FGF-8 in the ventral pharynx alters development of myocardial calcium transients after neural crest ablation.

Cardiac neural crest ablation results in depressed myocardial calcium transients and elevated proliferation in myocardium at a stage when cardiac neural crest cells are not in contact with the myocardium. To test the hypothesis that cardiac neural crest-derived cells, which migrate into the caudal, ventral pharynx at stage 14, block a signal from the ventral pharynx, we cultured stage 12 chick heart tube or myocardial strips in the presence or absence of ventral pharynx. We found that myocardium cultured with ventral pharynx that had not yet contacted neural crest cells had significantly reduced calcium transients and an increased rate of proliferation. Ventral pharynx from intact embryos at a stage when neural crest-derived cells had reached the pharynx had no effect on myocardial calcium transients. Ventral pharynx from neural crest-ablated embryos continued to suppress myocardial calcium transients at this later stage. Myocardium cultured with FGF-2 also showed a significant reduction in calcium transients. An FGF-2-neutralizing Ab reversed the deleterious effect of the ventral pharynx on myocardial calcium transients and proliferation. We therefore examined the expression of FGF-2 and similar FGFs in the ventral pharynx. Only FGF-8 was expressed in a temporospatial pattern that made it a viable candidate for altering the myocardial calcium transient during stages 14-18. In explant cultures, neutralizing Ab for FGF-8 rescued development of the myocardial calcium transient in neural crest-ablated chick embryos.

Colocalization of dihydropyridine and ryanodine receptors in developing heart with a neural crest-associated defect.

BACKGROUND: Neural crest-associated congenital heart defects in humans are among the most lethal and costly to treat. In avian and mouse embryos with persistent truncus arteriosus (PTA), the most severe of the neural crest anomalies, there is poor cardiac function because of impaired excitation-contraction coupling. One possible explanation for poor excitation-contraction coupling is that peripheral junctions, composed of closely associated sarcoplasmic reticulum Ca(2+) release channels (ryanodine receptors) and surface membrane L-type Ca(2+) channels (dihydropyridine receptors), are not well colocalized. OBJECTIVE: To compare the degree of colocalization of these two Ca(2+) channel proteins in isolated ventricular myocytes from normal hearts and hearts with PTA. ANIMALS AND METHODS: PTA was induced in the embryonic chick by laser ablation of the cardiac neural crest before migration from the neural tube. Immunofluorescent staining of dihydropyridine and ryanodine receptors along with computer-assisted image analysis was used to measure relative colocalization. RESULTS: Dihydropyridine and ryanodine receptor colocalization was greater by 20% in embryos with PTA. Much of the increase appeared to result from a 14% increase in the area stained for ryanodine receptors. A third observation was that a high level of colocalization was maintained even after enzymatic dissociation in which the embryonic myocytes had typically lost their elongated appearance and assumed a spherical shape. CONCLUSIONS: The increased colocalization of dihydropyridine and ryanodine receptors in hearts with PTA may be a compensatory response to a defect at the level of single Ca(2+) channel proteins. These results indicate the high degree of stability of sarcoplasmic reticulum junctional complexes.

Action potentials that mimic fibrillation activate sodium current.

Ventricular fibrillation (VF) has brief action potentials (50-70 ms) with short diastolic intervals (10-30 ms). Under these conditions ion channel activity may be grossly different to normal sinus rhythm (NSR). In particular, sodium channel activation may not contribute to the generation and propagation of action potentials during VF. This study determined if sodium channels can be activated when action potentials mimic VF. Isolated chick ventricular myocytes (n=7) were voltage-clamped to quantitate fast inward sodium current. The voltage clamp protocol simulated VF with a 10 pulse train at 10 Hz (100 ms cycle length (CL)) and depolarization interval (action potential duration) ranging from 90 to 20 ms. After each train a test pulse was delivered from holding (-80 mV) in 10-ms steps. The train preceded each step pulse. Peak sodium current for control and each VF protocol occurred at a membrane potential (V(m)) of -10 mV. Sodium current was evident during brief resting intervals as short as 20 ms, albeit 10-20% of baseline. Resting intervals less than 60 ms shifted the sodium conductance activation curve from Vm(0.5)-30 mV to -22 mV membrane potential. Similar findings occurred when resting potential was at -65 mV, although there was less sodium current with all tested protocols. There was significantly less inactivation of sodium current when the prepulse was shorter (100 v 1000 ms). There was approximately 20% greater sodium current when the test pulse followed a short v long depolarized (>-80 mV) prepulse. Although the longer depolarization pulses produce approximately 20% greater sodium current at membrane potentials more negative than -80 mV. Lastly the time for half recovery of sodium current from activation was significantly less when the inactivating prepulse was short v long (45.9+/-9 v 118+/-20 ms, P<0.05). In conclusion, sodium current is evident when the diastolic rest interval is as brief as 10-20 ms. Rest interval, length of membrane depolarization and membrane potential interact to affect sodium channel activation, inactivation and recovery from inactivation. These data demonstrate that the brief action potentials at more depolarized membrane potentials seen during VF allow for inward sodium current upon depolarization, less sodium channel inactivation, and a faster recovery from inactivation, thereby compensating for a short diastolic rest interval. Therefore, it is likely that the inward sodium channel contributes to wave front propagation during ventricular fibrillation.

A novel role for cardiac neural crest in heart development.

Ablation of premigratory cardiac neural crest results in defective development of the cardiac outflow tract. The purpose of the present study was to correlate the earliest functional and morphological changes in heart development after cardiac neural crest ablation. Within 24 hours after neural crest ablation, the external morphology of the hearts showed straight outflow limbs, tighter heart loops, and variable dilations. Incorporation of bromodeoxyuridine in myocytes, an indication of proliferation, was doubled after cardiac neural crest ablation. The myocardial calcium transients, which are a measure of excitation-contraction coupling, were depressed by 50% in both the inflow and outflow portions of the looped heart tube. The myocardial transients could be rescued by replacing the cardiac neural crest. The cardiac jelly produced by the myocardium was distributed in an uneven, rather than uniform, pattern. An extreme variability in external morphology could be attributed to the uneven distribution of cardiac jelly. In the absence of cardiac neural crest, the myocardium was characterized by somewhat disorganized myofibrils that may be a result of abnormally elevated proliferation. In contrast, endocardial development appeared normal, as evidenced by normal expression of fibrillin-2 protein (JB3 antigen) and normal formation of cushion mesenchyme and trabeculae. The signs of abnormal myocardial development coincident with normal endocardium suggest that the presence of cardiac neural crest cells is necessary for normal differentiation and function of the myocardium during early heart development. These results indicate a novel role for neural crest cells in myocardial maturation.

Role of cardiac neural crest cells in cardiovascular development.

The discovery in the chick embryo that a specific region of the neural crest, termed the cardiac neural crest, is essential for septation of the cardiac outflow tract and for aortic arch artery development has led to the classification of a whole series of human cardiac defects as neural crest-associated. Recently, several mouse genetic models have been effectively employed to yield new insights into the relationship between cardiac neural crest and structural heart development. In all the animal models of neural crest-related heart defects, prenatal mortality is too high to be attributed to structural defects of the heart alone, and there are obvious signs of severe cardiac dysfunction. The evidence indicates that poor viability is from impaired cardiac excitation-contraction coupling and contractile function at the myocyte level. The continued study of experimental and genetically defined models with neural crest-associated heart defects will prove useful in identifying the common pathways by which the neural crest contributes to normal heart development.

Excitation-contraction coupling in the day 15 embryonic chick heart with persistent truncus arteriosus.

Ca2+ transients were examined in embryonic chick hearts with an experimentally induced cardiac neural crest-related outflow tract defect known as persistent truncus arteriosus (PTA). In all of the animal models of neural crest-related heart defects, prenatal mortality is too high to be attributed to structural defects of the heart alone, suggesting that there is altered development of the myocardium. Earlier reports indicating reduced L-type Ca2+ current in hearts with PTA suggest that poor viability may be related to impairment of cardiac excitation-contraction coupling. To test this hypothesis, direct measurements of the systolic Ca2+ transient in fura-2-loaded myocytes from normal hearts and hearts with PTA were carried out. We found that Ca2+ transients were severely depressed in hearts with PTA and difficult to measure above background noise unless signal averaged or treated with isoproterenol (ISO). We confirmed that the reduced Ca2+ transients were due, at least partly, to a reduction in L-type Ca2+ current. In addition we found that although ISO could raise the L-type current in hearts with PTA to the level found in normal hearts in the absence of ISO, it could not fully restore the Ca2+ transient. Furthermore, caffeine-stimulated Ca2+ transients were diminished in size and the time-to-peak and the decaying phase were significantly slowed. Interestingly, these observations were not accompanied by a reduction in the number of Ca2+ release channels. These results indicated an impairment of SR function in addition to the reduction in L-type Ca2+ current. These results strongly support our hypothesis that the poor viability of embryos with PTA is due to impaired cardiac excitation-contraction coupling.

Neural crest is involved in development of abnormal myocardial function.

Around 85% of embryos homozygous for the splotch (Sp2H) allele (Sp2H/Sp2H), a Pax3 mutation, develop persistent truncus arteriosus (PTA), a defect related to the cardiac neural crest. These embryos die by 14.5 days post coitum. In an investigation of the cause of lethality in these embryos, we used digital video imaging microscopy to examine beating embryonic hearts in situ at 13.5 dpc. The hearts of Sp2H/Sp2H embryos with PTA clearly showed poor function when compared with normal litter mates. Contractile force was examined in detergent-skinned ventricular muscle strips from Sp2H/Sp2H embryos at ages 12.5 and 13.5 dpc. There was no significant difference in the maximum force or in myosin content between Sp2H/Sp2H and control groups, indicating no significant dysfunction of the contractile apparatus in hearts from Sp2H/Sp2H embryos. Ca2+ transients were examined in enzymatically-dissociated ventricular myocytes and were significantly reduced in defective hearts, indicating that reduced cardiac function in Sp2H/Sp2H embryos with PTA was due to impaired excitation-contraction (EC) coupling. Ca2+ currents were examined using the perforated patch clamp technique. The magnitude of the Ca2+ current was found to be reduced by approximately 3.2-fold in Sp2H/Sp2H hearts with PTA compared to normal. Since the sarcoplasmic reticulum is sparse or absent in the embryonic heart, the impaired EC coupling was due to the reduction in Ca2+ current. These observations suggest that neural crest abnormalities result in a defect in EC coupling, causing depressed myocardial function and death in utero from cardiac failure. Interestingly, Sp2H/Sp2H hearts without PTA had normal EC coupling. These results indicated that impaired EC coupling was secondary to the Pax3 mutation. The findings in this report indicate an important role for the neural crest in the development of normal myocardial function, and represent the first demonstration of impaired excitation-contraction coupling in a genetically-defined embryonic mammalian model of a cardiac structural defect.

Beta 1 integrin ligation stimulates tyrosine phosphorylation of phospholipase C gamma 1 and elevates intracellular Ca2+ in pancreatic acinar cells.

We have recently reported increased tyrosine (TYR) phosphorylation of a number of pancreatic acinar cell proteins following antibody ligation of beta 1 integrins (Wrenn and Herman, Biochem, Biophys. Res. Commun. 208, 1995, 978-984). Concurrent with this TYR phosphorylation was a marked activation of protein kinase C (PKC). This led us to investigate phospholipase C gamma 1 (PLC gamma 1), a key enzyme responsible for diacylglycerol generation, as a target for integrin-mediated TYR phosphorylation. Staining with antiphosphotyrosine antibodies revealed increased TYR phosphorylation of immunoprecipitated PLC gamma 1 prepared from beta 1 integrin-ligated acinar cells. Subsequent stripping and reprobing of Western blots with polyclonal anti-PLC gamma 1 was confirmatory. Over this same time period, intracellular [Ca2+] increased from < 100 nM to 600 nM, further suggesting a functional relevance of integrin-linked phosphorylation as a regulatory mechanism in exocrine pancreas.

Elevation of intracellular Ca2+ concentration in rabbit nonpigmented ciliary epithelial cells by allicin.

A previous study has shown that allicin produces changes in aqueous humor dynamics, and this study was conducted to examine possible cellular mechanisms. In rabbit nonpigmented ciliary epithelial cells, basal levels of [Ca2+]i were determined to be 164 +/- 34 nM. Allicin, a sulfhydryl-reactive agent, induced Ca2+ transients at 0.01 mM and at 0.2 mM, the Ca2+ transient peaked at 732 +/- 35 nM. Allicin-induced Ca2+ transients were prevented by pretreatment with dithiothreitol which did not affect the basal Ca2+ levels. Allicin had only a slight, insignificant, effect on L-type Ca2+ currents, and allicin-induced Ca2+ transients were also present under extracellular Ca(2+)-free conditions. These data suggest that intracellular Ca2+ stores are the most probable source of allicin's effect. Pretreatment of cells with ryanodine, an inhibitor of Ca(2+)-induced-Ca(2+)-release, inhibited allicin-induced Ca2+ transients, but the basal Ca2+ levels were unaffected by ryanodine. Thus, allicin-induced Ca2+ transients are most likely mediated through ryanodine-sensitive intracellular Ca2+ stores.

Ca2+ transients in embryonic chick heart: contributions from Ca2+ channels and the sarcoplasmic reticulum.

In the embryonic mammalian heart, virtually all the Ca2+ available for the Ca2+ transient comes through sarcolemmal Ca2+ influx. However, several studies in avian species indicate that the sarcoplasmic reticulum (SR) is functional relatively early in development. For the present report we studied fura 2 Ca2+ transients elicited by field stimulation in single isolated ventricular myocytes from the day 11 embryonic chick heart to ascertain directly the roles of the SR and Ca2+ channels. A positive staircase phenomenon was observed at higher frequencies of stimulation (1 Hz). Isoproterenol (Iso) increased the peak of the transient in a dose-dependent manner with a maximum increase of 93% in 100 microM Iso. Nifedipine (10 microM) reduced the transient such that is was not observable above background noise. However, Ca2+ transients were visible when the myocytes were stimulated by Iso. These were blocked by approximately 70% with nifedipine, suggesting that most, but not all, of the transient is associated with L-type Ca2+ current. Thus a portion of the transient may result from T-type Ca2+ channels and/or reverse Na+/Ca2+ exchange. Calculations based on integration of the Ca2+ currents and cell volume indicate that as much as one-fourth of the Ca2+ entering via sarcolemmal Ca2+ channels is from T-type channels. Ryanodine at high concentrations (10-100 microM) inhibited the transients by 30%. Both Iso and ryanodine reduced the time to peak, the time constant of the exponential decay, and the total duration of the transients. Depolarizing the myocytes with high KCl induced a large and partially sustained transient when the external solution contained 1.8 mM CaCl2. CaCl2 (10 mM) in the external solution induced large cyclic Ca2+ oscillations. These results suggest that the SR is functional in the embryonic chick heart well before hatching at day 22, although most of the Ca2+ associated with the transient comes through the sarcolemmal Ca2+ channels and possibly reverse Na+/Ca2+ exchange.

The effect of 2,3-butanedione 2-monoxime (BDM) on ventricular trabeculae from the avian heart.

2,3-butanedione 2-monoxime (BDM, 3-30 mM) decreased twitch force of intact ventricular trabeculae isolated from 19-day embryonic chick hearts in a dose-dependent manner. The responses to BDM were rapid and reversible. In an attempt to determine the cellular basis for the inhibitory effect of BDM, experiments were carried out on skinned muscle fibres and isolated myocytes. In trabeculae skinned with Triton X-100, BDM depressed maximum calcium activated force (Fmax) with an IC50 of 14 mM. At 3 mM BDM, the proportional decrease in twitch force in intact tissue was similar to that of Fmax in skinned tissue. At higher BDM concentrations (10 and 30 mM), however, the proportional decrease in twitch force was greater than that of Fmax. BDM (up to 10 mM) had no effect on the normalized force-pCa relationship. In saponin-skinned preparations, BDM (3 and 30 mM) released calcium from the fully loaded sarcoplasmic reticulum to a slightly greater extent in the absence of calcium (pCa 8.5) than in the presence of a fixed level of free calcium (pCa 5.5). Whole cell patch clamping of freshly isolated chick myocytes demonstrated that BDM caused a dose-dependent decrease in the T- and L-type calcium current. Therefore, at low BDM concentrations (3 mM), the decrease in twitch force can be ascribed predominantly to depression of the contractile apparatus while, at higher concentrations of BDM, there is an additional inhibitory effect of BDM on excitation-contraction coupling.

Myocardial enlargement in defective heart development.

BACKGROUND: The cardiac neural crest (neural crest extending from the mid-otic placode to the caudal region of somite 3) provides ectomesenchymal cells that contribute to aortic arch development and are essential for aortico-pulmonary septation of the outflow tract. Bilateral ablation of the cardiac neural crest in the chick embryo, prior to migration, leads to aortic arch anomalies and failure of septation of the cardiac outflow tract, which produces a severe defect known as persistent truncus arteriosus (PTA). Altered hemodynamics resulting from abnormal aortic arch artery development and PTA and other unknown factors related to the absence of neural crest, are likely to alter the developmental history of the myocardium. METHODS: In this study the wet and dry weights of ventricles and whole embryos, the total number of myocytes per ventricle and the myocyte density (number of myocytes per unit volume of ventricular myocardium) were compared in control (unwindowed eggs), sham-operated and cardiac neural crest ablated chick embryos at day 11 of incubation. RESULTS: We found that the wet and dry weights of ventricles from hearts with PTA were not different from normal hearts in control and sham-operated embryos. However, the embryos with PTA weighed less than embryos with normal hearts. Thus, the ventricle to embryo weight ratios were greater in embryos with PTA compared to control and sham-operated embryos for both wet (14 and 20%, respectively) and dry (30 and 59%) weights. The data further implied that more water was present with respect to body weight in comparison with sham-operated and control embryos which indicated that the embryos with PTA were edematous. The total number of myocytes and the number of myocytes per unit volume were not different when comparing sham-operated with PTA. Further, there was no indication that the myocardium from hearts with PTA was abnormal despite the small size and edema of the embryos. CONCLUSIONS: It appears that hemodynamic stresses, resulting from the structural defects produced by neural crest ablation, are insufficient to increase heart growth, although cardiac function is depressed as evidenced by edema and failure of the embryo to thrive.

Comparison of the number of dihydropyridine receptors with the number of functional L-type calcium channels in embryonic heart.

We compared the density of dihydropyridine (DHP) receptor sites with the density of functional L-type calcium channels in ventricular myocytes from chick heart at embryonic day 11. DHP receptors were quantified by using the DHP antagonist (+)-[3H]PN200-110 and by competition binding with the agonist Bay K 8644. The number of agonist and antagonist binding sites per ventricle was similar (250 +/- 15 and 244 +/- 8 fmol, respectively; mean +/- SEM). The mean number of myocytes per ventricle was 8.57 +/- 0.65 x 10(6), as determined by histological methods. From these data, the number of DHP receptors was calculated to be approximately 17,000 per myocyte or 25 to 26 DHP receptors per square micron of cell membrane, based on a mean myocyte membrane capacitance of 6.7 pF and a specific membrane capacitance of 1 microF/cm2. We next determined the number of functional L-type calcium channels by nonstationary fluctuation analysis with whole-cell patch clamp. The mean number of functional L-type channels per cell was 291 +/- 49 and 131 +/- 10 with Ca2+ and Ba2+ as the charge carriers, which yielded a channel density of 0.50 +/- 0.08 and 0.28 +/- 0.02 per square micron of cell membrane, respectively. From these data, the density of DHP receptor binding sites was determined to be from 50 to 100 times the density of functional L-type calcium channels. The function of the "excess" DHP receptors and the determination of whether the proportion of functional channels increases with development require further investigation.

Calcium currents in hearts with persistent truncus arteriosus.

We have used the neural crest model of defective heart development to characterize both L- and T-type Ca2+ currents (ICa,L and ICa,T) in ventricular myocytes from embryonic chick hearts with a severe outflow tract anomaly known as persistent truncus arteriosus (PTA). Because of smaller whole embryo weights but no significant change in the weights of ventricles with PTA, the ventricle to whole embryo weight ratios from hearts with PTA were 61% larger than normal at day 11 of incubation. There was a 51% reduction in the peak magnitude of ICa,L at a test potential of +10 mV (-1.4 vs. -0.7 microA/cm2), whereas ICa,T, a proportionately large fraction of the total Ca2+ current in the embryonic chick ventricle, was unaffected. In comparison to sham-operated controls, ICa,L was otherwise not different. Half-activation occurred at about -1 and -41 mV, whereas half-inactivation occurred at -19 and -61 mV for ICa,L and ICa,T, respectively. The time for half-recovery from inactivation were not different and were 200 and 230 ms for ICa,L and ICa,T, respectively. The time for half-decay of the currents and their responses to BAY K 8644 were also similar in both sham-operated and experimental hearts. Although the dihydropyridine receptor binding experiments suggest that the total number of L-type Ca2+ channels was not different, the results from the physiological experiments indicate that the number of functional L-type channels available for opening and/or the single-channel conductance may be reduced in hearts with PTA. Finally, our results with the neural crest model indicate that it is unlikely that the development of Ca2+ currents is influenced by the onset of cholinergic innervation in the heart.

Reduced L-type calcium current in the embryonic chick heart with persistent truncus arteriosus.

Calcium currents were examined in an experimental model in the embryonic chick heart with a congenital malformation known as persistent truncus arteriosus. This is a severe defect characterized by failure of conotruncal and aorticopulmonary septation of the embryonic heart tube. As a result, no separation into the aortic and pulmonary arteries occurs, and there is a common outflow tract. The hearts with persistent truncus arteriosus had a 26% greater ventricular to whole embryo weight, which indicated that the ventricles were enlarged. Both the low-threshold T-type (ICa.T) and the 1,4-dihydropyridine-sensitive L-type (ICa.L) Ca2+ currents were present in the ventricular myocytes from hearts at day 11 of incubation. However, day 11 hearts with persistent truncus arteriosus showed a twofold reduction in the peak magnitude of ICa.L at a test potential of + 10 mV without a concomitant reduction in the number of L channels detected by 1,4-dihydropyridine antagonist [(+)[3H]PN200-110] and agonist (Bay K 8644) receptor binding. The results indicated that an L channel regulatory mechanism other than protein synthesis was affected. These changes are consistent with responses to conditions of excessive hemodynamic burden that have been characterized in adult hearts.

Chronotropic responses of chick atria to field stimulation after various neural crest ablations.

Three models of altered autonomic innervation of the chick heart have been developed in the last few years. These include sympathetically aneural heart, parasympathetically aneural heart, and heart with cholinergic innervation reconstituted from the nodose placodes. The neural status of these hearts has been assessed by a variety of morphological and biochemical methods, but the functional status of innervation is not known. In the present study, we have used electrocardiography and field stimulation to determine the functional neural status of the three different innervation models. The RR and QTc intervals were measured to assess the dominant autonomic tone and autonomic dysfunction in the heart. Even though the RR and QTc intervals were found to be identical in sham and experimental embryos, field stimulation of superfused atria showed that the sympathetically aneural heart has functional cholinergic innervation but lacks any sympathetic response. Hearts from embryos which were parasympathetically aneural lacked a cholinergic response to field stimulation and were judged to be functionally parasympathetically aneural. Hearts with cholinergic ganglia reconstituted from the nodose placodes have normal RR and QTc intervals as well as a normal cholinergic response to field stimulation. The results indicate that these neurons are functionally indistinguishable from neural crest-derived neurons.

Increased muscarinic receptor binding in heart membranes by an inhibitor of protein kinase C.

The number of muscarinic acetylcholine receptors (MAChRs), as detected by binding of [3H]quinuclidinyl benzilate (QNB), was investigated under conditions which promote protein phosphorylation. Incubation of a crude heart membrane preparation in the presence of ATP/Mg2+ reduced MAChR number by 50%. Incubation with polymyxin B, an inhibitor of protein kinase C, blocked the effect of ATP/Mg2+ and increased MAChR number by 74%.

Reduction of muscarinic acetylcholine receptor number and affinity by an endogenous substance.

We examined the soluble fraction from homogenates of 12-day embryonic chick heart for the presence of an endogenous modulator of muscarinic acetylcholine receptors (mAChR). Homogenates were separated into 100,000 g soluble and crude membrane fractions by differential centrifugation. Aliquots of membranes were incubated in the presence or absence of the soluble fraction and the muscarinic antagonist, [3H]quinuclidinyl benzilate ( [3H]QNB), and the data subjected to Scatchard analysis. In the presence of the soluble fraction, mAChR number decreased up to 70% and the affinity for [3H]QNB decreased six- to eightfold. These results suggested that an endogenous soluble factor (ESF) affected cholinergic ligand binding to the receptor. The amount of ESF extracted from less than 10 mg of brain was sufficient to reduce by 50% [3H]QNB binding to 50 fmol mAChR. ESF activity was partially purified by heat and acid treatment. The loss of receptors was dependent upon the amount of ESF added and was time dependent. QNB protected some receptors from loss due to ESF. The change in mAChR affinity for [3H]QNB was observed only if ESF was present continuously during the [3H]QNB binding assay. Ultrafiltration and gel filtration showed that ESF was less than 10,000 daltons and probably less than 700 daltons. ESF activity was blocked by EDTA. However, ESF was not a divalent cation since it was base labile, and removal of divalent cations with Chelex-100 did not inhibit ESF activity. ESF activity was also blocked by catechol, catecholamines, ascorbate, and dithiothreitol. ESF was present in embryonic but not in adult heart.