro contained a 14-amino-acid sequence (LMSGPVPPSACSPR) next to the 3' end of the M region of HuD (27). HuDmex had 13 amino acids (LDNLLNMAYGVKR) deleted from HuD (27). It has been reported that the expression of HuDmex is increased in human SCLCs (5). In addition, it has also been reported that an alternatively spliced form of rat HuD has an identical 13-amino-acid sequence deleted, as seen in HuDmex (18). We obtained a short form of the mouse HuD sequence that had 42 nucleotides deleted, identical to 13 amino acids deleted in human HuDmex, in the result of amplification of cDNA that originated from the RNA of mouse brain. Northern blot analysis of our mouse HuD antigen gene revealed a 3.7-kb transcript, as seen in the rat HuD gene (18). Western blot analysis of transient and stable transfectants of the HuD gene revealed the presence of an approximately 41-kD band, almost the same size as seen in recombinant human HuD (3). In this study, the plasmid based on mouse, but not human, HuD sequences was used to immunize mice because it would avoid the immune response directed to human transgene-encoded proteins in mice. In fact, it has been reported that human erythropoietin, which is 79% identical at the amino-acid level, induced antibody production in mice (28).

Immunization with purified recombinant HuD protein based on the sequence of human HuD antigen can generate a high titer of anti-HuD antibody but not cause any pathologic findings in the brain (29, 30). Of particular interest in this study, then, was whether DNA vaccination with the mouse HuD gene, with or without challenge of HuD-expressing tumor cells, caused neurologic disease in mice. It is known that DNA vaccination elicits cellular and humoral immunity (15, 16). In fact, DNA vaccination with HuD-expressing plasmid induced de novo DNA synthesis and anti-HuD antibody production in this study. However, evidence of encephalomyelitis was not detected in the brains of mice vaccinated with HuD-expressing plasmid before or after tumor challenge (data not shown). Although there were questions about the role of anti-Hu antibodies in the pathogenesis of PEM/PSN, the strength of the immune response in the host may influence the onset of paraneoplastic neuronal syndromes, on the basis of the observation that patients with PEM/PSN had a higher titer of anti-Hu antibodies than did patients without neurologic disorders (2). It is suggested that the distribution of HuD antigen within neurons and SCLC is not restricted to the nucleus but is also detected on the cell surface (25). The surface-expressed HuD antigen would be recognized by the host immune system and be related to the onset of a paraneoplastic syndrome and the suppression of tumor spread in patients with SCLC. Because cell-surface expression of HuD antigen depends on nuclear expression of HuD antigen, it would be important to increase total expression of HuD antigen in the cells to activate the immune system. To potentiate antitumor immunity, we have to improve the efficacy of HuD expression in tumor cells or enhance the host immune response against HuD. This improvement will also be important in the analysis of the mechanisms of PEM/PSN. Expression of HuD is high in SCLC cells (2, 3, 11), and anti-Hu immunity was a favorable prognostic indicator for patients with SCLC (26). In addition, the immunogenecity of HuD antigen is expected to be identical between humans and BALB/c mice (30). Together with these observations, our murine experiments support the hypothesis that HuD antigen is a possible candidate molecule for antitumor vaccination against SCLC. However, it is also possible that the neurologic paraneoplastic syndromes associated with anti-HuD immunity may abrogate this therapeutic approach.

In summary, DNA vaccination with mouse HuD antigen suppressed the growth of HuD-expressing tumors in a murine SCLC model.