INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL METHODS RESULTS DISCUSSION REFERENCES

ONE OF THE FUNDAMENTAL TASKS required of implantable medical devices is accurate real-time determination of relevant functional physiological needs. For example, a cardiac pacemaker must determine the pacing rate required to supply the body with adequate cardiac output. Biosensors, which transduce biological actions or reactions into signals amenable to processing, are well suited for such monitoring (16). However, typical in vivo biosensors only approximate physiological function via the measurement of surrogate signals and thereby introduce a prime source of error in biological monitoring; e.g., cardiac pacemakers that use such signals often lack a high degree of dynamic fidelity with chronotropic requirements (3, 15).

A novel alternative approach is to use a biologically based system that can sense physiological signals directly, thereby avoiding the approximation errors associated with surrogate-signal sensing. To this end, we recently reported the development of such a tissue-based biosensor exploiting the endogenous signaling pathways of excitable tissue to couple the detection of in vivo circulating physiological inputs to a functionally responsive electrical output (5). Specifically, these studies focused on the activity and regulation of remotely engrafted neonatal cardiac tissue in a murine model system. Indeed, the chronotropic dynamics of the exogenous excitable cardiac allografts were highly correlated with the activity of the endogenous heart. Moreover, pharmacological trials showed that the transplanted allografts were regulated by circulating catecholamines, suggesting that this approach may offer a foundation for the development of tissue-based biosensors for the detection of a range of blood-borne substances.

The present study was conducted to promote the functional biosensory utility of excitable tissue-based biosensors. We investigated the enhanced vascularization of the transplanted tissues as a means of promoting the chronotropic competence and biosensory potential of this approach. Moreover, to facilitate the possible translation of such biologically based biosensors into experimental or clinical in vivo tools, we advanced this approach significantly by utilizing stem cell-derived myocyte aggregates in place of whole cardiac tissue.

	EXPERIMENTAL METHODS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL METHODS RESULTS DISCUSSION REFERENCES

Cardiac transplant model. Whole heart tissue-based in vivo biosensors were generated, employing a neonatal (24 h old) murine heart transplanted into the pinna of an isogeneic adult murine host, as our laboratory has previously described (1, 7). Sets of mice were pretreated subcutaneously by pinnal injections of recombinant platelet-derived growth factor (PDGF) AB (100 ng/20 µl PBS; R&D Systems), vascular endothelial growth factor (VEGF; 100 ng/20 µl PBS; R&D Systems), or vehicle alone 1 day before mice received cardiac allograft transplants. The following day, a small pocket between the skin and cartilage was dissected toward the tip of the ear with delicate curved forceps. The total donor neonatal heart was excised without the pericardial sac and inserted into the ear pocket. Gentle pressure with the tips of the forceps was applied to the ear to express air from the pocket and facilitate the adherence between donor and recipient tissues. Two days after transplantation, functional blood flow to the transplanted cardiac tissue was assessed by laser-Doppler with a laser flowmeter (model ALF21/21D, Advance, Tokyo, Japan) similar to as has been previously described (17). The PDGF AB dose-response curve of chronotropic activity was also tested by pretreatment of the murine pinnae with a range of PDGF AB concentrations (1, 10, and 100 ng/20 µl PBS) or vehicle alone 1 day before the mice received cardiac allograft transplants. The ears were allowed to heal for 2 days before data acquisition, and chronotropic activity within the first week posttransplantation was measured as described below. The n value was 10 for each experimental set.

Embryonic stem cell-derived cardiac myocyte transplant model. Cardiac cell-based in vivo biosensors were developed with embryonic stem cell-derived cardiac myocytes in place of whole neonatal cardiac tissue. Spontaneously beating cardiac myocytes were derived from E9 murine pluripotent embryonic stem cells (American Type Collection Tissue, Rockville, MD), as previously described (14). Briefly, embryonic stem cells were cultivated on a feeder layer of primary mouse embryonic fibroblasts in DMEM supplemented with nonessential amino acids, L-glutamine, -mercaptoethanol, 20% fetal calf serum, and 100 IU leukemia-inhibiting factor. Droplets of cells (10⁴ cells in 30 µl of culture media without leukemia-inhibiting factor) were pipetted onto the lids of 3-cm bacteriological petri dishes filled with PBS and cultivated for 2 days. The resulting aggregates were transferred from the hanging drops into 6-cm dishes, cultivated for 5 days, and then transferred to 12-well plates. Spontaneous chronotropic myocyte aggregates formed between 5 and 10 days after transfer and were subsequently employed in the murine pinnal transplant model in place of the neonatal cardiac tissue. Mice were pretreated with PDGF (20 ng in 20 µl PBS, n = 37; or vehicle alone, n = 20), as described above. The following day, myocyte aggregates were physically dissociated and suspended in PBS (5 × 10⁴ cells in 20 µl). These suspensions were transferred into the pinnal transplant pocket, which was then sealed via gentle pressure with forceps. Data acquisition for chronotropic activity assessment was performed 3-7 days posttransplantation, as described below.

Electrocardiograms. Between 3 and 7 days posttransplantation, electrocardiogram (ECG) activity of the endogenous heart and exogenous tissue or aggregates were measured after intraperitoneal anesthetization with avertin. ECGs were acquired for a minimum of 30 min via an A-M Systems model 1700 four-channel differential alternating-current amplifier. Signals were band-pass filtered between 3.0 and 100.0 Hz, notch filtered at 60.0 Hz, amplified ×1,000, and sampled at 500 Hz by a National Instruments AT-MIO-16E-10 data acquisition board on a 266 MHz Intel Pentium-II computer running Real-Time Linux (4). Transplant chronotropic activity was defined by two criteria. Sustained activity was characterized by consistent, monomorphic, periodic waveforms that continued for at least 200 s. Sporadic activity was characterized by a range of activity including short-lived, irregular, multimorphic activity, regular activity lasting <200 s, and slow, scattered, monomorphic waveforms that recurred multiple times throughout the recording period.

Quantitative rate analysis. Postacquisition automatic (with manual correction as needed) ECG excitation annotation was performed by using custom Linux C++ software. Excitations were defined as the R waves for the endogenous and exogenous hearts and the myocyte aggregate action potentials for the myocyte aggregates. Mean interexcitation intervals (RR) were computed every 2 s so that the dynamics of the endogenous and exogenous signals, which have different inherent rates, could be compared quantitatively at synchronized time slices.

Endogenous-exogenous cardiac chronotropic correlation. Recordings from the exogenous and endogenous tissue were analyzed for relative (ability of the exogenous myocyte aggregate to sense increasing and decreasing endogenous heart rate trends) and absolute (ability to sense absolute heart rate, i.e., one-to-one correspondence) chronotropic tracking. Discrete data sets of at least 200 s were fit (by using Matlab 5.3.1) to a continuous-time polynomial function as previously described (5). The concordance of the endogenous and exogenous signals was computed as the fraction of the time that their derivatives (computed analytically from the fitted polynomial function) had the same sign. A concordance of >0.70 was employed as a measure of the ability of the exogenous tissue or aggregate to track the increases and decreases in endogenous rate. Absolute chronotropic correlation was measured by the correlation coefficient computed between each exogenous and corresponding endogenous time series.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL METHODS RESULTS DISCUSSION REFERENCES

Enhanced vascularization and optimization of tissue-based chronotropic kinetics. The potential of proangiogenic cytokines to enhance the vascularization and chronotropic biosensing by the cardiac allografts was assessed. Rheological measurements revealed that hearts transplanted into the pinnae pretreated with PDGF received over twice as much blood flow as those allografts implanted in control and VEGF-pretreated mice (Fig. 1). On the basis of these results, we evaluated the role of PDGF pretreatment in the development of chronotropic activity in the transplanted hearts. These studies revealed that almost twice as many exogenous hearts in the mice injected with 10-100 ng of PDGF had spontaneous chronotropic activity compared with control and 1-ng pretreatment groups (Fig. 2A). In addition to measuring spontaneous beating, the ability of the transplanted cardiac tissue to emulate the chronotropic dynamics of the endogenous heart rate was also measured. These studies demonstrated similar average relative tracking concordance in all the transplants with spontaneous chronotropic activity regardless of the pretreatment groups (Fig. 2B), suggesting that the vascular threshold for the development of chronotropic activity is linked to the biosensory potential of the intact cardiac tissue transplants.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Platelet-derived growth factor (PDGF) AB induction of cardiac allograft rheology. Laser-Doppler measures of blood flow in the transplanted cardiac tissue after pinnal pretreatment with PDGF AB, vascular endothelial growth factor (VEGF), or vehicle alone are shown. *P < 0.05 vs. control and VEGF pretreatments.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   PDGF enhancement of cardiac tissue chronotropic activity. Open bars, percentage of neonatal cardiac allografts demonstrating spontaneous electrocardiographic activity after pinnal pretreatment with PDGF (absolute percentage of positive trials); filled bars, relative tracking concordance of chronotropically competent allografts and endogenous electrocardiographic dynamics (means ± SE). * P < 0.05 vs. control and 1-ng pretreatments.

Embryonic stem cell-derived cardiac myocyte biosensory potential. Defining a set of host manipulations that enhance the kinetics of tissue-based chronotropic dynamics offered a foundation for developing a cell-based, as opposed to a whole heart-based, system to act as a biosensor of physiological activity. To this end, embryonic stem cell-derived cardiac myocytes were generated and implanted into the murine pinnae in place of the neonatal cardiac tissue. Chronotropic dynamics were recorded as described above for the cardiac allograft experiments. The majority of cellular transplants in both pretreatment groups demonstrated spontaneous or sustained electropotential activity (19 of 20 control transplants; 30 of 37 PDGF transplants). Moreover, approximately one-quarter of both of these electrically viable cellular transplants demonstrated sustained depolarizations (Fig. 3). However, unlike the whole heart allografts, pretreatment of the hosts with PDGF did not alter the development of chronotropic activity of the transplants, suggesting that the myoctyes received a sufficient vascular supply to maintain rhythmic electopotentials in the intact host.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Embryonic stem cell-derived cardiac myocyte transplants demonstrated sustained electropotential activity after pinnal pretreatment with PDGF or vehicle as a percentage of all allografts with electropotential activity (filled bars). Percentage of cardiac myocyte transplant spontaneous electropotential activity demonstrated concordant relative tracking (open bars; 70%; absolute percentage of positive trials). *P < 0.05 vs. control pretreatment.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Cardiac myocyte-based relative chronotropic biosensing activity. Representative example of the mean interexcitation intervals (RR) vs. time for the endogenous heart (A) and embryonic stem cell derived-cardiac myocyte aggregate transplant (B) in a mouse pretreated with PDGF is shown. Insets, segments of the respective endogenous and aggregate electrocardiograms. First-order derivatives vs. time (dRR/dt) of the polynomial fits of the RR dynamics (C) demonstrated an 80% concordance in sign for the trial and, therefore, a high degree of relative sensing ability. endo, Endogenous; exo, exogenous.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Representative example of embryonic stem cell-derived cardiac myocyte aggregate RR vs. endogenous RR in a mouse pretreated with PDGF. For this trial, the correlation coefficient (r = 0.92) indicated a high degree of absolute tracking ability.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Representative example of embryonic stem cell-derived cardiac myocyte aggregate RR vs. endogenous RR before and after administration of clonidine. For this trial, the linear relationship of endogenous RR and exogenous RR was 2.8 before clonidine (dashed line) and 1.8 after clonidine (solid line), indicating a more profound inhibition of the signals regulating endogenous chronotropic activity.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL METHODS RESULTS DISCUSSION REFERENCES

The results of our studies demonstrate the feasibility of a new class of biosensors: in vivo biologically based biosensors that utilize excitable cells for direct physiological measurement of the circulating inputs of endogenous chronotropic demand. Importantly, we found that specific proangiogenic host interventions that increased the vascularization of the transplanted tissue markedly enhanced the development of biosensor activity. Moreover, the extension of our biosensor approach to employ genetically plastic stem cell technology to detect the blood-borne signals that regulate the endogenous heart rate should facilitate the development and potential clinical translation of this approach for the direct biological detection of physiological as well as pathophysiological signals.

Our studies exploited the endogenous PDGF-dependent communication between cardiac myocytes and endothelial cells (8) to promote the neovascularization of the engrafted biosensors as a means of increasing detection of the blood-borne physiological signals regulating heart rate. PDGF pretreatment of the host implantation sites specifically enhanced cardiac allograft rheology and improved the chronotropic competence of the neonatal heart transplants. The relative tracking potential of the active cardiac allografts, however, was not altered by the increased blood flow, revealing that the vascular threshold for allograft chronotropic concordance was directly linked to electropotential activity. Moreover, these results suggested that similar host tissue engineering could facilitate the advancement of more elemental cellular approaches to develop biological biosensors. Indeed, vascular enhancement allowed the embryonic stem cell-derived cardiac myocytes to act as high-fidelity in vivo biosensors of the circulating humoral inputs of chronotropic dynamics of the endogenous heart.

One potential application of biologically based biosensors is to serve as the chronotropic-sensing element for implantable pacemakers. By utilizing the inherent ability of cardiac myocytes to regulate chronotropy by setting electronic pacing rate according to sensed humoral signals, such a pacemaker would avoid the approximation errors associated with the surrogate-signal rate estimates utilized by current rate-adaptive pacemakers.

Additionally, the interaction between the enhancement in myocyte aggregate neovascularization and biosensory potential may be of increased importance in individuals with a significant prevalence of chronotropic incompetence (12, 20). Because previous studies have demonstrated that the angiogenic development of new blood vessels decreases with aging (18), specific strategies targeted at enhanced vascular activity may be critical in the translation of biologically based biosensor approaches to improve chronotropic treatments.

We recognize that developing the utility of cardiac myocyte-based biosensors will require that the cardiac myocytes be derived from autologous sources of stem cells, such as the endogenous bone marrow. Recent murine studies have demonstrated that cardiac myocytes can be derived from bone marrow cells (13). These techniques may allow for the potential clinical translation of cell-based chronotropic biosensor systems. Further advances in in vivo biosensors might employ such cells on silicon chips or other defined biocompatible materials (6, 9, 11) to ensure that the longevity of the cell-based sensors is similar to that of whole heart transplants, which can remain viable for the life span of the host (10, 19).

We project the utility of this approach will extend beyond that of chronotropic regulation biosensing. Molecular engineering may offer a means for the detection of physiological and pathophysiological signals that do not routinely alter cardiac chronotropy. Indeed, excitable cell biosensor systems could lead to the development of long-term, physiologically tuned, functionally integrated bioprocessing interfacing with a range of external or implantable devices to facilitate the rapid initiation of appropriate actions.