N-acetylglucosaminyltransferase V (Mgat5 or GnT-V) is an enzyme that catalyzes β1–6 branching of N-acetylglucosamine on asparagine (N)-linked oligosaccharides (N-glycan) of cell proteins. The levels of Mgat5 glycan products commonly are increased in malignancies. Although Mgat5 is known to be important in tumor metastases, the effects of Mgat5 on host immune responses are not fully defined. In this study, a Mgat5 specific-short hairpin RNA (shRNA) vector was transfected into murine mammary adenocarcinoma MA782 cells to assess the effects of Mgat5 on tumor cell growth, T cells, and macrophages following inoculation of mice with shRNA-transfected cancer cells. The results showed that blocking expression of Mgat5-modified glycans in MA782 cells significantly suppressed tumor progression both in vivo and in vitro, strongly stimulated Th1 cytokine production, and enhanced opsonophagocytic capability of macrophages in vivo. Importantly, reduction of complex N-glycans on MA782 tumor cells by Mgat5-shRNA resulted in significantly increased proliferation and CD45 surface expression of CD4+ T cells. Our data suggest Mgat5-shRNA could serve as a useful tool to treat breast cancer as well as a powerful tool for the functional investigation of N-glycans and glycoprotein synthesis. Our data suggest that knockdown of Mgat5 inhibits breast cancer cells’ growth with activation of CD4+ T cells and macrophages.
Carbohydrates are attached to proteins in two major linkages called N-glycans and O-glycans, respectively. N-glycans have N-acetylglucosamine (GlcNAc)3 linked to the amide group of asparagine residues in the sequon Asn-X-Ser/Thr in the nascent polypeptide (1, 2, 3). Carbohydrate chains in the mature N-glycans include the typical high-mannose, hybrid, and complex forms. N-acetylglucosaminyltransferase V (Mgat5), also known as GnT-V, plays an important regulatory role in the synthesis of complex N-glycans and this process is highly conserved between yeast and mammals (4).
Mgat5 and its glycan products β1, 6 GlcNAc branched N-glycans are tumor-associated glycoproteins commonly increased in malignancies, and these higher levels correlate with disease progression (5, 6, 7). The secreted form of GnT-V (soluble GnT-V), which is regulated by γ-secretase, could promote angiogenesis (8). The plant lectin leukoagglutinin (l-PHA) binds specifically to mature Mgat5 products (9) and, therefore, has been used as a probe for Mgat5-modified glycans. Many studies have demonstrated the association of increased l-PHA binding and Mgat5 activity with increased tumor cell invasiveness. Cancers of breast, colon, and also melanomas show increased levels of β1, 6 GlcNAc branched N-glycan measured by l-PHA immunohistochemistry (10, 11), as l-PHA has been shown to bind specifically to this structure (11). l-PHA reactivity also is increased in atypical hyperplasia and carcinomas of breast compared with normal and benign lesions. β1,6-branched oligosaccharides are increased in lymph node metastases and are predictors of poor prognosis of breast carcinoma (10).
It has been reported that Mgat5 knockout mice displayed reduced tumor growth and metastasis in vivo (12), suggesting inhibitors of Mgat5 might be useful in the treatment of malignancies. Morgan et al. (2004) (13) reported that Mgat5-mediated N-glycosylation negatively regulates Th1 cytokine production. Anti-CD3-activated splenocytes and naive T cells from Mgat5−/− mice produce more IFN-γ and less IL-4 compared with wild-type cells (14). In addition, deficiency in Mgat5 lowers the T cell activation threshold by directly enhancing TCR clustering (15). Furthermore, depletion of the Mgat5-modified glycans by swainsonine can enhance Ag-dependent T cell proliferation in vitro (16), and similar results were observed in Mgat5 gene knockout mice (13). Recently it was reported that Mgat5 directed integrin stability (17) and regulated expression of cytokine receptors, which balances their surface retention against loss via endocytosis (18). These reports suggest that stimulated IFN-γ production and enhanced T cell activation and differentiation might be achieved by down-regulation of Mgat5.
Despite the fact that Mgat5 is known to be critical in tumor development and the host immune response, the mechanisms involved are not fully defined. Gene therapy strategies such as knockdown of Mgat5 by shRNA in vivo may provide new therapeutic approaches for tumor treatment. In this study, we found that blocking expression of Mgat5-modified glycans in murine mammary adenocarcinoma MA782 cells using short hairpin RNA (shRNA) specific against Mgat5 suppressed tumor progression both in vivo and in vitro and also activated CD4+ T cells and macrophages.
Materials and Methods
Construction of Mgat5 shRNA expression vectors
The pSilencer shRNA expression vectors, pSilencer1.0-U6 and pSilencer2.0-U6 used in this study, were purchased from Ambion. The 21-nt target sequences located between nucleotide 836 and 856 of the Mgat5 gene (GenBank accession number: AF474155) were selected and two pairs of oligonucleotides were synthesized: pair 1: 5′-TCGGGCTTCAAGATTGCGGTTCA-3′ (forward) and 5′-AGCTTGAACCGCAATCTTGAAGCCCGAGGCC-3′ (reverse); and pair 2: 5′-AGCTTCCGCAATCTTGAAGCCCGATTTTTT-3′ (forward) and 5′-AATTAAAAAATCGGGCTTCAAGATTGCGGA-3′ (reverse). Underlined italic letters CG were mutated to GC in the shRNA1m-Mgat5. Pair 1 oligonucleotides were annealed and inserted into pSilencer1.0-U6 (or pSilencer2.1-U6neo) digested with ApaI and HindIII to produce an intermediate plasmid. The annealed product of pair 2 oligonucleotides was subcloned into the HindIII and EcoRI sites of the intermediate plasmid to generate plasmid shRNA1-Mgat5. Plasmid shRNA2-Mgat5, with the 21 target sequence located between nucleotides 1789 and 1809 of the Mgat5 gene, was constructed in a similar manner. Plasmids shRNA1-Mgat5 and shRNA2-Mgat5 were designated as shRNA1 and shRNA2, respectively. The cDNA of GnT-V was cloned into eukaryotic expression vector pcDNA3.1A to yield plasmid pcDNA3.1-GnT-V (19).
Transient and stable transfections
For transient transfection experiments, MA782 cells derived from mouse mammary adenocarcinoma (20) were grown in DMEM (Invitrogen Life Technologies) supplemented with 10% FBS. Plasmids expressing shRNA1, shRNA2, or the pSilencer1.0-U6 were purified using endotoxin-free DNA extraction kit (Qiagen). A total of 4.5 × 105 cells/ml were plated in 60-mm plates at a density appropriate to achieve 60% confluency 24 h before transfection. Cells were transfected with 4 μg of each plasmid using Lipofectin2000 (Invitrogen Life Technologies). Cells were harvested 48 h posttransfection and suspended in PBS.
The transient transfection efficiency of the cells was measured by using red fluorescent protein (RFP) as a reporter molecule. The percentage of transfected cells was determined by using pASRed-C1 plasmid, which encodes RFP, cotransfected with shRNA1-Mgat5 into MA782 cells. Histogram quality and the percentage of the transfected cells were determined (based on the number of fluorescent cells counted) with the flow cytometer. The numbers of RFP-positive cells showed ∼35–50% of plasmid expression (data not shown).
Stably transfected cell lines harboring silencing plasmids were generated by culturing transfected cells in 0.5 mg/ml G418 (Sigma-Aldrich) added to the media 48 h posttransfection. Stably transfected cells were isolated within 2 wk of selection.
Total RNA was extracted from transfected MA782 cells using TRIzol reagent (Invitrogen Life Technologies) and reverse transcribed using the First Strand cDNA synthesis kit (Fergment). Two micrograms of RNA were used for cDNA synthesis using 0.5 μg of oligo(dT)18 primer, 1 μl of RNase inhibitor, 1 μl of Moloney murine leukemia virus reverse transcriptase, and 2 μl of 10X RT buffer (Ambion) in a final volume of 20 μl. The reaction was allowed to proceed for 45 min at 42°C followed by 10 min at 95°C, and then reactions were cooled to 4°C. The cDNA was amplified by PCR using primers: Magt5-F: 5′-AGCAGCTCCATGTTACGG-3′; Magt5-R: 5′-GACCAGATTGTCCACCTTT-3′; G3PDH-F: 5′-ACCACAGTCCATGCCATCAC-3′; and G3PDH-R: 5′-TCCACCACCCTGTTGCTGTA-3′. A total of 1 μl of each RT reaction product was amplified in a 24-μl volume PCR premix, containing 1 pmol of each primer, 2.5 mM dNTP mix, and 1 U of Taq polymerase. The samples were heated to 94°C for 2 min, followed by 28 cycles of 95°C for 10 s, 56°C for 20 s, and 72°C for 30 s. The PCR products were analyzed by electrophoresis in 2.0% agarose gels.
l-PHA binding assay
MA782 cells transfected with shRNA1-Mgat5, shRNA2-Mgat5, or pSilencer1.0-U6 were harvested 60 h posttransfection and then washed twice with PBS (Mg2+Ca2+ free) containing 1% BSA. FITC-conjugated l-PHA (Vector Laboratories) was added to a final concentration of 2.5 μg/ml in PBS-BSA, and cells were incubated for 30 min at 4°C. The cells were collected by centrifugation and then suspended in 1 ml PBS-BSA for FACS analysis.
Tumor cell proliferation in vitro
MA782 cells were transfected with plasmids shRNA1-Mgat5, shRNA2-Mgat5, or pSilencer1.0-U6. Forty-eight hours or more posttransfection, cells were cultured in the presence of 20 μCi [3H]thymidine (Beijing Hi-Tech Atomic Technology) for 20 h. Cells were collected on glass fiber filler mats and dried at 80°C. [3H]Thymidine incorporation was measured by a liquid scintillation counter.
Proliferation of CD4+ and CD8+ T cells
CD4+ or CD8+ T cells were purified from splenocytes isolated from mice using the BD IMag mouse CD4+ or CD8+ T lymphocyte enrichment set-DM and the BD IMagnet (BD Biosciences) by a negative selection procedure. Purified CD4+ or CD8+ T lymphocytes at a density of 2 × 104 cells per well were prestimulated with 1 μg/ml anti-mouse CD3 (eBioscience) and 1 ng/ml IL-2 (Sigma-Aldrich) for 72 h. MA782 cells transfected with plasmids shRNA1-Mgat5, shRNA1m-Mgat5, shRNA2-Mgat5, pcDNA3-GnT-V, pcDNA3-GnT-V, and shRNA1-Mgat5 together, or pSilencer1.0-U6. Seventy-two hours posttransfection, cells were treated with 30 μg/ml mitomycin C for 1 h at 37°C. Cells then were washed twice with PBS and then incubated with stimulated CD4+ or CD8+ T lymphocytes at a density of 2 × 104 cells per well. A total of 1 μCi [3H]thymidine was added to each well and mixtures were incubated for 20 h. A total of 1 μM paclitaxel-treated MA782 cells (cells were mostly dead after treatment) and nontransfected cells were used as controls. Cells were collected on glass fiber filler mats and dried at 80°C. [3H]Thymidine incorporation was measured in a liquid scintillation counter.
Tumor formation in vivo
Female BALB/c mice ∼6–8-wk old were purchased from the Animal Center of Wuhan University. Mice were divided into four groups of nine mice per group. MA782 cells were transfected with plasmids shRNA1, shRNA2, or pSilencer1.0-U6 and, at 72 h posttransfection, were diluted to a density of 4 × 106 cells/ml in PBS. A total of 100 μl of cells were injected into one side of the back of each mouse. Tumor size was measured with calipers daily. Eight days after tumor cell implantation, mice were sacrificed by cervical dislocation and tumors were excised. Tumor volume was calculated by the formula: volume (mm3) = (width)2 × length/2.
Index of phagocytosis
Murine peritoneal macrophages were obtained from the nine mice in each group, mixed, and then washed and suspended in PBS to a final concentration of 2.5 × 107 cells/ml. A total of 100 μl of macrophage suspension was mixed with 0.1 ml Bacillus coli suspension and incubated for 30 min, with slow shaking, at 37°C. The extracellular bacteria were removed by washing twice with ice-cold PBS. A total of 100 μl of the final suspension was dropped onto each slide and cells were stained with Giemsa-Wright stain. Bacteria were detected under oil immersion. The index of phagocytosis (IP) was calculated as follows: IP = (percentage of macrophages containing at least one bacterium) × (mean number of bacteria per positive cell) (21).
Splenocytes were isolated from each mouse injected with genetically modified tumor cells and diluted to 5.0 × 106 cells/ml. These cells were plated in 24-well plates in 1 ml per well and stimulated with 0.5 μg mouse anti-CD3 plus 1 μg mouse anti-CD28 for 72 h. Cytokines IFN-γ, IL-4, or TNF-α in supernatants were measured by ELISA assay kit (eBioscience). All data were analyzed for statistical significance.
For assessment of surface CD45, CD25, CD69, or Foxp3 expression, splenocytes were placed in 24-well plates at a density of 5.0 × 106 cells/ml per well. Cells were prestimulated with 0.5 μg/ml anti-mouse CD3 and 1 μg/ml anti-mouse CD28 for 72 h, washed, and suspended in PBS. A total of 5 μl of anti-CD45-FITC, anti-CD25-FITC, anti-CD69-FITC, anti-CD3-PE-Cy5, or anti-CD4-PE was incubated with the cells for 30 min at 4°C. In other experiments, 5 μl of anti-Foxp3-PE-Cy5, anti-CD25-FITC, and anti-CD4-PE was incubated with the cells for 30 min at 4°C. The stained cells were then analyzed in a Beckman Coulter EPICS ALTRA II flow cytometer. FITC-, PE-, or PE-Cy5-conjugated rat IgG1was used as isotype matched control.
Experimental data were analyzed by ANOVA or an unpaired Student’s t test. The values of p < 0.05 were considered statistically significant.
ShRNA1-Mgat5 specifically blocks expression of Mgat5 and Mgat5-modified glycan products
Inhibition of Mgat5 and Mgat5-modified N-glycan product expression by shRNA1 and shRNA2 in MA782 cells was examined by RT-PCR and l-PHA binding assay. The results from RT-PCR assays (Fig. 1,A) and normalized Mgat5 mRNA levels based on the expressions of housekeeping gene G3PDH (Fig. 1,B) showed that shRNA1-Mgat5 significantly reduced Mgat5 expression in MA782 cells (Fig. 1,A, lane 4), whereas Mgat5 expression was not reduced in cells transfected with shRNA2 (Fig. 1,A, lane 3), control vector (Fig. 1,A, lane 2), or in untransfected control cells (Fig. 1,A, lane 1). Flow cytometry analysis (Fig. 1 C) showed that l-PHA binding to Mgat5-modified glycans on MA782 cells was significantly decreased in shRNA1-transfected cells (64%) compared with shRNA2 (96.7%) or pSilencer1.0-U6-transfected (95.7%) or nontransfected control cells (97%). These results suggest that shRNA1 efficiently blocked expression of Mgat5 and Mgat5-modified glycans products in MA782 cells.
ShRNA1-Mgat5 inhibits MA782 cell growth in vitro
MA782 cells transfected with shRNA1-Mgat5, shRNA2-Mgat5, or the pSilencer1.0-U6 control vector were tested for [3H]thymidine incorporation in vitro to determine the effects of reduced Mgat5 expression on cell proliferation. Fig. 2 A shows shRNA1 significantly inhibited MA782 cell growth in vitro compared with any of the other groups at day 3 of posttransfection (*, p < 0.05).
The growth of MA782 cells stably transfected with shRNA1-Mgat5 also was examined and cell death occurred within 2 wk of G418 selection. In contrast, cells stably transfected with shRNA2-Mgat5 or pSilencer2.1-U6neo vector remained viable (data not shown). Similar results were obtained in four separate experiments. These data indicate that shRNA1-Mgat5 inhibits MA782 cell growth in vitro and MA782 cells stably transfected with shRNA1-Mgat5 do not survive more than 2 wk. Transient transfections thus were used in the remainder of the experiments in the present study.
Time course experiment of [3H]thymidine incorporation in vitro showed that shRNA1 transiently transfected MA782 cancer cells, which had deduction of β1, 6 GlcNAc branched N-glycan proliferated at ∼58, 62, 67, and 80% of the growth rate of the nontransfected control groups in vitro at days 3, 6, 8, and 11 posttransfection, respectively (Fig. 2,B). After ∼14 days posttransfection, the growth rate of transiently transfected cells was similar to that of the nontransfected cells (Fig. 2 B).
ShRNA1-Mgat5 significantly inhibited MA782 cells growth in vivo
To determine whether shRNA1 inhibits growth of MA782 cells in vivo, cells transiently transfected with shRNA1, shRNA2, or pSilencer1.0-U6 vector for 72 h were thus implanted into the flank of syngeneic BALB/C mice. Murine mammary tumors were observed at day 2 postinoculation and were measured each day thereafter. As shown in Fig. 3,A, cells treated with shRNA1 demonstrated significantly (**, p < 0.01) lower oncogenicity compared with any of the other groups over the course of the experiment. Eight days after cell implantation, mice were euthanized and tumors were harvested and measured. Tumor sizes were much smaller in shRNA1-treated group than those in shRNA2, pSilencer1.0-U6, or control groups (Fig. 3,B). shRNA1 transiently transfected MA782 cancer cells’ growth was at ∼38% of the growth rate of the nontransfected control groups in vivo at day 8 postimplantation (implantation was 3 days later after transfection) (Fig. 3 B).
Comparing the data from both in vitro (Fig. 2, A and B) and in vivo (Fig. 3,A), the results showed that shRNA1-transfected cancer cells, which had deduction of β1, 6 GlcNAc branched N-glycan, proliferated at ∼62% of the growth rate of control groups in vitro (Fig. 2,B) at day 6 posttransfection and to ∼45% of the growth rate of control group at day 6 posttransfection (equal to day 3 postimplantation) in vivo (Fig. 3,A). With time, shRNA1-transfected cancer cells proliferated at ∼80% of the growth rate of control groups in vitro (Fig. 2,B) at day 11 posttransfection and to ∼38% of the growth rate of control group at day 11 posttransfection in vivo (corresponding to day 8 postimplantation) (Fig. 3,A). Comparing the data of time course experiments both in vitro (Fig. 2,B) and in vivo (Fig. 3 A), we conclude that the tumor growth inhibition in vivo after shRNA1 transfection might be due to shRNA1 direct effect at early stage and T cell and macrophage effects at later stage, especially after ∼3 days of postimplantation.
Eight days after tumor cell implantation, tumors were excised and further digested to single-cell suspension by collagenase I and incubated with FITC-conjugated l-PHA for FACS analysis. Similar to in vitro results (Fig. 1,C), Fig. 3,C showed that shRNA1-Mgat5-treated tumor cells growing in vivo also had two populations of cells, shRNA1-Mgat5-positive cells (PHA-negative cells) were ∼20.9% and shRNA1-Mgat5-negative cells (PHA-positive cells) were ∼79.1%. shRNA1-Mgat5-negative cells (PHA-positive cells) were ∼97.1, 97, and 97% for shRNA2, pSilencer1.0-U6, and nontransfected cells, respectively (Fig. 3 C).
ShRNA1-Mgat5 transfection significantly enhanced production of IFN-γ and TNF-α in vivo
We further determined the levels of IFN-γ, TNF-α, and IL-4 synthesized by splenocytes isolated from the animals injected with the various transfected tumor cells. Splenocytes were stimulated with anti-CD3 and anti-CD28 for 72 h. Supernatants were collected, and the secreted cytokines levels were measured by ELISA. The results showed that splenocytes from animals injected with shRNA1-Mgat5-transfected cells produced significantly higher levels of TNF-α (**, p < 0.01) (Fig. 4,A) and IFN-γ (*, p < 0.05) (Fig. 4,B) than any other group. ShRNA1-treated mice produced ∼7-fold more TNF-α and ∼2-fold more IFN-γ than shRNA2 and empty vector-transfected or nontransfected controls (Figs. 4, A and B). However, the production of IL-4 was not significantly altered in any of the groups (Fig. 4 C). These results suggest shRNA1-Mgat5 could stimulate the expressions of Th1 cytokines such as IFN-γ and TNF-α.
Reduction of Mgat5 in tumor cells enhanced phagocytic capability of murine peritoneal macrophages
To evaluate whether peritoneal macrophages contribute to the anti-tumor effect of Mgat5 knockdown in vivo, macrophages were isolated from tumor-bearing mice and incubated with bacteria as described in Materials and Methods. The opsonophagocytic capability IP was highly increased in shRNA1-treated mice (n = 9) compared with all other groups (Fig. 4 D).
ShRNA1-Mgat5-modified glycan Ags on MA782 cells stimulated proliferation of CD4+ T cells
To determine whether a deficiency in Mgat5 and its N-glycan products on tumor cells influence the function of CD4+ or CD8+ T cells, CD4+ or CD8+ cells were isolated and purified from splenocytes harvested from normal BALB/C mice by a negative selection procedure. Purified CD4+ or CD8+T cells were prestimulated with 1 μg/ml anti-mouse CD3 and IL-2. MA782 cells transfected with shRNA1-Mgat5, shRNA1m-Mgat5, shRNA2-Mgat5, pcDNA3.1-GnT-V, pcDNA3.1-GnT-V plus shRNA1-Mgat5, or control vector were inactivated with mitomycin C and then mixed with stimulated CD4+ or CD8+ T cells for assessment of [3H]thymidine incorporation.
The results in Fig. 5,A show that proliferation of CD4+ T cells was significantly increased by shRNA1-transfected MA782 cells compared with all other groups (*, p < 0.05). The levels of [3H]thymidine incorporation in CD4+ cells in the shRNA1-Mgat5-treated group were nearly 3-fold higher than those of shRNA1m-Mgat5, pSilencer1.0-U6-transfected groups, or nontransfected controls, and nearly 2-fold higher than those of shRNA2-Mgat5, GnT-V, or GnT-V plus shRNA1-Mgat5 groups (Fig. 5,A). Compared with shRNA1-treated group, transfection of pcDNA3.1-GnT-V or cotransfection of pcDNA3.1-GnT-V with shRNA1 significantly decreased proliferation of CD4+ T cells (*, p < 0.05) (Fig. 5,A), suggesting GnT-V (or Mgat5) could compensate for reduced Mgat5 expression by shRNA1-Mgat5. However, there were no statistically significant differences in the proliferation of CD8+ T cells among all the groups (Fig. 5 B). These data suggest that Mgat5-modified glycan Ags on MA782 cells by shRNA1 treatment could preferentially stimulate proliferation of CD4+ T cells.
To determine whether CD4+ T cells are responsive to alterative glycans or respond to dead cells, the following experiments were performed. MA782 cells transfected with shRNA1-Mgat5 (for transient transfection for 3 days), and 1 μM paclitaxel-treated MA782 cells or nontransfected cells were incubated with purified CD4+ T lymphocytes. As shown in Fig. 5 C, CD4+ T cells were not responsive to paclitaxel-treated dead cells or nontransfected cells but significantly responded to shRNA1 transiently transfected cells.
We have already observed that MA782 cells stably transfected with shRNA1-Mgat5 death occurred within 2 wk of G418 selection. Accordingly, the shRNA1 stably transfected cells cultured in G418 for 8 or 11 days were injected to mice at the density of 4 × 106 per mouse. Then tumor formation could not been observed (data not shown), which means these stably transfected cells were dead. We have also detected that cells stably transfected with shRNA1 (cultured in G418 for 8 days or 11 days) could only weakly activate CD4+ T cells (data not shown) because most of these cells were dead. All of these results suggest that shRNA1-treated tumor cells can stimulate CD4+ T cells in vivo, owing to alterative glycans of cancer cells but not dead cells.
ShRNA1-Mgat5 enhanced CD45 surface expression
Splenocytes isolated from animals injected with transfected or control tumor cells were stimulated with anti-CD3 and anti-CD28 for 72 h, followed by staining with TRI-color-anti-CD3, PE-anti-CD4, and FITC-anti-CD45. FACS analysis showed the CD45+ mean fluorescence intensity (MFI) on CD3+CD4+ T cells from shRNA1-Mgat5-treated group was significantly higher (MFI: 8 ± 1.0) than those from shRNA2-Mgat5 (MFI: 2.3 ± 0.3), empty vector control (MFI: 2.35 ± 0.1), and nontransfected control groups (MFI: 2.3 ± 0.2) (*, p < 0.05) (Fig. 6,A). The percentage of CD3+CD4+CD45+ T cells was 10.1% in the shRNA1-treated group, 3.6% in the shRNA2-treated group, 1.8% in the pSilencer1.0-U6 vector group, and 1.6% in nontransfected control group (Fig. 6,C). There were no statistically significant differences between surface expression of T cell activation markers CD25 and CD69 on CD3+CD4+ T cells (Fig. 6,C). Development and function of regulatory CD4+ T cell (Tr) are controlled by Foxp3, and no statistically significant differences in Foxp3 expression on CD4+CD25+ T cells were observed among all groups (Fig. 6 B).
In the last several years, many genes that control glycan or carbohydrate production and diversification on the cell surface have been identified and cloned. Data from several recent reports demonstrate that specific alteration in N-glycan structure could disable some processes in the immune system and immune-related diseases. However, the roles of many glycans or genes involved in their synthesis in the immune system function and immune-related diseases remain largely elusive (1, 2, 3).
We observed that shRNA1-Mgat5 may by itself exhibit direct cytotoxicity as well as specific inhibitory effects on MA782 cell growth and proliferation (Fig. 2). The tumor growth inhibition in vivo after shRNA1 transfection, which had deduction of β1, 6 GlcNAc branched N-glycan, might be due to shRNA1 direct effect on tumor cells at early stage and T cell and macrophage activities at later stage (Fig. 3 A). The tumor inhibition over time may be related to immune responses indicated by the presence of activated CD4+ T cells and macrophages.
Importantly, we found that shRNA1-Mgat5-transfected cancer cells significantly stimulated CD4+ T cell proliferation (Fig. 5,A), and transfection of pcDNA3.1-GnT-V or cotransfection of pcDNA3.1-GnT-V with shRNA1 significantly decreased proliferation of CD4+ T cells (*, p < 0.05) (Fig. 5 A). These data suggest GnT-V can compensate for the effects of reduced Mgat5 expression by shRNA1-Mgat5, and that knockdown of either GnT-V or Mgat5 had a stimulatory effect on CD4+ T cells.
The mechanism by which shRNA1-transfected tumor cells regulate TNF-α and IFN-γ production and CD4+ T cell proliferation might be due to deduction of β1, 6 GlcNAc branched N-glycan on tumor cells and/or alteration or exposure of some Ag epitopes, which in turn activate APC, such as DC cells and Langerhans cells in s.c. tissue, and eventually up-regulate CD4+ T cell proliferation and TNF-α and IFN-γ production.
CD45 is a protein tyrosine phosphatase, a leukocyte common Ag, a type I transmembrane glycoprotein, and plays multiple important roles in immune functions and leukocyte receptor signal transduction (22). Our results suggest that Mgat5 reduction resulting in increased surface expression of CD45 might activate signal transduction in CD4+ T cells or in monocytes/macrophages and stimulate proliferation of CD4+ T cells (Fig. 5) and phagocytic capability of macrophages (Fig. 4 D).
We also observed that activated splenocytes from animals injected with shRNA1-Mgat5-transfected MA782 tumor cells produced higher levels of Th1 cytokines such as IFN-γ and TNF-α (Fig. 4). IFN-γ and TNF-α both are critical cytokines that protect against tumor development and likely have some inhibitory effects on tumor cells growth in vivo (23). Morgan et al. (2004) (13) reported that Mgat5-mediated N-glycosylation negatively regulates Th1 cytokine production by T cells. Similar results were observed in this study. In addition, macrophages isolated from the mice injected with shRNA1-transfected MA782 cells displayed stronger phagocytic capability (Fig. 4 D) that might exhibit direct cytotoxicity toward tumor cells (24). Taken together, our results represent the first study to show that shRNA1-Mgat5 could serve as a useful tool to treat breast cancer as well as for the functional investigation of N-glycan and glycoprotein synthesis both in vitro and in vivo. Our results suggested that reduced expression of Mgat5 not only inhibited breast cancer cell growth, but also had stimulatory effects on CD4+ T cells and macrophages, and enhanced Th1 cytokine production in vivo.
The amino acid sequences of Mgat5 are highly conserved among diverse species. The murine Mgat5 mRNA sequence is similar to other mammalian sequences with 95, 91, and 87.8% overall sequence identity to rat, hamster, and human mRNAs, respectively (25). It is therefore possible that shRNA1-Mgat5 directed toward mouse mammary tumor growth in vivo could be applied for the treatment of breast cancer in other species, including humans.
Current therapeutic regimens for breast cancer include surgery, radiotherapy, chemotherapy, and endocrine treatment as adjunctive therapy (26). Therapeutic efficacy strongly correlates with initiating treatment early in the disease course (27). The newly developed Mgat5-specific shRNA1 described in this study may directly serve as a potent early immunotherapeutic agent in vivo for the treatment of established solid tumors. In fact, cancers of breast, colon, and melanomas have much higher expression levels of Mgat5 product than normal cells, and Mgat5 is positively associated with cancer progression and metastasis (10, 11). Consequently, the treatments using shRNA1, which is Mgat5 specific shRNA, would have less effect on normal cells. Furthermore, the shRNA1-Mgat5 strategy could also be used in combination with tumor-specific Ab or designed with a tissue-specific promoter to specifically augment the efficacy of current breast tumor regimens. Our data indicate that shRNA1 has the potential to be developed as a novel therapeutic drug for tumor patients.
The plasmid pcDNA3.1-GnT-V was provided by Prof. Shen ZH.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by grants to X.-L.Z. from the National Natural Science Foundation of China (30370310, 20532020, and 30670098), 973 Program 2006CB504300, the Ministry of Education Scientific Research Foundation for Outstanding New Century Scholars and Outstanding Young Teachers (NCET-04- 0685), Hubei Province Science Technology Department (2006ABD007 and 2005AA304B04), Hubei Ministry of Public Health (JX1B074), and the Open Research Fund Program of the State Key Laboratory of Virology of China (III-2005007).
Abbreviations used in this paper: GlcNAc, N-acetylglucosamine; shRNA, short hairpin RNA; Mgat5 or GnT-V, N-acetylglucosaminyltransferase V; l-PHA, leukoagglutinin; RFP, red fluorescent protein; IP, index of phagocytosis; MFI, mean fluorescence intensity.