Improved Efficacy of Dendritic Cell–Based Immunotherapy by Cutaneous Laser Illumination

With the ability of generating large amounts of dendritic cells (DC) ex vivo in the past 2 decades, therapeutic potential of DCs has been extensively explored in preclinical and clinical studies, in particular, in the field of tumor immunotherapy (1–3). In the clinic, DCs can be generated from human peripheral blood monocytes, matured in the presence of appropriate growth factors and cytokines, pulsed with tumor antigens, and given back to patients to elicit antitumor immunity (1, 3). This DC-based immunotherapy has been broadly evaluated in a variety of tumors, like malignant melanoma, prostate cancer, breast cancer, renal cell carcinoma, and so on (3, 4). However, only a small fraction (<1%) of DCs could migrate to the dLNs following intradermal (i.d.) or subcutaneous injection and the majority of DCs remained and died over time at the site of injection, limiting their antitumor efficacy (3, 5). A novel strategy to improve DC migration from the skin to the dLNs would potentially boost DC-based immunotherapy.

In the skin, initial lymphatic vessels are localized beneath the epidermis and serve as a conduit for DC migration from the skin to the dLNs. Migration of DCs to and through the lymphatic vessels after i.d. administration is primarily guided by chemokine gradients of CCL19/CCL21 secreted by lymphatic endothelial cells under a physiologic condition (6). Apart from chemokines, gene knockout studies indicate that intrinsic DC migratory ability and dermal tissue microenvironment are also pivotal in the control of DC migration from the skin to the dLNs. In this regard, a lack of MHC II–associated invariant chain (li or CD74) or junctional adhesion molecular A (JAM-A) on DCs was associated with increased DC motility in vitro and sufficient migration of the cells to the dLNs in vivo due to decreased adhesion to the matrix of the skin connective tissue (7, 8). In contrast, absence of matrix metalloproteinase (MMP)-2 and -9 on DCs diminished their ability to degrade extracellular matrix proteins, resulting in reduced DC migration to the dLNs after subcutaneous injection (9, 10). An importance of dermal tissue microenvironment in the control of DC migration was also stressed by deletion of a matricellular protein named SPARC (secreted protein, acidic, rich in cysteine, or osteonectin/BM-40). Bone marrow–derived DCs migrated more sufficiently to the dLNs in SPARC-deficient mice than in wild-type mice following i.d. injection (11).

The dermal connective tissue consists of densely aligned collagen fibers and extracellular matrixes that serve as tissue scaffold restricting DC migration. We recently developed a laser-based vaccine adjuvant capable of safe and efficient enhancement of vaccine-induced immune response (12). The vaccine adjuvant effect is induced primarily by enhancing velocity and motility of in situ DCs in the skin as well as sufficient transportation of antigen-captured DCs from the immunization site to the dLNs. In the present study, we show that the laser illumination can also greatly enhance migration of i.d. injected DCs to the dLNs and potentiate DC-based immunotherapy in both preventive and therapeutic tumor models.

BALB/c mice at 6 to 8 weeks of age were purchased from Charles River Laboratories. C57BL/6 (B6, CD45.2) and B6.SJL (CD45.1) mice at 6 to 8 weeks of age were purchased from the Jackson Laboratory. MHC II-EGFP mice expressing MHC class II molecule infused into enhanced GFP (EGFP) were a kindly gift of Drs. Boes and Ploegh, Department of Pathology, Harvard Medical School (Boston, MA; ref. 13). All mice were housed in conventional cages at the animal facilities of Massachusetts General Hospital in compliance with institutional guidelines. All animal procedures were reviewed and approved by the Subcommittee on Research Animal Care at the Massachusetts General Hospital.

A Q-switched 532-nm Nd:YAG laser (Spectra-Physics Inc.) was used with a pulse width 6 ns, a beam diameter approximately 7 mm, and a frequency 10 Hz. The average output power was measured by a power meter (Ophir Optics, Inc.) and temperature (T_m) at the illumination center of the skin was monitored by an infrared camera (FLIR Systems).

Bone marrow–derived DCs were prepared as described (14). Briefly, femur and tibia bone marrow cells were cultured for 6 days in the presence of recombinant murine granulocyte macrophage colony—stimulating factor (GM-CSF; 15 ng/mL) and interleukin (IL)-4 (10 ng/mL). Nonadherent cells were replated and matured by recombinant murine TNF-α (50 ng/mL) and lipopolysaccharide (LPS; 100 ng/mL) for another 2 days. Mature DCs (mDC) were harvested on day 8 and stained for 20 minutes at 37°C with 10 μmol/L CFSE [5-(and-6)-carboxyeosin diacetate, succinimidyl ester; Invitrogen] according to the manufacturer's instruction. GM-CSF, IL-4, and TNF-α were all purchased from R & D Systems and LPS was obtained from Sigma.

CFSE-labeled DCs were i.d. injected to the lower dorsal skin of mice, and dLNs were collected at indicated times after DC administration. Single-cell suspensions were prepared as our previous report (12) and stained with phycoerythrin (PE)-anti-cd11c (N418), PerCP-Cy5.5-anti-CD45.2 (104), or APC-anti-CD45.1 (A20) followed by flow cytometric analysis on BD FACSAria. Migrated DCs marked as CFSE⁺ cd11c⁺ cells were calculated relatively to the total cell number of the lymph node. To analyze DC maturation, DCs were stained with PE-anti-CD80 (16-10A1), APC-anti-MHC II (M5/114.15.2), PE-anti-MHC I (28-8-6), and APC-anti-CD40 (3/23) and subjected to flow cytometric analysis of mean fluorescence intensity (MFI) of CD80, MHC II, CD40, and MHC I expression in gated CFSE⁺ cd11c⁺ cells.

Autologous tumor lysates were used to pulse mDCs (15). Briefly, 4T1 breast tumor cells [American Type Culture Collection (ATCC), CRL-2539] and renal adenocarcinoma cells (Renca; ATCC, CRL-2947) were cultured in RPMI-1640 medium and B16F10 melanoma cells (ATCC, CRL-6475) were cultured in Dulbecco's Modified Eagle's Medium. The culture media were supplemented with 10% FBS, 2 mmol/L l-glutamine, and 1% penicillin-streptomycin. The cells were harvested and subjected to 4 freeze-thaw cycles followed by centrifugation at 5,000 rpm for 30 minutes. Supernatants were collected and added to the DC culture for 18 hours in a tumor to DC ratio of 3:1. Tumor lysates–pulsed mDCs (5 × 10⁵) were i.d. injected into syngeneic BALB/c or C57BL/6 mice at the site of laser illumination (0.4 W for 1.5 minutes except otherwise specified) right after laser treatment or a corresponding sham site once a week for 3 consecutive weeks.

4T1 cells (3 × 10⁴) and B16F10 cells (5 × 10⁴) per mouse were subcutaneous inoculated to the right flank of BALB/c and C57BL/6 mice, respectively. Tumor volume measurement and tumor resection were referred to our previous report (16). Lung metastatic foci were counted under a dissection microscope by an investigator blind to the sample groups after careful dissection and separation of the lung tissue into single lobes.

Student t test and one-way ANOVA were used to analyze a difference between 2 groups or among multiple groups, respectively. Kaplan–Meier method and log-rank test were used to analyze the survival data. P values were calculated by PRISM software (GraphPad) and considered significant if a P value was less than 0.05.

Additional methods and materials can be found in online supplements including whole-mount ear immunofluorescent staining, transmission electron microscopy (TEM), CTL assays, and intracellular cytokine staining.

Our previous study showed that cutaneous laser illumination at a specific setting increased the motility of antigen-presenting cells (APC) in the skin without incurring any overt side effects (12). To address how the increased motility affected lymphatic transportation of the cells, the right ear of MHC II-EGFP transgenic mice was exposed to laser at 80 mW for 15 seconds, a low laser dose that is well tolerable to the ear, followed by intravital confocal imaging 1 hour later as previously described (12). We found a number of dermal cords in the laser-illuminated area but not in the control ear, presumably resulting from a high number of GFP⁺ cells flowing into lymphatic vessels (Fig. 1A). These GFP⁺ cells encompass mainly MHC class II⁺ DCs and macrophages defined broadly as APCs (13, 17). Confocal imaging of collagen IV–stained whole-mount ear of MHC II-EGFP transgenic mice revealed few, if any, GFP⁺ cells inside the lymphatic vessels under a physiologic condition (Fig. 1B). Enumeration of approximately 250 GFP⁺ cells in 15 randomly selected fields found only 0.8% GFP⁺ cells inside lymphatic vessels in control ear, which increased by 20-fold, 30 minutes after laser illumination (Fig. 1B and C). Meanwhile, percentage of GFP⁺ cells in association with the lymphatic vessel wall was increased from 10.9% in control ear to 38.2% in laser-illuminated ear (Fig. 1B and C). These lymphatic vessel–associated cells crawled on the vessel wall and some of them extended their body into the channel (Fig. 2A). In contrast, few GFP⁺ cells crawled or extended their body into the channel in the control ear (Figs. 1A and 2A). Moreover, laser illumination appeared to enlarge preformed portals in perilymphatic membrane as previously described (18). The average diameter of the channels was 1.57 ± 0.06 μm in the control ear but 2.93 ± 0.20 μm after laser illumination, a representative of an 87% increase (Fig. 2A and B). The size of the channels was further enlarged to 4.34 ± 0.29 μm with cell occupancy, corroborating previous investigation showing that the preformed channels could be passively dilated by entering cells (18).

We next investigated whether laser illumination also disarrayed the densely packed dermal connective tissue, permitting relatively free cell migration in the interstice. To this end, TEM was used to analyze the ultrastructure of the dermis following laser illumination at 0.3 W for 2 minutes. Dermal connective tissue was composed of well-aligned bundles of collagen fibers in control skin and the cells were intimately contacted and surrounded by dense extracellular matrix and fibers with which cell migration might be restricted (Fig. 2C). Upon laser illumination, however, dermal tissue became loosely compacted with apparent gaps between fibers and between fibers and cells. The cells in laser-illuminated skin were disconnected with collagen fibers, concurrent with extending their long dendrites to remote fibers, suggestive of cell migration. These data, in line with our previous investigation (12), suggest that laser facilitates DC migration from the skin to the lymphatic vessels potentially by at least 3 mechanisms: enlargement of the preformed channels in the initial lymphatic vessels, disarray of the dense protein network, and increased motility of APCs.

Laser-induced alterations in the dermal microstructure described earlier promoted us to investigate whether cutaneous laser illumination could enhance migration of i.d. injected DCs to the dLNs. As shown in Fig. 3A and B, laser illumination significantly increased the number of DCs migrating into the dLNs at all times tested, with cumulative migration peaking on day 2 after DC inoculation. At the peaking level, the inoculated DCs arriving at the dLN increased from 1,376 ± 176 in control group to 2,735 ± 256 in 0.3 W laser group or 3,482 ± 188 in 0.4 W laser group in BALB/c mice or from 1,668 ± 100 in control group to 2,947 ± 210 in 0.3 W laser group or 3,738 ± 262 in 0.4 W laser group in C57BL/6 mice (Fig. 3C). No CFSE⁺ DCs were found in the dLNs of PBS-injected control mice (data not shown). Besides mDCs, laser also enhanced migration of immature DCs similarly (data not shown), in agreement with a physics-, rather than chemokine-, involved mechanism in laser-augmented migration of DCs. The migration efficiency appeared to be laser power dependent with laser illumination at 0.4 W for 1.5 minutes significantly stronger than that at 0.3 W for 2 minutes in both mice and thus laser illumination at 0.4 W for 1.5 minutes was chosen in the subsequent studies. In addition to DC migration, how laser illumination affected the maturation of i.d. injected mDCs was also evaluated 2 days after DC injection. As shown in Fig. 3D, MFIs of CD80 and MHC I expression, but not of CD40 and MHC II expression on injected DCs were slightly but significantly higher in laser-treated skin than in sham-treated skin. The result suggests that laser illumination does not compromise, but augments DC maturation following i.d. injection even though the microarchitecture of the skin connective tissue is significantly altered by laser exposure.

To address whether laser-enhanced DC migration gave rise to more vigorous CD8⁺ T-cell responses, the number and function of these cells were evaluated 1 week after the last immunization. It was found that the percentage of IFN-γ–secreting CD8⁺ T cells relative to a total number of CD8⁺ T cells was 1.12% ± 0.15% in control group but increased to 3.32% ± 0.31% in DC group, or 4.73 ± 0.08 in laser + DC group, which represents a 42% increase in the presence as compared with the absence of laser illumination (Fig. 4A and B). Similarly, the total number of IFN-γ–secreting CD8⁺ T cells per lymph node was 2.3-fold higher with laser treatment than without it (P < 0.001, Fig. 4C), and this trend was extended to the spleen (Fig. 4D). No significant increases were seen in the number of IL-4- or IFN-γ–secreting CD4⁺ T cells or IL-4–secreting CD8⁺ T cells by laser illumination in spite of the slightly high levels of their induction as compared with control groups (data not shown). Furthermore, T cells isolated from mice immunized with 4T1 tumor lysates–pulsed mDCs showed specific killing of 4T1 tumor cells, but not irrelevant Renca cells, with the killing activity significantly stronger in the presence than in the absence of laser illumination (Fig. 4E and F). These results strongly suggest that laser illumination followed by DC immunization selectively expands tumor-specific IFN-γ–secreting CD8⁺ T cells and increases their specific cytotoxicity.

We next validated whether more vigorous expansion of tumor-specific CD8⁺ T cells and enhanced CTL activity could be translated into enhanced antitumor immunity. As shown in Fig. 5A, no 4T1 tumor growth was seen for 17 days after tumor inoculation in laser + DC group. In sharp contrast, tumor grew fast in the control and DC groups: tumor volume reached 131 mm³ in control group and 64 mm³ in DC group on day 17. The tumor in laser + DC group gradually became visible thereafter and was 150 mm³ 4 weeks after tumor inoculation, whereas tumor was 637 mm³ in control group and 399 mm³ in DC group at the same time. At the end of the study, tumor mass was 74% smaller in mice receiving DC immunization at the site of laser illumination or 40% smaller in mice receiving DC immunization alone than that in nonimmunized mice (P < 0.01, Fig. 5B). Moreover, laser plus DC immunization reduced lung metastases significantly, as manifested by significantly fewer lung metastatic foci when compared with control group (P < 0.05, Fig. 5C). No significant difference was found in lung metastatic foci between control and DC immunization alone under similar conditions. No macroscopic metastases were found in other organs like liver, kidney, intestine, colon, and brain. Laser-mediated improvement of DC-based immunotherapy was also showed with B16F10 melanoma in C57BL6 mice. All mice in control group (6 of 6) and two thirds of the mice in DC group (4 of 6) developed visible tumors within 2 weeks, whereas none of the mice in laser + DC group (0 of 6) developed tumor in the entire experimental period (Fig. 5D). Log-rank test indicated that the percentage of tumor-free mice in laser + DC group was significantly higher than all other groups (P < 0.05 vs. DC group and P < 0.001 vs. control group). In accordance with this, tumor volume on average was significantly smaller in laser-treated group than in sham-treated group (Fig. 5E). Strikingly, whereas DC immunization reduced tumor volume, laser plus DC immunization completely abrogated tumor growth in the entire observation period. All mice in control and laser + unpulsed DC groups, and 4 of 6 mice in DC group, died of cancer within 50 days, but all mice in laser + DC group survived beyond the end of the experiment (P < 0.001 vs. control group; P < 0.05 vs. DC group). Laser-enhanced DC immunotherapy in both 4T1 and B16F10 models critically depended on tumor antigen used to pulse DCs as no significant protection was attained when unpulsed DCs were administered into the site of laser illumination (Fig. 5A–F, laser + unpulsed DC).

DC-based immunotherapy would be most valuable in prevention of tumor relapse or tumor resection–induced metastases for patients with primary tumor surgically removed. In an attempt to mimic this clinic setting, BALB/c mice were subcutaneously inoculated with 4T1 tumor cells and primary tumor was surgically removed after reaching 100 mm³. One week later, tumor-resected mice were i.d. immunized with tumor lysates–pulsed mDCs at the site of laser or sham illumination once a week for 3 weeks. All mice in control and DC groups succumbed to severe lung metastases within 28 days after tumor resection. In contrast, 50% of the mice in laser + DC group were still alive at this time point, although the mice eventually died in the following weeks. Log-rank test verified that laser plus DC, but not DC alone, significantly extended the survival of tumor-resected mice (P = 0.017 vs. control group and P = 0.042 vs. DC group, Fig. 6), suggesting that brief laser illumination before DC immunization could give rise to more effective therapy.

The present investigation introduces a novel technology to improve DC-based immunotherapy by using cutaneous laser illumination to enhance DC migration following i.d. injection. We observe dermal cord formation in MHC II-EGFP mice within 1 hour after laser illumination, which, to the best of our knowledge, is the first report on the visualization of in vivo dermal cord formation. Active entry of skin APCs into lymphatic vessels was seen 30 minutes after laser illumination, in contrast to dermal cord formation in ex vivo skin explants that occurs after 1 to 3 days of culture in a manner dependent on maturation and chemoattraction of APCs (19–22). The observation argues that laser illumination induces dermal cord formation by a different mechanism as synthesis and release of CCL19/CCL21 chemokines by lymphatic endothelial cells takes at least 6 to 8 hours (23, 24). Moreover, when immature DCs were administered into the site of laser illumination, similar enhancement in DC migration from the skin to the dLNs was also observed (data not shown). These immature DCs did not express a significant level of the CCR7 receptor, a chemokine receptor for CCL19/CCL21, on the surface, and laser-treated skin did not produce detectable levels of CCL19/CCL21 chemokines either at this time point when measured by real-time PCR in homogenates prepared from the skin (data not shown). This chemokine-independent, instant effect of laser on DC migration would be significant in the clinic where laser illumination and DC administration can be conducted at the same visit. In contrast, use of TNF-α or imiquimod to enhance DC migration requires more than 6-hour treatment of the injection site before DC administration. In addition, laser illumination could enhance DC migration similarly and significantly in a broad dose range from 10⁵ to 10⁶ cells injected, which would be a key because 10⁶ DCs are usually needed for effective DC-based immunotherapy (25). Unlike laser illumination, the efficacy of inflammatory cytokine-enhanced DC migration varied with the number of DCs, with a maximal effect when less than 5 × 10⁵ DCs were injected and a minimal effect if more than 10⁶ DCs were injected (24). Repeated imiquimod treatment of DC injection site has been reported to synergize with DC-based immunotherapy in an intracranial tumor model by increasing migration and survival of injected DCs and enhancing priming of tumor-specific CD8⁺ T cells (26). While the effect of laser illumination is similar to repeated imiquimod treatment in enhancement of DC migration, expansion of tumor-specific CD8⁺ T cells, and inhibition of tumor growth, single laser treatment is much more convenient when compared with multiple imiquimod treatments at different time points. Further studies are needed to determine whether combination of laser with imiquimod can further enhance DC-based immunotherapy.

One mechanism underlying laser-facilitated DC migration may be alterations of the dermal microarchitecture, such as loosely compacted dermal tissues, enlarged gaps, and disarrayed collagen fibers, as suggested by the TEM study. These alterations would reduce DC migratory resistance in the interstice and increase DC motility as shown in our previous study (12). It is also clear that laser illumination results in significant enlargement of preformed channels in the basement membrane of the initial lymphatic vessels (18), adding another attributable factor to a sufficient entrance of DCs into the lymphatic vessel. Moreover, it is well known that fever can accelerate interstitial flow by 10-fold and promote lymphatic transportation of APCs during infection (27, 28). Although it has not been tested, increased interstitial flow presumably contributes to the effect as well, as local skin temperature was increased to 41°C when measured by an infrared camera focusing on the center of the illumination site. A pulsed temperature inside the skin may be higher at its peak as a high-repetition short-pulsed laser was used (29), although further investigation is required to unravel the photothermal effect. Besides, we found by injection of mDCs from congeneic B6.SJL mice into C57BL/6 mice that CFSE⁺cd11c⁺ cells in the dLNs were primarily the injected CD45.1⁺ DCs, not CD45.2 host cells that took up of CFSE-stained cells, ruling out that skin resident DCs were activated after engulfment of CFSE-stained DCs and migrated into the dLNs, contributing to the observed therapeutic improvement. The finding is consistent with little improvement of DC-based immunotherapy when laser illumination was followed by injection of unpulsed DCs. In brief, our data suggest that laser-augmented migration of DCs involves a complex physical mechanism, which is consistent with the quick formation of dermal cords after laser illumination and its independence of DC maturation status.

DC-based immunotherapy has been extensively evaluated in the clinic in a variety of tumor models in the past 2 decades. However, it is until very recently that this approach shows significantly prolonging patient survival with advanced prostate cancer and melanoma (30, 31). But the efficacy is far from satisfying, requiring improvement in multifacets, among which the major efforts have been made to optimize DC culture conditions, cytokine cocktails for maturation, and tumor antigen loading. Laser modulation of the skin at the injection site to enhance DC migration can be complementary to any of these efforts to further improve therapeutic efficacy of DCs. Although intravenous infusion is also used to deliver therapeutic DCs in the clinic, this route of administration has been found to favor humoral immune response, rather than cell-mediated immune response that is more pivotal for tumor therapy (3, 32–34). In addition, intravenous delivery of DCs induced memory T cells primarily in the spleen, whereas i.d. route of DC delivery induced memory T cells not only in the spleen but also in the lymph nodes, enabling more targeted protection against tumors at specific anatomic locations like cutaneous tumors and breast cancer in which metastatic tumor cells first reach cutaneous lymph nodes (3, 35). Furthermore, because laser only illuminates a small area of the skin, this procedure is unlikely to generate any undesirable systemic side effects, it is expected to gain fast approval by U.S. Food and Drug Administration (FDA), and it thus warrants further investigation.

No potential conflicts of interest were disclosed.

Conception and design: X. Chen, M.X. Wu.

Development of methodology: X. Chen, Q. Zeng, M.X. Wu.

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Chen, Q. Zeng, M.X. Wu.

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Chen, M.X. Wu.

Writing, review, and/or revision of the manuscript: X. Chen, M.X. Wu.

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M.X. Wu.

Study supervision: M.X. Wu.

The authors thank Dakai Mu and Zhanliang Wei for harvesting and culturing of bone marrow–derived DCs; Peggy Sherwood for TEM sectioning, staining, and photograph taking; the member in Drs. Wu's groups for stimulating discussion; and Drs. Boes and Ploegh for MHC II-EGFP mice.

This work was supported in part by the NIH grants AI070785 and RC1 DA028378 and Grand Challenges Explorations grant # 53273 from the Bill and Melinda Gates Foundation (to M.X. Wu).

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.