Intratumoral Expression of Mature Human Neutrophil Peptide-1 Mediates Antitumor Immunity in Mice

Human neutrophil peptides (HNP1-4), or -defensins, are known as potent antimicrobial peptides of 30 amino acids in length. HNP1-3 have activity against a wide variety of bacteria, fungi, and some enveloped viruses (1, 2). HNP1-3 show high similarity and this is shown among mature HNP1 and HNP3, which differ only in their NH₂-terminal amino acids due to a single nucleotide difference. HNP2 is identical within the last 29 amino acids of both HNP1 and HNP3. HNP2 is presumably produced from proHNP1 and/or proHNP3 by post-translational proteolytic cleavage rather than being encoded by a unique gene (3). HNP4 is significantly different from HNP1-3 in its amino acid sequence and shows more effective protective activity from infection by both X4 and R5 HIV-1 strains (4). HNP1-3 exhibit cytotoxicity in vitro to a wide range of normal and malignant mammalian target cells (5,9). A series of clinical investigations have revealed the presence of HNP1-3 in the tissues of epithelial tumors (10,13) and that intratumoral HNP1-3 in renal cell carcinoma tissues is related to tumor necrosis (13). Furthermore, upregulation of HNP1-3 in tumor tissue and plasma has been thought to be a potential marker for prognostic assessment (14, 15). We recently also reported that intratumoral expression of HNP1 within human A549 lung carcinoma xenografts resulted in increased cell apoptosis and inhibition of angiogenesis in nude mice (16).

In fact, HNP1-3 provide an important link between innate responses and adaptive immunity by chemoattracting immature dendritic cells, monocytes, and T lymphocytes (2, 3, 17, 18). It has been reported that HNP1-3 act as potent adjuvants that promote cellular and humoral immune responses in mice to a lymphoma idiotype and carrier antigens (19). This was supported by another study that found HNP1-3 enhanced both proliferative responses and T-helper cytokine secretion profiles of naive CD4 T cells resulting in enhanced systemic antigen-specific IgG production in mice (20). Furthermore, a recent research has shown that HNP2 induces the recruitment of dendritic cells in cervical human papillomavirusassociated (pre)neoplastic lesions and likely to be involved in host immune recognition and presentation of tumor antigen by dendritic cells (11). Activation of dendritic cells within tumors can enhance tumor antigen presentation and consequently elicit specific antitumor immunity (21^,23). A similar antimicrobial peptide, murine -defensin 2, has been shown to mediate specific antitumor immunity through dendritic cells mediating antigen presentation (24 26). Recently, we also found that murine -defensin 2 can mediate specific antitumor immunity in situ through recruiting and activating immature dendritic cells.³

HNP1 and HNP2 have been shown to be chemotactic for immature dendritic cells and murine T cells, indicating the immune activity of HNP1-3 in mice (27). We have also found that secreted HNP1 mature peptide within tumors can recruit lymphocyte-like cells in situ (16) and that portions of these infiltrated cells are CD11c⁺, indicating their chemotactic activity to monocytes or dendritic cells (Supplementary Fig. S1). In a recent investigation, we found that HNP1-3 are also expressed in gastric carcinoma and breast cancer accompanied with an increased number of lymphocytes. Furthermore, a high level of HNP1-3 expression was detected in gastrointestinal mucosa-associated lymphoid tissue lymphoma with a good prognosis (Supplementary Fig. S1). Therefore, we hypothesized that, in addition to the direct cytotoxicity of HNP1 against tumor cells, the presence of these multifunctional peptides within tumor tissues may also be involved in host immune responses against tumors.

In this study, we attempted to investigate the possible roles of HNP1-3 within the tumor milieu. Given the toxicity of HNP1-3 to adenovirus, high similarity in their amino acid sequence, and upregulated levels in tumors, we used a recombinant plasmid expressing a secretable form of mature HNP1 to explore whether intratumoral expression of HNP1 is involved in host immune responses to tumors. Here, we found that the intratumoral expression of HNP1 resulted in recruitment and activation of dendritic cells within tumors, which subsequently triggered host immune responses to tumors. Our data suggest that human -defensins may mediate antitumor immunity in situ.

Female BALB/c mice, 6 to 8 weeks old, were purchased from the West China Experimental Animal Center. Both CT26 mouse colon carcinoma and 4T1 mouse breast carcinoma cells were obtained from the American Type Culture Collection and cultured in RPMI 1640 supplemented with 10 heat-inactivated fetal bovine serum and antibiotics.

To test the expression of HNP1 and the activity of the products from transfected cells, CT26 cells were transfected with a pSec-HNP1 or pSecTag vector. Supernatants were harvested 48 h after transfection and quantified using a commercial ELISA kit (Hbt, HK317) as described previously (16). The chemotactic activity of HNP1 in cell supernatants to chemoattract immature dendritic cells was determined as described previously (25). Maturation-inducing activity was evaluated as described previously (28).

To show the functional expression of HNP1 within tumors, CT26 colon cancer and 4T1 breast carcinoma models were established. In brief, 3 10⁵ CT26 or 4T1 cells were subcutaneously injected into the right dorsal flank of 6- to 8-week-old female BALB/c mice. A recombined vector (pSec-HNP1) was prepared as described previously (16). Cationic liposomes were used to enhance the efficiency of transfection. When the tumor volume reached 50 mm³ (diameter 4-5 mm), cationic liposome-DNA complexes [DNA (100 g)/liposome (300 g) = 1:3] were injected intratumorally and around the tumor once every 3 days in a volume of 100 L for a total of five times and the control injection in a volume of 100 L normal saline.

To detect early HNP1 expression and dendritic cell infiltration, tumors were harvested after two injections (3 days after the first injection). For the detection of CTLs, tumors were harvested 1 week after the last injection. Subsequently, tumor tissues were analyzed by H&E staining and immunostaining. Tumor growth was evaluated by measurement of tumor diameters every 3 days and tumor volume was calculated as length width² 0.52. For ethical reasons and experimental consideration, mice were sacrificed when the tumor volume reached 4,000 mm³ in the CT26 model and 2,000 mm³ in the 4T1 model.

Expression of HNP1 was determined by immunohistochemical staining with an anti-HNP1 monoclonal antibody (1:1,000, MCA1465; Serotec). To determine whether the antitumor effects of HNP1 are involved in the inhibition of angiogenesis, microvessel density in tumor tissues was detected with an anti-CD31 antibody (Abbiotec). Microvessel density was determined by counting the number of microvessels per high-power field.

Immunofluorescence staining was used for the detection and analysis of immune cells infiltrated within tumors. Anti-CD11c (PE-labeled; BD Bioscience), CD40 (PE-labeled; Biolegend), and CD86 (FITC-labeled; BD Bioscience) monoclonal antibodies were used to detect dendritic cells, whereas anti-CD8 (Cy5PE conjugate) and CD3 (FITC conjugate; eBioscience) monoclonal antibodies were used to determine CTLs. Tumors were snap-frozen and 8-m sections were prepared in Tissue Tek (Sakura Finetek) for immunofluorescence analysis. Fluorescence was visualized and images were captured with an Olympus BX60.

The cytokine production in tumors was determined after receiving two administrations 3 days after the first injection as described previously. Nonnecrotic tumors were harvested from the mice in the various treatment groups described previously for cytokine evaluation (29). Tumor necrosis factor- (TNF-), interleukin (IL)-12(p70), IL-10, transforming growth factor- (TGF-), and IFN- levels were measured in the supernatant of tumor homogenates by a cytokine-specific ELISA kit.

An adoptive transfer and a ⁵¹Cr release assay, as we described previously (25), were used to determine specific cytotoxicity mediated by CTLs. Freshly isolated spleen cells (2 10⁷) from different mice treated with pSec-HNP1, pSecTag, or normal saline were injected into recipient BALB/c mice via the tail vein on the second day after CT26 challenge. For the ⁵¹Cr release assay, T cells were isolated and restimulated with mitomycin Ctreated CT26 cells. CTL activity was calculated using the formula: lysis = [(experimental release - spontaneous release) / (maximum release - spontaneous release)] 100.

Antibodies in the sera were identified with lysates of CT26 or 4T1, respectively. The lysates were separated by SDS-PAGE. Gels were electroblotted onto polyvinylidene difluoride membranes. Subsequently, the membranes were probed with mouse sera at 1:500.

Immune cell subsets were depleted from BALB/c mice as described (25). Briefly, BALB/c mice were injected intraperitoneally with anti-CD4 (clone GK1.5; rat IgG), anti-CD8 (clone 2.43; rat IgG), or anti-NK (clone PK136) monoclonal antibody with 500 g/kg/mouse 1 day before treatment and then twice weekly for 2 weeks. Tumor size was measured 25 days after inoculation.

CT26 cells including both attached cells and floating cells were harvested 48 h after transfection of pSec-HNP1. Flow cytometric analysis was done to identify sub-G₁ cells/apoptosis cells.

Tumor tissues were removed from tumor-bearing mice 48 h after the last treatment. The presence of apoptotic cells within tumor tissues was evaluated by terminal deoxynucleotidyl transferasemediated dUTP nick end labeling technique using the DeadEnd Fluorometric TUNEL System (Promega) following the manufacturer's protocol.

SPSS 11.5 was used for statistical analysis. The statistical significance of results in all of the experiments was determined by Student's t test and ANOVA. Survival curves were compared by the log-rank test. The findings were regarded as significant if P < 0.05.

We reported previously that the intratumoral expression of HNP1 accompanied increased infiltration of lymphocyte-like cells in a nude mouse model (16). Because nude mice lack T cells owing to thymus defects, we first investigated in a recent study whether these infiltrated lymphocyte-like cells were dendritic cells or monocytes through immunostaining with an anti-CD11c⁺ monoclonal antibody. The positive CD11c⁺ staining suggests that mature HNP1 show chemotactic activity in mice (Supplementary Fig. S1).

To determine the expression and activity of HNP1 in immunocompetent mice, tumor models were established with CT26 or 4T1 cells. Tumor-bearing mice were divided into three groups that received an intratumoral injection twice with pSec-HNP1, pSecTag, or normal saline, respectively. The mice were then sacrificed and the tumors were harvested. Immunohistochemical staining was done to detect intratumoral expression of HNP1. The results indicate that mature HNP1 can be effectively expressed and secreted from tumor cells in BALB/c mice. Although some HNP1-positive cells showed abnormal cellular morphologies, suggesting apoptotic events, the secreted HNP1 was found in interstitial spaces and binding blood vessel walls (Fig. 1A).

To determine whether dendritic cells were recruited into the tumor sites, tumor tissues were analyzed with the antibodies to the dendritic cell marker CD11c⁺. Significantly increased CD11c⁺ cells coexpressing the costimulatory molecules, CD86 and CD40, were found in tumors from pSec-HNP1treated mice, indicating the mature status of infiltrated dendritic cells. Cells copositive for CD11c⁺, CD86, and CD40 were rare within tumors from control groups (Fig. 1A and B). Moreover, pSec-HNP1 treatment resulted in 3-fold increase of dendritic cells in tumor-draining lymph nodes in mice compared with that of other groups (Fig. 1C). Although it is difficult to determine whether the maturation of dendritic cells occurred before or after being recruited into tumors, a previous investigation showed that HNP1-3 exhibit chemotactic activity to immature dendritic cells rather than mature dendritic cells, supporting the infiltration and subsequent maturation of immature dendritic cells (18). It is obvious that, regardless of when or where immature dendritic cells were activated, these mature dendritic cells recruited into tumors contributed to enhancing antigen presentation.

To further understand the roles of HNP1 in recruitment and activation of immature dendritic cells, we tested the activity of HNP1 secreted from tumor cells in vitro. We further detected early intratumoral expression of cytokines, including TNF-, IL-12, IL-10, TGF-, and IFN-, after treatment. Supernatants containing HNP1 (90 ng/mL) not only exhibited similar chemotaxis as commercial mature HNP1 (PeproTech) at the same concentration (Fig. 2A) but also promoted maturation of immature dendritic cells in vitro (Fig. 2B). HNP1 alone did not exhibit this capability, and pretreatment with anti-HNP1 antibody failed to impair maturation of immature dendritic cells. These findings indicate that HNP1 has direct chemotaxis to immature dendritic cells but may exert an indirect effect on the maturation of immature dendritic cells. We observed increased apoptosis of pSec-HNP1transfected CT26 cells (Supplementary Fig. S2) in vitro but failed to detect changes of TNF-, IL-4, and granulocyte/macrophage colony-stimulating factor (data not shown). Given that apoptotic cells can maturate immature dendritic cells (30), it was speculated that the indirect maturation-inducing activity of HNP1 is dependent on the resulting apoptotic cells. Nevertheless, TNF- and IL-12 were significantly upregulated within tumors from the pSec-HNP1treated group, whereas IL-10 and TGF- only showed minute downregulation, and IFN- did not show significant difference at the early stage (Fig. 2C). This suggested that HNP1, in vivo, plays more complicated roles in host immune responses.

The in vivo effects were explored in CT26 and 4T1 tumor models (Fig. 3A and B). Although liposome-DNA complex also exhibits cytotoxicity to tumor cells as described previously (31, 32), significant inhibition of tumor growth was observed in pSec-HNP1treated mice compared with other groups. Six of 10 of the pSec-HNP1treated tumors in the CT26 model and 4 of 10 in the 4T1 model were eradicated after the treatment. There were two pSec-HNP1treated mice that exhibited stable disease in the two models. In addition, increased survival benefits were also gained in the pSec-HNP1 groups. These results indicate that intratumoral expression of mature HNP1 may effectively inhibit or even eradicate established tumors.

4T1 breast cancer cells have a high metastasis potential and can metastasize to the liver, lymph nodes, bone, and brain, and with high frequency to the lung, in BALB/c mice as early as 2 weeks after inoculation (33, 34). Previous studies have also suggested that 4T1 tumor lethality is due to early metastases, among which lung metastases could better represent the extent of 4T1 cells spreading throughout the mouse (33). Based on these considerations, when the tumor volume reached 2,000 mm³ in the current study, mice were statistically regarded as imminent deaths and sacrificed. The lung weight was measured and the metastatic lung nodules were counted under a dissecting microscope (Fig. 4). Finally, 4 mice from the pSec-HNP1treated group were sacrificed when the tumor volume reached 2,000 mm³ regardless of the delay in tumor growth. However, significantly fewer lung metastatic nodules were still observed, which further supported that pSec-HNP1 treatment significantly inhibited lung metastases.

Because HNP1 can attract immature dendritic cells to the tumor site and mediates their maturation, the enlisted dendritic cells may mediate specific antitumor immunity. In fact, the inhibition of lung metastases in the 4T1 model already suggested the existence of antitumor immunity. To confirm this, we first performed H&E and immunofluorescence staining of tumor tissue to detect whether there were CTLs existing within tumors with anti-CD3 and CD8 antibodies. Increased local necrosis was found in both CT26 and 4T1 tumors from pSec-HNP1treated mice (data not shown). Abundant lymphocytes were observed at the margins and local interspaces of tumor tissues from some pSec-HNP1treated mice (Fig. 5). Anti-CD8 and CD3 monoclonal antibodies staining showed numerous CD8⁺ T cell infiltrations. The results revealed that HNP1 possibly triggered specific CTL response.

To further document specific CTL responses, spleen cells transfer and a ⁵¹Cr release assay were done. Both tumor formation and growth were significantly restrained in the group that received T cells from pSec-HNP1treated mice (Fig. 6A). T cells from pSec-HNP1treated mice exhibited increased cytotoxicity to tumor cells compared with other groups in the CT26 model (Fig. 6B). These findings indicate elicitation of specific antitumor cellular immunity.

In addition, to explore whether there exists humoral immunity against tumors, we used lysates of CT26 or 4T1 as antigens to detect possible antibodies in the sera from treated mice by Western blot analysis. A conspicuous increase in antibodies was shown in the sera of both CT26 and 4T1 models treated with pSec-HNP1. Interestingly, mice with stable tumors from the pSec-HNP1treated group showed more intensive antibody responses than the mice with tumor repression (Fig. 6C and D). However, the increased antibodies may perform a subordinate role in HNP1-mediated antitumor immunity because tumor size did not response to the intensity of the antibodies.

We reported previously that HNP1 inhibits human lung carcinoma A549 cell growth in nude mice via the induction of apoptosis (16). HNP1-3 can regulate angiogenesis by affecting endothelial cell adhesion and migration in a fibronectin-dependent manner as well as affecting endothelial cell proliferation (35, 36). Synthetic HNP1 inhibits pathologic retinal neovascularization by binding to new blood vessel walls with ₅-integrin in mice (35). Therefore, we investigated the proapoptotic and antiangiogenesis effects of HNP1 on CT26 and 4T1 cells. Increased apoptosis and decreased activity of proliferation were observed in vitro with a flow cytometry assay (Supplementary Fig. S2) and MTT (data not shown) in pSec-HNP1 and pSecTagtransfected CT26 cells when compared with untreated CT26 cells. The terminal deoxynucleotidyl transferasemediated dUTP nick end labeling assay also indicated that pSec-HNP1 treatment resulted in significantly increased apoptosis in vivo (Supplementary Fig. S2). In addition, we also showed the binding of HNP1 on tumor blood vessel walls, indicating similar angiogenesis inhibition as described previously (35). Decreased microvessel density accompanying local necrosis was observed in tumor tissues from mice treated with pSec-HNP1. In addition, compensated hyperplasia of new microvessels was not found within the region without necrosis in pSec-HNP1treated tumors (Supplementary Fig. S3). It has been confirmed that dead tumor cells can induce maturation of immature dendritic cells (30, 37); therefore, HNP1-mediated apoptosis and inhibition of angiogenesis in vivo may play another important role besides direct antitumor effects.

To further confirm that HNP1-mediated immune responses also contributed to, and may be required for its antitumor effects, we performed the same treatment in CT26-bearing nude mice (n = 5) and found that pSec-HNP1 treatment did not result in complete regression of established tumors in spite of significant tumor inhibition (Fig. 7A and B). Furthermore, the therapeutic effects of HNP1 were significantly attenuated in the CT26 tumor model when depleted of CD8⁺ or CD4⁺ T cells. In contrast, treatment with monoclonal antibody against NK cells had no discernable effect on the antitumor activity (Fig. 7C). Therefore, the antitumor responses induced by HNP1 were dependent on both CD8⁺ cytotoxic T cells and CD4⁺ helper T cells. Furthermore, we investigated the long-term effects using these mice cured by pSec-HNP1 >4 months after tumors regression. Low tumorigenesis rates and delayed tumor growth were observed, suggesting a lasting immunity (Fig. 7D). These findings further document that HNP1-mediated immune responses are required to completely eradicate established tumors and to prevent relapse.

Expression of HNP1-3 in various tumors is associated with tumor necrosis, prognosis, and dendritic cell infiltration (11, 13, 14). We recently reported that intratumoral expression of HNP1 inhibits human lung carcinoma A549 cells xenografts in nude mice (16). Because HNP1-3 are chemotactic for human T cells and dendritic cells in vitro and in vivo (11, 18, 27), HNP1-3 expressed intratumorally may play a role in the immunity and progression of tumors. In the present study, we have shown that HNP1 induces specific antitumor immunity using two different tumor models by recruiting and activating dendritic cells at tumor sites in immunocompetent mice. This, together with the direct cytotoxicity and antiangiogenesis activity of HNP1, results in inhibition, and even eradication, of established tumors.

To date, highly conserved peptides similar to HNP1-3 have not been identified in mice. A few studies have proven that HNP1-3 can also exert biological activity in mice as xenogenic peptides (16, 18, 35). In our study, the expression and secretion of HNP1 were observed within tumors accompanied with increased dendritic cell infiltration and morphologic changes of some HNP1-positive tumor cells. In addition, binding of HNP1 on tumor blood vessel walls was shown as described previously (35). In the two models, HNP1 was biologically active in vivo and able to attract dendritic cells into the tumor core, suggesting diffusion and action of HNP1 within the tumor milieu. The ability of HNP1 to attract dendritic cells into and throughout tumors is important, because, in most studied cancer patients, dendritic cells have been found mainly at the periphery of tumors (38, 39), which may limit their interaction with tumor cells. Importantly, the capacity of HNP1 to recruit circulating dendritic cells to the tumor site exposes tumor cells to freshly generated dendritic cells (as opposed to naturally occurring tumor-infiltrating dendritic cells), which may be less affected by an immunosuppressive tumor milieu. However, infiltration of immature dendritic cells within tumors may have an adverse effect by inducing a tolerance to the tumor cells (40), whereas maturation of dendritic cells is required for induction of an antitumor response against tumors. In fact, the mature status of dendritic cells enlisted within tumors was confirmed via immunostaining for CD11c⁺, CD40, and CD80. HNP1-mediated recruitment and activation of immature dendritic cells at the tumor site was sufficient to elicit of both cellular and humoral antitumor immunity against parental CT26 and 4T1 tumors.

It has been reported that murine -defensin 2 acts directly on immature dendritic cells as an endogenous ligand for Toll-like receptor 4, inducing dendritic cell maturation for mediating tumor immunity (28). However, the mechanism underlying HNP1-mediated dendritic cell maturation remains to be explored. Both apoptosis and necrosis were observed in two pSec-HNP1treated tumor models owing to biological activity of HNP1-3 to induce apoptosis and inhibit angiogenesis (13, 16, 35, 36, 41). There is accumulating evidence that dead tumor cells, whether apoptotic or necrotic cells, contribute to the maturation of dendritic cells and subsequent antigen presentation (30, 37, 42, 43). In the current study, although HNP1 alone did not maturate immature dendritic cells in vitro, the supernatant containing HNP1 showed maturation-inducing activity. We thus concluded that products from apoptotic cells trigger the maturation of immature dendritic cells and exclude the expression of cytokines, including TNF-, IL-4, and granulocyte/macrophage colony-stimulating factor. Therefore, the apoptosis effects of HNP1 in vivo are critical to the followed immune response. However, changes in additional cytokines were detected in vivo. The increased production of TNF- perhaps related to the effects of HNP1 on monocytes and the activation of immature dendritic cells (30, 44). In turn, the upregulated level of TNF- at the tumor site contributes to activation of immature dendritic cells. The maturation of immature dendritic cells should be responsible for the upregulation of IL-12. Although it was difficult for us to identify whether it is dead tumor cells or release of cytokines mediated by mature HNP1 in tumor milieu-induced maturation of immature dendritic cells, the resulting activation of immature dendritic cells is important for triggering antigen presentation.

In addition, HNP1 mediates T-cell activity as a T-cell chemoattractant (27). Previous studies have shown that depletion of rat neutrophils or suppression of neutrophil activity reduced subsequent delayed-type hypersensitivity responses (45,47). These results suggest that neutrophil-derived peptides may play an important role in promoting T-celldependent immune responses by activation of T cells. Moreover, it has been reported that HNP1-3 have potent immunoadjuvant activity in promoting antigen-specific immunoglobulin responses and tumor immunity (19). A previous study has shown that defensins promote release of CD4 Th1- and Th2-type cytokines to enhance systemic IgG responses and foster B- and T-cell interactions, which link innate immunity with the adaptive immune system (20). In our study, the inhibition of lung metastases in the 4T1 model followed local treatment of pSec-HNP1 reflects systematic antitumor effects and possible antitumor immunity. Then, both specific CTL and antibody responses were observed in pSec-HNP1treated tumor-bearing mice, suggesting activated specific cellular and humoral immunity. The different antibody responses were observed in pSec-HNP1treated mice. This may be attributed to the level of expression of HNP1, the number of dendritic cells accumulated within the tumors, the continuous antigens stimulation, and the resulting difference antigen presentation. Furthermore, depletion of CD4⁺ or CD8⁺ T cells with their monoclonal antibodies attenuated HNP1-mediated antitumor effects, which indicated HNP1-mediated antitumor immunity is dependent on both CD4⁺ helper and CD8⁺ cytotoxic T cells. Tumor suppression was also observed when pSec-HNP1cured mice in CT26 model were rechallenged with CT26 cells, exhibiting lasting immunity or evocation of immune memory.

We also observed minute downregulation of IL-10 and TGF- in pSec-HNP1treated groups at an early stage. The underlying mechanism is unknown; however, HNP1-associated decrease in the production of IL-10 by monocytes may explain the decrease of IL-10 (44). IL-10 and TGF- within tumors are thought to promote CD4⁺ CD25 T cells to transform CD4⁺CD25⁺ regulatory T cells. Given that HNP1 can chemoattract T cells into the tumor site, we detected infiltration of CD4⁺ T cells within tumors and did not observe obvious infiltration of CD4⁺ T cells in all groups (data not shown). This can also explain the lack of obvious increase of CD4 Th1- and Th2-type cytokines within tumors. Nevertheless, the decrease in IL-10 and TGF- may attenuate IL-10mediated immune suppression and reduces generation of CD4⁺ CD25⁺ regulatory T cells (48, 49); this may indirectly help produce host immune responses against tumors.

Activation of HNP1 peptide requires proteolytic removal of an anionic propiece of 40 residues, which protects HNP1-producing cells (2, 50). HNP1-3 mediates lysis of tumors in a concentration-dependent manner (13, 51, 52). In the current study, we used the mature peptide gene fragment of HNP1 without an anionic propiece directly fused with the Ig-chain leader sequence in pSecTag plasmid to obtain secretory mature HNP1 with cytotoxic activity (16). In this way, intracellularly expressed HNP1 was insufficient in the early stages to generate lethal effects owing to low concentrations, which contributed to the secretion of HNP1. The secreted HNP1 subsequently exerted chemotaxis or entered the blood and was bound to blood vessel walls. When the level of intracellular HNP1 reaches a lethal concentration, however, these resulting dead cells may help in the maturation and antigen presentation of immature dendritic cells recruited within tumor (30, 37, 42, 43). Therefore, under circumstances of gaining effective antigen presentation, high efficiency of gene delivery based on HNP1 to elicit specific antitumor immunity is certainly not indispensable for further cancer gene therapy.

Previous studies have also shown that low concentrations of HNP1-3 enhance proliferation of untransformed human epithelial cells, fibroblasts, and tumor cells, whereas higher concentrations of HNPs are cytotoxic to these cells in vitro (13, 53). However, HNP1-3, especially HNP1, do not enhance but inhibit the proliferation of human umbilical vein endothelial cells by interfering with ₅₁-integrindependent endothelial cell migration and adhesion to fibronectin in response to vascular endothelial growth factor even at a low concentration (35). Moreover, in vivo expression of HNP1-3 has been found to be related to tumor apoptosis and necrosis (13, 16), whereas HNP1-3 purified from colon tumors show lytic activity to canine kidney cells (MDCK cells; ref. 51). These observations indicate that multiple mechanisms, such as possible receptor signals or the mature status of HNP1-3, may be involved in roles of HNP1-3 in tumor microenvironment and angiogenesis. Our study suggests that the mature HNP1 expressed and secreted within tumors tends to generate more adverse effects to tumor cells. Given the multiple roles of HNP1-3 in comparison with the default roles of -defensins in apoptosis induction and angiogenesis inhibition, we think that -defensins may be a better choice for this type of treatment.

In summary, HNP1-mediated inhibition, and even eradication of tumors, may involve three different mechanisms: mediating antitumor immunity, inducing apoptosis, and inhibiting angiogenesis. This study showed that HNP1-mediated antitumor immunity may play a more important role. Therefore, gene therapy based on HNP1 may provide a promising candidate to be explored for further clinical application.

No potential conflicts of interest were disclosed.