Indirect evidence suggests that fluctuations, or oscillations, in the intracellularfree calcium concentration ([Ca2+] (4)(5)(6).Oscillations in force, membrane potential, and related parameters can also occur spontaneously-i.e., not coupled to any prior electrical stimulation (see references in refs. 3, and 7-12).A large body of evidence also points to Caei+ as the cause of these oscillations: spontaneous mechanical oscillations correlate closely with those in membrane potential (or current) (7); both mechanical and electrical fluctuations occur most readily under conditions of increased Ca2" loading; conversely, both are attenuated by intracellular EGTA injection, presumably by che-

The systolic and diastolic properties of single myocytes and intact papillary muscles isolated from hearts of adult rats and rabbits were examined at 37'C over a range of stimulation frequencies and bathing [Ca 21] . (Ca.) . In both rabbit myocytes and intact muscles bathed in 1 mM Ca., increasing the frequency of stimulation from 6 to 120 min-' resulted in a positive staircase of twitch performance . During stimulation at 2 min-', twitch performance also increased with increases in Ca . up to 20 mM. In the absence of stimulation, both rabbit myocytes and muscles were completely quiescent in G15 mM Ca . . Further increases in Ca . caused the appearance of spontaneous asynchronous contractile waves in myocytes and in intact muscles caused scattered light intensity fluctuations (SLIF), which were previously demonstrated to be caused by Ca 2 '-dependent spontaneous contractile waves . In contrast to rabbit preparations, intact rat papillary muscles exhibited SLIF in 1 .0 mM Cao . Two populations of rat myocytes were observed in 1 mM Cao : -85% of unstimulated cells exhibited low-frequency (3-4 min-') spontaneous contractile waves, whereas 15%, during a I-min observation period, were quiescent . In a given Ca., the contractile wave frequency in myocytes and SLIF in intact muscles were constant for long periods of time . In both intact rat muscles and myocytes with spontaneous waves, in 1 mM Ca., increasing the frequency of stimulation from 6 to 120 min -' resulted, on the average, in a 65% reduction in steady state twitch amplitude . Of the rat myocytes that did not manifest waves, some had a positive, some had a flat, and some had a negative staircase ; the average steady state twitch amplitude of these cells during stimulation at 120 min-' was 30% greater than that at 6 min-' . In contrast to rabbit preparations, twitch performance during stimulation at 2 min-' saturated at 1 .5 mM Cao in both intact rat muscles and in the myocytes with spontaneous waves. We conclude Address reprint requests to Dr . Edward G.

In quiescent rat ventricular myocardium, bathed in solution of 2 mM Ca++ or less, it has been previously demonstrated that spontaneous microscopic oscillatory cell motion is present and interacts with an incident laser beam to produce scattered light intensity fluctuations which can be monitored to quantify the underlying motion. The present study shows that scattered light intensity fluctuations are not present under any conditions in frog atrial or ventricular preparations, but do occur in each type of mammalian cardiac tissue studied in the unstimulated state. The magnitude of scattered light intensity fluctuations in mammalian tissues varies with species and cellular Ca++ loading. In some tissues, e.g., rabbit or ferret ventricle, either an increase in the Ca++ concentration in the perfusate [( Ca++]e), reduction of perfusate Na+ concentration [( Na+]e), or addition of cardiac glycosides was required to elicit scattered light intensity fluctuations; in other tissues, however, e.g., the canine Purkinje fiber, atria, and ventricle, and guinea pig atria, scattered light intensity fluctuations were present at 2 mM [Ca++]e in the absence of experimental Ca++ loading. Scattered light intensity fluctuations were not affected by LaCl3, or verapamil, and were reversibly abolished by caffeine. When the pCa in the myofilament space is kept constant in detergent "skinned" fibers, scattered light intensity fluctuations are not present during contractile activation. We conclude: that scattered light intensity fluctuations are due to spontaneous intracellular Ca++ oscillations that require a functional sarcoplasmic reticulum; that the potential to exhibit these oscillations is a fundamental property of mammalian excitable cardiac cells; and that, in many mammalian tissues, these oscillations are present in the unstimulated state, even in the absence of experimental perturbations to enhance cell Ca++ loading.

Laser light scattered by nonstimulated rat cardiac muscle bathed in physiological saline containing a [Ca"] of 0.4-2 .5 mM displays scattered-light intensity fluctuations (SLIF) ; the frequencies of both SLIF and resting force are Ca" dependent. Direct inspection of these muscles by phase-contrast microscopy under incoherent illumination revealed the presence of spontaneous asynchronous cellular motions that are also Ca" dependent. The physical properties of the scattered light are compatible with the hypothesis that SLIF are due to the diastolic motion, except for the dependence on scattering angle, which may be perturbed because the muscles are optically thick. To determine whether diastolic SLIF and motion are an intrinsic property of activated myofilaments, photon-counting autocorrelation of the scattered light was performed both in rat right-ventricular papillary muscles skinned with the detergent Triton X-100 (1%) and in muscles with intact membranes under conditions that alter cellular Ca++ fluxes . In skinned muscles activated over a range of Ca++ from threshold to maximum force production, neither SLIF nor asynchronous motion was observed when Ca++ was buffered to constant values . In intact muscles the frequency of SLIF and the amplitude of diastolic motion were (a) markedly increased by substituting K+ or Li+ for Na+ in the bath ; (b) not altered by verapamil (1 AM); and (c) reversibly abolished by caffeine (?10 mM). These properties are exactly those of mechanical oscillations that have been observed in isolated cardiac cell fragments, which are the result Ca++ oscillations caused by Ca++ release from the sarcoplasmic reticulum (SR). We infer that mechanical oscillations caused by spontaneous Ca++-induced Ca++ release from the SR occur in intact nonstimulated cardiac muscle even in the absence of Ca++ overload and are the principle cause of SLIF, and that myoplasmic [Ca++] in "resting" muscle is not in a microscopic steady state.