Lymphatic Vessel Hypertrophy in Inflamed Human Tonsils

Introduction

The anatomic detail of the human palatine tonsil lymphatic vascular system is not well understood despite its potential contribution to fluid balance, immunity, and tumor metastasis. The palatine tonsil is a partially encapsulated component of Waldeyer's ring of lymphoid tissue in the oropharynx. The tonsils are one of the first lymphoid aggregates to encounter pathogens that enter via the oral cavity. The function of the tonsil is not well understood, although it is presumed to contribute to immunity as it contains leukocytes and secretes immunoglobulin. Similar to lymph nodes, the tonsil contains a well-organized arrangement of follicles containing B cell-rich germinal centers adjacent to marginal or stromal regions. The tonsil is fundamentally distinct from other lymphoid aggregates, such as lymph nodes, in that there are no readily identifiable sources of afferent lymph. This anatomy suggests that lymphatic capillaries might be specialized to optimize fluid and leukocyte transport from within the microenvironment of the tonsil. This is different from the well-accepted paradigm that illustrates afferent lymphatic vessels providing lymph to regional lymph nodes en route to the venous circulation. ¹ Using casting techniques and electron microscopy the tonsil lymphatic vascular system has been described in rabbits, dogs, and in humans. Lymphatic vessels have been identified in the marginal or stromal regions, adjacent to the follicles, and under the epithelium,^2,3 This anatomy has not been defined using lymphatic specific approaches.

The association between inflammatory and malignant diseases and the lymphatic vasculature system has generated an interest in understanding the structure and function of the lymphatic system. Although it has not been demonstrated in clinical studies, the results of several preclinical models have demonstrated that the lymphatic system grows and remodels, a process termed lymphangiogenesis, in response to at least certain types of inflammatory stimuli,^4,5

In the last 10 years, antibodies reagents and microscopy techniques have been developed to detect the lymphatic vasculature in preclinical models. For example, rabbit antibodies to the hyaluronic acid receptor, LYVE-1, have been shown to detect mouse or human lymphatic vessels reliably.^6,7 The adhesion molecule PECAM-1 (CD31) is expressed on lymphatic and more significantly on vascular endothelial cells at the junctions between adjacent cells. Immunoreactivity with antibodies to LYVE-1 or CD31 and conventional immunofluorescence (IF) or confocal laser scanning microscopy (LSM) techniques were used to identified the lymphatic or blood vasculature, respectively, in human pediatric tonsil tissue. Remarkably conserved lymphatic vessel architecture was observed in noninflamed tonsil tissue. Lymphatic capillaries surrounded but did not extend into germinal centers.

Confocal LSM and image analysis acquires serial images in the ‘z’ plane that are stacked to create a three-dimensional reconstruction useful for evaluating vessel morphology and cell surface molecules that have spatially restricted expression. Discontinuous expression of the adhesion molecules CD31 and VE-cadherin is a phenotype of initial lymphatic capillaries.^8,9 This expression pattern is hypothesized to facilitate fluid and cellular transport in lymphatic endothelial cells. This phenotype was detected in some of lymphatic vessels, suggesting the presence of initial lymphatic capillaries in human tonsil tissue. A dramatic increase in lymphatic but not blood vessel density and complexity was identified in inflamed tonsil tissue. The results of this study describe the spatial organization of the human tonsil lymphatic system, the discontinuous expression of CD31 and VE-cadherin in human lymphatic capillaries, and a remarkable change in lymphatic morphology in inflamed tonsil tissue.

Materials and Methods

Results

Germinal centers and marginal or stromal regions are well visualized in human tonsil tissue stained with standard hematoxylin and eosin techniques (Fig. 1A). The lymphatic and blood vasculature was detected in noninflamed tonsil tissue using IF microscopy. Staining with antibodies to CD31 (Fig. 1B) or LYVE-1 (Fig. 1C) identified the blood or lymphatic vasculature, respectively, within the marginal zone of the tonsil and surrounding the germinal center.

OPEN IN VIEWER

FIG. 1.

(A) Lymphatic vessels identified within human tonsil using IF microscopy. Human tonsil was stained with standard hematoxylin and eosin techniques. (B) Multiple germinal centers (GC) and follicles were identified. Marginal and stromal rich regions can be seen separating the follicles (low power). Noninflamed tonsil tissue was stained with antibodies to CD31 to identify the blood vasculature. Blood vessels were detected within the marginal zone of the tonsil and surrounding the germinal centers (100 ×). (C) Noninflamed tonsil tissue was stained with antibodies to LYVE-1 to identify the lymphatic vasculature. Lymphatic vessels were detected within the marginal zone of the tonsil and surrounding but not extending into the germinal centers (100 ×). (See Lymphatic Research and Biology at www.liebertonline.com for color figure.)

Confocal LSM and image analysis was used to visualize the complex morphology of the blood and lymphatic vasculature within the noninflamed tonsil tissue. Lymphatic vessels identified by LYVE-1 staining in noninflamed tonsil tissue were abundant, irregularly shaped, and in many cases contoured to the edge of the germinal center (Fig. 2). Blood vessels were detected in the marginal regions, and thin CD31-positive LYVE-1 negative vessels-like structures were detected within the germinal centers. They are presumably endothelial extensions of blood vessels and their significance is unknown. High magnification demonstrated the expression of LYVE-1 at the perimeter of the lymphatic endothelial cells outlining an ‘oak leaf’ pattern (Figs. 3A and 3B). ¹⁰ Discontinuous expression of CD31 (Fig. 3A) and VE-cadherin (Fig. 3B) was identified in both marginal lymphatics and in lymphatic vessels adjacent to the germinal center, suggesting the presence of initial lymphatic capillaries in noninflamed tonsil tissue. A restricted spatial pattern of CD31 or VE-cadherin expression on lymphatic vessels or within the tonsil could not be identified.

OPEN IN VIEWER

FIG. 2.

Lymphatic vessels in noninflamed tonsil tissue are relatively large, irregular, and surround the germinal center (GC) periphery. 100 μm noninflamed tonsil specimens were stained with antibodies to LYVE-1 or CD31 to detect the lymphatic or blood vasculature system, respectively. Confocal LSM and image analysis was used to acquire the data. Isotype specific control samples generated no specific signal (data not shown). (See Lymphatic Research and Biology at www.liebertonline.com for color figure.)

OPEN IN VIEWER

FIG. 3.

Discontinuous expression of CD31 and VE-cadherin in LYVE-1 positive lymphatic vessels. 100 μm noninflamed tonsil specimens were stained with antibodies to LYVE-1 (red) and co-stained with antibodies to CD31 (green; A) or VE-cadherin (green; B). Confocal LSM microscopy and image analysis were used to visualize expression (630 ×). Isotype specific control samples generated no specific signal (data not shown). (See Lymphatic Research and Biology at www.liebertonline.com for color figure.)

A qualitative increase in lymphatic vessels in inflamed compared to noninflamed tonsil tissue was detected by IF microscopy (Figs. 4A–4D). This analysis revealed a remarkable degree of stromal hypertrophy in the inflamed compared to noninflamed tonsil tissue. Confocal LSM and image analysis demonstrated a similar qualitative increase in lymphatic density and complexity in inflamed (Figs. 5A–5C) compared to noninflamed (Figs. 5D–5F) tonsil tissue. Quantitative analysis using Matlab software demonstrated a statistically significant increase in lymphatic microvascular density in inflamed tonsil tissue compared to noninflamed tissue without inflammation (p = 0.03). A similar analysis demonstrated no significant difference in the blood microvascular density in inflamed tonsil tissue compared to noninflamed tonsil tissue without inflammation (Fig. 6).

OPEN IN VIEWER

FIG. 4.

Lymphatic hypertrophy within the marginal zone of inflamed tonsil tissue. 8 micron noninflamed (A and C) or inflamed (B and D) tonsil specimens were stained with antibodies to CD31 (A and B) or LYVE-1 (C and D) and reactivity was detected with IF microscopy (100 ×). Representative images are shown of more than 6 tonsil samples. (See Lymphatic Research and Biology at www.liebertonline.com for color figure.)

OPEN IN VIEWER

FIG. 5.

Lymphatic hypertrophy in inflamed tonsil tissue. 100 μm specimens were obtained from inflamed (A–C) or noninflamed (D–F) tonsils and stained with antibodies to LYVE-1 (red) and CD31 (green) to detect the lymphatic or blood vasculature system, respectively. Confocal LSM microscopy and image analysis were used to visualize expression. There is a qualitative increase in lymphatic density in inflamed compared to noninflamed tonsil tissue. Representative images used in quantitative Matlab evaluation (Fig. 6) are shown. (See Lymphatic Research and Biology at www.liebertonline.com for color figure.)

OPEN IN VIEWER

FIG. 6.

A statistically significant difference was detected in lymphatic but not blood vasculature density in inflamed tonsil tissue. The percentage of red (lymphatic) or green (blood) pixels for each tonsil image was calculated and is shown. The statistical difference between the two groups was calculated using Student's t test.

Conclusions

The results of this study demonstrate that the lymphatic vasculature is abundant and well organized within the marginal regions of the tonsil tissue but not within the germinal centers. The function of the tonsil is not known with certainty; however, it is suspected to contribute to immunity. The tonsil is one of the first lymphoid aggregates to encounter pathogens that enter the oral cavity and it is composed of leukocytes organized in germinal centers and stromal tissue. Interestingly, the tonsils do not have afferent lymphatics, a feature that suggests that tonsil lymph is generated intrinsically. The tonsil anatomy may be conceptualized as distal or terminal lymphatic capillaries positioned in intimate contact with lymphoid tissue. This is dissimilar to the conventional schemata of afferent lymphatic vessels draining into lymphoid aggregates (lymph nodes). The discontinuous expression of CD31 and VE-cadherin in human tonsil lymphatic capillaries suggested the presence of initial lymphatics capillaries, vessels hypothesized to have attributes of fluid and solute transport.^8–10 Efferent tonsil lymph may generated intrinsically by passage of fluid, antigen, or cells conditioned by the constituents of the germinal centers into the initial lymphatic capillaries within the tonsil. This unique anatomic relationship could contribute to local and peripheral immunity. ¹¹

Preclinical corneal and tracheal models have demonstrated lymphangiogenesis in response to inflammatory stimuli.^4,5 Inflammation-mediated lymphangiogenesis is hypothesized to occur clinically; although this principle has not been tested in human subjects. Lymphatic morphology and density was increased in inflamed pediatric tonsil tissue compared to noninflamed tonsil tissue. IF microscopy and confocal LSM and image analysis techniques demonstrated a qualitative increase in lymphatic density in inflamed compared to noninflamed tissue. These results were consistent with a quantitative analysis that demonstrated a statistically significant increase in lymphatic but not blood vasculature in inflamed compared to noninflamed tonsil tissue. Additionally, remarkable stromal hypertrophy was detected, a common event in progressive inflammatory conditions. The findings reported here are consistent with the proposed model based on the results of preclinical studies demonstrating that certain inflammatory conditions stimulate lymphangiogenesis. Interestingly, no detectable increase in blood vascular density was identified in inflamed tonsil tissue. This was unexpected, as in many preclinical experimental systems angiogenesis occurs in concert with lymphangiogenesis.^4,5 Inflammation-mediated lymphangiogenesis but not angiogenesis may be unique to the tonsil microenvironment or the pathogenesis of tonsillitis which in practice and in this study is often not identified specifically. Important questions to address are 1) whether lymphangiogenesis is common to all types of inflammatory responses or represents a subset of inflammatory processes; and 2) how lymphangiogenesis alters host immunity and the course of clinical disease.

One potential limitation of this study is subject group allocation and the disease state that is represented in each group. Nonhypertrophic noninflamed pediatric tonsil tissue is difficult to obtain, and this study does not include such a group. This study was designed such that tonsil inflammation (erythema and exudate) was identified at the time of tonsillectomy in individuals with a history of tonsillitis. A correlation between Streptococcal infection or a history of Streptococcal infection could not be made reliably. Noninflamed hypertrophic tonsils were identified by the absence of erythema, and mucopurulence. These patients had obstructive symptoms, for instance snoring, sleep apnea, and potentially a history of hypoxia. It is recognized that the etiology of tonsil hypertrophy is poorly understood and potentially could involve inflammation. Collectively, these issues may explain the variance in our statistical analysis.

The results of these studies establish useful microscopy techniques to visualize the human lymphatic microvascular system in lymphoid tissue. A highly organized lymphatic vasculature comprised in part of initial lymphatic capillaries was demonstrated to increase in complexity in inflamed tonsil tissue. These clinical findings are a prerequisite to understanding how the lymphatic vasculature interfaces with lymphoid tissue at the cellular and molecular level. The effects of lymphangiogenesis on the extracellular matrix, immunity, and end organ function are important unresolved questions. Understanding these principles may provide important insight towards the mechanism of progressive inflammatory disease.