Background
De novo lymphangiogenesis influences different pathological courses via modulating tissue fluid homeostasis, macromolecule absorption, and leukocyte transmigration [
1]. In addition, lymphatic vessels play a crucial role in a variety of human cancers [
2]. Invasion of lymphatic vessels by tumor cells and subsequent development of lymph node metastases significantly influences the prognosis of cancer patients and, therefore, represents an integral part of tumor staging. Increasing knowledge of the tumor's biological significance in lymphatics within the tumors and at the tumor periphery has greatly promoted understanding of tumor access into the lymphatic system by inducing lymphangiogenesis or by co-opting preexisting lymphatics [
2]. In contrast, impaired functioning of lymphatic vessels results in lymphedema as observed during breast cancer diagnosis and treatment [
3‐
5].
During cancer progression, a bi-directional communication is established between the tumor microenvironment (TME) and lymphatic vessels. In one direction, the lymphatic vasculature alters TME by draining the interstitial protein-rich exudate fluid (lymph) into the bloodstream. In another direction, inflammation influences the composition and pressure of TME leading to altered lymphatic vessel function.
We choose to study bladder cancer because it represents 2% of all human malignancies. Urothelial carcinoma is one of the most common cancers – it ranks fifth among all cancers in the Western world, and there are 336,000 new cases and 132,000 deaths annually worldwide [
6]. In the US alone, the American Cancer Society estimates that 50,040 men and 17,120 women will be diagnosed, and 13,060 men and women will die of cancer of the urinary bladder in 2007 [
7]. Although the role of lymphatic vessels during bladder cancer progression is remarkably unknown, invasion of lymphatics during bladder cancer has been reported [
8], whereas in prostate cancer there is a decrease in intratumoral lymphatic vessel density [
9]. More recently, Fernandez and collaborators published the first manuscript suggesting the existence of proliferating lymph vessels and, therefore, of lymphangiogenesis in bladder transitional cell carcinoma (TCC), and proposed strong correlation of higher peritumoral LVD with the presence of lymph nodes in clinically localized invasive bladder TCC [
10]. However, up to now, no animal model was available for a systematic study of lymphatic vessel density and function during bladder cancer progression.
We previously described that a transgenic mouse (κB-
lacZ) with a reporter gene (
lacZ) for NF-κB presented constitutive β-galactosidase (β-gal) activity in all lymphatic endothelial cells [
11] and that these mice serve a dual purpose by permitting both visualization of lymphatics and detection of constitutive and inducible NF-κB activity [
11]. To study in-depth the role of lymphatic vessels in bladder cancer progression, we generated a double transgenic mouse (
SV40-lacZ) by crossing the κB-
lacZ mice with a well established model of bladder cancer (UPKII/SV40T) [
12,
13]. Here, we demonstrate that these SV40-lacZ mice present an increased lymphatic vessel density during bladder cancer progression.
We also show that a new compound, coined Gd-Cy5.5, which corresponds to a recently described contrast agent (biotin-BSA-Gd-DTPA) [
14‐
16] conjugated to Cy5.5 permits dual imaging by near-infrared fluorescent (NIRF) and MRI. We, therefore, provide proof-of-concept that this contrast agent can be used for determination of lymphatic vessel function during bladder cancer progression.
Discussion
Our results with SV40-
lacZ mice indicate that in the normal bladder, a rich lymphatic vessel network is visible from the adventitia through the detrusor smooth muscle. It is characterized as a vascular network of blind ended, thin-walled capillaries that merge to larger collecting ducts, all positively stained with LYVE-1 antibody. In sharp contrast with the dense blood vessel vascularization of the bladder mucosa, we found that lymphatic vessels are absent of the sub-urothelium being located much deeper in the lamina propria (Figures
6A, 6B,
7A, and
7B) which seems to parallel the clinical findings in the human bladder [
35,
36].
Our analysis indicates an increased lymphatic density during cancer progression and a co-localization of Ki-67 with some of LYVE-1-positive lymphatics, suggesting cancer-induced lymphangiogenesis (Figure
4), as it was recently suggested in human bladder cancer [
10,
37]. However, the mechanisms for bladder lymphangiogenesis are not clear. In contrast to blood vessel angiogenesis, the mechanisms of lymphangiogenesis in general are still relatively vague [
38]. Vascular endothelial growth factor-C (VEGF-C) and VEGF-D have been implicated as specific regulators of lymphangiogenesis [
39‐
43]. Both growth factors mediate their biological activity mainly by VEGF receptor-3 (VEGFR-3, Flt-4) [
44,
45]. It remains to be determined whether VEGF-C and VEGF-D along with VEGFR-3 play a role in lymphangiogenesis during bladder cancer development. Another interesting line of research involves LYVE-1-positive tumor associated macrophages [
46] that have indeed been associated with tumor lymphangiogenesis [
47,
48]. In the eye, a stepwise mechanism of inflammation-associated
de novo lymphangiogenesis involves potential lymphatic progenitor cells [
49] derived from circulation that transmigrate through the connective tissue stroma, presumably in the form of macrophages [
49‐
51], and finally incorporate into the growing lymphatic vessels [
52]. Our present findings indicate that in addition to LYVE-1-positive lymphatic endothelial cells, during bladder cancer development, LYVE-1 positive macrophages were found in the bladder detrusor muscle isolated from SV40-
lacZ mice. Although this may be only a circumstantial finding, it opens a testable hypothesis on the role of macrophages and other inflammatory cells in bladder lymphangiogenesis. The introduction of SV40-
lacZ along with the visualization and quantification techniques described here will permit further investigation on this subject.
It has been proposed that lymphangiogenesis is correlated with tumor metastasis. Increasing knowledge of the tumor's biological significance in lymphatics within the tumors (intratumoral lymphatics, ITLs) and at the tumor periphery (peritumoral lymphatics, PTLs) has greatly promoted understanding of tumor access into the lymphatic system by inducing lymphangiogenesis or by co-opting preexisting lymphatics [
2]. Indeed, peritumoral lymphatics have also been associated with both regional metastasis and survival in bladder [
37], lung [
53], breast [
54], and prostate cancer [
55]. But the question still remains as to whether pre-existing vessels are sufficient to serve this function, or whether tumor cell dissemination requires
de novo lymphatic formation or an increase in lymphatic size. In this regard, Fernandez and collaborators reported that in the human bladder, higher intratumoral LVD correlates significantly with poor histological differentiation, and that higher peritumoral LVD showed a significant association with the presence of lymph node metastasis [
10]. Although peritumoral lymphatic vessels contribute to tumor metastasis, opposite views exist as to whether intratumoral lymphatics have any role in tumor metastasis [
56,
57]. Padera and collaborators examined functional lymphatics associated with mouse tumors expressing normal or elevated levels of VEGF-C [
58]. Although VEGF-C over-expression increased lymphatic surface area in the tumor margin and lymphatic metastasis, these tumors contained no functional lymphatics, as assessed by four independent functional assays and IHC staining [
58]. These findings suggest that the functional lymphatics in the tumor margin alone are sufficient for lymphatic metastasis and should be targeted therapeutically [
58]. Our MRI results are in agreement with those described by Pandera and collaborators [
58] in the sense that intratumor lymphatic vessels have a reduced function when compared to normal areas of the bladder.
Another question answered by the present work was whether an increase in the number of lymphatic vessels leads to increased function. For this purpose, we followed the strategy recently reviewed by Neeman and collaborators [
59]. The contrast agent introduced here was originally described by Dafni and collaborators [
14‐
16], and subsequently by Pathak and collaborators [
15], for visualization of lymphatics and determination of their function. The modification of conjugating biotin-BSA-Gd-DTPA [
15] to Cy5.5 permitted us to use NIRF and MRI to follow the dynamics of the same compound and calculate lymphatic vessel function. The advantage of using NIRF is the faster time for data acquisition. NIRF information permitted us to narrow the number of time points for subsequent MRI studies.
The proposed mechanism of visualization of BSA-Gd-Cy5.5 (~82 kDa) is based on the presence of BSA which prolongs its lifetime in circulation. Initially the probe is confined to blood vessels and it is systemically distributed, and as the time passes it extravasates to the extracellular space in points of increased vascular permeability. Both NIRF and MRI are able to detect when BSA-Gd-cy5.5 starts to accumulate in the extracellular space and when its accumulation reaches a plateau. This was done at the MRI level by following the longitudinal relaxation rates (1/T1) that correlates with increased uptake of the Gd-probe from the blood vessels into the urinary bladder extra vascular space. This phase was called the "early phase" in this manuscript. After 80 minutes of the plateau, a second series of MRI were started and followed the clearance of BSA- Gd-Cy5.5 from the extravascular by the lymphatic vessels. This was supported by ex vivo images which indicated that, at this point, most of the BSA-Gd-Cy5.5 is drained from the tissue by lymphatic vessels.
Although MRI lymphangiography does measure areas of draining and pooling instead of directly evaluating lymphatic vessel function, the delayed enhancement observed at later time points from the Gd-probe may also reflect uptake of Gd-probe in the lymphatics. In addition, this information can not be obtained any other way and clinical studies attest the validity of using lymphangiography for assessment of lymphatic vessel function [
60‐
64]. A step forward in this direction was introduced here by the use of a single probe that can be imaged by MRI and NIRF. The first indicates areas of draining and pooling and the second permitted the temporal association of images with cross-sections indicating the relative distribution of the probe between CD31-positive blood vessels and LYVE-1-positive lymphatic vessels.
As pointed out by Neeman and collaborators [
59], contrast MRI data are typically dominated by vascular permeability, which often masks the relatively slow lymphatic drain. To separate vascular leakage from the lymphatic drain, an avidin chase needs to be introduced [
14,
16] which through the rapid clearance of intravenously administered biotin-BSA-Gd-DTPA, allowed them to experimentally track interstitial convection and lymphatic drain in the absence of continuing vascular leakage [
14,
16]. Our present MRI results are not corrected for vascular leakage. Therefore, future experiments will take into consideration this point for a more detailed analysis of lymphatic vessel function using the advantage of the presence of biotin in the Gd-Cy5.5 molecule.
Another advantage of NIRF was to permit the collection of tissues exhibiting fluorescence. In this regard, we were able to further pursue the morphology of lymph sacs using dextran-Cy5.5. This will allow us in the near future to determine the contractility of lymph sacs and abdominal lymphatic vessels. The rhythmic activity observed in the abdominal lymphatics draining the urinary bladder is in agreement with recent results that indicated that mesenteric lymphatic vessels have a true pacemaker mechanism [
65,
66]. The experiment described in Figures
10,
11,
12 will permit the capture and evaluation of large lymphatic function in health and disease.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors read and approved the final manuscript. MRS conceived study and experimental design, participated in NIRF experiments, performed IHC analysis, interpreted the results, and drafted the manuscript; RT performed MRI experiments, participated on the design, results interpretation, and helped drafting the manuscript; NS synthesized the d-Cy5.5; AA performed MRI experiments and interpretation of MRI results; MN consulted MRS regarding the proper contrast agents, participated in the design of MRI experiments and interpretation of the results, and helped drafting the manuscript; CAD maintained the genotyping and animal colony, and performed image analysis; CS performed animal experiments and whole mount IHC; JM participated on the design of immunohistochemistry protocols and choice of antibodies, performed the immunohistochemistry and confocal microscopy analysis; SM developed kB-lacZ mice, participate in the design of the double transgenic, and helped drafting the manuscript; X-RW developed UPKII-SV40T mice, participate in the design of the double transgenic, and helped drafting the manuscript; and RS participated in the experimental design, performed NIRF experiments and data interpretation.