An Apparent Deficiency of Lymphatic Capillaries in the Islets of Langerhans in the Human Pancreas

Type 1 diabetes (T1D) is caused by a continuing destruction of the insulin-producing cells that occurs over a period of several years after diagnosis (1). Autoantibodies with affinity for β-cells and for exocrine antigens usually appear several years before the diagnosis (2,3). The incidence of insulitis, defined as the presence of more than two or five T cells infiltrating at least three islets in children (≤14 years of age) dying within 1 month after diagnosis, is 73% (4). These observations have led to the conclusion that T1D is an immune-mediated disease.

Immune cells are constantly circulating in the body to detect and evoke an immune response against invading organisms and cells recognized as nonself. Antigen-presenting cells leave the extracellular space of the affected organ via the lymphatic capillaries and accumulate in regional lymph nodes, where the encountered antigens are presented to stimulate clonal expansion of T cells with affinity for the foreign peptides presented on the individual’s own HLA.

The β-cell is one of the most metabolically active cells in the body and is critically dependent on a high supply of oxygen and nutrients from the blood. Almost every β-cell is in direct contact with a capillary that has a fenestrated endothelial cell lining to allow optimal transport through the capillary wall, and in rodents, islet blood perfusion is ∼10 times higher than in the exocrine pancreas (5). In all organs, there is a net surplus in fluid transport over the wall of the blood capillary that is correlated with the level of blood perfusion and the permeability of the capillary. This extracellular interstitial fluid (EIF) is transported from the interstitial space via the lymphatic capillaries, and disturbances in this system can lead to the formation of edema (6).

A well-organized lymphatic system is crucial for efficient immune surveillance and induction of T-cell–mediated immune responses as well as for the maintenance of physiological pressure in the interstitial space (6). An absence of lymphatic capillaries has previously been reported within the islets in rodents (7,8) and in the fetal human pancreas (9); however, to the best of our knowledge, no information is available concerning the lymphatic drainage of the islets of Langerhans in the human pancreas.

Lymph capillaries can be distinguished from blood capillaries by the expression of specific antigens (e.g., podoplanin and LYVE-1) on the endothelial cells (10,11). The present morphological study was conducted on human pancreases from subjects without diabetes and from subjects with T1D or type 2 diabetes (T2D) to describe the extent of the intraislet lymph capillaries and their relation to the insulin-producing cells and intraislet blood capillaries.

All work involving human tissue was conducted according to the principles expressed in the Declaration of Helsinki and in the Council of Europe’s Convention on Human Rights and Biomedicine. Consent for organ donation (for clinical transplantation and for use in research) was obtained from the relatives of the deceased donors by the donor’s physicians and documented in the medical records of the deceased patient. The Regional Ethics Committee in Uppsala, Sweden, approved the study according to the Act Concerning the Ethical Review of Research Involving Humans (2003:460; permit number Dnr 2009/043, 2009/371).

Biopsy specimens from 35 human pancreases were included in the study. Before islet isolation, a clamp was used to compress the main pancreatic duct at the head of the pancreas, and the tissue proximal to the clamp was taken as a specimen and stored in formalin. Donors were chosen based on factors such as weight, age, and health (Table 1). Two pancreases were obtained at the onset of T1D, as previously described in detail (12); seven were obtained from patients with longstanding T1D; eight were obtained from patients with longstanding T2D; and the remaining pancreases were collected from multiorgan donors without any known pancreatic disease, divided according to age and BMI (Table 1).

Formalin-fixed and paraffin-embedded tissues were cut into 5-μm sections. Consecutive sections were processed and labeled using a standard immunoperoxidase technique, as previously described in detail (12). With the exception of insulin, all other antigens were unmasked by heat-induced epitope retrieval. Antibodies against chromogranin A (1:150, clone LK2H10; NeoMarkers, Thermo Fisher Scientific, Inc., Fremont, CA), insulin (1:200, clone A564; Dako, Glostrup, Denmark), and synaptophysin (1:50, clone DAK-SYNAP; Dako) were used to identify islets. A monoclonal antibody against CD34 (1:50, clone EP373Y; Dako) was used to detect blood endothelial cells and antibodies against podoplanin (1:50, clone D2-40; Dako) or LYVE-1 (1:200, clone ab 36993; Abcam, Cambridge, U.K.) were used to detect lymphatic endothelial cells. Bound antibodies were visualized using EnVision or an EnVision DuoFLEX Doublestain System (both Dako) and diaminobenzidine-based substrate or 3-amino-9-ethylcarbazole (Dako). Sections were counterstained with hematoxylin, scanned, and analyzed by Aperio ImageScope and by light microscopy by two investigators who were blinded to their origin.

Optimization of the staining for lymph and blood capillaries was performed on sections of the intestinal wall and on human pancreases. Double staining of lymph and blood capillaries revealed that the two types of capillaries were distinctly and specifically stained, without overlap or background.

Larger collecting lymphatic vessels (Fig. 1A, B, M, and O) were often located in fibrotic septa between the exocrine lobules and adjacent to the ductal system of the pancreas. Lymphatic capillaries were frequently found in the exocrine pancreas (Fig. 1E, F, K, and L) located close to and around acini.

A total of 4,365 islets from 35 subjects were examined (Table 1). Lymphatic capillaries were found in only 24 islets from 9 different subjects. In 1 subject with longstanding T2D, a total of 15 islets with lymphatic capillaries were found. The pancreas of this subject showed significant fibrosis in the islets and in the exocrine parenchyma. The islets with lymphatic capillaries also showed signs of fibrosis in the other subjects, including 1 in the high BMI group, 3 in the high age group, 1 in the group with long-standing T1D, and 1 with recent-onset T1D. The lymphatic capillaries were located in fibrotic strands within the islet and were often close to blood capillaries. Fibrotic strands in islets frequently occur, and these strands are usually devoid of lymphatic capillaries (Fig. 1F) (i.e., less than 5% of these islets contain lymph capillaries). However, the rare islets with extensive fibrosis or hyalinized islets consistently contained lymphatic capillaries within these strands of nonendocrine tissue (Fig. 1I).

No lymphatic capillaries were found in the remaining 4,341 islets examined, but an extensive network of blood capillaries was seen (Fig. 1N). Lymphatic capillaries were, however, often found at the islet-exocrine interface (Fig. 1C, F–H, and J), frequently located along blood capillaries and other fibrotic structures within or close to the islet capsule.

There are few tissues in the human body lacking lymphatic capillaries and, hence, lymph drainage. The observations reported here show that in addition to the brain, the islets of Langerhans also show a pronounced deficiency of lymphatic capillaries. This finding has implications for immune surveillance and regulation of interstitial fluid transport in the endocrine pancreas.

Notably, there are marked similarities in the morphological structure of the blood capillaries in the central nervous system and in human pancreatic islets, with a unique paravascular space surrounded by a double basal membrane structure (13,14). This peculiar anatomical structure, which functions as the prevailing transport system for EIF, has been described in the central nervous system (13,15,16). Disturbances in the glymphatic pathway have been associated with accumulation of metabolic degradation products and the development of neurodegenerative diseases (15,16). Even if a similar paravascular space surrounded by a double basal membrane is present in human islets (14), no report thus far has described glymphatic transport of interstitial fluids and waste products in human islets.

EIF is continuously formed by filtration from blood capillaries (6), collected by lymphatic capillaries, and finally recirculated back to the blood via the thoracic duct. The extracellular volume in human islets has been estimated to be ∼14% of the total islet volume (17). Ligation of the thoracic duct in rats induces edema and the accumulation of inflammatory cells in the exocrine parenchyma (18), demonstrating the importance of the lymphatic system in maintaining physiological levels of EIF volume and pressure in the exocrine pancreas. Notably, no comment was made concerning a similar edema in the islets (18). Disturbances in a tentative glymphatic transport system within the islets would have implications for the development of both T1D and T2D. Remarkably, there is a fivefold increase in the accumulation of hyaluronan and hyaladherins in the pericapillary space in islets from subjects with T1D compared with control subjects without diabetes (19).

In our study, the human exocrine pancreas showed a well-developed lymphatic system, as evidenced by the frequent lymph capillaries close to the acini and the presence of larger lymph vessels in the interstitial septa of the pancreas. Drainage of the thoracic duct has previously been applied to reduce the number of circulating lymphocytes in order to induce systemic immunomodulation in autoimmune disorders and after organ transplantation; the amount of lymph collected per day was in the range of 1.5 to 2 L. When patients were examined after a meal or during a secretin test, the levels of the exocrine enzymes in the lymph rose prior to and to higher concentrations than those simultaneously measured in the blood, supporting the notion of a substantial direct delivery of these enzymes into the lymph capillaries adjacent to the acini (20,21). In line with the morphological observations reported here, corresponding measurements of insulin during an intravenous glucose tolerance test gave no support for a direct delivery of insulin into lymphatic capillaries (22).

The absence of lymphatic capillaries in most subjects examined also makes the islets a “locus minoris resistentiae” (23) for infections and dysregulated immune responses resulting from the lack of normal trafficking of immunocompetent antigen-presenting cells from the parenchyma via the lymphatic capillaries to the regional lymph nodes. The few lymphatic capillaries identified within the islets were in most cases not in direct contact with the endocrine cells; instead, the capillaries detected were found in conjunction with fibrotic strands entering the islets. Lymphatic capillaries were instead often found in the peri-islet area, close to the blood vessels supporting the islet (i.e., the site into which a tentative glymphatic transport system would empty) (14). The importance of this observation for the development of T1D is unknown and beyond the scope of this study. However, the observation of an extensive peri-islet network of lymphatic capillaries is consistent with the frequently reported predominant accumulation of immune cells in this area in subjects with recent-onset T1D (4).

Even if lymphatic capillaries in the exocrine pancreas were readily detected using antibodies to podoplanin or LYVE-1, the apparent absence of lymphatic capillaries within the islets could be a result of the absence of these proteins specifically on the lymphatic endothelial cells within the islets. It should, however, be noted that these antisera readily detected lymphatic capillaries in pancreatic endocrine tumors, including benign and malign insulinomas (24), and transplanted islets (25).

In summary, we report that there is a marked deficiency of lymphatic capillaries in the islets of Langerhans with the human pancreas, a finding that may have important implications for our understanding of the physiology of the endocrine pancreas as well as for the pathophysiology of both T1D and T2D.

Acknowledgments. The authors are grateful to Sofie Ingvast, Department of Immunology, Genetics and Pathology, Uppsala University, for excellent technical assistance.

Funding. This study was supported by grants from the Swedish Medical Research Council (VR K2011-65X-12219-15-6, K2015-54X-12219-19-4), the Nordic Insulin Fund, European Foundation for the Study of Diabetes/Novo Nordisk Foundation, the Ernfors Family Fund, Barn diabetes fonden, the Swedish Diabetes Association, the Diabetes Wellness Foundation, and JDRF.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. E.K. and O.K. designed the study, analyzed and interpreted the obtained data, and wrote and approved the manuscript. O.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.