September 2015
Volume 56, Issue 10
Free
Anatomy and Pathology/Oncology  |   September 2015
A Novel Model of Metastatic Conjunctival Melanoma in Immune-Competent Mice
Author Affiliations & Notes
  • Simona L. Schlereth
    Department of Ophthalmology University of Cologne, Cologne, Germany
  • Sandra Iden
    Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases, and Center for Molecular Medicine Cologne, University of Cologne, Cologne, Germany
  • Melina Mescher
    Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases, and Center for Molecular Medicine Cologne, University of Cologne, Cologne, Germany
  • Bruce R. Ksander
    Department of Ophthalmology, Schepens Eye Research Institute, Massachusetts Eye and Ear Infirmary and Harvard Medical School, Boston, Massachusetts, United States
  • Jacobus J. Bosch
    Department of Internal Medicine 5, Hematology and Oncology, University of Erlangen-Nuremberg, Erlangen, Germany
  • Claus Cursiefen
    Department of Ophthalmology University of Cologne, Cologne, Germany
  • Ludwig M. Heindl
    Department of Ophthalmology University of Cologne, Cologne, Germany
  • Correspondence: Simona L. Schlereth, University of Cologne, Department of Ophthalmology, Kerpenerstr. 62, 50924 Cologne, Germany; Simona.schlereth@uk-koeln.de
Investigative Ophthalmology & Visual Science September 2015, Vol.56, 5965-5973. doi:10.1167/iovs.15-17290
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Simona L. Schlereth, Sandra Iden, Melina Mescher, Bruce R. Ksander, Jacobus J. Bosch, Claus Cursiefen, Ludwig M. Heindl; A Novel Model of Metastatic Conjunctival Melanoma in Immune-Competent Mice. Invest. Ophthalmol. Vis. Sci. 2015;56(10):5965-5973. doi: 10.1167/iovs.15-17290.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Conjunctival melanoma (CM) is an ocular surface tumor that can lead to fatal metastases. Patients developing, tumor-associated lymphangiogenesis have a significantly increased risk of metastatic disease, because tumor spread primarily occurs via lymphatic vessels to the draining lymph node. Here, we describe a novel immune-competent mouse model of CM that displays tumor-associated lymphangiogenesis with development of metastatic tumors.

Methods: C57BL/6N mice received C57BL/6N-derived dermal melanoma cells (hepatocyte growth factor [HGF] cyclin dependent kinase-4 [Cdk4]+) or B16F10 via subconjunctival injection. A clinical score quantified primary tumor growth and metastases were identified by macroscopic examination of the draining lymph nodes, lung, and spleen. Confirmation of tumors and metastases was achieved by immunohistochemical staining for markers of pigmented cells (tyrosinase related protein-2 [TRP2]) and S-100, and of cell proliferation (Ki67). The intra- and peritumoral CD31+ blood and lymphatic vessel endothelium hyaluronan receptor-1 (LYVE-1)+ lymphatic vessels were quantified immunohistochemically.

Results: All mice rapidly developed aggressive TRP2+, S100+, and Ki67+ CM. Metastatic tumors were found in the lymph node (9%) and lung (6%) of HGF-Cdk4R24C–treated mice and in the spleen (8%) and lung (17%) of B16F10-treated mice. The amount of peri- and intratumoral blood vessels was significantly increased compared with lymphatic vessels.

Conclusions: This CM model in immune-competent animals offers new possibilities to study the pathobiology of tumor growth, invasion, and mechanisms of metastatic tumor spread, and provides a robust model to explore new immune-based and antilymphangiogenic treatment modalities of this malignancy.

Conjunctival melanoma (CM) is the second most common malignancy of the ocular surface. The current treatment of primary CM is complete surgical removal of the tumor combined with adjuvant therapy in terms of radio-, cryo-, topical chemo-, and/or immunotherapy. However, even though surgical specimens indicate intact tumor-negative margins and apparent tumor removal, CM patients still display an unusually high rate of local recurrence (36%–62% of patients16). Tumor recurrence coincides with a high risk for the development of metastases, which occurs in 12% to 25% of patients.13 
Metastatic tumors in CM patients are typically first detected in the lymph nodes draining the tumor-containing eye, indicating the primary route of tumor cell dissemination occurs via the lymphatic vessels. In support of this, we previously demonstrated that lymphangiogenesis is associated with the transition of premalignant precursors into CM.7 Moreover, the extent of lymphatic vessels within CM coincided with an increased risk of local recurrence, lymphatic spread of tumor cells, and development of metastatic disease.810 
Melanomas are derived from the malignant transformation of melanocytes, which are found in a variety of different tissues, including the skin and, within the eye, the uveal tract and conjunctiva. Although all these types of melanoma share a common cellular origin, they can display significantly different disease progression pathways. For example, uveal melanoma spreads from the choroid of the eye via the blood vasculature and metastatic tumors develop almost exclusively in the liver. By contrast, skin and CM spread primarily via lymphatic vessels, and metastatic tumors first develop in local tumor-draining lymph nodes. This suggests that CM is more similar to skin than to uveal melanoma, a pattern that is further supported by data on the genetic mutations associated with these tumors. Uveal melanoma exhibits mutations in GNAQ and GNA11, whereas skin melanoma and CM typically possess BRAF and NRAS mutations.11,12 In addition, the incidence of CM is increasing,1,1315 which is similar to the increasing frequency observed in skin melanoma. 
Primary cutaneous melanomas develop spontaneously in the skin of hepatocyte growth factor (HGF)-cyclin dependent kinase 4 (Cdk4)-transgenic mice (on the C57BL/6 background).16 The basis for this established mouse model of skin melanoma is that the overexpression of a Cdk4R24C mutant impairs the cell cycle control, which furthermore leads to insensitivity of inhibitory signals by INK4A proteins, combined with overexpression of HGF. Hepatocyte growth factor allows maintaining melanocytes within the interfollicular epidermis, next to hair follicles. Adult mice usually do not harbor melanocytes in the interfollicular epidermis, thus this model recapitulates the human situation. Combined HGF-Cdk4R24C expression leads to spontaneous melanoma development and can be accelerated by 7,12-dimethylbenzanthracene (DMBA) treatment or UV irradiation. Subsequently, macroscopically darkly pigmented dermal tumors and metastasis to the lymph nodes and lung are detectable.16 
In this study, we used the B16F10 or HGF-Cdk4R24C melanoma cells to develop a novel mouse model of CM that displays tumor-associated lymphangiogenesis with development of metastatic tumors in immune-competent mice, thus allowing exploration of novel immune-mediated therapies. 
Materials and Methods
Animals and Subconjunctival Injections
Four groups, each containing 11 to 12 female C57BL/6N mice (8–12 weeks; Charles River Laboratories, Sulzfeld, Germany) received subconjunctival injections of melanoma cells (group 1: 5 × 105 HGF-Cdk4R24C cells, group 2: 1 × 105 HGF-Cdk4R24C cells, group 3: 5 × 103 HGF-Cdk4R24C, group 4: 5 × 103 B16F10 cells). For subconjunctival injection, mice were anaesthetized with an intraperitoneal injection of a combination of 8 mg/kg body weight ketamine (Ketanest S; Godecke AG, Berlin, Germany) and 0.1 mL/kg body weight xylazine (Rompun; Bayer, Leverkusen, Germany), and both eyes were topically anaesthetized with oxybuprocain (conjucain EDO; Dr. Gerhard Mann, Berlin, Germany). Tumor cells (5 × 103/1 × 105 or 5 × 105 cells/10 μL PBS) were subconjunctivally injected using a Hamilton needle (Hamilton Company, Reno, NV, USA). Correct location of the subconjunctival injection was achieved by appearance of a temporary conjunctival edema, without tumor cell reflux. Injections were performed only on the temporal bulbar conjunctiva in all animals (Fig. 1A). After injection, mice received a single preventive topical ofloxacin treatment (Floxal EDO; Dr. Gerhard Mann). 
Figure 1
 
Conjunctival tumor growth of HGF-Cdk4R24C melanoma cells. (A) Scheme of the subconjunctival injection into the bulbar conjunctiva to imitate penetration through the basement membrane by the tumor mass. (BD) Representative clinical pictures of CM at different days postinjection when injecting 500,000 HGF-Cdk4R24C cells. (E) Clinical picture for intense neovascularization in a less-pigmented area of the tumor. (F) Increase in the clinical score (including chemosis, redness, lid distance from bulb, and proptosis) is increasing with time. Here, different numbers of cells were compared in their time course to induce clinical signs. (G) Survival of the animals, defined by proptosis as termination criterion.
Figure 1
 
Conjunctival tumor growth of HGF-Cdk4R24C melanoma cells. (A) Scheme of the subconjunctival injection into the bulbar conjunctiva to imitate penetration through the basement membrane by the tumor mass. (BD) Representative clinical pictures of CM at different days postinjection when injecting 500,000 HGF-Cdk4R24C cells. (E) Clinical picture for intense neovascularization in a less-pigmented area of the tumor. (F) Increase in the clinical score (including chemosis, redness, lid distance from bulb, and proptosis) is increasing with time. Here, different numbers of cells were compared in their time course to induce clinical signs. (G) Survival of the animals, defined by proptosis as termination criterion.
Mice were examined under a stereomicroscope (Nikon SMZ800; Nikon GmbH, Düsseldorf, Germany) using a standardized clinical scoring system (Table) that allows discrimination of melanoma growth in detail and thus to grade disease progression. Photographs were taken using a digital sight DS-Vi1 camera (Nikon). Chemosis, redness, and lid distance from bulb were differentiated in three different grades, whereby only proptosis was defined as a termination criterion; therefore, appearance of proptosis was counted 5 points. Mice were killed once proptosis occurred. The tumor-bearing eye and the contralateral eye were enucleated and weighted using a micro scale (Sartorius, Göttingen, Germany). 
Table
 
Clinical Score to Classify Tumor Growth: Mice Were Examined for Chemosis, Redness, and Lid Distance From Bulb
Table
 
Clinical Score to Classify Tumor Growth: Mice Were Examined for Chemosis, Redness, and Lid Distance From Bulb
All experiments were conducted in compliance with the ARVO Statement of the Use of Animals in Ophthalmic and Visual Research and with institutional animal care committee policies and procedures and animal license of the State Office, North Rhine-Westphalia, Germany. 
Cell Culture
The HGF-Cdk4R24C melanoma cells, isolated from cutaneous melanoma of HGF-Cdk4R24C mice on a C57Bl/6 background, were kindly provided by Thomas Tüting (Department of Dermatology, University of Bonn, Bonn, Germany) and originally described by Landsberg et al.17 The B16F10 cells were kindly provided by Sandra Iden. Cells were kept in RPMI medium (Gibco, Life Technologies, Darmstadt, Germany), containing 10% fetal calf serum, Pen Strep (both PAA laboratories, Coelbe, Germany), and nonessential amino acids at 37%, 5% CO2 and passaged using trypsin and trypsin inhibitor (both PromoCell, Heidelberg, Germany) when reaching 80% to 90% confluency. Before injection, cells were collected using trypsin, washed in PBS, counted, suspended in PBS, and kept on ice until injection. Cells were resuspended directly before each injection. 
Histology and Immunostaining
Immunohistochemistry was performed as described previously.1820 Briefly, tissues were dissected and snap frozen in tissue tek O.C.T. compound (Sakura Finetek Europe, Alphen aan den Rijn, The Netherlands); 10-μm-thick cryosections were prepared using a cryotome CM3050S (Leica, Wetzlar, Germany), and transferred on Superfrost slides (Thermo Scientific, Braunschweig, Germany). All specimens were fixed with acetone (Carl Roth, Karlsruhe, Germany). Nonspecific binding was blocked by incubation in PBS containing 5% calf serum (1 hour, room temperature) and the following primary antibodies were used: S100B/S100 (monoclonal rabbit anti-mouse; Biozol, Munich, Germany), tyrosinase related protein 2 (TRP2) (polyclonal goat anti-mouse antibody; Santa Cruz Biotechnology, Heidelberg, Germany), Ki67 (rabbit anti-mouse; DCS innovative diagnostic systeme, Hamburg, Germany), CD31-FITC (BD Pharmingen, Heidelberg, Germany), and lymphatic vessel endothelium hyaluronan receptor 1 (LYVE1) (AngioBioCo, Del Mar, CA, USA).21 The primary antibodies were added in a dilution of 1:100 in PBS (containing 2% fetal calf serum) (overnight incubation, 4°C). The next day, samples were washed three times for 5 minutes with PBS with continuous shaking. Primary antibodies were detected with corresponding secondary antibodies (2 hours at 20°C), diluted in 2% normal goat serum (Dako, Glostrup, Denmark) in PBS to avoid nonspecific binding. 
The following secondary antibodies were used in a 1:400 dilution: Alexa Fluor 488 conjugate donkey anti-goat (Abcam, Cambridge, UK), Alexa Fluor 555 goat anti-rat (Invitrogen, Darmstadt, Germany), Alexa Fluor 488 goat anti-rabbit (Life Technologies, Eugene, OR, USA) or goat anti-rabbit Cy3 (Dianova, Hamburg, Germany). 
Samples were rinsed again three times for 5 minutes on the shaker, then incubated with 4′,6-diamidino-2-phenylindole (DAPI) diluted 1:2000 in PBS (Carl Roth, Karlsruhe, Germany) (10 minutes, 20°C), washed again and embedded in fluorescent mounting medium (Dako), and stored at 4°C. Slides were examined using a fluorescent microscope (Olympus BX53; Olympus Optical, Hamburg, Germany), and micrographs were taken using a digital camera (XM10; Olympus). Negative controls were included in the analysis by omission of the primary antibodies and resulted in no distinct staining. 
Blood and lymphatic vessels were quantified as previously published.19 Briefly, from each eye three photographs were taken at different locations using a fluorescent microscope (Olympus BX53) and digital pictures of each probe layer were taken using the digital camera mentioned above at 10-fold magnification. At peri- or intratumoral location, 2.7mm2-sized central pictures were taken, and the area covered by CD31+ FITC+ vessels was measured with an algorithm established in the imaging Cell-F software (Olympus Deutschland GmbH, Hamburg, Germany). Before analysis, gray value images of the whole-mount pictures were modified by several software filters, and vessels were detected by the appropriate threshold setting including the bright vessels and excluding the dark background. 
Statistical Analyses
Statistical analysis was performed using Prism 6, V.6.02 (GraphPad Software, San Diego, CA, USA). A two-tailed unpaired t-test was used for detection of differences between the groups. For the survival curve, a log-rank test was performed. For the analysis of blood and lymphatic vessels, a Kruskal-Wallis test was used. P values less than 0.05 were considered statistically significant. 
Results
Subconjunctival Injection of HGF-Cdk4R24C Melanoma Cells Induced Rapid and Aggressive CM Growth in All Animals
A major aim was to develop a novel robust mouse model for CM. First, we tested different amounts of injected tumor cells and followed the kinetics of tumor formation and progression. We injected 5 × 105, 1 × 105, or 5 × 103 cells of C57BL/6N-derived HGF-Cdk4R24C melanoma cells into female C57BL/6N mice. All animals showed a rapid and aggressive local tumor growth with darkly pigmented solid tumors (Figs. 1B–D). At the end of the first week, an intense vascularization was visible (Fig. 1B), and after 7 to 10 days a brown tumor was visible. The mice had to be killed due to tumor growth causing massive unilateral proptosis with lid closure difficulties. For example, when 5 × 105 cells were injected, the mice had to be euthanized within approximately 9 days. In less pigmented areas, an intense neovascularization was visible (Fig. 1E). The more HGF-Cdk4R24C cells were injected, the faster was the tumor growth. The clinical findings were quantified using a clinical score (Table; Fig. 1F), demonstrating that high numbers of tumor cells led to earlier and higher clinical scores. Figure 1G shows the survival rates of the HGF-Cdk4R24C groups, using proptosis as an endpoint. The survival curve similarly shows that high numbers of tumor cells led to earlier termination of the experiments (P < 0.0001): all mice were killed by day 9 in the 5 × 105 group, whereas in the low-dose group (5 × 103 cells) mice could be kept until day 21. 
HGF-Cdk4R24C– and B16F10-Induced CM Show a Similar Ocular Phenotype
In a next step, we compared the tumor growth kinetic of the HGF-Cdk4R24C cells with B16F10 cells. In these experiments, we injected 5000 HGF-Cdk4R24C cells as well as 5000 B16F10 cells and compared tumor growth, clinical scores, and survival (defined by incomplete lid closure due to large tumors). Exenterating the tumor-bearing eye revealed tumor infiltration of the whole orbit in B16F10- or HGF-Cdk4R24C– induced CM (Figs. 2A–C). Regarding the clinical scores, the B16F10-induced CM was slightly faster than the HGF-Cdk4R24C–induced CM (Fig. 2D). The survival curves did not significantly differ from each other (Fig. 2E) (P = 0.075). These data indicate significant and aggressive growth of both HGF-Cdk4R24C and B16F10 tumor cells in the conjunctiva of the eye with a higher speed of tumor growth for B16F10 compared with the HGF-Cdk4R24C–induced CM. 
Figure 2
 
Clinicopathological characteristics of CM. (A) Conjunctival melanoma infiltrates the orbit of a HGF-Cdk4R24C mouse, detectable only after (B) exenteration. (C) Clinical picture of a CM induced by B16F10 cells. (D) Clinical score development shows an earlier clinical phenotype in the B16F10-induced CM compared with the HGF-Cdk4R24C–induced CM, associated with an earlier termination of the experiments: (E) survival curve of B16F10- and HGF-Cdk4R24C–induced CM.
Figure 2
 
Clinicopathological characteristics of CM. (A) Conjunctival melanoma infiltrates the orbit of a HGF-Cdk4R24C mouse, detectable only after (B) exenteration. (C) Clinical picture of a CM induced by B16F10 cells. (D) Clinical score development shows an earlier clinical phenotype in the B16F10-induced CM compared with the HGF-Cdk4R24C–induced CM, associated with an earlier termination of the experiments: (E) survival curve of B16F10- and HGF-Cdk4R24C–induced CM.
HGF-Cdk4R24C and B16F10 Melanoma Cells In Vivo Express Tumor Markers S100, Tyrosinase-Related Protein 2 (TRP2), and Ki67
After development of the tumor, we verified the melanoma properties of the HGF-Cdk4R24C and B16F10 cell line and validated the known melanoma profile (i.e., expression of TRP2, S100, and the proliferation marker Ki67). For this, tumor-containing eyes were harvested, cryosectioned, and stained by immunohistochemistry for these markers. The tumors were contained within the conjunctiva and the globe was microscopically free of melanoma cells in any of the animals tested (Fig. 3A). At the posterior part of the eye, the tumors surrounded the extraocular muscles and the optic nerve, whereby the tumors, despite their size, did not infiltrate the eye, as the sclera was macro- and microscopically intact. In some cases, the bulbus was compressed by the tumor mass (data not shown). HGF-Cdk4R24C tumor cells expressed high levels of TRP2, Ki67, and S100 (Figs. 3B–D). Similar results were obtained for the B16F10-induced tumors (Figs. 3E–G), whereby S100 was weakly expressed here. 
Figure 3
 
HGF-Cdk4R24C and B16F10 CM cells express Ki67, TRP2, and S100 in vivo. (A) The exenterated eyes were prepared for cryo sections followed by immunohistochemistry: the overview indicates that nearly all the tumor cells proliferate as determined by Ki67 staining. (B) The border between conjunctiva and tumor demonstrating that only the tumor is positive for TRP2 expression. (C) Magnification of Ki67-positive tumor cells. (D) Intratumoral section of S100-positive cells. Similarly, B16F10-induced CM is positive for Ki67 (E) and TRP2 (F) and weakly positive for S100 (G). Conj, conjunctiva.
Figure 3
 
HGF-Cdk4R24C and B16F10 CM cells express Ki67, TRP2, and S100 in vivo. (A) The exenterated eyes were prepared for cryo sections followed by immunohistochemistry: the overview indicates that nearly all the tumor cells proliferate as determined by Ki67 staining. (B) The border between conjunctiva and tumor demonstrating that only the tumor is positive for TRP2 expression. (C) Magnification of Ki67-positive tumor cells. (D) Intratumoral section of S100-positive cells. Similarly, B16F10-induced CM is positive for Ki67 (E) and TRP2 (F) and weakly positive for S100 (G). Conj, conjunctiva.
In HGF-Cdk4R24C– and B16F10-Induced CM, the Tumor Is Enriched With CD31+ Blood and LYVE1+ Lymphatic Vessels
Progressive tumor growth coincided with intra- and peritumoral LYVE1+ lymphatic vessels and CD31+ blood vessels (Fig. 4). Quantification of the areas covered by CD31+ blood vessels or LYVE1+ lymphatic vessels showed that the intra- and peritumoral amount of blood vessels was significantly increased compared with lymphatic vessels, both in the HGF-Cdk4R24C– (intratumoral P < 0.05 and peritumoral P < 0.01, respectively; Figs. 4A–C) and B16F10-induced CM (intratumoral P < 0.0001 and peritumoral P < 0.0001, respectively; Figs. 4D–F). Comparing the location within the tumor, lymphatic vessels were higher in the tumor periphery than in the tumor center, both in the HGF-Cdk4R24C– (not significant) and in the B16F10-induced CMs (P < 0.01). Blood vessels were detected throughout the tumor (Figs. 4B, 4E). There was a tendency for more lymphatic vessels in the periphery of the HGF-Cdk4R24C group compared with the B16F10 group. However, these results were not significantly different (P > 0.05). 
Figure 4
 
Quantification of peri- and intratumoral blood and lymphatic vessels: (A) In HGF-Cdk4R24C–induced CM, more peri- and intratumoral blood vessels (BVs) than lymphatic vessels (LVs) are detectable (P < 0.05 and P < 0.01). No significant differences were seen between peri- and intratumoral lymphatic vessels. (B) Example of intratumoral BVs (white arrowheads) and LVs (red arrows). (C) Example of a large peritumoral LV. (D) In B16F10-induced CM peri- and intratumoral clearly more BVs than LVs were measured (P < 0.0001). Intratumorally, very few lymphatic vessels were detected compared with peritumoral (P < 0.01). (E) Example of intratumoral BVs and LVs. (F) Example of peritumoral BVs and LVs, showing LVs mainly located at the tumor margin.
Figure 4
 
Quantification of peri- and intratumoral blood and lymphatic vessels: (A) In HGF-Cdk4R24C–induced CM, more peri- and intratumoral blood vessels (BVs) than lymphatic vessels (LVs) are detectable (P < 0.05 and P < 0.01). No significant differences were seen between peri- and intratumoral lymphatic vessels. (B) Example of intratumoral BVs (white arrowheads) and LVs (red arrows). (C) Example of a large peritumoral LV. (D) In B16F10-induced CM peri- and intratumoral clearly more BVs than LVs were measured (P < 0.0001). Intratumorally, very few lymphatic vessels were detected compared with peritumoral (P < 0.01). (E) Example of intratumoral BVs and LVs. (F) Example of peritumoral BVs and LVs, showing LVs mainly located at the tumor margin.
Lymphatic Metastasis in the Draining Lymph Node and Distant Metastasis to the Lung in HGF-Cdk4R24C–Treated Mice
In total, 23% of mice treated with 5 × 103 or 1 × 105 HGF-Cdk4R24C cells showed macroscopically detectable darkly pigmented spots in the draining lymph nodes (lymph nodes: 14% of the lower dosed groups or a total of 9% in all groups; lung: 9% of the lower dosed groups or a total of 6% in all groups) (Fig. 5). Compared with the normal contralateral lymph nodes (Fig. 5A), the tumor-containing eye draining lymph nodes were pigmented and enlarged (Fig. 5B). Mice that had been injected with 5 × 105 or more cells showed intense pigmentation of the draining lymph nodes, but without a clear demarcation zone (data not shown). None of the animals that received 5 × 105 cells showed metastasis, likely because the rapid growth of the primary tumor required early termination of the experiment. Decreasing the tumor cell dose from 5 × 105 cells to 5 × 103 or 1 × 105 and thereby prolonging the experiment led to metastases (see Fig. 5B as an example for a large metastasis in the draining lymph node). The presence of metastatic tumor cells was confirmed by expression of S100 (Fig. 5C) and TRP2 (Figs. 5D, 5E). Dark-pigmented spots on the lung were observed macroscopically, and immunohistochemical analysis confirmed positivity for TRP2 and S100 (Figs. 5F–K). 
Figure 5
 
HGF-Cdk4R24C CM metastasizes to the draining lymph node and lung: (A) Control lymph nodes with no apparent brown pigmentation. (B) Lymphatic metastasis to the draining lymph nodes associated with slow-growing conjunctival melanoma, induced by injection of 5000 or 100,000 cells. (C) Immunohistochemical analysis of cryosectioned lymph node metastases revealed positivity for S100 and (D, E) TRP2. Scale bar: 50 μm. (F, G) Macroscopic appearance of distant lung metastasis. (HK) Immunohistochemistry confirmed the expression of S100 and TRP2 in the lung metastasis of CM. These data are representative for two animals with a total of 21 lung metastases.
Figure 5
 
HGF-Cdk4R24C CM metastasizes to the draining lymph node and lung: (A) Control lymph nodes with no apparent brown pigmentation. (B) Lymphatic metastasis to the draining lymph nodes associated with slow-growing conjunctival melanoma, induced by injection of 5000 or 100,000 cells. (C) Immunohistochemical analysis of cryosectioned lymph node metastases revealed positivity for S100 and (D, E) TRP2. Scale bar: 50 μm. (F, G) Macroscopic appearance of distant lung metastasis. (HK) Immunohistochemistry confirmed the expression of S100 and TRP2 in the lung metastasis of CM. These data are representative for two animals with a total of 21 lung metastases.
Lymphatic Metastasis to the Spleen and Distant Metastasis to the Lung in B16F10–Treated Mice
Mice treated with 5 × 103 cells of B16F10 mice developed metastasis to the spleen (8.3%) and to the lung (17%). The metastatic tumor was confirmed histologically by positive staining for TRP2 (Figs. 6A–E). The S100 staining was nearly absent. In the spleen, only one large-sized brown pigmentation was detectable (Fig. 6A), whereas in the lung, multiple small pigmented dots (Fig. 6D), all positive for TRP2 (Fig. 6E), were observed. Metastasis to the lymph node was not observed in the B16F10 group. 
Figure 6
 
B16F10 CM metastasizes to the spleen and lung: (A) Spleen with a brown pigmentation (arrow). (B) Hematoxylin and eosin staining of the pigmented area in the spleen. (C) TRP2 positivity of the spleen cells. (D) Lung metastasis seen as multiple small brown spots. (E) Pigmented cells are positive for TRP2.
Figure 6
 
B16F10 CM metastasizes to the spleen and lung: (A) Spleen with a brown pigmentation (arrow). (B) Hematoxylin and eosin staining of the pigmented area in the spleen. (C) TRP2 positivity of the spleen cells. (D) Lung metastasis seen as multiple small brown spots. (E) Pigmented cells are positive for TRP2.
Discussion
We have previously shown the importance of peri- and intratumoral lymphatic vessels for recurrence, metastasis, and prognosis in different ocular malignancies, specifically in CM.8-10,2226 Therefore, antilymphangiogenic therapies are needed to develop new therapeutic options. To study them, we need murine immune-competent models because for lymphangiogenesis, immune cells are closely interlinked.27 However, to date there are no immune-competent mouse models to study the biology of CM and to test new therapeutic approaches.28,29 In this study, we aimed to develop a new model for CM in immune-competent mice. In patients, CM growth is initiated within the epithelium and invasion across the basement membrane is a major indicator of malignancy. In our model, we are directly injecting tumor cells underneath the basement membrane, to mimic the clinicopathological situation. Similar to existing tumor models30 in other anatomical sites of the eye,3133 this mouse model has its pros and cons. A major disadvantage might be the use of cutaneous and not primary CM cells, as it is not suitable for testing human antibodies or immune-mediated human therapies. However, due to the current lack of mouse CM cell lines and considering the fact that epidermis and conjunctiva show many similarities regarding embryonic origin, development, stratification, lymphatic drainage, and outer barrier to the environment, we consider cutaneous cell lines are currently one of the best options for tumor induction in immune-competent syngeneic mice. A similar approach for uveal melanoma is used in a mouse model developed by Lattier et al. and Dithmar et al.,3133 in which B16F10 cutaneous melanoma cells are injected in the posterior compartment of the eye. Another approach has been performed by De Waard et al.30 Hereby, the authors inject human CM cells into severe combined immunodeficient (SCID) mice. 
As noted above, an advantage of our allogeneic model reported here is that the tumor develops in healthy, immune-competent animals, and does not require genetic modifications or inactivation of the immune system, in contrast to the xenograft approaches, in which human tumor cells are orthotopically transplanted in SCID mice3436 or athymic nude mice.3740 It is well accepted that the tumor microenvironment is essential for progression of tumors,41,42 therefore our model presented here allows studying the effect of the tumor microenvironment on the host's immune system, neovascularization, and metastasis.43 
The role of different murine immune cells, cytokines, or growth factors can be studied in detail in this model (e.g., cancer-associated fibroblasts or macrophages, chemokine receptors, or the origin of tumoral lymphatic vessels). We have optimized the protocol for a mouse strain that is also commonly used to generate genetically modified mouse strains, opening the possibility to apply our model using specific mutant mice in future studies. In humans, genetic causes underlying CM are poorly investigated. 
Another advantage is that tumor growth is fast and aggressive, allowing quick and cost-reduced experiments. The aggressive tumor growth imitates the clinical situation in patients exemplified by a full infiltration of the orbit, which is comparable to a T3 stage in humans. However, orbital infiltration is clinically seen much less frequently than in our murine model, which underlines a disadvantage of this model: it is not directly comparable to the clinical situation. This is also seen in the metastasis behavior: our model provides faster manifestation of metastasis compared with humans. 
The tumor growth and metastases in our model developed within approximately 14 days in the draining lymph nodes or the lung, a faster process than seen in an immune-deficient conjunctival mouse model.30 In our model, tumor growth is much faster, although only 0.005% of the number of cells were injected compared with the immune incompetent model. This emphasizes the importance of a functional immune system that is used by the tumor to promote and support the malignancy. 
Still, both models are usable for different questions and complement one another. The reason for this quick metastatic spread is unclear. Clinically, metastases are often detected months or years after the resection of the primary malignancy.44,45 In human studies, regional lymph node metastasis appeared within 34 ± 10 months after primary diagnoses of CM.8 Distant metastases were observed with a mean time interval of 42 ± 10 months after primary treatment.8 
In the current study, we used two skin melanoma tumor cell lines to develop a new mouse model of CM, which is critical for the development and testing of new therapies as well as studying the mechanisms of disease progression. A key requirement of a useful CM model is that the growth of primary tumors coincides with lymphangiogenesis, lymphatic tumor spread, and development of local draining lymph node metastases, to mimic human disease. 
In conclusion, we established a mouse model of CM in immune-competent C57/BL6N mice, with reproducible primary tumor growth as well as locoregional lymphatic and distant metastasis. This allows further studies of the biology, microenvironment, hem- and lymphangiogenesis, and immunity in this important form of ocular melanoma. Furthermore, this immune-competent in vivo model provides valuable tools to identify effective antitumor therapies both locally in the conjunctiva as well as at locoregional and distant metastatic organ sites based on immunomodulatory and antilymphangiogenic strategies. However, all future conclusions resulting from this mouse model will still need verification in the clinical situation to improve the outcome for patients who suffer from CM. 
Acknowledgments
We thank Roland Ullrich from the University of Cologne, Department of Internal Medicine, for helpful scientific discussions. 
This work was presented in part as a poster session at the annual meeting of the Association for Research in Vision and Ophthalmology, Orlando, Florida, United States, May 2014. 
Supported by German Research Foundation (FOR 2240 “[Lymph]Angiogenesis and Cellular Immunity in Inflammatory Diseases of the Eye” [JJB, CC, LMH]; HE 6743/2-1 and HE 6743/3-1 [LMH]; BO 4000/2-1 and BO 4000/3-1 [JJB]; CU 47/6-1 [CC]); the Excellence Initiative by the German federal and state governments (CECAD Cologne) (SI); DFG Grants ID79/1-1, ID79/1-2, and CRC829/A10 (SI); the Stiftung Kölner Krebsforschung (Exzellenz Initiieren) (SI); CMMC project grant A3 (SI); SFB 829 ID 79/1-1 (SI), German Cancer Aid (LMH, CC), GEROK Program University of Cologne (SLS, LMH), and EU COST BM1302 “Joining Forces in Corneal Regeneration” (CC). The authors alone are responsible for the content and writing of the paper. 
Disclosure: S.L. Schlereth, None; S. Iden, None; M. Mescher, None; B.R. Ksander, None; J.J. Bosch, None; C. Cursiefen, None; L.M. Heindl, None 
References
Tuomaala S, Eskelin S, Tarkkanen A, Kivela T. Population-based assessment of clinical characteristics predicting outcome of conjunctival melanoma in whites. Invest Ophthalmol Vis Sci. 2002; 43: 3399–3408.
Shildkrot Y, Wilson MW. Conjunctival melanoma: pitfalls and dilemmas in management. Curr Opin Ophthalmol. 2010; 21: 380–386.
Shields CL, Shields JA, Gunduz K, et al. Conjunctival melanoma: risk factors for recurrence, exenteration, metastasis, and death in 150 consecutive patients. Arch Ophthalmol. 2000; 118: 1497–1507.
Seregard S. Conjunctival melanoma. Surv Ophthalmol. 1998; 42: 321–350.
Tuomaala S, Toivonen P, Al-Jamal R, Kivela T. Prognostic significance of histopathology of primary conjunctival melanoma in Caucasians. Curr Eye Res. 2007; 32: 939–952.
Seregard S, Kock E. Conjunctival malignant melanoma in Sweden 1969-91. Acta Ophthalmol (Copenh). 1992; 70: 289–296.
Heindl LM, Hofmann-Rummelt C, Adler W, et al. Tumor-associated lymphangiogenesis in the development of conjunctival squamous cell carcinoma. Ophthalmology. 2010; 117: 649–658.
Heindl LM, Hofmann-Rummelt C, Adler W, et al. Prognostic significance of tumor-associated lymphangiogenesis in malignant melanomas of the conjunctiva. Ophthalmology. 2011; 118: 2351–2360.
Heindl LM, Hofmann-Rummelt C, Adler W, et al. Tumor-associated lymphangiogenesis in the development of conjunctival melanoma. Invest Ophthalmol Vis Sci. 2011; 52: 7074–7083.
Zimmermann P, Dietrich T, Bock F, et al. Tumour-associated lymphangiogenesis in conjunctival malignant melanoma. Br J Ophthalmol. 2009; 93: 1529–1534.
Feng X, Degese MS, Iglesias-Bartolome R, et al. Hippo-independent activation of YAP by the GNAQ uveal melanoma oncogene through a trio-regulated rho GTPase signaling circuitry. Cancer Cell. 2014; 25: 831–845.
Shoushtari AN, Carvajal RD. GNAQ, and GNA11 mutations in uveal melanoma. Melanoma Res. 2014; 24: 525–534.
Triay E, Bergman L, Nilsson B, All-Ericsson C, Seregard S. Time trends in the incidence of conjunctival melanoma in Sweden. Br J Ophthalmol. 2009; 93: 1524–1528.
Inskip PD, Devesa SS, Fraumeni JF Jr. Trends in the incidence of ocular melanoma in the United States 1974-1998. Cancer Causes Control. 2003; 14: 251–257.
Yu GP, Hu DN, McCormick S, Finger PT. Conjunctival melanoma: is it increasing in the United States? Am J Ophthalmol. 2003; 135: 800–806.
Tormo D, Ferrer A, Gaffal E, et al. Rapid growth of invasive metastatic melanoma in carcinogen-treated hepatocyte growth factor/scatter factor-transgenic mice carrying an oncogenic CDK4 mutation. Am J Pathol. 2006; 169: 665–672.
Landsberg J, Kohlmeyer J, Renn M, et al. Melanomas resist T-cell therapy through inflammation-induced reversible dedifferentiation. Nature. 2012; 490: 412–416.
Schlereth SL, Neuser B, Caramoy A, et al. Enrichment of lymphatic vessel endothelial hyaluronan receptor 1(LYVE-1)-positive macrophages around blood vessels in the normal human sclera. Invest Ophthalmol Vis Sci. 2014; 55: 865–872.
Schlereth S, Lee HS, Khandelwal P, Saban DR. Blocking CCR7 at the ocular surface impairs the pathogenic contribution of dendritic cells in allergic conjunctivitis. Am J Pathol. 2012; 180: 2351–2360.
Schwartzkopff J, Schlereth SL, Berger M, et al. NK cell depletion delays corneal allograft rejection in baby rats. Mol Vis. 2010; 16: 1928–1935.
Schroedl F, Kaser-Eichberger A, Schlereth SL, et al. Consensus statement on the immunohistochemical detection of ocular lymphatic vessels. Invest Ophthalmol Vis Sci. 2014; 55: 6440–6442.
Heindl LM, Hofmann TN, Adler W, et al. Intraocular tumor-associated lymphangiogenesis a novel prognostic factor for ciliary body melanomas with extraocular extension? Ophthalmology. 2010; 117: 334–342.
Heindl LM, Hofmann TN, Knorr HL, et al. Intraocular lymphangiogenesis in malignant melanomas of the ciliary body with extraocular extension. Invest Ophthalmol Vis Sci. 2009; 50: 1988–1995.
Heindl LM, Hofmann TN, Schrodl F, Holbach LM, Kruse FE, Cursiefen C. Intraocular lymphatics in ciliary body melanomas with extraocular extension: functional for lymphatic spread? Arch Ophthalmol. 2010; 128: 1001–1008.
Heindl LM, Schrodl F, Lutjen-Drecoll E, Cursiefen C. Ciliary body lymphangiogenesis. Ophthalmology. 2013; 120: e41–e42.
Koch KR, Refaian N, Hos D, et al. Autocrine impact of VEGF-A on uveal melanoma cells. Invest Ophthalmol Vis Sci. 2014; 55: 2697–2704.
Cursiefen C, Chen L, Borges LP, et al. VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J Clin Invest. 2004; 113: 1040–1050.
Bock F, Maruyama K, Regenfuss B, et al. Novel anti(lymph)angiogenic treatment strategies for corneal and ocular surface diseases. Prog Retin Eye Res. 2013; 34: 89–124.
Hos D, Schlereth SL, Bock F, Heindl LM, Cursiefen C. Antilymphangiogenic therapy to promote transplant survival and to reduce cancer metastasis: What can we learn from the eye? Semin Cell Devel Biol. 2015; 38: 117–130.
de Waard NE, Cao J, McGuire SP, et al. A murine model for metastatic conjunctival melanoma. Invest Ophthalmol Vis Sci. 2015;.
Lattier JM, Yang H, Crawford S, Grossniklaus HE. Host pigment epithelium-derived factor (PEDF) prevents progression of liver metastasis in a mouse model of uveal melanoma. Clin Exp Metastasis. 2013; 30: 969–976.
Dithmar S, Adams AW, Rusciano D, Grossniklaus HE. Metastatic uveal melanoma. Development of neoadjuvant treatment in animal experiments [in German]. Ophthalmologe. 2001; 98: 761–765.
Dithmar S, Albert DM, Grossniklaus HE. Animal models of uveal melanoma. Melanoma Res. 2000; 10: 195–211.
Cruz-Munoz W, Man S, Xu P, Kerbel RS. Development of a preclinical model of spontaneous human melanoma central nervous system metastasis. Cancer Res. 2008; 68: 4500–4505.
Munoz R, Man S, Shaked Y, et al. Highly efficacious nontoxic preclinical treatment for advanced metastatic breast cancer using combination oral UFT-cyclophosphamide metronomic chemotherapy. Cancer Res. 2006; 66: 3386–3391.
Francia G, Emmenegger U, Lee CR, et al. Long-term progression and therapeutic response of visceral metastatic disease non-invasively monitored in mouse urine using beta-human choriogonadotropin secreting tumor cell lines. Mol Cancer Ther. 2008; 7: 3452–3459.
Kubota T. Metastatic models of human cancer xenografted in the nude mouse: the importance of orthotopic transplantation. J Cell Biochem. 1994; 56: 4–8.
Furukawa T, Kubota T, Watanabe M, Kitajima M, Hoffman RM. A novel “patient-like” treatment model of human pancreatic cancer constructed using orthotopic transplantation of histologically intact human tumor tissue in nude mice. Cancer Res. 1993; 53: 3070–3072.
Kiguchi K, Kubota T, Aoki D, et al. A patient-like orthotopic implantation nude mouse model of highly metastatic human ovarian cancer. Clin Exp Metastasis. 1998; 16: 751–756.
Price JE, Zhang RD. Studies of human breast cancer metastasis using nude mice. Cancer Metastasis Rev. 1990; 8: 285–297.
Brandner JM, Haass NK. Melanoma's connections to the tumour microenvironment. Pathology. 2013; 45: 443–452.
Schlereth SL, Refaian N, Iden S, Cursiefen C, Heindl LM. Impact of the prolymphangiogenic crosstalk in the tumor microenvironment on lymphatic cancer metastasis. BioMed Res Int. 2014; 2014: 639058.
Goubran HA, Kotb RR, Stakiw J, Emara ME, Burnouf T. Regulation of tumor growth and metastasis: the role of tumor microenvironment. Cancer Growth Metastasis. 2014; 7: 9–18.
Weiss L. Metastasis of cancer: a conceptual history from antiquity to the 1990s. Cancer Metastasis Rev. 2000; 19: I–XI 193–383.
Fidler IJ. Models for spontaneous metastasis. Cancer Res. 2006; 66: 9787.
Figure 1
 
Conjunctival tumor growth of HGF-Cdk4R24C melanoma cells. (A) Scheme of the subconjunctival injection into the bulbar conjunctiva to imitate penetration through the basement membrane by the tumor mass. (BD) Representative clinical pictures of CM at different days postinjection when injecting 500,000 HGF-Cdk4R24C cells. (E) Clinical picture for intense neovascularization in a less-pigmented area of the tumor. (F) Increase in the clinical score (including chemosis, redness, lid distance from bulb, and proptosis) is increasing with time. Here, different numbers of cells were compared in their time course to induce clinical signs. (G) Survival of the animals, defined by proptosis as termination criterion.
Figure 1
 
Conjunctival tumor growth of HGF-Cdk4R24C melanoma cells. (A) Scheme of the subconjunctival injection into the bulbar conjunctiva to imitate penetration through the basement membrane by the tumor mass. (BD) Representative clinical pictures of CM at different days postinjection when injecting 500,000 HGF-Cdk4R24C cells. (E) Clinical picture for intense neovascularization in a less-pigmented area of the tumor. (F) Increase in the clinical score (including chemosis, redness, lid distance from bulb, and proptosis) is increasing with time. Here, different numbers of cells were compared in their time course to induce clinical signs. (G) Survival of the animals, defined by proptosis as termination criterion.
Figure 2
 
Clinicopathological characteristics of CM. (A) Conjunctival melanoma infiltrates the orbit of a HGF-Cdk4R24C mouse, detectable only after (B) exenteration. (C) Clinical picture of a CM induced by B16F10 cells. (D) Clinical score development shows an earlier clinical phenotype in the B16F10-induced CM compared with the HGF-Cdk4R24C–induced CM, associated with an earlier termination of the experiments: (E) survival curve of B16F10- and HGF-Cdk4R24C–induced CM.
Figure 2
 
Clinicopathological characteristics of CM. (A) Conjunctival melanoma infiltrates the orbit of a HGF-Cdk4R24C mouse, detectable only after (B) exenteration. (C) Clinical picture of a CM induced by B16F10 cells. (D) Clinical score development shows an earlier clinical phenotype in the B16F10-induced CM compared with the HGF-Cdk4R24C–induced CM, associated with an earlier termination of the experiments: (E) survival curve of B16F10- and HGF-Cdk4R24C–induced CM.
Figure 3
 
HGF-Cdk4R24C and B16F10 CM cells express Ki67, TRP2, and S100 in vivo. (A) The exenterated eyes were prepared for cryo sections followed by immunohistochemistry: the overview indicates that nearly all the tumor cells proliferate as determined by Ki67 staining. (B) The border between conjunctiva and tumor demonstrating that only the tumor is positive for TRP2 expression. (C) Magnification of Ki67-positive tumor cells. (D) Intratumoral section of S100-positive cells. Similarly, B16F10-induced CM is positive for Ki67 (E) and TRP2 (F) and weakly positive for S100 (G). Conj, conjunctiva.
Figure 3
 
HGF-Cdk4R24C and B16F10 CM cells express Ki67, TRP2, and S100 in vivo. (A) The exenterated eyes were prepared for cryo sections followed by immunohistochemistry: the overview indicates that nearly all the tumor cells proliferate as determined by Ki67 staining. (B) The border between conjunctiva and tumor demonstrating that only the tumor is positive for TRP2 expression. (C) Magnification of Ki67-positive tumor cells. (D) Intratumoral section of S100-positive cells. Similarly, B16F10-induced CM is positive for Ki67 (E) and TRP2 (F) and weakly positive for S100 (G). Conj, conjunctiva.
Figure 4
 
Quantification of peri- and intratumoral blood and lymphatic vessels: (A) In HGF-Cdk4R24C–induced CM, more peri- and intratumoral blood vessels (BVs) than lymphatic vessels (LVs) are detectable (P < 0.05 and P < 0.01). No significant differences were seen between peri- and intratumoral lymphatic vessels. (B) Example of intratumoral BVs (white arrowheads) and LVs (red arrows). (C) Example of a large peritumoral LV. (D) In B16F10-induced CM peri- and intratumoral clearly more BVs than LVs were measured (P < 0.0001). Intratumorally, very few lymphatic vessels were detected compared with peritumoral (P < 0.01). (E) Example of intratumoral BVs and LVs. (F) Example of peritumoral BVs and LVs, showing LVs mainly located at the tumor margin.
Figure 4
 
Quantification of peri- and intratumoral blood and lymphatic vessels: (A) In HGF-Cdk4R24C–induced CM, more peri- and intratumoral blood vessels (BVs) than lymphatic vessels (LVs) are detectable (P < 0.05 and P < 0.01). No significant differences were seen between peri- and intratumoral lymphatic vessels. (B) Example of intratumoral BVs (white arrowheads) and LVs (red arrows). (C) Example of a large peritumoral LV. (D) In B16F10-induced CM peri- and intratumoral clearly more BVs than LVs were measured (P < 0.0001). Intratumorally, very few lymphatic vessels were detected compared with peritumoral (P < 0.01). (E) Example of intratumoral BVs and LVs. (F) Example of peritumoral BVs and LVs, showing LVs mainly located at the tumor margin.
Figure 5
 
HGF-Cdk4R24C CM metastasizes to the draining lymph node and lung: (A) Control lymph nodes with no apparent brown pigmentation. (B) Lymphatic metastasis to the draining lymph nodes associated with slow-growing conjunctival melanoma, induced by injection of 5000 or 100,000 cells. (C) Immunohistochemical analysis of cryosectioned lymph node metastases revealed positivity for S100 and (D, E) TRP2. Scale bar: 50 μm. (F, G) Macroscopic appearance of distant lung metastasis. (HK) Immunohistochemistry confirmed the expression of S100 and TRP2 in the lung metastasis of CM. These data are representative for two animals with a total of 21 lung metastases.
Figure 5
 
HGF-Cdk4R24C CM metastasizes to the draining lymph node and lung: (A) Control lymph nodes with no apparent brown pigmentation. (B) Lymphatic metastasis to the draining lymph nodes associated with slow-growing conjunctival melanoma, induced by injection of 5000 or 100,000 cells. (C) Immunohistochemical analysis of cryosectioned lymph node metastases revealed positivity for S100 and (D, E) TRP2. Scale bar: 50 μm. (F, G) Macroscopic appearance of distant lung metastasis. (HK) Immunohistochemistry confirmed the expression of S100 and TRP2 in the lung metastasis of CM. These data are representative for two animals with a total of 21 lung metastases.
Figure 6
 
B16F10 CM metastasizes to the spleen and lung: (A) Spleen with a brown pigmentation (arrow). (B) Hematoxylin and eosin staining of the pigmented area in the spleen. (C) TRP2 positivity of the spleen cells. (D) Lung metastasis seen as multiple small brown spots. (E) Pigmented cells are positive for TRP2.
Figure 6
 
B16F10 CM metastasizes to the spleen and lung: (A) Spleen with a brown pigmentation (arrow). (B) Hematoxylin and eosin staining of the pigmented area in the spleen. (C) TRP2 positivity of the spleen cells. (D) Lung metastasis seen as multiple small brown spots. (E) Pigmented cells are positive for TRP2.
Table
 
Clinical Score to Classify Tumor Growth: Mice Were Examined for Chemosis, Redness, and Lid Distance From Bulb
Table
 
Clinical Score to Classify Tumor Growth: Mice Were Examined for Chemosis, Redness, and Lid Distance From Bulb
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×