Novel GM-CSF-based vaccines: One small step in GM-CSF gene optimization, one giant leap for human vaccines

ABSTRACT

Granulocyte macrophage-colony stimulating factor (GM-CSF) is a potent immunomodulatory cytokine that is known to facilitate vaccine efficacy by promoting the development and prolongation of both humoral and cellular immunity. In the past years we have generated a novel codon-optimized GM-CSF gene as an adjuvant. The codon-optimized GM-CSF gene significantly increased protein expression levels in all cells tested and helped in generating a strong immune responses against HIV-1 Gag and HPV-associated cancer. Here, we review the literature dealing with the adjuvant activity of GM-CSF both in animal models and clinical trials. We anticipate that the codon-optimized GM-CSF gene offers a practical molecular strategy for potentiating immune responses to tumor cell-based vaccinations as well as other immunotherapeutic strategies.

Introduction

Immunotherapies have been utilized to treat cancer for more than a century. A few clinical studies have successfully demonstrated immune activation in the treatment of malignant tumors. Granulocyte–macrophage colony stimulating factor (GM-CSF) is one of the most potent molecules used in these studies.

Granulocyte–macrophage colony stimulating factor (GM-CSF) is a glycoprotein comprising 127 amino acids with 2 potential N-linked glycosylation sites.^1,2 GM-CSF was first identified to regulate cell survival, proliferation, and differentiation in granulocyte-macrophage populations.³ GM-CSF also serves as a haematopoietic growth factor for erythroid, megakaryocyte and eosinophil progenitors. The effects of this growth factor on immune system have been studied in murine, ex vivo, and human models.⁴

GM-CSF is an inflammatory cytokine produced by various cell types, including T cells, B cells, macrophages, mast cells, endothelial cells, fibroblasts, and adipocytes, in response to cytokine or inflammatory stimuli. GM-CSF enhances the function of antigen presenting cells (APCs) by maturing, activating, and conscripting dendritic cells (DCs), macrophages, monocytes and eosinophils.^5,6

Persistent inflammation, immunosuppression, and catabolism syndrome (PICS) describes the development of chronic critical illness (CCI)⁷, which is characterized by increased risks of secondary infections, extended ICU and long-term acute facilities stays, long-term functional or mental impairment, and unexpectedly high post-hospital discharge mortality rates.⁴ GM-CSF can stimulate and enhance the production and function of neutrophils and monocytes and is used in immunostimulating adjuvant therapies to fight immunosuppressive diseases,^4,8,9,10 such as late sepsis. ⁴ GM-CSF has been studied for its effects on sepsis. A GM-CSF-treated group showed remarkably greater increases in total leukocyte counts, enhanced clearance rates and better clinical outcomes. Furthermore, membrane-associated human leukocyte antigen receptors (mHLA-DR), a biomarker for immune suppression ¹¹, expression levels increased.⁴ Lower mHLA-DR expression levels typically indicate inefficient immune cell function.¹² This study showed that treatment with GM-CSF resulted in significantly improved clinical outcomes without harmful side effects, including a more rapid recovery from infections, decreased admission lengths, decreased mechanical-ventilation days, and decreased financial costs.¹³

Disturbed GM-CSF expression has been attributed to the pathogenesis of autoimmune inflammatory diseases.¹⁴ Several studies have shown that GMCSF has important therapeutic potential in several inflammatory diseases, especially rheumatoid arthritis (RA).¹⁵ RA is characterized by chronic pain and destruction in afflicted joints and is induced by macrophages that produce a variety of inflammatory cytokines, such as TNFα, IL-1β, and IL-6.¹⁶ GMCSF was one of the first cytokines detected at high levels in the synovial fluid of inflamed joints.^17,18 The correlation between GM-CSF and RA was observed after GM-CSF injections in a mouse model of collagen-induced arthritis (CIA) exacerbated the disease. Treatment of these arthritic mice with neutralizing antibodies against GM-CSF prevented disease progression.¹⁹ Thus, the benefits for targeting GMCSF are obvious. The rapid and sustained clinical responses from early phase clinical trials of GMCSF antagonism in patients with RA seem assuring.¹⁵

In 1991, GM-CSF was approved for neutropenia associated with stem cell transplantation by the US Food and Drug Administration. Available forms of GM-CSF include sargramostim and molgramostim. Because GM-CSF promotes the function of APCs, GM-CSF has been considered an immunostimulatory adjuvant in numerous clinical trials and has been used to speed up recovery in patients suffering from non-Hodgkin's lymphoma, acute lymphoblastic leukemia and Hodgkin's disease undergoing autologous stem cell transplantation.²⁰ In 1990, a study of Gray collie dogs with cyclic neutropenia, i.e., recurring episodes of neutropenia caused by mutations in the endocytosis gene AP3B1,²¹ demonstrated the involvement of 3 cytokines, G-CSF, GM-CSF, and IL-3. The study revealed that GM-CSF and G-CSF increased neutrophil counts.²² GM-CSF was also studied as a potential adjunct to immunosuppressive therapy²³ for severe aplastic anemia, which is characterized by deficiencies in haematopoietic cell lines due to damaged bone marrow stem cells. GM-CSF immunosuppressive therapy lowered infection complication rates and hospital days than standard therapy alone,^24,25 especially for severe aplastic anemia.¹⁴

Adverse effects of GM-CSF are dose related. Appropriate and usual dose range being 5–10 micrograms/kg/day either by 4–6 h intravenous infusion or by subcutaneous injection. At such doses, adverse effects are almost reversible, mild-to-moderate in nature, occur in 20–30% of patients and usually comprise fever, myalgia, fatigue, diarrhea and injection site reaction (swelling and tenderness). A rare but significant side effect is thrombosis which rarely can result in pulmonary embolism or stroke. Similarly, a so-called “first-dose effect,” defined as a syndrome of hypoxia, hypotension and tachycardia after the initial but not subsequent doses of GM-CSF. Another serious, but very uncommon side effect of GM-CSF is “capillary leak syndrome” or “vascular leak syndrome.” It is characterized by the presence of 2 or more of the following 3 symptoms: hypotension, edema, and hypoalbuminemia due to the fluids within the vascular system (veins and capillaries) leaks into the tissue outside the bloodstream. Certain patient groups, for example those with myelodysplastic syndrome, acute myeloid leukemia, inflammatory disease and autoimmune thrombocytopenia, require careful clinical monitoring in order to avoid potential complications following the administration of GM-CSF.

GM-CSF in cell-based immunotherapy

GM-CSF advances the function of APCs.^5,6 Cytokines, such as IL-2^26,27 and GM-CSF,^28,29 have been applied in cancer treatment to elicit strong antitumor immune responses.³⁰ Treatments with irradiated GM-CSF-expressing B16 melanoma cells generated more potent and long-lasting anti-tumor responses than other immune modulating gene products. Subcutaneous or intradermal injections of GM-CSF-secreting tumor cells contributed to enhanced infiltration of APCs, such as DCs, macrophages, and granulocytes, which allowed CD4+ and CD8+ T cells to recognize circulating tumor-associated antigens.^29,31 DCs are considered as the most potent APCs within the immune system. The recruitment of professional APCs by GM-CSF augmented the intracellular environment for tumor antigen presentation.³¹

GVAX, a cell-based immunotherapy, is a GM-CSF gene-transduced tumor vaccine. The application of the GVAX system (Cell Genesys, now Aduro Biotech, Berkeley, CA) in preclinical studies has shown no remarkable local or systemic toxicities at clinically relevant doses.³¹ Murine tumor models and human clinical trials showed that GM-CSF-secreting tumor cells could be used to vaccinate patients with several tumor types, such as non-small cell lung carcinomas³², prostate tumors³³, pancreatic tumors³⁴, leukemia³⁵, melanomas³⁶, gliomas³⁷, cervical tumors³⁸, and renal cell carcinomas.³⁹

In a GVAX prostate vaccine phase II study, 34 hormone refractory prostate cancer patients with metastatic bone diseases at baseline were treated. Nine of the 24 patients at lower boost dosages (41%) and 7 of the 10 patients at the higher boost dosages (70%) survived the disease at the 2-year follow up. Additionally, there was a longer median time to disease progression as measured by bone scans in patients receiving the higher doses of vaccine compared with those at the lower doses (140 d vs. 85 d).⁴⁰ In a renal cell carcinoma study, 4 patients received GVAX for 48 vaccinations without showing severe adverse events. Two of the 4 patients survived at 58+ and 40+ months after the initial vaccination. These results indicated that GVAX fundamentally enhanced the antitumor cellular and humoral immune responses and may have contributed to the relatively long survival rates of patients.⁴¹

Clinical investigations with GVAX continue. While phase II studies have shown significant effects, many steps remain prior to therapeutic approval from the FDA. Prostate cancer has also been considered for tumor immunotherapy; these treatment principles could also be applied to many other tumor types. ⁴²

Positive results from phase II trials will likely lead to larger and more expensive phase III studies. However, unlike other medical disciplines, phase III trials often end in failure for cancer patients⁴³, e.g., GVAX immunotherapy for prostate cancer. Previous phase II studies have shown that GVAX71 immunotherapy was safe. Immunological studies have developed dosing regimens of tumor antigen-specific antibodies.⁴⁴ Phase II studies have not yet been compared with chemotherapy treatments. Two phase III trials have used chemotherapy comparator groups.⁴²

In the first trial, VITAL-1, a GVAX immunotherapy, was directly compared with chemotherapy in men with asymptomatic, castration-resistant prostate cancer (despite few radiographically detectable responses in phase II GVAX studies). However, VITAL-1 was unsuccessful. Most patients in the trial were considered unlikely to reach the primary survival endpoint and subsequently, the trial was stopped in 2008. In the second study (VITAL-2), a combination of immunotherapy and chemotherapy was hypothesized to increase survival rates in men with more advanced symptoms. No phase II trials were previously carried out to confirm the effects of a combined chemotherapy and immunotherapy treatment in animal or humans. Furthermore, no prior studies determined appropriate dosing schedules or dosing amounts. The VITAL-2 trial resulted in a greater number of deaths in patients treated with the combination of GVAX and chemotherapy, and the trial was subsequently terminated.^42,45 Although the positive predictive values of phase II trials for phase III outcomes are not especially vigorous, the negative predictive values are strong. A negative phase II trial in oncology clearly predicts a negative phase III result.⁴⁶

Combination immunotherapies could be used to maximize cancer patient benefits. Traditional cancer treatments include a combination of chemotherapy drugs or a combination of radiation therapy and chemotherapy.⁴² GVAX and CRS-207 are cancer vaccines that have been assessed in pancreatic ductal adenocarcinomas (PDAs). GVAX is composed of 2 irradiated, granulocyte-macrophage colony stimulating factor (GM-CSF)–secreting allogeneic PDA cell lines administered 24 hours after treatment with low-dose cyclophosphamide (Cy) to inhibit regulatory T cells. In a prior study, patients with previously treated advanced PDA who received Cy/GVAX had better induction of mesothelin-specific CD8+ T cells than those treated with GVAX alone. Respectively, the median survival rates were 4.3 and 2.3 months.³⁴ In a CRS-207 phase I study in 2012, patients with PDA who received GVAX before entering the study (n = 3) survived for a median of 17 months compared with the 5 months survival rate for those who did not receive prior GVAX (n = 4).⁴⁵ These observations led to a phase II randomized multicenter study where Cy/GVAX followed by CRS-207 treatments were shown to significantly improve the overall survival rates by 56% (2.2 months) of patients with metastatic PDAS when compared with Cy/GVAX treatments alone. The stable disease rate was 31%, the 1-year survival rate was 24%, and a stabilization or reduction in CA19-9 levels linked to survival was observed. This study encouraged the following combination trial.⁴⁶

The most common GVAX-related adverse events were local vaccine injection site reactions (90%) followed by fatigue (16%), nausea (12%), and pain and arthralgia (each at 5%). Two grade 4 (pericardial effusion) and 6 grade 3 (dyspnea, fatigue, injection site reaction, hypokalemia, malignant ascites, and pulmonary embolism) events were reported. There was no association between GVAX dose and the total number of adverse events or grade 3 and 4 adverse events.⁷⁹

A plasmid encoding both GM-CSF and bi-shRNA furin DNA was transfected into harvested tumor cells via electroporation as part of a vaccination termed FANG (Gradalis, Dallas, TX), which provides the afferent arm of the immune system with a full tumor antigen matrix.^47,48 This vaccine is a combination immune therapy that produces intra and extra-cellular adjuvant GM-CSF and simultaneously expresses an innovative RNA interference (RNAi) moiety and a bifunctional short hairpin RNA-furin (bi-shRNA-furin).⁴⁸

In 2013, a long term follow-up phase I study of FANG in advanced cancer cases⁴⁹ showed encouraging results. The resulting survival rates demonstrated an advantage of FANG over non-FANG treated populations (576 to 205 d and 604 to 228 d). Additionally, no long-term adversely toxic events were reported or observed with FANG.⁴⁹ The study also showed a remarkable knockdown (>90%) of both TGF-β1 and TGF-β2.⁴⁸ TGF-β affects a variety of cell types and has been shown to stimulate or inhibit cell growth, induce apoptosis and increase angiogenesis.^50,51 The overexpression of TGF-β has been associated with tumor progression and poor prognoses.^52,53 Consequently, the inhibitory effects of TGF-β isoforms suggested an immune-regulating function of GM-CSF. The observed FANG-induced immune response was a basic combinatorial TGF-β-suppressing/GM-CSF-expressing immune modulating therapy, which has since advanced to phase II evaluations for treating different tumor types and stages.⁴⁹

Novel codon-optimized GM-CSF gene

Native GM-CSF protein expression is poor in tissue-specific and activation-dependent forms.^54-56 Highly expressed genes are typically biased toward particular codons. These codons are species-dependent.^57,58 A number of studies have demonstrated that a good association between biased codon genes and their levels of expression.^59-61 The wild-type human and murine GM-CSF cDNA sequences were created by total gene synthesis and optimized by first identifying codons within the cDNA which were not associated with the codon usage in highly expressed genes in humans and mice, respectively. Each suboptimal codon was replaced with those identified from highly expressed genes of the same species. The novel GM-CSF sequences would improve the large-scale protein production without altering the amino acid sequences and biological function. This molecular strategy to enhance immune responses will prove effective for both DNA vaccines and GVAX vaccine. ^{62, 68}

The mechanism underlying the improvement of gene expression and induction of immune responses with codon-optimized, GM-CSF-adjuvanted vaccination is still largely unclear.

In a study of DNA vaccines against HIV-1 Gag using codon-optimized GM-CSF, we showed strong antibody, CTL, and protective immune responses against infection with a recombinant vaccine virus expressing HIV-1 Gag.⁶² To generate pcGM-CSF (plasmid codon-optimized GM-CSF), the codon-optimized form of murine GM-CSF, the researchers used codons of highly expressed human genes.^63,64 In initial tests on the protein expression levels of wild-type and codon-optimized murine GM-CSF vectors transfected into mice (NIH3T3), monkey (COS-7) and human (HeLa) cells, the GM-CSF protein was detected in transfected cells with the pcGM-CSF constructs. The pcGM-CSF-transfected cells produced greater amounts of secreted GM-CSF proteins (11,605 pg/ml in NIH3T3, 12,048 pg/ml in COS-7 and 13,250 pg/ml in HeLa cells) than plasmid wild type GM-CSF (pwtGM-CSF)-transfected cells (158 pg/ml, 205 pg/ml and 210 pg/ml, respectively). These results indicated that the novel codon-optimized murine GM-CSF coding sequences increased GM-CSF expression and significantly elevated the release of GM-CSF in NIH3T3, COS-7 and HeLa cells. Using the ELISPOT assay, we measured the number of antigen-specific T cells per million splenocytes and found that pcGM-CSF at least twofold increased the number of Ag-specific responders as measured by IFN-γ production. (1200 SFU versus 500 SFU). Additionally, We observed a high level of Gag-specific CD8+ memory cells reactivated in the spleens of mice immunized with pcGM + Gag (1213 CD8+/IFN-γ) within 4 d of vP1287 challenge compared with pwt-GM+Gag (564 CD8+/IFN-γ).⁶² The locally sustained release of this cytokine was successfully shown in murine models.^62,65

A similar study on codon-optimizing the human IL-15 gene was also recently investigated.^66,67 Additional research is required to investigate the effects of homologous and heterologous codon optimizations in mammalian hosts.⁶²

A study explored codon-optimized murine GM-CSF as a prophylactic vaccine adjuvant in cancer immunotherapy for an HPV-16 E6/E7-transformed cell line, TC-1.⁶⁸ Peak DC recruitment was observed at 72 h post-inoculation in draining lymph nodes (dLNs). This level was significantly increased (p < 0.05) in mice vaccinated with codon-optimized murine GM-CSF in TC-1 cells (TC-1/cGM) when compared with control mice or mice inoculated with wild-type GM-CSF in TC-1 cells (TC-1/wt). Mice vaccinated with irradiated TC-1/cGM cells exhibited increased levels of functional GM-CSF, enhanced immunosurveillance against TC-1 tumors, increased numbers of Ag-specific IFN-γ-producing CD8+T cells, and enhanced recruitment of macrophage-like cells into dLNs and IFN-γ-producing CD8+T cell compared with mice vaccinated with TC-1/wtGM. Using the ELISPOT assay, we measured the number of antigen-specific T cells per million splenocytes and found that TC-1/cGM at least 5-fold increased the number of Ag-specific responders as measured by IFN-γ production. (2800 SFU vs. 480 SFU). Additionally, we observed a high level of E7-specific IFN-γ producing CD8+ cells progressed through cell cycle upon restimulation (20.3% of TC-1/wt versus 30.5% of TC-1/cGM). These results demonstrated that cell-based vaccines secreting the novel c-GM-CSF gene product could prevent the growth of tumors.⁶⁸

The sustained local releases of GM-CSF at vaccination sites by GM-CSF-secreting cells in the induction of tumor immunity in many animal models were achieved using cells designed to release 90–300 ng/10⁶ cells per 24 h.³¹ Between 240 and 3100 ng/10⁶ irradiated cells was released over 24 h in transgenic TC-1/cGM cancer cells.⁶⁸ Significantly increased levels in macrophage and DC populations and tumor-specific effector T cell immunity were noted in the immune response. These results were associated with enhanced antitumor therapeutic benefits. Codon-optimized GM-CSF genes offer utilitarian molecular strategies for potential immune responses to tumor cell-based vaccinations and other immunotherapeutic strategies. Here, we review the literature dealing with the adjuvant activity of GM-CSF both in animal models and clinical trials (Table 1).

Major clinical studies

major preclinical and clinical studies to date with GM-CSF
year	Trial	Cancer (stage)/Phase study	Number patients Enrolled/treated	Safety/Response	Vector/Dose(vaccine cells)	GM-CSF production(ng/106 cells/24 h)	Concurrent treatment	Route/Number of injection/frequency
1997	Simons73	Renal cell (IV)/phase I	33/18	Yes/1 PR	Retrovirus/4 × 106, 4 × 107, 4 × 108 Dose escalation	17 – 149	IL2	50% i.d. and 50% s.c./total 4–21 injection/monthlyx3, 3-monthlyx3 then yearly
1999	Mastrangelo74	Melanoma (IV)/phase I	7/7	Yes/1 PR, 1 CR 3 MR	Vaccinia virus	—	None	i.t. / 12 at least/Biweekly until CR or PD
1999	Simons75	Prostate (R.P with LN mets)/phase I	11/8	Yes/No response	Adenovirus/1 × 107 or 5 × 107	150 – 1500	None	i.d. / total 3–6/3-weekly, until exhaustion of the supply of vaccine.
2000	Chang76	Melanoma (IV)/phase I	7/5	Yes/1 CR	Retrovirus/1 × 107	56 – 100	IL-2, 3.6 × 106μg/kg x15 doses	i.v. VPLN cells/1 to until exhaustion of supply of vaccine/Every 8 hour
2003	Salgia77	NSCLC (IV)/phase I	38/25	Yes/5 SD 2 CR (NED 43m, 42m)	Adenoivirus/1 × 106 or 4 × 106 or 1 × 107	Mean 513 (6–3017)	Surgery	50% i.d. and 50% s.c./6 injection at least /Weekly x3 then 2-weekly until the exhaustion of the supply of vaccine
2003	Soiffer78	melanoma (metastatic)/phase I	35/34	Yes/1 CR 1 PR 1 MR	Adenovirus/1 × 106, 4 × 106, 1 × 107	Mean 745	None	50% i.d. and 50% s.c./ND/weekly and biweekly
2004	Nemunaitis79	NSCLC (20 early stage Ib-IIa; 63 advanced III-IV)/phase I/II	83/43	Yes/3 CR in 33 (stage III-IV) 7/10 DFS >6 months (stage IB/II)	Adenovirus/Mean 23 × 106 Dose escalation 5–10 × 106, 10–30 × 106 30–100 × 106	Mean 104 (50 – 1871)	None	i.d./3–6/biweekly
2005	Luiten80	melanoma (IV/metastatic)/phase I/II	64/28	Yes/ 3 NED 1 NED s/p surgery 2 SD 9 NA	5 × 106 or 5 × 107	41–738	Surgery	2 i.d. and 2–3 s.c. 3 vaccinations/3-weekly
2006	Simons81	Prostate (isolated PSA recurrence after R.P) phase I/II	21/21	Yes/1 PR (PSA); 76%: descreased PSA slope	rAAV transduced/2 Allogeneic cells 6 × 107+ 6 × 107	150—450	None	i.d./8 doses/Weekly
2008	Laheru82	Pancreatic cancer(III-IV) phase I/II	cohort A, 30 pts. cohort B, 20 pts	Yes/SD: 26%	Plasmid transfected/Allogeneic cells 2.5 × 108– 5 × 108	80	cohort B: +CYC i.v. One day before vaccine	cohort A/B: i.d./6 doses/3 weekly
2009	Emens83	breast cancer (metastatic) phase I Dose-ranging Factorial design	28/28	Yes/NA	Plasmid transfected/Allogeneic cells 5 × 107 or 5 × 108	305	Chemotherapy (CYC, Dox)	28 patients received at least one immunization, 16 patients received 4 immunizations.
2012	Eertwegh84	Prostate Cancer (metastatic) phase I	12/12	Yes/7 patients >50% decline of PSA from baseline	rAAV transduced/Allogeneic cells 3×108	150—450	i.v. q28d ipilimumab at 0.3, 1.0, 3.0, or 5.0 mg/kg	i.d./total of 13 injections/bi-weekly
2013	VUKY85	Prostate cancer(local advanced)phase II	6/6	Yes/83%:PR17%:SD	rAAV transduced/Allogeneic cells first dose 5 × 108 cells; subsequent 3 × 108	150–450	Neoadjuvant docetaxel (4 cycles)	i.d./ GVAX (5 cycles), before R.P. and six courses of immunotherapy after R.P/ 3-weekly
2015	Le DT86	Pancreatic cancer (metastatic) phase II	93/90	Yes/OS:9.7 month GVAX+CRS-207 v.s 4.6 month in 6 doses GVAX	Plasmid transfected/2 Allogeneic cells 2.5 × 108 cells	80	CYC/GVAX priming, CRS207 (1 ×109cfu ) boosting	i.d./2 or 6 doses/3-weekly
2015	Lipson87	Melanoma (stage IIB-IV) phase I (post-surgery adjuvant)	20 /19	Yes/16 Pts: no recurrence in 6-month study period	Plasmid electroporated transfected/Allogeneic cells 5 × 107 or 2 × 108 or 2 × 108 with CYC	200–400	SURGERY; CYC 200 mg/m2	i.d./ 4 vaccinations/monthly

I.d., Intradermal injection; s.c., subcutaneous injection; i.v., intravenous infusion; i.t, Intratumoral injection

VPLN: vaccine-primed lymph node cells; rAAV: recombinant adeno-associated virus

DFS, disease-free survival; OS: overall survival

CR, complete response; PR, partial response; MR, mixed response; NED, no evidence of disease, PD, progression disease

Na, Nonassessable disease, ND, not described; NSCLC, non-small-cell lung cancer, IL-2, recombinant human interleukin-2; s/p, status post.

CYC, cyclophosphamide; Dox, Doxorubicin

R.P: radical prostatectomy

B codon-optimization GM-CSF. Major preclinical studies

Year	Major preclinical studies	Key findings
2007	Qiu J-T, Novel codon-optimized GM-CSF gene as an adjuvant to enhance the immunity of a DNA vaccine against HIV-1 Gag.62	first paper in DNA vaccine
2016	Lin C-C, The efficacy of a novel vaccine approach using tumor cells that ectopically express a codon-optimized murine GM-CSF in a murine tumor model.68	first paper in GVAX vaccine

Major clinical studies

pending

Table 1.

A: GMCSF.

Major preclinical studies
Year	Major preclinical studies	Key findings
1993	Dranoff G, Vaccination with irradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and long-lasting anti-tumor immunity.70	Comparison multiple cytokine genes
2001	Parker, S E, Safety of a GM-CSF adjuvant-plasmid DNA malaria vaccine71	GMCSF DNA vaccine is safe in GLP safety study
2002	Barouch DH, Potent CD4+ T cell responses elicited by a bicistronic HIV-1 DNA vaccine expressing gp120 and GM-CSF72	GMCSF DNA vaccine is potent

Conclusions

GM-CSF plays a critical role in eliciting tumor-specific immune responses, including activating both antigen-specific CD4+ and CD8+ T cells.⁶⁹ Our study is the first to demonstrate that changes in GM-CSF gene optimization may be a breakthrough of the anti-HIV and anti-tumor vaccine. Further clinical study involving larger numbers of patients is needed to assess the utility of this novel gene. We anticipate that the development of GM-CSF treatments will lead to a number of effective cancer therapies. The codon-optimized GM-CSF gene offers a practical molecular strategy for potentiating immune responses to tumor cell-based vaccinations as well as other immunotherapeutic strategies.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We are grateful to Long-Ji Chang and Becky Chen for their helpful discussions.

Funding

This study was supported by CMRPG390611 and CMRPG3E1281, which were provided by Chang Gung University Hospital (Taoyuan, Taiwan).