Growth Control
As reviewed in the first chapter, the clinical use of externally generated electrical currents to enhance healing or retard tumor growth was common in the latter half of the nineteenth century. While the technique rapidly fell into disfavor in the early decades of the present century with the mounting evidence against electrical properties of living things, some laboratory studies were continued. Frazee, for example, studied the effect of passing electrical current through the water in which salamanders were kept. In 1909 he reported that this appeared to increase the rate of limb regeneration in these animals (27). In their long series of investigations extending from the 20's through the 40's, both Burr and Lund reported growth effects of applied electrical currents on a variety of plants and animals. Some of their observations were confirmed and extended by Barth at Columbia University (28).
In 1952 Marsh and Beams reported on an interesting series of experiments on Planaria, a species of relatively simple flatworm with a primitive nervous system and simple head-to-tail axis of organization (29). As expected, electrical measurements had indicated a simple head-tail dipole field. This animal had remarkable regenerative powers; it could be cut transversely into a number of segments, all of which would regenerate a new total organism. Even more remarkable, the original head-tail axis would be preserved in each regenerate, with that portion nearest the original head end becoming the head of the new organism. Marsh and Beams postulated that the original head-tail electrical vector persisted in the cut segments and that it provided the morphological information for the regenerate. If this was so, then reversal of the electrical gradient by exposing the cut surface to an external current source of proper orientation should produce some reversal of the head-tail gradient in the regenerate. While performing the experiment they found that as the current levels were increased the first response was to form a head at each end of the regenerating segment. With still further increases in the current the expected reversal of the head-tail gradient did occur, indicating that the electrical gradient which naturally existed in these animals was capable of transmitting morphological information.
A few years later Humphrey and Seal attempted a scientific evaluation of the old clinical techniques of electrical control of tumor growth (30). It had been observed many times that rapidly growing tissues were electrically negative in polarity, with tumors being the highest in magnitude. Many of the old clinical techniques therefore applied positive potentials and currents on the theory that the opposite polarity should slow or stop the growth. Using rats with implanted malignant tumors, Humphrey and Seal applied anodes of copper or zinc over the tumor masses and passed currents averaging mamp for a period of 3 hours per day. In their most impressive series of 18 control and 18 experimental animals, the mean volume of the tumors in the controls was 7 times greater than in the experimentals after 24 days of treatment, and all of the controls died by day 31 while 7 of the treated animals demonstrated complete tumor regression and survived for more than a year thereafter. All of these experiments were based on the postulated generalized total body bioelectric field and the fact that this concept was not generally accepted by the body of science precluded any serious clinical consideration of these findings.
The only work that applied Szent-Gyorgyi's concepts in part was that of Huggins and Yang who showed quite conclusively that carcinogenic (cancer-producing) agents produced their effect by a combination of their steric organization and their capacity for electron transfer (31). In their view these agents were, because of their size and shape, able to attach to certain areas of the cell surface and then effect the electron transfer. Compounds of the same steric property but lacking the electron transfer capability were not carcinogenic. Except for this work of Huggins and Yang all other reports of growth effects of actual electrical currents could not be placed in a frame of reference that was acceptable to the scientific establishment.
The demonstration of the existence of intrinsic electrical currents and particularly their localization to the nervous system permitted the problem to be viewed in a new light. It had been a long-standing clinical observation that healing was delayed and often defective in areas that were deficient in innervation. With the idea that the nervous system in some ways possessed a growth-controlling function, Hasson reasoned that there should be a similar relationship between denervation and tumor formation (32). In 1958 he reported that tumor induction by carcinogenic agents was facilitated by denervation, and these observations were subsequently confirmed in 1967 by Pawlowski and Weddell (33).
By the mid-1950's Singer published a series of papers in which he demonstrated the dependence of limb regeneration in the salamander on the presence of a threshold amount of nerve tissue in the amputation stump (34). He was even able to produce a small amount of limb regeneration in the adult frog (normally a nonregenerator) by transplanting additional functioning nerve tissue into an amputation stump. The evidence suggested that the nerves somehow controlled normal growth; in their absence normal growth was inhibited and abnormal growth was facilitated. None of these observations made much sense under the nerve impulse concept of neural functioning-in fact nerves to a healing area are usually "silent," with very little action potential traffic. However, when the previous reports of the growth controlling properties of direct currents were combined with the localization of the intrinsic DC system to the nerves, the relationships between the nervous system and growth began to make some sense.
This relationship is particularly clear in the area of regenerative growth, where considerable research has now been conducted over the past two decades. Regeneration is the most dramatic and important of the growth/ healing processes, being the actual regrowth of missing parts in full anatomical detail. It is most common in the lower animals-the salamander limb regeneration preparation being the one most frequently used in research- and it diminishes as one ascends the evolutionary scale. In the human, true regeneration is limited to the healing of fractures of the long bones; other processes commonly called regenerative (i.e., skin and peripheral nerve fiber) are simply increased rates of cellular multiplication or growth. The essence of the true regenerative process is the appearance at the site of injury of a mass of primitive, presumably totipotent cells, called the blastema. After reaching a critical size this cellular mass begins to grow in length and to redifferentiate to produce the multicellular, multitissue, complex missing structure. The capability of the process is best indicated by noting that the salamander foreleg is anatomically equivalent to the complex human arm.
Regeneration was first formally reported by Spallanzini in 1768 and has been the subject of study ever since, with many attempts being made to restore the process to animals normally lacking it. The first successful attempt was reported by Rose in 1944 when he produced a small measure of regeneration of the amputated foreleg of the frog (a species that, despite folk-lore, cannot regenerate an extremity) by dipping the extremity daily in hypertonic saline (35). Two years later, a similar result was reported by Polezhayev (36) in the same animal, by repeatedly needling the stump daily. While it was the intent of both of these investigators to delay the overgrowth of skin over the end of the amputation stump, the procedure used in each experiment was obviously repeatedly traumatic. Ten years later Singer, again using the amputated foreleg of the adult frog, obtained the same amount of regeneration by surgically augmenting the nerve supply to the extremity. There seemed to be little relationship between these two stimulating factors-increased injury and increased nerve-until in 1958 when Zhirmunskii reported that the current of injury was directly related to the extent of innervation (37). This observation, coupled with the much earlier work of Matteucci that indicated a direct relationship between the magnitude of the current of injury and the extent of the injury itself, suggested that the current of injury was the factor common to both Rose's and Singer's experiments.
On this theoretical basis we measured the current of injury following foreleg amputations in salamanders compared to the same amputation in frogs (38). While the immediate postamputation potentials were positive in polarity and about the same in magnitude in both species, the frog's potential slowly returned to the original slightly negative potential as simple healing by scarification and epithelialization took place. In the salamander, the positive potential very quickly (3 days) returned to the original base line but then became increasingly negative in polarity, coinciding with blastema formation and declining thereafter as regeneration occurred. These observations regarding polarity and duration of the potentials have recently been confirmed by Neufeld using the same techniques (39).
Fig. 2.6. Measurements of the current of injury following forelimb amputation in the frog (not capable of regeneration) and in the salamander (capable of regeneration). The immediate effect is a shift to a highly positive polarity in both animals. The frog slowly decreases this polarity as healing by scarification occurs, while the salamander reverses the polarity, shifting negatively at about the third day. Following this the blastema appears and regeneration occurs over a 3-week period, during which the negative polarity slowly subsides.
Two observations were immediately pertinent to Bernstein's original theorization that the current of injury was simply an expression of the transmembrane potential of damaged cells. The first was the polarity reversal in the salamander at 3 days and the second was the persistence of the potentials for several weeks until the injury was either healed closed or regenerated. Neither observation is compatible with Bernstein's hypothesis (the polarities of all damaged cells should be the same and they should persist no longer than the time required to repair or replace the damaged cells), but they are compatible with the concept of an organized neural DC control system with actual current flow.
On the basis of these observations we theoretically divided regeneration into two separate but sequential phases; the first being the formation of a blastema in response to a signal that is stimulating to the local cells and through their dedifferentiation produces the blastema. The information content of the signal responsible for the first phase is obviously sparse and the signal may be correspondingly simple, whereas the signal responsible for the second phase must be capable of carrying an enormous amount of information (what structure is to be formed, what its orientation with respect to the rest of the body is to be, and finally all of the details of its complex structure).
In our view, the DC potentials and currents generated at the site of injury by the DC control system were quite suitable as the signal for the first phase, whereas their information content was totally inadequate for the second phase. This concept meant that there could be two mechanisms at fault in those animals normally incapable of regenerative growth. First the initial phase may fail to produce a blastema because of either an inadequate signal or an inability of the cells to respond to an adequate signal by dedifferentiation. If an adequate blastema was formed, the second phase informational signal might be missing or inadequate to produce the subsequent redifferentiation and growth. Since it is common knowledge that nonregenerating animals fail in the first phase and do not produce blastemas, and in view of our finding of the polarity differences between regenerators and nonregenerators in the first phase, we postulated that the initial stimulating signal was missing in the nonregenerating animals. Simulation of this signal by external means was technically quite feasible; however, one could not predict whether the cells would be capable of responding to it or if they did, and a blastema was formed, whether the complex informational signal that controlled the second phase would be present.
The first test of this hypothesis was provided by Smith, who implanted simple bimetallic electrical generating devices (a short length of platinum wire soldered to a short length of silver wire) in amputated forelegs of adult frogs (40). In 1967 he reported the successful stimulation of partial limb regeneration by this technique. Theorizing that the failure to regenerate completely was due to the device being fixed in position at the original amputation level, he repeated the experiment using a device that had extensible electrode leads and in 1974 he reported securing regeneration of a complete extremity in the same animal (41). Meanwhile, we applied a modification of Smith's device to the foreleg amputation in the rat, reporting in 1972. the regeneration of the forelimb from the amputation level midway between the shoulder and elbow, down to and including the elbow joint complete in all anatomical detail (42). This was the first successful stimulation of the regeneration of a complex extremity by artificial means in a mammal. It has subsequently been substantiated by Libben and Person in 1979 (43) and by Smith in 1981 (44), all using similar techniques.
Fig. 2.7. Implanting a small device in the amputation stump of the rat foreleg results in a major amount of limb regeneration if the device is oriented so that the end of the stump is made negative, similar to the salamander current of injury. if the device is implanted with the distal end positive there is no regeneration.
It seemed now abundantly clear that artificially generated electrical currents of appropriate polarity and magnitude could stimulate regeneration in a variety of animals not normally possessed of this facility. Nevertheless, the identification of these currents as being the analog of those currents normally produced by the nervous system was lacking. Growth could be produced by this technique, but this did not necessarily mean that the neurally related currents measured in animals that were normally capable of regeneration were the real cause of their regenerative growth. This missing factor was supplied by one of the latest experiments of Rose (45).
In this experiment he carefully denervated the forelimbs of a number of salamanders, some of whom received daily applications of negative polarity electrical current to the amputation stump. Normal complete regeneration occurred in this group and subsequent examination demonstrated no ingrowth of nerve fibers. The control group demonstrated no regeneration whatsoever. Rose was therefore able to conclude that the factor supplied by the nerve that is normally responsible for limb regeneration in the salamander is the flow of an electrical current of the proper polarity and magnitude. However, the story is not quite over yet as the situation is actually somewhat more complex. However, it is better understood in the light of our most recent findings.
As early as 1962 Rose had called attention to the importance of a peculiar relationship between the epidermis and the nerves in the salamander limb regeneration process. The first event in such regeneration is the overgrowth of the epidermis alone (not the dermis) over the cut end of the amputation stump. Following this the cut ends of the nerves remaining in the amputation stump begin to grow into this epidermal "cap" where they form peculiar "synapse-like" junctions with the epidermal cells. This "neuroepidermal junction" (NEJ) is apparently the primary structure in the regenerative process, since following its formation the blastema appears, and if the formation of the NEJ is prevented by interference with either the nerve or the epidermis, or by simply interposing a layer of the dermis under the epidermis, blastema formation does not occur and regeneration is absent. In experiments in which limb regeneration was stimulated by electrical means no NEJ formed and we postulated that its function had been taken over by the applied electrical currents. Therefore, the NEJ could be postulated to be the single structure that produced the "regeneration type" potentials, not the nerve, nor the epidermis acting alone.
Fig. 2.8. The electrical mechanism producing regeneration of the salamander limb. After the epithelium alone grows over the end of the amputation stump, and the nerve fibers that were cut regrow, these two tissues grow together at the end of the stump. The nerve fibers attach themselves to the epithelial cells, producing a neuro-epidermal junction. This strucrure is then responsible for producing the specific electrical current that causes the cells left in the stump to dedifferentiate. If the neuro-epidermal junction fails to form for any reason, regeneration will not occur.
In an attempt to evaluate this concept we attempted to surgically produce such neuroepidermal junctions in animals that normally lacked regenerative ability (46). Hind limb amputations were done in a series of adult rats. Experimental animals had the sciatic nerve surgically inserted into the epidermis, while control animals had the nerve similarly mobilized but not inserted into the epidermis. The skin was sewn closed over the amputation stump and representative animals were sacrificed and the area examined daily. We found that, by the third postoperative day, the sciatic nerve in the experimental animals had grown laterally into the epidermis where it made junctions with the epidermal cells similar to the NEJ of the salamander. In the same group, blastemas appeared by the fifth day and regeneration of the major part of the hind limb ensued. No junctions were formed in the control group and neither blastemas nor regenerative growth occurred. Of most interest, however, were electrical measurements that were made daily on all animals. The control group demonstrated a series of potential changes identical to those of the nonregenerating frog or normal rat, while in the experimental group the changes paralleled those of the normal regenerating salamander, with a negative potential appearing concurrent with the formation of the junction between the nerve and the epidermis. It would now appear fairly certain that the specific sequence of changes in electrical potential that produce regenerative growth are themselves produced by the neuro-epidermal junction and not by either the nerves or the epidermis alone.
Intrinsic electromagnetic energy inherent in the nervous system of the body is therefore the factor that exerts the major controlling influence over growth processes in general. The nerves, acting in concert with some electrical factor of the epidermis, produce the specific sequence of electrical potential changes that cause limb regenerative growth. In animals not normally capable of regeneration this specific sequence of electrical changes is absent. However, it can be simulated by artificial means, resulting in blastema formation and major regenerative growth even in mammals.
The two effects of the intrinsic DC system previously described-biasing the activity level of the neurones and the stimulation of growth processes-obviously require exquisite sensitivity of certain cells to extremely low levels of current. In the case of the neurone, the cellular response is an alteration presumably in the properties of the membrane, rendering it more or less active in generating action potentials. In the case of growth stimulation, however, the cellular response is much more complex. In regenerative growth, for example, the blastema is formed by the dedifferentiation of mature specialized cells at the injury site into primitive, possibly totipotent cells; a profound alteration in both function and morphology. This may require some explanation.
In normal embryogenesis, the original fertilized egg cell contains all of the genetic programs (genomes) for the total adult organism, including all of the various cell types, each expressed as a separate genome. As the organism grows, these various specialized cells appear when the genomes for all of the other cell types are repressed. Thus the nucleus of a muscle cell for example has the genome for muscle unrepressed and operating and the genomes for all of the other cell types present but repressed. The genome produces the specialized cell type by governing the production of specific proteins which make up the cell itself. Dedifferentiation consists of derepressing these repressed genomes so the cell returns to a more primitive, less differentiated level and now has the option to redifferentiate into a new cell type, depending upon its local circumstances.
In the previous sections of this chapter experiments were described in which low levels of DC were administered to groups of cells within the organism with growth responses as predicted by theory. This type of phenomenological experiment is useful to define the functions of the DC l system and the general cellular responses, but it tells us nothing about the cellular-level mechanisms involved. At this time there are unfortunately few studies reported in the literature at this level. This is due in part to the enormous complexities of the living cell and our lack of knowledge in this area. Also, since the total organism is a complex of interrelating systems of biochemical and biophysical factors, one cannot assume with any degree of confidence that a cellular change following the application of DC energy to the intact organism is due primarily to the electrical current or to some second-order effect.
The only viable experiments, therefore, are those carried out on isolated cell populations where the only factor changed is the electrical current. If the electrical factors are within the levels observed in the living organism, the cellular changes observed may be inferred to be the same as those occurring within the organism in response to the normal operation of the DC system. Even here, however, an unavoidable artifact is introduced by the presence of the metallic electrode. The passage of current, even at low levels, through such an electrode produces electrochemical alterations that are not present in the living system. Another artifact is produced by the tendency of almost all normal cells in culture to change their morphology and function (the culture circumstance is presumably "sensed" by the cells as being different from that of their normal position within the organism). Therefore such in vitro experiments must utilize normal cells in culture before such changes even begin. These constraints require a normal cell type that can easily be harvested from the normal animal, that can be paced immediately within a culture system approximating the normal internal milieu, that demonstrates a definite alteration in a short time after exposure to currents simulating those found in the living animal. Therefore all standard-type tissue and cell culture experimental situations are theoretically not capable of producing unequivocal results.
In the course of studying the electrical factors associated with fracture healing, a regenerative-type growth process in the frog, such an ideal cell system was inadvertently discovered by our group in 1967 (47). We found that the red blood cells in the blood clot that formed between the fracture ends underwent a dedifferentiation process, transforming into the fracture blastema and eventually becoming bone. It should be noted that the red blood cells of all vertebrates, other than mammals, are complete cells, retaining their nuclei in the adult circulating state. This nucleus is, however, quite inactive; the cytoplasm contains relatively few subcellular organelles and in general the total cell is in a quiescent state. The cytoplasm of course contains a large amount of hemoglobin, and the cell is considered, despite the presence of the nucleus, as analogous to the mammalian red cell, which is non-nucleate and totally inactive. Therefore the dedifferentiation process, while being all the more remarkable in view of the inactive state of the cells, can be followed with ease by the light microscope.
Once the observation was made, the obvious question was whether this cellular process is caused directly by the electrical factors at the fracture site. These were measured and found to be similar to those found at the site of the regenerating salamander limb. Since the red cells exist naturally as discrete cells circulating freely in the blood stream, they are an ideal cell population for evaluating this hypothesis; they can be readily harvested, immediately introduced into an appropriate in vitro situation, and exposed to the electrical factors without delay, without complicating biochemical factors such as hormones or enzymes, or the other factors introduced by long term cell culture.
This experiment was carried out in a chamber that permitted both application of electrical current and direct visualization by the light microscope as the currents were being administered. Morphological changes typical of the dedifferentiation sequence were observed to take place within a few hours at total current ranges between 300 and 700 pamp. The lucite chamber in which the cells were suspended was 1 cm in diameter, approximating the size of the fracture hematoma, and from the measured voltages and resistances in the animal this amount of current was calculated to be identical to that present in vivo at the fracture site. Currents below and above this range were progressively less efficient in producing the morphological change until all effects ceased below 1 and above 1000 pamp for this size chamber. In the original experiments the changed cells were observed to have a markedly increased uptake of tritiated mixed amino acids and to survive for several days in cell culture media (normal unchanged red cells die under such circumstances).
Fig. 2.9. Sequence of morphological changes in a single frog nucleated erythrocyte exposed to very low levels of electrical current. The same cell was photographed at intervals of 5 minutes, demonstrating a change from the normal red cell type to a cell that has become round, lost all its hemoglobin, and has major phase changes in its nucleus. These cells are quite alive, surviving in cell culture and chemically demonstrating a marked increase in RNA content and a complete alteration in protein composition.
In later experiments, Harrington (48) confirmed the original observation and determined that while the RNA content of the changed cells was markedly increased, the DNA content remained constant, and the protein composition changed markedly compared to the original normal cells- changes that were consistent with the dedifferentiation process. He was also able to show that the direct action of the electrical current was in the nature of a "trigger" stimulus at the level of the cell membrane which then effected the dedifferentiation by means of the messenger RNA system. (Cells in which the RNA protein mechanism was inhibited from acting by exposure to puromycin would not undergo morphological alteration when exposed to adequate electrical current for an adequate amount of time. However, if the current was then stopped and the puromycin washed out by several changes of media the cells would then undergo dedifferentiation at the expected rate.) Pilla later confirmed these observations and determined that the same dedifferentiation could be caused by exposure of the cells to an appropriate pulsed magnetic field. In his view the primary effect of the electrical factors is a perturbation of the Helmholtz outer layer (such as produced by a change in the ionic concentration or the absorption of specific ions) on the cell membrane, with transmission of this information across the membrane by charge transfer through molecules that span the membrane.
Therefore, in at least this one cell system, the nucleated erythrocyte, specific effects of low-level direct currents have been observed and part of the mechanism involved in producing the effects has been determined. The effect is a most profound one, involving the basic machinery of the cell and resulting in major alterations in cell function (49).