INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CHAMEIDES ET AL. (8) and McCue et al. (18) noted more than 20 years ago that women who gave birth to children with congenital heart block (CHB) often had autoimmune diseases. It is now well established that CHB detected before or at birth, in the absence of structural cardiac abnormalities, is strongly associated with maternal antibodies to SSA/Ro and/or SSB/La ribonucleoproteins (6). Though more prevalent in the presence of autoimmune disorders, CHB in offspring is independent of whether the mother has systemic lupus erythematosus, Sjögren's syndrome, or is totally asymptomatic (6, 26). Damage to the cardiac conduction system occurs in an otherwise normal fetus and is presumed to arise from transplacental passage of these maternal autoantibodies (immunoglobulin; IgG) (25). Other neonatal abnormalities affecting the skin, liver, and blood cells have also been reported (26) and are associated with these maternal IgG. Although varying degrees of conduction block can occur, third-degree atrioventricular (AV) block, which is irreversible, carries substantial morbidity and mortality that approaches 30%. More than 60% of the children affected by AV block require lifelong pacemakers (26). Noncardiac manifestations are transient, resolving ~6 mo after birth, coincident with the disappearance of maternal IgG from infant circulation (7).

The candidate antigens and their cognate antibodies have been extensively characterized at the molecular level. The 60-kDa SSA/Ro contains a putative zinc finger and an RNA-binding protein consensus motif (2), both of which could account for its direct interaction with small cytoplasmic hY-RNAs, a class of low-molecular-weight molecules of 83-112 bases, which are small uncapped cytoplasmic RNA (16). This protein may function as part of a novel quality control or discard pathway for 5S rRNA precursors in Xenopus oocytes (22). Many sera, which recognize the 60-kDa SSA/Ro protein, also react with another protein of 52 kDa (1). Anti-SSB/La antibodies recognize a 48-kDa polypeptide, which does not share antigenic determinants with either 52-kDa or 60-kDa SSA/Ro (9). SSB/La facilitates the maturation of RNA polymerase III transcripts, directly binds a spectrum of RNAs, and associates at least transiently with 60-kDa SSA/Ro (13).

We recently (3, 4) provided evidence that IgG-enriched fractions and anti-52-kDa SSA/Ro antibodies affinity purified from sera of mothers whose children have CHB induce complete AV block in Langendorff perfused hearts and inhibit the L-type Ca²⁺ current (I_Ca,L). In addition, immunization of female BALB/C mice with recombinant ribonucleoproteins generated high-titer antibodies, which crossed the placenta during pregnancy and were associated with varying degrees of AV conduction abnormalities in the pups, including complete AV block (4, 17, 20). We unexpectedly found significant sinus bradycardia in mouse pups born to mothers injected with human maternal antibodies (17) or immunized with recombinant antigens (4, 20). These findings brought clinical attention not only to AV nodal disorders but also to sinoatrial (SA) nodal conduction abnormalities. In this regard, Brucato et al. (5) recently reported similar sinus bradycardia in infants born to mothers seropositive to anti-SSA/Ro antibodies, suggesting that the spectrum of conduction abnormalities associated with maternal IgG extend beyond the AV node.

The electrophysiological mechanisms by which maternal antibodies affect electrical propagation throughout the intact heart remain unclear. This may be due in part to the inability to map electrical activity patterns by using conventional microelectrode techniques in isolated multicellular preparations (3). Furthermore, whereas the primary electrophysiological consequence of maternal antibodies is AV block, little, if anything, is known about the effects of these maternal IgG on both atrial and ventricular activation and repolarization patterns. Sinus bradycardia might well be an early manifestation of CHB and may prove to be a critical clinical marker preceding the onset of advanced AV conduction disorders in autoimmune-associated CHB. Driven by these observations, we mapped cardiac activation patterns in Langendorff-perfused young rabbit hearts (atria and ventricles) exposed to maternal IgG. Optical mapping techniques, which reveal the two-dimensional spread of activation, were used to study propagation abnormalities by using a 124-element photodiode array during sinus and paced rhythms.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surgical procedure and experimental setup. The rabbit model was chosen for several reasons. First, the dimensions of the heart are well suited for the spatial resolution of the optical system and permit imaging of large planar surfaces of the heart and specific cardiac structures (15, 24). Second, the optical resolution of the system can be adjusted from 400 to 1,800 µM/pixel, depending on the dimension of the region of interest, and electrophysiological observations can be correlated with high resolution with the anatomic features of the heart (10). Third, the electrophysiological properties of the ionic currents that constitute the rabbit action potential (AP) have been well characterized and may better represent the human heart rather than smaller animals (21).

1

1

View larger version (91K): [in this window] [in a new window]   Fig. 1.   Experimental setup. A: view of right atrium and ventricle. B: superimposition of optical grid on heart. 1, left ventricle; 2, right ventricle (RV); 3, left atrium; 4, right atrium (RA); 5, superior vena cava (SVC); 6, inferior vena cava (IVC); and 7, aorta.

Stimulation and electrocardiogram. Stimulation electrodes were placed at the right atrial appendage near the superior vena cava. The bipolar stimulation electrodes were composed of Teflon-coated silver wires (75-µm diameter) with an exposed length of 500 µm. The hooked ends were inserted into the myocardium with an interpolar distance of <0.5 mm. Stimulation was applied from a constant current source at two times diastolic threshold.

IgG purification. Purification of IgG was performed as previously described (3, 4, 17, 20). Briefly, immunoglobulin fractions containing IgG were purified from serum by protein A-sepharose columns and confirmed pure by electrophoresis. Throughout this study, we tested IgG from mothers (n = 3) whose children have CHB and contain antibodies against 48-kDa SSB/La, 52-kDa SSA/Ro, and 60-kDa SSA/Ro, as tested by enzyme-linked immunosorbent assay (ELISA) and immunoblot (3, 4, 17, 20). This will be referred to in the text as positive IgG. Control IgG (negative IgG) was purified from sera of healthy mothers (n = 3) with healthy children who tested negative for anti-SSA/Ro and anti-SSB/La antibodies by ELISA and immunoblot, as described above (3, 4, 17, 20). An IgG concentration of 800-1,200 µg/ml was used in the experiments. The selected doses were based on our previous observations (3), in which a complete AV block was induced at 800 µg/ml in the rat heart.

Data analysis. All data are presented as means ± SE. Statistical tests were performed by using Systat for Windows, version 8 (SPSS; Chicago, IL). AP characteristics were compared by Student's t-test for paired and unpaired data where appropriate. Differences in activation time (total atrial/ventricular activation time) and cardiac intervals (P-R interval and R-R interval) were compared by ANOVA. Data before and after IgG application were compared by using ANOVA for repeated measures. If ANOVA failed to reach statistical significance, Bonferroni's correction for paired data was also applied to test for effect.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, 15 isolated hearts were used with 3 hearts used for sham study. Ten experimental hearts were exposed to 800 µg/ml positive IgG, and in two experiments a dose of 1,200 µg/ml positive IgG was used. Of the 10 hearts exposed to 800 µg/ml positive IgG, 4 hearts were studied by using a high data acquisition rate (0.64 ms/sample) to compare dF/dt and APD₉₀ before and after IgG. In the sham experiments, hearts were exposed to 1,200 µg/ml negative IgG (i.e., control IgG obtained from healthy mothers with healthy children) and considered as controls for positive IgG. As shown in Table 1, there was no significant change in ECG parameters from control in these hearts. In 9 of 12 hearts exposed to positive IgG, complete heart block (3° AV block) was obtained. Only partial recovery was observed when the hearts were perfused with positive IgG-free modified Tyrode solution for up to 45 min. The summary of ECG measurements is shown in Table 1. Table 2 compares dF/dt and APD₉₀ before and after positive IgG for four hearts. In these experiments, the same six sites were selected and the data were pooled (n = 24). The right atrial sites were chosen from the free right atrial wall, distant from the vicinity of the SA. There was a modest increase of 5% in APD₉₀ in the right atrium and an increase of 15% in the imaged portion of the right ventricle. Both effects were statistically significant. On the other hand, there was no statistically significant effect of positive IgG on dF/dt in both the right atrium and ventricle.

Figure 2 shows ECG recordings from a typical experiment that produced CHB in the presence of positive IgG (800 µg/ml). The control ECG shows normal sinus rhythm with a P wave, followed by a QRS complex at a spontaneous heart rate of 160 beats/min (Fig. 2A). The P-R interval was 67 ms. Fifteen minutes after infusion of positive IgG, the heart rate slowed to 113 beats/min and the P-R interval increased to 180 ms (Fig. 2, trace B). Note that the QRS complex duration was essentially unaffected by positive IgG infusion. Fig. 2, traces C and D, shows 2° AV block developed 19 min after start of positive IgG infusion and 2° AV block with a 4:1 pattern at 24 min, respectively. Figure 2E shows partial recovery of heart block 45 min after perfusion with positive IgG-free Tyrode solution. However, sinus bradycardia and 1° AV block at a heart rate of 138 beats/min were maintained.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Electrocardiogram (ECG) recordings from Langendorff perfused rabbit heart before and after positive immunoglobulin (IgG) (800 µg/ml). A: control; B: slowing of sinus rate after positive IgG; C: second-degree heart block in the same heart with a 2:1 pattern, which later evolved into 4:1 AV block (D). After washout, sinus bradycardia persisted (E). Arrows, P waves.

Figure 3 shows atrial and ventricular activation maps, along with selected optical APs during fixed atrial pacing (A₁A₁ = 250 ms) in the presence of positive IgG (800 µg/ml). Optical AP were mapped using a photodiode array (124 sites) and di-4-ANEPPS. Stimulation was applied in the sinus node region (see asterisk in Fig. 3). The map projections shown in the top panels are right-sided projections of the epicardial heart surface in control (Fig. 3, left) and in the presence of positive IgG (Fig. 3, right). A rough schematic of the right atrium is outlined in the lower portion of the maps with anatomic landmarks indicated. Optical AP were acquired to simultaneously record from the right atrium and ventricle. In the activation maps, each shaded region correspond to an isochronal area activated each successive millisecond. The arrows indicate the direction of activation wave front propagation. In the control right atrial map, total activation time was 4 ms; total activation time for the imaged portion of the right ventricle was 5 ms. Complete heart block was obtained after positive IgG infusion (Fig. 3, right). In the right atrial activation map, there was considerable delay at the SA junction, as indicated by the crowded isochrones. Whereas right atrial activation time was 9 ms, most of the delay was at the SA junction. The majority of the remaining atrium was activated within 3 ms, as in control. The right ventricular activation time for the ventricular escape beat was 4 ms. The ECG and a selected atrial and ventricular optical action potential during control (left) and positive IgG infusion (right) are shown at the lower portion of Fig. 3.

View larger version (43K): [in this window] [in a new window]   Fig. 3.   Optical maps of atrial and ventricular activation along with selected optical action potentials (AP) during fixed atrial pacing at a rate of 240 beats/min. Stimulation (*) was applied in the sinus node region. Rough schematic of RA is outlined in the lower portion of the maps with anatomic landmarks indicated. Activation maps are right-sided projections of epicardium in control (left) and in the presence of 800 µg/ml positive IgG (right). Each isochronal zone (shaded region) corresponds to an isochronal time of 1 ms. Color code key for atrial and ventricular isochrone intervals is shown in between the 2 maps.

The role of atrial conduction disturbances during spontaneous sinus rhythm in control and after positive IgG perfusion (1,200 µg/ml) is shown in Fig. 4. The topology of these maps was similar for all six hearts that developed 3° AV block. The control right atrial map (Fig. 4, left) shows that the right atrium was activated within 5 ms. The spontaneous atrial rate was 180 beats/min with 1:1 AV conduction. The atrial isochrones were evenly spaced, indicating that there were no conduction disturbances in the right atrium. The right ventricle also activated within 5 ms. The activation pattern was from apex to base, indicating normal sinus activation of the ventricles. Perfusion of positive IgG resulted in sinus bradycardia, followed by complete AV block corresponding to an atrial rate of 144 beats/min. Examination of the atrial activation map (Fig. 4, right) shows that sinus bradycardia was associated with marked conduction delay at the SA junction. The activation isochrones were crowded in the SA region, similar to the paced example shown in the Fig. 3. Total right atrial activation time increased to 7 ms. The right ventricle activated within 6 ms and was similar in activation sequence and timing to the control map. In this and all other experiments, sinus bradycardia was associated with only minor changes in overall atrial and ventricular activation patterns.

View larger version (41K): [in this window] [in a new window]   Fig. 4.   Optical maps of atrial and ventricular activation along with selected optical AP during spontaneous sinus rhythm in control and after positive IgG (1,200 µg/ml) perfusion. Left: control RA and ventricular activation maps. No conduction disturbances were evident in the right atrium in control. RV had a normal sinus activation pattern. Right: positive IgG caused third-degree heart block with a ventricular escape rhythm. Sinus bradycardia was associated with marked conduction delay at the sinoatrial junction as indicated by the crowded isochrones in the sinoatrial region. Maps and abbreviations are the same as in Fig. 3.

To understand the electrophysiological effects of positive IgG better, optical APs were obtained at a higher magnification in the region of the sinus node and SA junction. The results from a representative experiment, in which complete AV dissociation was obtained in the presence of positive IgG (800 µg/ml), are shown in Fig. 5. Optical APs are from one column of the grid. Figure 5, traces A-D, shows that most of the conduction occurred in the sinus node to SA junction region. In this zone, there was a reduced upstroke in optical AP. Optical AP from the free right atrial wall before and after positive IgG were similar, indicating that positive IgG had little effect on the rest of the atrium. Although there was a modest increase (15%) in APD₉₀, there was no statistically significant effect on dF/dt at this concentration of positive IgG, indicating that IgG had little effect on the conduction in the imaged portion of the right ventricle.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Optical AP obtained at a magnification of 0.4 mm/pixel along one column of the grid in the presence of positive IgG (800 µg/ml) going from A (sinus node region) to L (free RV wall).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The primary finding of this study is that positive maternal IgG causes bradycardia in the isolated Langendorff, perfused young rabbit heart. Whereas complete heart block is a hallmark of this potentially fatal congenital disorder, bradycardia was a consistent finding in this study. Along with slowing of the intrinsic sinus rate, optical mapping revealed marked conduction delays for propagation of the sinus impulse to the atrium. Comparison of the activation maps and dF/dt measurements indicates that there was no effect on conduction in other parts of the right atrium and the mapped portion of right ventricle. The fact that conduction is affected at the SA junction is not unlike the documented concept that these IgG target specific cardiac conduction systems in the heart, such as the AV node. The reason that sinus node and SA dysfunction have been overlooked until recently may be related to the severity of the disease during high-level AV nodal block and the limited number of electrophysiological studies in this relatively rare congenital disorder.

Whereas optical mapping of hearts with CHB need to be investigated in detail at many levels, our attention to atrial and SA conduction disorders in the present study was precipitated by the recent and somewhat unexpected findings from our laboratory, that sinus bradycardia was more common in animal models of CHB than previously realized (4, 17). These findings have been extended to the clinical settings, where sinus node dysfunction in infants of mothers with anti-Ro antibodies has been documented (5, 19). It is tempting to speculate that because of the significant mortality and morbidity associated with AV node dysfunction, the role of sinus node dysfunction may well have been overlooked for years.

There are two implications to the findings of this study. First, sinus bradycardia may have long been an early electrophysiological consequence in the CHB syndrome. Our results, along with the observations of others (5, 19), hopefully will draw more attention of health care providers to the potential critical role of sinus bradycardia for the early diagnosis of CHB. Second, because it appears that maternal IgG target specific conduction systems in the heart, there is a continued need for basic scientific investigation into the histopathogical mechanism by which these antibodies interact with systems responsible for the initiation of the cardiac impulse in the SA node and its propagation to the atria and ventricles.

Proposed mechanisms of tissue injury. Whereas the molecular and biochemical mechanisms of CHB have been explored to a certain extent (see Ref. 7 for review), the electrophysiological mechanisms of CHB are just emerging (3, 4, 12, 17, 20). Work from our laboratory was based on the initial observation of Garcia et al. (12), who showed that the use of a single-surface ECG that conduction disorders associated with neonatal lupus could be reproduced in an isolated adult rabbit heart. They further showed that anti-SSA/Ro positive sera induced a reduction in I_Ca,L. Subsequently, and unlike Garcia et al. (12), we used a large panel of sera only from mothers whose children have CHB and tested their electrophysiological effects on human fetal heart (4). We provided evidence that IgG-enriched fractions and anti-52-kDa SSA/Ro antibodies affinity purified from sera of mothers whose children have CHB induce complete AV block in the fetal heart perfused by the Langendorff technique and inhibit I_Ca,L at the whole cell and single channel level (3, 4). In addition, immunization of female BALB/C mice with recombinant proteins generated high-titer antibodies, which crossed the placenta during pregnancy and were associated with varying degrees of AV conduction abnormalities, including complete AV block, in the pups (4, 17, 20). Because conduction in the SA and AV node is essentially dependent on Ca electrogenesis, sinus bradycardia, and AV block would be expected to result from interventions that reduce Ca²⁺ currents. The possibility that these IgG interact with other currents in the SA or AV node cannot be ruled out. We propose that chronic exposure of Ca²⁺ channels to maternal antibodies during pregnancy could lead to Ca²⁺ channel internalization, degradation, cell death, and eventually fibrosis and calcification, as has been reported from the autopsies of affected children (14). Because cardiac myocytes do not undergo any significant mitosis after birth, affected cells are not replaced, and thus this may explain the irreversibility of the 3° AV block in human infants. The partial recovery of AV block reported in this study is likely due to the acute exposure of the antibodies, unlike chronic exposure during the pregnancy term. Altogether, these findings strongly suggest that maternal IgG are causally related to a number of electrophysiological abnormalities in the development of CHB. However, it should be emphasized that not all mothers with anti-SSA/Ro-SSB/La antibodies have affected children. This discordance may be related to undocumented electrophysiological effects or subclinical aspects of autoimmune-related CHB indicating that additional factors are required to promote disease expression.