This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Marron, K. Articles by Wharton, J. Search for Related Content PubMed PubMed Citation Articles by Marron, K. Articles by Wharton, J.

(Circulation. 1996;94:368-375.)
© 1996 American Heart Association, Inc.

Articles

Innervation of Human Atrioventricular and Arterial Valves

Kevin Marron, BSc; Magdi H. Yacoub, FRCS, DSc; Julia M. Polak, MD, DSc, FRCP; Mary N. Sheppard, MD, MRCP; David Fagan, MD, FRCP; Bruce F. Whitehead, MB, MRCP; Marc R. de Leval, MD, FRCS; Robert H. Anderson, MD, FRCP; John Wharton, PhD

the Department of Histochemistry (K.M., J.M.P., J.W.), Royal Postgraduate Medical School, Hammersmith Hospital, London; Heart Science Centre (M.H.Y.), Harefield Hospital, Middlesex; the Department of Histopathology (D.F.), Queens Medical Centre, University of Nottingham Medical School; the Cardiothoracic Unit (B.F.W., M.R. de L.), Great Ormond Street Hospital for Children, London; and the Departments of Pathology (M.N.S.) and Paediatrics (K.M., R.H.A.), National Heart and Lung Institute, London, United Kingdom.

Correspondence to Dr John Wharton, Department of Histochemistry, Royal Postgraduate Medical School, Du Cane Rd, London W12 ONN, UK.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background Limited information exists on the innervation of human cardiac valves and the relationship of nerve fibers and terminals with functional elements within leaflets.

Methods and Results We examined human AV and arterial valves, obtained postmortem and at surgery, using quantitative immunohistochemical, histochemical, and confocal microscopic techniques. Significant differences in nerve density and distribution were found both between and within cardiac valves. Nerve density within the anterior leaflet of the mitral valve, for example, was twofold greater than that in the posterior leaflet (P<.001). Nerves within the AV valves were situated in the atrial layer and extended over the proximal and medial portions of the leaflets, whereas those in the arterial valves were situated in the ventricular layer. No nerves reached either the free edge or the fibrous core of the leaflets. The arterial valves displayed a similar density of innervation, except for the noncoronary leaflet of the aortic valve in which the innervation was attenuated (P<.01). The innervation of aortic valvar leaflets was age dependent. Nerve terminal arborizations, arising from myelinated nerves and exhibiting variable morphology, were detected in all four cardiac valves and in some tendinous cords. Nerve terminals exhibited either acetylcholinesterase activity or tyrosine hydroxylase and neuropeptide Y immunoreactivity. Varicose nerve fibers occurred in close physical proximity to valvar endothelial, smooth muscle, and fibroblast cells.

Conclusions Human cardiac valves have distinct patterns of innervation that comprise both primary sensory and autonomic components. The presence of distinct nerve terminals and the close association of varicose nerve fibers with endothelial, smooth muscle, and fibroblast cells suggest a possible neural involvement in the control of valvar function.

Key Words: valves • nervous system, autonomic • receptors • peptides • aging

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
The valves of the heart have generally been considered passive structures, moving only in response to the hemodynamic forces generated by cardiac contractions.^{1 2} The valves, particularly the mitral and tricuspid valves, however, are now known to be capable of independent contraction during various stages of the cardiac cycle and may therefore be capable of compensatory adjustments that influence the timing and effectiveness of valvar closure.^{3 4 5 6 7 8} This control may be achieved by a combination of movement of the surrounding myocardium and annulus relative to the valve,^{3 4 5} by contraction and relaxation of papillary muscles,⁷ or through neural mechanisms.^{8 9 10 11 12} Myocardium within the proximal portion of the valvar leaflet often is reported in the AV valves and may help to control movement.^{9 12 13 14} Smooth muscle cells also have been demonstrated in mammalian valve leaflets.^{12 13 15} Valvar fibroblasts may express a smooth muscle phenotype and exhibit a contractile response to both adrenaline and angiotensin II.¹¹ Similarly, canine mitral valves have been shown to possess an intrinsic contractility with appropriate responses to administration of noradrenaline, acetylcholine, and electric stimulation.^{8 9} Evidence for the presence of nerve fibers in the AV valves was provided by the anatomic studies of Woollard¹⁶ and Nettleship,¹⁷ who used histochemical techniques such as methylene blue and silver staining to examine cardiac innervation, and by more recent investigations that used formaldehyde-induced fluorescence, acetylcholinesterase (AChE) staining techniques,^{14 18 19 20} immunocytochemistry,²¹ and electron microscopy.^{9 10 11} Although all of these studies demonstrated the presence of a network of nerve fibers in either the AV or arterial valves, only a minority noted the presence of multipolar nerve processes or terminals.^{18 19 20} Furthermore, the innervation of individual leaflets within each valve was not distinguished. We have now determined the distribution and neurochemistry of the innervation of the individual valvar leaflets in the human heart, investigated the association of nerve fibers and terminals with immunohistochemically defined cell types, and examined the innervation of the tendinous cords associated with leaflets of the AV valves.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Tissue Processing
Human mitral, tricuspid, pulmonary, and aortic valves were obtained postmortem, within 24 hours of death, from a total of 15 individuals with normal hearts (age, 3 months to 80 years; Table 1). Other AV and arterial valves were obtained at surgery from the recipient hearts of 2 children undergoing transplantation for cystic fibrosis. Valves obtained from patients 3 months to 21 years of age (n=12) were used to assess variations in innervation density between leaflets. The effect of age on nerve density was assessed quantitatively in both AV (3 months to 61 years; n=10; Table 1) and arterial (3 months to 80 years; n=8; Table 1) valves. To collect tissues, we followed the ethical standards of the institutions in which they were obtained. All hearts were fixed by immersion in modified Bouin's solution for 24 hours at 4°C and subsequently rinsed in several changes of PBS (0.1 mol/L, pH 7.2) containing 15% (wt/vol) sucrose and 0.1% (wt/vol) sodium azide as previously described.²² Entire valves, including the tendinous cords and papillary muscles, were excised from surrounding tissues and processed for indirect immunofluorescence and AChE staining as whole-mount preparations, thus allowing the nerve plexus and nerve terminal arborizations to be preserved intact and analyzed in their entirety.

View this table: [in this window] [in a new window]   Table 1. Clinical Features

Immunofluoresence and AChE Staining
A modified indirect immunofluoresence method was used to demonstrate nerves, endothelial and smooth muscle cells, and fibroblasts either alone or in combination in whole-mount preparations.^{22 23} The innervation as a whole was demonstrated by use of antisera to the general neuronal marker protein gene product 9.5 (PGP 9.5), which is a ubiquitin carboxyl-terminal hydrolase isozyme expressed in both autonomic and sensory nerves.^{22 23} Subpopulations of nerves and nerve terminals were distinguished immunohistochemically with specific antisera to peptides and enzymes that predominate in postganglionic sympathetic (neuropeptide Y and tyrosine hydroxylase), parasympathetic (vasoactive intestinal polypeptide), or sensory nerves (substance P).^{22 23} Myelinated afferent nerves were also identified by their myelin basic protein immunoreactivity.²³

After incubation in buffer containing 0.2% Triton X-100 and impregnation with Pontamine sky blue, the preparations were incubated for 16 to 24 hours at room temperature in diluted primary antisera.²³ After being washed several times in buffer, the preparations were incubated for 60 minutes in FITC-labeled goat anti-rabbit IgG (Tago Inc) or donkey anti-sheep IgG at a dilution of 1:100. Combined immunofluorescence staining of CD31- (VM64; 1:100),²⁴ -smooth muscle actin– (IA4; 1:1000),²⁵ vimentin (V9; 1:300),²⁶ or -sarcomeric muscle actin– (5C5; 1:1000)²⁵ positive cells and nerves was achieved by sequential incubation with a monoclonal antiserum to these markers (Sigma); FITC-labeled goat anti-mouse IgG (Sigma); polyclonal antisera to PGP 9.5 or specific neuropeptides²³ and TRITC-labeled goat anti-rabbit IgG (Sigma). Preparations were mounted in Vectashield (Vector Laboratories) and examined with Olympus AH2 and BX60 microscopes equipped for epi-illumination with filters selective for FITC and TRITC fluorochromes. The spatial relationship between nerves and cells was assessed with dual-channel laser scanning confocal microscopy.²² Control experiments included omission of the primary antisera, use of inappropriate secondary antisera, and preabsorption with their respective antigens (10^-5 to 10^-6 mol/L) for 2 hours at room temperature, all of which prevented immunostaining.

AChE-positive nerves, which represent a distinct nerve subpopulation of cardiac nerves,^{22 23} were demonstrated by use of a method adapted by Tago et al.²⁷ The simultaneous visualization of immunostaining and AChE activity was demonstrated as described previously.²² Negative controls included incubation of preparations without acetylthiocholine iodide and in the presence of 10^-4 mol/L eserine, an active inhibitor of the enzyme.

Quantitative Analysis of Nerve Staining
The distribution and relative density of specific nerve subpopulations in individual leaflets were determined by computer-assisted analysis of immunofluorescent and AChE-stained whole-mount preparations with an image analysis system (Seescan). In complete leaflets, antisera to PGP 9.5 were used to determine the overall density and distribution of nerves and nerve terminals. The total field area and the area occupied by stained nerves were obtained for an average of 10 randomly selected fields, and the proportional area occupied by stained nerves was derived for each leaflet. When several antibodies were applied, each leaflet was subdivided longitudinally from base to edge into six segments. One in every three fields was then randomly selected, resulting in some five measurements per leaflet for each antibody. Terminal arborizations were interactively discriminated, and the immunostained nerve area occupied by each terminal, its dispersion, and the diameter of the supplying nerve fiber were measured (Table 2).

View this table: [in this window] [in a new window]   Table 2. Quantitative Summary of Nerve Terminals Identified

Statistical Analysis
Data are presented as the mean and 95% CI. Variation in the proportion of stained nerve subtypes and comparisons between leaflets were assessed, following logit transformation of the data, by Student's t test and paired two-way ANOVA. Correlation between variables in the valves was assessed in specific hearts (Table 1) by Pearson's (parametric) and linear regression (least squares) analyses. Values of P<.05 were considered significant.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Mitral and Tricuspid Valve Innervation
The overall innervation of the mitral and tricuspid valves, demonstrated with antisera to the general neuronal marker PGP 9.5, was localized to the atrial layer, or upper region, of the valvar leaflets. In the anterior leaflet of the mitral valve, the nerves arose from several elongated nerve trunks at the base of the leaflet, which branched to form an intricate network of anastomosing varicose and nonvaricose nerve fibers (Fig 1A). The innervation supplying the other regions of the mitral valve and that of the tricuspid valve arose as a direct continuation of the atrial endocardial plexus (Fig 1A and 1C). Nerve fibers extended approximately two thirds of the way to the free edge of the valve leaflets but were not observed in the outer or distal region itself. Quantitative analysis of PGP 9.5 immunoreactivity revealed that the anterior leaflet of the mitral valve showed a greater density of innervation compared with the posterior leaflet (Fig 2A), whereas no difference was seen between the anterior leaflet and the leaflets of the tricuspid valve. The density and distribution of innervation did not vary significantly with age (3 months to 61 years) in either the mitral or tricuspid valve leaflets.

View larger version (33K): [in this window] [in a new window]   Figure 1. Schematic of the anatomic features of the AV valves (A and B, mitral; C and D, tricuspid) and the relative density and distribution of nerve fibers (A, C) and nerve terminal arborizations (B, D). A, The posterior leaflet of the mitral valve has been subdivided into lateral, middle, and medial leaflets or scallops. Annotations in parentheses indicate alternative nomenclature used to describe the position of the valve leaflets in their in situ position. Also illustrated are the papillary muscles and their cordal attachments to the valve leaflets. B, APM indicates anterolateral papillary muscle; PPM, posteromedial papillary muscle. C, The anterior leaflet of the tricuspid valve is also known as the anterosuperior leaflet; the posterior leaflet, as the inferior or mural leaflet. D, APM indicates anterior papillary muscle; PPM, posterior papillary muscle; and IPM, inferior papillary muscle.

View larger version (19K): [in this window] [in a new window]   Figure 2. Distribution and relative density of the overall protein gene product 9.5–immunoreactive innervation (A) and of nerve subtypes in the mitral valve leaflets (B), displaying immunoreactivity for either neuropeptide Y (NPY), tyrosine hydroxylase (TH), vasoactive intestinal polypeptide (VIP), or substance P (SP) and showing acetylcholinesterase (AChE) activity. Each bar represents the mean and 95% CI for eight individuals 3 months to 21 years of age. The lateral (Lat), medial (Med), and middle leaflets or scallops of the posterior leaflet contained a similar density of nerve fibers, but this was significantly less dense (*P<.001) than that observed in the anterior leaflet (A). Compared with the anterior leaflet, significantly fewer nerve subtypes (*P<.001) were also identified in the posterior leaflet (B).

The mitral and tricuspid valves both contained extensive nerve terminal arborizations that exhibited significant differences in morphology and distribution. The anterior leaflet of the mitral valve contained large nerve terminals with multiple anastomosing fine fibers arising from a single thick parent fiber (Fig 3A and Table 2). Also evident were distinct arborizations again arising from a single thick fiber (Fig 3B and Table 2) and smaller, extensively distributed terminals arising directly from fine fibers within the valvar plexus (Table 2). Nerve terminals were not found in the posterior leaflet of the mitral valve (Fig 1B). In contrast, nerve terminals were identified relatively frequently in the leaflets of the tricuspid valve (Fig 1D and Table 2). All three leaflets contained terminals, some of which were present as individual arborizations or groups of distinct arborizations arising from a single fiber. In addition, some nerve terminals also were found to be closely associated with valvar endothelial cells (Table 2).

View larger version (93K): [in this window] [in a new window]   Figure 3. Photomicrographic montages showing (A) part of a large diffuse nerve terminal arborization (mean area and dispersion, 15 972 and 59 110 µm2, respectively; Table 2) in a whole-mount preparation of tricuspid valve anterior leaflet displaying immunoreactivity for protein gene product (PGP) 9.5 and (B) the predominant type of nerve terminal found in the cardiac valves (mean area and dispersion, 1565 and 4352 µm2, respectively; Table 2) shown in the anterior leaflet of the mitral valve exhibiting tyrosine hydroxylase (TH) immunoreactivity. Five distinct nerve terminals (numbered 1 through 5) are shown arising from a single myelinated fiber (arrow). Bar=100 (A) and 50 µm (B).

In the hearts of children (4 to 8 years of age), the tendinous cords were differentially innervated. Only the cords inserted into the free edge and outer one third of the leaflets contained nerves that continued close to the ventricular surface of the leaflet (Fig 4). In adolescent and adult hearts, however, the density and distribution of innervation appeared to decline, with only the main branches of the major cords containing nerves. All the nerve fibers seen within the endocardium of the tendinous cords originated from the endocardial lining of the papillary muscles (Fig 4A through 4C). Several nerve terminal arborizations were detected in the major fixing cords that support the various leaflets of the mitral valve and the anterior leaflet of the tricuspid valve (Fig 4D and 4E and Table 2).

View larger version (127K): [in this window] [in a new window]   Figure 4. Photomicrographs showing the relative density and distribution of protein gene product 9.5–immunoreactive nerve fibers and terminals in the endocardium of the tendinous cords (A through E). Nerves arise from the endocardium lining the papillary muscles (A through C) and decline in density as the cords approach the valve leaflet (A). Also shown are two types of morphologically and quantitatively distinct nerve terminals (D, E; Table 2). The terminal illustrated in D was found in the tendinous cords of both the mitral and tricuspid valves, whereas the other (E) was found only in the cords attached to the leaflets of the tricuspid valve. Arrow in A indicates orientation of the cords relative to the associated leaflet. PM indicates papillary muscle. Bar=200 (A, C) and 50 (B, D, E) µm.

Neuropeptide Immunoreactivity and AChE Staining
Nerves displaying either immunoreactivity for neuropeptide Y and tyrosine hydroxylase or AChE activity were identified within the plexus in all mitral and tricuspid valvar leaflets. They were most numerous in the anterior leaflet of the mitral valve and exhibited similar densities (Figs 2 and 5). Vasoactive intestinal polypeptide– and substance P–immunoreactive nerves were relatively sparse (Figs 2 and 5). Nerve terminal arborizations displaying PGP 9.5, tyrosine hydroxylase and neuropeptide Y immunoreactivity, or AChE activity (Figs 3 and 4) arose from thick fibers immunoreactive for myelin basic protein. AChE activity was found to be either colocalized with PGP 9.5 in the myelinated preterminal nerve or present in both the terminal arborization and preterminal nerve. Nerve terminals arising from nonmyelinated nerves were predominantly AChE positive.

View larger version (144K): [in this window] [in a new window]   Figure 5. Photomicrographs showing nerve fibers and fascicles in the atrial layer of the anterior leaflet of the mitral valve in individuals 3 months to 21 years of age displaying immunoreactivity for either protein gene product 9.5 (PGP; A), tyrosine hydroxylase (TH; B), neuropeptide Y (NPY; C), vasoactive intestinal polypeptide (VIP; D), or substance P (SP; E) and showing acetylcholinesterase staining (AChE; F). Bar=50 µm.

Pulmonary and Aortic Valve Innervation
The density of innervation in the leaflets of the pulmonary valve and the coronary leaflets of the aortic valve was similar to that observed in the tricuspid valve and the anterior leaflet of the mitral valve. The innervation of the arterial valves arose from the ventricular endocardial plexus (Fig 6), and the entire leaflet contained nerves, except for the coapting edge, localized to the ventricular layer or lower region of each leaflet. The noncoronary leaflet had an attenuated level of innervation compared with the two coronary leaflets (Fig 7), with the overall density and distribution of innervation in the aortic valve leaflets declining with age (Fig 7).

View larger version (55K): [in this window] [in a new window]   Figure 6. Confocal photomicrograph of the basal section of a pulmonary valve leaflet and the adjacent portion of right ventricular endocardium immunostained for protein gene product 9.5. The image constitutes a maximum projection of 40 optical slices in the Z plane, each of which is 5 µm apart. The valve innervation submerges 200 µm at the base of the valve and is continuous with the endocardial plexus of the right ventricle. The box and arrow in the adjacent schematic indicate the corresponding area in the photomicrograph and the orientation with which this is viewed. AT indicates arterial trunk; VW, ventricular wall. Bar=100 µm.

View larger version (16K): [in this window] [in a new window]   Figure 7. Relative density of the overall protein gene product 9.5–immunoreactive innervation in the aortic valve leaflets and the decline in nerve density with age. Data points represent the mean value derived from multiple fields of aortic valve leaflets from individual patients (n=8; Table 1). The innervation of the coronary leaflets was significantly denser than that of the noncoronary leaflet (P<.001) and exhibited a significant correlation (r=.9404, P<.0001) between nerve density and age. indicates noncoronary leaflet; , left coronary leaflet; and , right coronary leaflet.

Large nerve terminals, such as those seen in the AV valves, were not evident in the arterial valvar leaflets, and no terminal arborizations were observed arising from fine nonmyelinated fibers associated with the valvar plexus (Table 2). Groups of nerve terminals, however, were seen arising from thick individual nerves. These nerves and the nonmyelinated nerves associated with the plexus sometimes traversed the entire leaflet and appeared to supply more than one leaflet. Thus, the noncoronary leaflet, which is not associated with ventricular endocardium, received its innervation from the adjacent coronary leaflets. Nerve terminals were more numerous in the pulmonary than in aortic valve leaflets but were not associated with a particular leaflet.

Neuropeptide Immunoreactivity and AChE Staining
The density and neurochemical content of nerves in the pulmonary valvar plexus were similar to those of the AV valves. Nerve terminals expressed immunoreactivity for tyrosine hydroxylase and neuropeptide Y or showed AChE activity. AChE was found to colocalize with PGP 9.5 immunoreactivity in nerve terminals but was not detected in nerves expressing tyrosine hydroxylase or neuropeptide Y immunoreactivity. Vasoactive intestinal polypeptide and substance P immunoreactivities were not localized to nerve terminals.

Double Immunostaining of Neural and Cellular Components in Valves
Double-labeling experiments indicated morphological and spatial distinctions between smooth muscle cells expressing -smooth muscle actin immunoreactivity and fibroblasts expressing vimentin immunoreactivity (Fig 8B and 8C). Smooth muscle cells were distributed in the proximal and middle thirds of the AV valve leaflets, whereas vimentin-positive fibroblasts were seen throughout (Fig 8A). PGP 9.5–immunoreactive nerve fibers were observed closely apposed to both cell types (<0.2 µm; Fig 8D and 8E). -Smooth muscle actin immunoreactivity also was detected in a subpopulation of fibroblast-like cells in the medial and distal portions of the leaflet. Immunoreactivity for -sarcomeric muscle actin was localized to the basal portion of the AV valves and was continuous with atrial myocardium but was not detected in the arterial valve leaflets. Nerve fibers in the myocardium appeared to be distinct from those in the valvar plexus. Blood vessels were evident in the proximal portions of both mitral and tricuspid valves but not those of the arterial valves. -Smooth muscle actin immunoreactivity also was detected in a band of cells distributed across the middle portion of arterial valve leaflets that showed a close association with PGP 9.5–immunoreactive nerve fibers (<0.2 µm). Varicose nerve fibers were observed in close proximity (<0.2 µm) to CD-31–immunoreactive endothelial cells lining both AV and arterial valve leaflets (Fig 8F).

View larger version (118K): [in this window] [in a new window]   Figure 8. Schematic of the distribution of -smooth muscle actin–immunoreactive cells (hatched area) and vimentin-positive fibroblasts (hatched and clear areas) in human AV valves (A). Boxes labeled B and C correspond to the photomicrographs that display cell types expressing -smooth muscle actin and vimentin immunoreactivity, respectively. Double-immunostained confocal photomicrographs illustrate the spatial relationship between protein gene product 9.5–immunoreactive nerves (red and yellow) and cells (green) expressing -smooth muscle actin immunoreactivity (D) or vimentin immunoreactivity (E) in the anterior leaflet of the mitral valve. CD-31–immunoreactive endothelial cells (green) were also found in close proximity to varicose nerve fibers (red and yellow; F). Bar=50 µm.

No significant differences were found in the distribution or relative density of nerves in cardiac valves obtained either postmortem or during surgery from patients 3 months to 21 years of age.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The present findings show that the human AV and arterial valves have distinct patterns of innervation comprising a heterogeneous population of nerve fibers and terminals. Furthermore, this innervation is found in close physical proximity to valvar smooth muscle actin–immunoreactive, fibroblast, and valvar endothelial cells.

Previous Morphological Studies on Mammalian Valve Innervation
The studies thus far performed on AV valvar innervation have been largely incomplete. Nonetheless, those performed have demonstrated significant species differences. Nerve fibers in human AV valves were described as covering the entire leaflet and entering the tendinous cords,^{19 20} whereas studies on experimental animals, such as the guinea pig and rabbit, indicated a distinct innervation of the cords and valvar leaflets.^{28 29 30} We consistently found a clear demarcation between cordal and valvar innervation across a wide range of ages. Our study indicates a similar level of innervation in the pulmonary and aortic valves, except for the aortic noncoronary leaflet, and represents the first demonstration of sensory nerve terminals arising from myelinated fibers in this region. Pulmonary valvar innervation, like that of the AV valves, appears relatively unaffected by aging. The innervation pattern of the aortic leaflets, however, was age dependent and declined in density with age. Studies in experimental animals have indicated similar patterns of innervation in the arterial leaflets and a reduction in aortic valvar innervation with age.²¹

Contractile Elements in Valves
Myocardium has been reported to occur in the proximal one third of AV valve leaflets, close to the base or annulus.^{12 13 31} Several studies have suggested that myocytes may play a role in valvar function, either supporting the leaflet¹⁶ or being actively involved in contraction and relaxation of the valve during the cardiac cycle.^{8 13 31} Myocardial fibers, however, are often sparse, particularly in the tricuspid valve, and on average only extend 1 to 3 mm into the leaflet.¹³ Therefore, it is uncertain whether they could significantly affect valvar motion. In contrast, smooth muscle cells exhibit a more extensive distribution in the proximal and middle portions of porcine AV and arterial valve leaflets,^{10 14} canine aortic valve leaflets,¹⁵ and human cardiac valves. Fibroblasts also were shown to be widely distributed in human valve leaflets, and studies on valvar interstitial cells in culture indicate that they are contractile and may display smooth muscle characteristics.¹¹ Using confocal microscopy, we identified nerve fibers in close proximity to both smooth muscle cells and fibroblasts; at the ultrastructural level, nerve terminals have been identified within 30 to 80 nm of rodent interstitial cells⁹ and 300 nm of porcine smooth muscle bundles.¹⁰

Distribution and Neurochemistry of Nerve Terminals
The AV and arterial valves contained nerve terminal arborizations that correspond both morphologically and neurochemically with those identified in the endocardium and epicardium of the human heart.²³ Several differences also are apparent. The AV valves possess considerably larger nerve terminals than those found elsewhere in the heart, and these terminals may be considered to correspond to the "end net" previously described in valves.^{18 32 33} The number of arborizations arising from a single parent fiber also differed in valves compared with other regions of the heart. In all four cardiac valves, the dominant appearance of nerve terminals was in groups of five or more distinct nerve terminal arborizations arising from a single fiber. In the epicardium, in contrast, it was rare to find groups of terminals arising from a single myelinated fiber, and none were detected in the endocardium. In addition, the density of nerve terminals arising from myelinated nerves in the AV valves was approximately two to three times greater than that previously observed in other regions of the heart.²³

Possible Functional Role of Nerves and Nerve Terminals
Electric stimulation and exogenous noradrenaline induce marked changes in the tension of the excised canine mitral valve that are reversed by the addition of acetylcholine^{8 9} and augmented by tyramine.⁸ ß-Adrenoceptors in the AV valvar leaflets of the rat also have been described as being concentrated in the atrial layer at the base of the leaflets, further indicating that catecholamines may influence valvar function through stimulation of these receptors.³⁴

Valvar endothelial cells produce several vasoactive factors, including nitric oxide, endothelin, and prostaglandins, the release of which may be controlled by such neuromodulators as acetylcholine and substance P.^{35 36} Conversely, substances released from endothelial cells, such as prostaglandins, may stimulate local nerve terminals and elicit reflex responses.³⁷ A local autocrine or paracrine role for endothelium-derived factors also is supported by the concordant localization of endothelin-like immunoreactivity and specific endothelin binding sites, corresponding to the ET_B receptor subtype, in human fetal heart valves.³⁸

Conclusions
It has been suggested that cardiac valves may be capable of independent movement, although the mechanisms responsible are uncertain. In this study, we have presented morphological evidence of extensive valvar innervation in the human heart and demonstrated possible cellular targets for nerves in valve leaflets. Further studies are required to determine the extent to which valvar motion is influenced by the innervation and to examine the relation between distinct nerve terminals and mechanosensitive or chemosensitive receptor function.

	Acknowledgments

 
This work was supported by grants (PG/92053 and PG/940100) from the British Heart Foundation. Antisera to regulatory peptides were produced in conjunction with Prof S.R. Bloom; antisera to tyrosine hydroxylase and CD31 were provided by Drs D. Kuhn and A.V. Mazurov. We thank the staff of the Beaumont and Mater Misericordiae Hospitals, Dublin, Ireland, for assistance in the collection of human cardiac tissue.

Received December 18, 1995; revision received February 12, 1996; accepted February 16, 1996.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Henderson Y, Johnson FE. Two modes of closure of the heart valves. Heart. 1912;4:69-82.
   
2. Dean AL Jr. The movements of the mitral cusps in relation to the cardiac cycle. Am J Physiol. 1916;40:206-212.
   
3. Little RC. Effects of atrial systole on ventricular pressure and closure of the AV valves. Am J Physiol. 1951;166:289-295.
   
4. Sarnoff SJ, Gilmore B, Mitchell JH. Influence of atrial contraction and relaxation on closure of the mitral valve. Circ Res. 1962;11:26-35.
   
5. Taskiris AG. The physiology of the mitral valve annulus. In: Kalmanson D, ed. The Mitral Valve: A Pluridisciplinary Approach. London, UK: Edward Arnold; 1975:21-26.
   
6. Cooper T, Napolitano LM, Fitzgerald MJT, Moore KE, Daggett WM, Willman VL, Sonnenblick EH, Hanlon CR. Structural basis of cardiac valvular function. Arch Surg. 1966;93:766-771.
   
7. Braunwald E, Rockoff SD, Newland-Oldham H Jr, Ross J Jr. Effective closure of the mitral valve without atrial systole. Circulation. 1966;33:404-409.[Abstract/Free Full Text]
   
8. Sonnenblick EH, Napolitano LM, Daggett WM, Cooper T. An intrinsic neuromuscular basis for mitral valve motion in the dog. Circ Res. 1967;21:9-15.[Abstract/Free Full Text]
   
9. Fenoglio JJ, Pham TD, Wit AL, Bassett AL, Wagner BM. Canine mitral complex: ultrastructure and electromechanical properties. Circ Res. 1972,31:417-430.
   
10. De Biasi S, Vitellaro-Zucharello L. Intrinsic innervation of porcine semilunar heart valves. Anat Embryol (Berl). 1982;165:71-79.[Medline] [Order article via Infotrieve]
    
11. Filip DA, Radu A, Simionescu M. Interstitial cells of the heart valves possess characteristics similar to smooth muscle cells. Circ Res. 1986;59:310-20.[Abstract/Free Full Text]
    
12. Hibbs RG, Ellison JP. The atrioventricular valves of the guinea-pig, II: an ultrastructural study. Am J Anat. 1973;138:347-69.[Medline] [Order article via Infotrieve]
    
13. Basset AL, Fenoglio JJ Jr, Wit AL, Myerburg RG, Gelband H. Electrophysiologic and ultrastructural characteristics of the canine tricuspid valve. Am J Physiol. 1976;230:1366-1373.
    
14. DeBiasi S, Vitellaro-Zuccarello L, Blum I. Histochemical and ultrastructural study on the innervation of human and porcine atrioventricular valves. Anat Embryol (Berl). 1984;169:159-165.[Medline] [Order article via Infotrieve]
    
15. Bairaiti A, DeBiasi S. Presence of smooth muscle system in aortic valve leaflets. Anat Embryol (Berl). 1981;161:329-340.[Medline] [Order article via Infotrieve]
    
16. Woollard HH. The innervation of the heart. J Anat. 1926;60:345-373.[Medline] [Order article via Infotrieve]
    
17. Nettleship WA. Experimental studies on the afferent innervation of the cat's heart. J Comp Neurol. 1936;64:115-133.
    
18. Williams TH. Mitral and tricuspid valve innervation. Br Heart J. 1964:26;105-115.
    
19. Ferreira AL, Rossi MA. Innervation of human atrioventricular valves. Acta Anat (Basel). 1974;87:57-65.[Medline] [Order article via Infotrieve]
    
20. Kwano H, Kawai S, Shirai T, Okada R. Morphological study on vagal innervation in human atrioventricular valves using histochemical method. Jpn Circ J. 1993;57:753-759.[Medline] [Order article via Infotrieve]
    
21. Su H, Mao Q, Zhang Y, Huang W, Cheung A. Distribution of peptide containing nerves in the cardiac valves of rats of different ages. Cell Vision. 1995;2:40-44.
    
22. Marron K, Wharton J, Sheppard MN, Gulbenkian S, Royston D, Yacoub MH, Anderson RH, Polak JM. Human endocardial innervation and its relationship to the endothelium: an immunohistochemical, histochemical and quantitative study. Cardiovasc Res. 1994;28:1490-1499.[Medline] [Order article via Infotrieve]
    
23. Marron K, Wharton J, Sheppard MN, Fagan D, Royston D, Kuhn DM, de Leval MR, Whitehead B, Anderson RH, Polak JM. Distribution, morphology and neurochemistry of endocardial and epicardial nerve terminal arborizations in the human heart. Circulation. 1995;92:2343-2351.[Abstract/Free Full Text]
    
24. Mazurov AV, Vinogradov DV, Kabaeva NV, Antonova GN, Romanov YA, Vlasik AS, Antonov AS, Smirnov VN. A monoclonal antibody, VM64, reacts with a 130 KDa glycoprotein common to platelets and endothelial cells: heterogeneity in antibody binding to human aortic endothelial cells. Thromb Haemost. 1991;66:494-499.[Medline] [Order article via Infotrieve]
    
25. Skalli O, Gabbaini G, BabaI F, Seemayer TA, Pizzolato G, Schurch W. Intermediate filament proteins and actin isoforms as markers for soft tissue tumour differentiation and origin. Am J Pathol. 1988;130:515-531.[Abstract]
    
26. Adams JC, Watt FM. An unusual strain of human keratinocytes which do not stratify or undergo terminal differentiation in culture. J Cell Biol. 1988;107:1927-1938.[Abstract/Free Full Text]
    
27. Tago H, Kimura H, Maeda T. Visualization of detailed acetylcholinesterase fiber and neuron staining in rat brain by a sensitive histochemical method. J Histochem Cytochem. 1986;34:1431-1438.[Abstract]
    
28. Smith RB. Intrinsic innervation of the atrioventricular and semilunar valves in various mammals. J Anat. 1971;108:115-121.[Medline] [Order article via Infotrieve]
    
29. Ellison JP, Hibbs RG. The atrioventricular valves of the guinea-pig, I: a light microscopic study. Am J Anat. 1973;138:331-345.[Medline] [Order article via Infotrieve]
    
30. Somogyi GT, Keast JR, Vizi ES. Presynaptic modulation of the release of noradrenaline from electrically stimulated bicuspid valve leaflet of the rabbit heart. J Auton Nerv Syst. 1991;35:99-106.[Medline] [Order article via Infotrieve]
    
31. Fenoglio JJ Jr, Pham TD, Wit AL, Basset AL, Wagner BM. Canine mitral complex: ultrastructure and electromechanic properties. Circ Res. 1972;31:417-430.[Abstract/Free Full Text]
    
32. Johnston BD. Nerve endings in the human endocardium. Am J Anat. 1968;122:621-630.[Medline] [Order article via Infotrieve]
    
33. Miller MR, Kasahara M. Studies on the nerve endings in the heart. Am J Anat. 1964;115:217-233.[Medline] [Order article via Infotrieve]
    
34. Pinto JEB, Nazarali AJ, Torda T, Saavedra JM. Autoradiographic characterization of ß-adrenoreceptors in rat heart valve leaflets. Am J Physiol. 1989;256:H821-H827.[Abstract/Free Full Text]
    
35. Ku DD, Nelson JM, Caulfield JB, Winn MJ. Release of endothelium relaxing factors from canine cardiac valves. J Cardiovasc Pharmacol. 1990;16:212-218.[Medline] [Order article via Infotrieve]
    
36. Siney L, Lewis MJ. Nitric oxide release from porcine mitral valves. Cardiovasc Res. 1993;27:1657-1661.[Abstract/Free Full Text]
    
37. Hainsworth R. Reflexes from the heart. Physiol Rev. 1991;71:617-658.[Free Full Text]
    
38. Wharton J, Rutherford RAD, Gordon L, Moscoso G, Schiemberg I, Gaer JR, Taylor KM, Polak JM. Localization of endothelin binding sites and endothelin-like immunoreactivity in human fetal heart. J Cardiovasc Pharmacol. 1991;17:S378-S384.
    

This article has been cited by other articles:

				F. J. Schoen Evolving Concepts of Cardiac Valve Dynamics: The Continuum of Development, Functional Structure, Pathobiology, and Tissue Engineering Circulation, October 28, 2008; 118(18): 1864 - 1880. [Abstract] [Full Text] [PDF]

				K. B. Donnelly Cardiac Valvular Pathology: Comparative Pathology and Animal Models of Acquired Cardiac Valvular Diseases Toxicol Pathol, February 1, 2008; 36(2): 204 - 217. [Abstract] [Full Text] [PDF]

				S. A. Butler-Manuel, L. D. K. Buttery, J. M. Polak, R. A'Hern, and D. P. J. Barton Autonomic Nerve Trauma at Radical Hysterectomy: The Nerve Content and Subtypes Within the Superficial and Deep Uterosacral Ligaments Reproductive Sciences, January 1, 2008; 15(1): 91 - 96. [Abstract] [PDF]

				J. I. Fann, N. B. Ingels Jr., and D. C. Miller Pathophysiology of Mitral Valve Disease Card. Surg. Adult, January 1, 2008; 3(2008): 973 - 1012. [Full Text]

				L. Osman, A. H. Chester, P. Sarathchandra, N. Latif, W. Meng, P. M. Taylor, and M. H. Yacoub A Novel Role of the Sympatho-Adrenergic System in Regulating Valve Calcification Circulation, September 11, 2007; 116(11_suppl): I-282 - I-287. [Abstract] [Full Text] [PDF]

				H.-H. Sievers In vivo tissue engineering an autologous semilunar biovalve: Can we get what we want? J. Thorac. Cardiovasc. Surg., July 1, 2007; 134(1): 20 - 22. [Full Text] [PDF]

				T. A. Timek, D. T. Lai, P. Dagum, D. Liang, G. T. Daughters, N. B. Ingels Jr, and D. C. Miller Mitral Leaflet Remodeling in Dilated Cardiomyopathy Circulation, July 4, 2006; 114(1_suppl): I-518 - I-523. [Abstract] [Full Text] [PDF]

				R. P. Visconti, Y. Ebihara, A. C. LaRue, P. A. Fleming, T. C. McQuinn, M. Masuya, H. Minamiguchi, R. R. Markwald, M. Ogawa, and C. J. Drake An In Vivo Analysis of Hematopoietic Stem Cell Potential: Hematopoietic Origin of Cardiac Valve Interstitial Cells Circ. Res., March 17, 2006; 98(5): 690 - 696. [Abstract] [Full Text] [PDF]

				J. Ritchie, J. N. Warnock, and A. P. Yoganathan Structural Characterization of the Chordae Tendineae in Native Porcine Mitral Valves Ann. Thorac. Surg., July 1, 2005; 80(1): 189 - 197. [Abstract] [Full Text] [PDF]

				M. H. Yacoub and L. H. Cohn Novel Approaches to Cardiac Valve Repair: From Structure to Function: Part I Circulation, March 2, 2004; 109(8): 942 - 950. [Full Text] [PDF]

				T. A. Timek, D. T. Lai, P. Dagum, F. Tibayan, G. T. Daughters, D. Liang, G. J. Berry, D. C. Miller, and N. B. Ingels Jr. Ablation of mitral annular and leaflet muscle: effects on annular and leaflet dynamics Am J Physiol Heart Circ Physiol, October 1, 2003; 285(4): H1668 - H1674. [Abstract] [Full Text] [PDF]

				J. I. Fann, N. B. Ingels Jr., and D. C. Miller Pathophysiology of Mitral Valve Disease Card. Surg. Adult, January 1, 2003; 2(2003): 901 - 931. [Full Text]

				S. A. Butler-Manuel, L. D. K. Buttery, R. P. A'Hern, J. M. Polak, and D. P. J. Barton Pelvic Nerve Plexus Trauma at Radical and Simple Hysterectomy: A Quantitative Study of Nerve Types in the Uterine Supporting Ligaments Reproductive Sciences, January 1, 2002; 9(1): 47 - 56. [Abstract] [PDF]

				K. L. Weind, D. R. Boughner, L. Rigutto, and C. G. Ellis Oxygen diffusion and consumption of aortic valve cusps Am J Physiol Heart Circ Physiol, December 1, 2001; 281(6): H2604 - H2611. [Abstract] [Full Text] [PDF]

				G. R. Green, P. Dagum, J. R. Glasson, G. T. Daughters, A. F. Bolger, L. E. Foppiano, G. J. Berry, N. B. Ingels Jr, and D. C. Miller Mitral Annular Dilatation and Papillary Muscle Dislocation Without Mitral Regurgitation in Sheep Circulation, November 9, 1999; 100(90002): II-95 - 102. [Abstract] [Full Text] [PDF]

				F. Santini, C. Dyke, S. Edwards, G. Stavri, M. Feccia, H. Khan, E. Birks, and M. H. Yacoub PULMONARY AUTOGRAFT VERSUS HOMOGRAFT REPLACEMENT OF THE AORTIC VALVE: A PROSPECTIVE RANDOMIZED TRIAL J. Thorac. Cardiovasc. Surg., May 1, 1997; 113(5): 894 - 900. [Abstract] [Full Text]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal