The lymphatic system is widely distributed in the heart. It originates from the lymphatic capillaries in the subendocardium, passes through the lymphatics in the myocardium and collecting lymphatics in the subepicardium, and eventually forms the lymphatic trunks and joins the venous circulation (Shimada et al.,1990). Several lines of evidence have demonstrated that mechanical or reactive obstruction of the cardiac lymphatics causes pathological changes in the myocardium, endocardium, cardiac valves, and conduction system (Gloviczki et al.,1983).
Myocardial fibrosis is a significant primary change resulting from hypertension (Cooper et al.,1990; Koren et al.,1991; Brilla et al.,1993). Myocardial fibrosis involves deposition of dense collagen fibers among the myocardium and around the cardiac blood vessels, and results in subsequent symptoms including eventual heart failure. The mechanism for myocardial fibrosis is still controversial. However, there is now a wealth of evidence suggesting that lymphedema may be one important factor that is closely related to myocardial fibrosis (Laine and Allen,1991; Stolyarov et al.,2002; Kong et al.,2006). Lymphedema results in the increase of interstitial pressure (Laine and Allen,1991) and eventually causes synthesis and accretion of collagen fibers. TGF-β1 may be involved in this process (Lijnen et al.,2000; Kuwahara et al.,2002). Lymphedema is closely associated with the function of the cardiac lymphatics. However, though hypertensive cardiac blood vessels have been well studied, little is known about the relationship between the cardiac lymphatics and myocardial fibrosis during the condition of hypertension.
This study was undertaken to investigate the structural changes in the cardiac lymphatics during hypertensive fibrosis. Currently, there are many different types of animal models for hypertension; we chose salt-sensitive rats as follows: Dahl S rats as the experimental group and salt-resistant rats as follows: Dahl R rats as the control group. The hypertension animal model was developed by feeding the Dahl S rats with a high salt diet, and structural changes in the cardiac lymphatics under conditions of hypertension were investigated. A better understanding of the cardiac lymphatics will hopefully lead to treatment of myocardial fibrosis.
MATERIALS AND METHODS
Hypertension Animal Model
Ten Dahl S rats and 10 Dahl R rats, male, 8 weeks-old, bought from the Jiudong Company, were fed on 8% NaCl diets for 1 month. Systolic blood pressure in the tail artery was measured every week by the tail cuff method. After the rats were sacrificed under deep anesthesia with an injection of sodium pentobarbital (50 mg/kg), heart perfusion with 0.9% buffered saline solution was carried out under 103–120 mmHg from the apex of heart, the hearts were then removed.
For light microscopy observation, two samples of both atriums and ventricles of every rat were obtained horizontally and transmurally, in total 40 blocks were fixed in 4% paraformaldehyde to prepare tissue sections of 4 μm thickness. For transmission electron microscopy (TEM) observation, section of both the atriums and ventricles of every rat were obtained horizontally and divided into small blocks (1 mm × 1 mm × 2 mm), samples were all immersed in 2% glutaraldehyde solution at 4°C. For scanning electron microscopy (SEM) observation, two samples of both atriums and ventricles of every rat were obtained horizontally and transmurally, in total 40 blocks were immersed in 2% glutaraldehyde solution at 4°C.
Van Gieson Staining
Paraffin sections (4 μm) were deparaffinized in dimethylbenzene, rehydrated through a graded alcohol series, spent 15 min in Weigert's iron hematoxylin and 3 min in Van Gieson's solution, were counterstained in hematoxylin, dehydrated, and mounted.
Sections were treated with 0.3% H2O2 in PBS for 15 min to inhibit intrinsic peroxidase activity and 10% blocking solution for 20 min to prevent nonspecific antibody binding. They were then incubated overnight at 4°C with either rabbit anti rat atrial natriuretic peptide (ANP) polyclonal antibody (diluted 1:100; YII-Y330-EX, purchased from Cosmo Bio), rabbit anti rat connexin43 polyclonal antibody (diluted 1:100; ab66151, purchased from Abcam), or rabbit anti rat LYVE-1 monoclonal antibody (diluted 1:200; sc-80170, purchased from Santa Cruz Bio). After rinsing in PBS, slides were then incubated for 25 min at room temperature with polyHRP goat anti-rabbit IgG. Slides were rinsed again with PBS and were stained with DAB, counterstained in hematoxylin, dehydrated, and mounted. Control immunostaining was carried out by the same procedure in which the first antibody was replaced by nonimmunized serum.
Transmission Electron Microscopy
Specimens were postfixed for 30 min at room temperature in 1% OSO4, and dehydrated in a graded series of ethanol and embedded in epoxy resin. Semi-thin sections (1.0 μm thick) were stained with 1% toluidine blue for light microscopy. In different rats, the numbers of blocks containing lymphatics were not same, to insure the selection bias minimized, 6 blocks of ventricle containing lymphatic and 4 blocks of atrium containing lymphatic from every rat were chose randomly, in total 200 blocks were prepared for ultra-thin slice, ultra-thin sections (80–100 nm thick) were stained with uranyl acetate and lead citrate and examined under TEM.
Scanning Electron Microscopy
Specimens were further treated with 1% Tween for 3 hr at 37°C and were then washed thoroughly in distilled water, immersed in a 1% aqueous solution of tannic acid for 2 hr, postfixed in cacodylate buffered 1% osmium tetroxide for 2 hr, dehydrated in graded concentrations of ethanol, and then dried by means of the t-butyl alcohol freeze-drying method. The specimens were then sputter-coated with gold and examined in an H-800 scanning electron microscope.
The results of Van Gieson staining and immunohistochemical staining were examined by a professional pathological doctor using the Image-Pro Plus 5.0 image analyze system. Five views on every slides were chosen randomly and the area density of positive expression (positive area/total area) was quantified. Twenty views of TEM with the same magnification were randomly chosen from the Dahl S and Dahl R groups. The total numbers, area density (μm2/μm2), and number density (N/μm2) of uncoated vesicles in lymphatic endothelium were analyzed. All of the data are presented as mean ± SD and were analyzed by SPSS 10.0 statistical analysis software. A value of P < 0.05 indicates statistical significance.
Hypertension Animal Model
After the rats were fed on 8% NaCl diets for 1 month, the systolic blood pressure of rats was measured, the maximum, minimum, and mean systolic blood pressure in the Dahl S group were 244, 228, and 232 mmHg, respectively. The maximum, minimum, and mean systolic blood pressure in the Dahl R group were 135, 121, and 126 mmHg, respectively. The Dahl S rats had higher blood pressure than Dahl R rats; over 180 mmHg is regarded as the hypertension standard in rats by most researchers (Dal Canton et al.,1989). Therefore, Dahl S rats were designated as the hypertension group, and Dahl R rats were the control group.
Using Van Gieson staining, the myofibrils appeared yellow and the collagen fibers were red. Rare collagen fibers were distributed around the blood vessels and among myofibrils in the Dahl R group (Fig. 1a). In the Dahl S group, the collagen fibers were increased in the myocardium, especially around the coronary artery where the collagen fibers became dense and thick. In addition, dense collagen fibers were deposited among the myofibrils in a cord shape (Fig. 1b). The area density of collagen fibers in the Dahl S group (0.193 ± 0.012) was significantly higher than that in the Dahl R group (0.122 ± 0.007, Fig. 2), P < 0.05.
Expression of atrial natriuretic peptide in the Atrium
Expression of ANP was observed in the cytoplasm of myofibrils in the atrium but not the ventricle in the two groups. In the Dahl R group, the immunoreactive particles were mainly near the nucleus of myofibrils (Fig. 1c). In contrast, the immunoreactive particles in the Dahl S group were found in the whole cytoplasm of myofibrils and showed a patch-like shape (Fig. 1d). The area density of expression of ANP in the Dahl S group (0.121 ± 0.007) was higher than that in the Dahl R group (0.049 ± 0.032, Fig. 2), P < 0.05.
Expression of Connexin43 in the Intercalated Disk
Immunoreactive particles of connexin43 appeared brown color and were in the intercalated disks of adjacent myofibrils (Fig. 1e,f). In the Dahl S group, the immunoreactive particles were decreased in number and were absent in some myofibril borders when compared with the Dahl R group. The area density of expression of ANP in Dahl S group (0.0045 ± 0.087) was lower than that in the Dahl R group (0.0243 ± 0.042, Fig. 2), P < 0.05.
Distribution and Shape of the Cardiac Lymphatics
LYVE-1 is a special type of marker for the lymphatics. The cardiac lymphatic wall stains brown with LYVE-1 antibody. We found that blood vessels showed negative results for LYVE-1. The lymphatics are distributed in every layer of the rat heart, including the subendocardium, myocardium, and subepicardium. Most of the lymphatics in the Dahl S group lost their normal shape and were dilated (Fig. 3a–c), especially those near the coronary arteries, which always showed a round shape. The lymphatics in the Dahl R group still maintained an irregular shape and were represented by a narrow lumen (Fig. 3d–f).
Ultrastructure of the Cardiac Lymphatics
Using TEM, damage of cellular organelles such as the mitochondrion could be observed in the endothelium of lymphatic capillaries and collecting lymphatics in the Dahl S group (Fig. 4d,e). Two types of vesicles were present in both of the two groups; uncoated vesicles and coated vesicles. Coated vesicles were infrequently seen in the two groups. In contrast, plenty of uncoated vesicles with different diameters were found in the endothelial cytoplasm in the Dahl S group. Using 20 views with same magnification in the two groups, the number, area density, and number density of uncoated vesicles in the Dahl S group were higher than those in the Dahl R group (Table 1). In addition to the coated and uncoated vesicles, vacuoles with larger diameter occurred only in the collecting lymphatics (Fig. 4a,b) and lymphatic capillary (Fig. 4c) in Dahl(S) group, there were more uncoated vesicles in the lymphatic endothelium in the Dahl S group (Fig. 4f) than in the Dahl R group.
Table 1. Uncoated vesicles in lymphatic endothelium of Dahl(S) and Dahl(R) rats
Number of views
Number of vesicles
Area density (μm2/μm2)
Number density (N/μm2)
0.34 ± 0.05
41.42 ± 2.64
0.16 ± 0.06
17.67 ± 3.02
Increased fibroblast and collagen fibers were assembled near the lymphatics in the Dahl S group (Fig. 5a), and in some cases, collagen fibrils were connected with the lymphatic endothelium directly (Fig. 5b). With regard to collagen fibers around the lymphatics, the Dahl S group mostly had thin collagen fibrils with a diameter of ∼30–40 nm (Fig. 5c) and the Dahl R group had thick collagen fibrils with a diameter of ∼50–60 nm (Fig. 5d). In the Dahl S group, thin threads could be seen protruding from the collagen fibrils (Fig. 5e). In contrast, the border of collagen fibrils in the Dahl R group were smooth (Fig. 5f).
Active exocytosis to the lymphatic lumen (Fig. 6a) and a large amount of cytoplasmic processes towards the lymphatic lumen (Fig. 6b) could also be seen in the lymphatic endothelium in the Dahl S group. Anchoring filaments lay around the lymphatics and connected with the endothelium and collagen fibers in the Dahl S group, and damage in certain parts of the lymphatic endothelium was observed (Fig. 6c). There were mainly three types of intercellular junctions in the lymphatic endothelium in the two groups, including the end-to-end type, overlapping type, and interdigitating type. However, open type intercellular junctions were rarely seen and there was no difference in the number of these junctions between the two groups. Using SEM, similar results to the TEM were obtained. In the Dahl S group, lots of collagen fibrils that connected with the lymphatic endothelium directly and cytoplasmic processes towards the lumen were observed (Fig. 7a,b) when compared with sparse collagen fibrils around the lymphatics in the Dahl R group (Fig. 7c,d).
Cardiac Lymphatic Fibrosis in the Hypertensive Heart
Extensive cardiac fibrosis was found in the myocardium in the hypertensive group. Myocardial fibrosis is one type of significant primary changes that result from hypertension (Cooper et al.,1990; Koren et al.,1991; Brilla et al.,1993). The extracellular matrix of the heart is mostly composed of collagen fibers, and they consist of mainly collagen type I and type III (Souza,2002). Collagen fibers function to form the framework of the heart and resist tensile force. Under conditions of hypertension, the synthesis of collagen fibers increases and they are deposited among the myocardium and around the blood vessels. The reason for cardiac fibrosis remains to be determined, but lymphedema may be one important factor that is closely related to myocardial fibrosis (Laine and Allen,1991; Stolyarov et al.,2002; Kong et al.,2006). Cardiac fibrosis makes the heart “harder” and has a negative influence on systole and diastole of the heart (Wu et al.,2004). Little is known about the effect of fibrosis on the cardiac lymphatics. However, there have been some reports about its effect on other organs (Yamauchi et al.,1998; Herd-Smith et al.,2001; Ji et al.,2004). Our study demonstrated that the cardiac lymphatics in the Dahl (S) group also presented with fibrosis in the ventricle and atrium, which was similar to that around blood vessels. Increased fibroblasts and collagen fibers were found around the lymphatics in the Dahl S group. Moreover, the collagen fibers were mainly composed of thin collagen fibrils, in contrast to thick collagen fibrils in the Dahl R group. Thin threads protruded from the collagen fibrils in the Dahl S group, when compared with the smooth border of collagen fibrils in the Dahl R group. This suggests that hypertension also induces the fibrillogenesis around the cardiac lymphatics as well as in the blood vessels in the Dahl S group, it was in agreement with other researchers's reports that fibrillogenesis was marked by the collagen fibrils with increasing diameter (Rentz et al.,2007).
In normal tissue, only a few collagen fibrils are closely associated with the lymphatic endothelium. For the most part, they are separated from the lymphatic endothelium by anchoring filaments. However, TEM and SEM demonstrated that in the Dahl S group numerous collagen fibrils were connected with the abluminal surface of the lymphatics directly. When compared with anchoring filaments, collagen fibrils represent more durability. Therefore, the collagen fibrils that connect with the lymphatics may be more available on maintaining the framework of lymphatics. The lymphatics in the Dahl R group still appeared irregular in shape and had a narrow lumen in the myocardium. However, the lymphatics in the Dahl S group were always enlarged in a round or oval shape, especially those near the arteries. The reason for this finding may be due to the decrease in drainage of the lymph. On the other hand, it may be due to numerous collagen fibers that connected with the lymphatic endothelium, which would resulted in resistance to the gradually elevated interstitial fluid pressure. Because myocardial fibrosis was more intense near the arteries, the lymphatics near the arteries showed a round shape more frequently than other parts in cardium.
Structural Changes of the Cardiac Lymphatics in the Hypertensive Heart
Lymphatic intercellular open junction in which adjacent cells do not come closer than 30 nm to each other is more frequent in pathological conditions (Qu et al.,2003). An increase in open junctions is seen in the small intestinal lymphatics during thoracic duct blockage (Ji and Kato,2001). However, we found no trend in the Dahl S group for an increase of open junctions. Because of the increased intraluminal pressure, the tissue fluid pushed the intercellular junction open, and this process has always been considered to be involved in anchoring filaments (Tammela et al.,2005). On the other hand, the anchoring filaments play a more important role in closing the opened intercellular junctions. The anchoring filaments function like chordae tendineae in the cardiac valve and insure that tissue fluid flows into the lymphatics in only one direction (Tortora and Nielsen,2005). We found that numerous collagen fibrils were directly connected with lymphatic endothelium in the hypertensive group. As the anchoring filaments, these collagen fibrils could enhance limiting the amount of open junctions even in high interstitial fluid pressure, this may explain why open junctions were absent in the Dahl S group.
The lymphatic endothelium contains endocytotic vesicles of both the coated and uncoated types. Uncoated vesicles are associated with transport across the lymphatic endothelium (P'Morchoe et al., 1984). Our study investigated the number, number density and area density of uncoated vesicles in Dahl S and Dahl R rats, and found that uncoated vesicles were increased in both number and area in the Dahl S group. We speculate that the increase in uncoated vesicles would restore the flow of tissue fluid to the lymphatics. We also observed another type of large vesicle or vacuole that was different from the above mentioned vesicles. Vacuoles occurred in both the lymphatic capillaries and the collecting lymphatics in the Dahl S group. Large vesicles were also found in the endothelial cytoplasm of a blocked thoracic duct (Ji and Kato,2001) and were shown to promote the transport of lymph. It is unclear whether such vacuoles elevate the efficiency of transportation across the lymphatic endothelium in hypertensive cardium.
We observed an active exocytosis phenomenon in the lymphatic endothelium in the Dahl S group and numerous cytoplasmic processes towards the lymphatic lumen, but these results were not found in the Dahl R group. Cytoplasmic processes are thought to be involved with transportation of some granules. We believe that exocytosis and the cytoplasmic process would restore the draining of tissue fluid. The mechanism of these changes in the structure of lymphatics in the hypertensive heart needs further investigation.
Cardiac Lymphedema and Myocardial Fibrosis
Recent research has shown that the dominating motive force of cardiac lymph drainage originates from not only the external systole and diastole of heart but also the internal contraction of smooth muscle of the collecting lymphatics. We demonstrated that expression of ANP was increased in the atrium in the Dahl S group compared with the Dahl R group. With regard to the effect of ANP on the lymphatics, some studies have reported that ANP seems to inhibit lymph transport through a reduction of spontaneous contractions and a marked relaxation of lymphatic smooth muscles (Ohhashi et al.,1990; Anderson et al.,1991; Atchison and Johnston,1996). We observed extensive dilatation of the collecting lymphatics in the subepicardium in Dahl S rats, which may be due to an effect of ANP on the smooth muscle of the collecting lymphatics. The dilatation of the collecting lymphatics leads to an inability of the valves to prevent retrograde lymph flow (Ji,2005), and consequently, enhanced lymphedema. Connexin43 is a type of gap junction protein, which is found in the intercalated disk of the myocardium. It is associated with the transfer of impulses between adjacent cardiac muscles (Lin et al.,2006; Kostin,2007). In our study, we found a decrease in expression of connexin43 in the Dahl S group when compared with the Dahl R group, which indicated that impulse transfer was impeded. Transmission electron microscopy of the cardiac muscle also showed serious damage such as the dissolution and rupture of myofibrils, as well as swelling and vacuole changes in the mitochondrion. These findings demonstrated obstruction of both the transmission of impulses and the contraction ability of myofibrils.
The findings in this study indicate the obstruction of lymph drainage in the hypertensive heart. Increased collagen fibers around the lymphatics also limited the opening of intercellular junctions. The drainage of lymph to the vein as well as tissue fluid to the lymphatics were all inhibited, which eventually led to lymphedema. A series of structural changes occurred in cardiac lymphatic endothelium to restore the drainage of lymph and tissue fluid, such as an increase in vesicles and occurrence of vacuoles, and active exocytosis and cytoplasmic processes. These changes maintained the balance between input and output of tissue fluid in the myocardium. If this balance is disrupted, myocardial fibrosis will become more severe.
In conclusion, our study demonstrates widespread cardiac fibrosis in the hypertensive heart of the Dahl S group. This resulted in changes in the structure and function of the cardiac lymphatics, which should restore tissue fluid drainage that was impeded by myocardial fibrosis. The cardiac lymphatics were involved in myocardial fibrosis and were affected by myocardial fibrosis. The lymphatics may be a promising target for treatment of fibrosis in the hypertensive heart.
The authors thank Ji Ruicheng, Yasuda and Kawazato of Oita university for their excellent assistance.