Temperature and Heart-Rate in Fundulus Embryos

In a large number of instances the effects of temperature on the duration of biological processes have been described objectively by the equation of Arrhenius (1915),

In this statement, R is the gas constant, and K₁ and K₀ the velocity constants of a specific chemical reaction occurring respectively at the absolute temperatures, T₁ and T₀. The term μ, introduced to account for the changes in rate associated with changes in temperature, is designated by Crozier (1924, a, b), as the temperature characteristic or critical thermal increment. As plotted in the graphs, p, is identical with the slope of our curves, but physically it is assumed to represent the energy requirement of a given type of molecule when passing from a quiescent to a chemically reactive condition. Thus our formula involves theoretical considerations although its fitness to describe biological data is a mere matter of fact (Hecht, 1926).

The observed rates at which vital activities proceed at different temperatures, e.g. movement in Paramecium (Glaser, 1924), and the rates calculated from the equation, agree too closely to be accidental. Yet even over a short range of temperatures, movements of this kind can hardly be attributed to the activation of only one species of molecule. How then can the values of μ that emerge from the behaviour of such complicated systems carry any specific meaning?

It has been shown, especially by Crozier (1924, b; 1926), that acts of the greatest biological diversity yield increments that actually fall into a small number of classes. Conceivably these activities are the outward expression of a series of underlying chemical reactions, the latter necessarily linked in definite sequences. When translated into dynamic terms, the economic law of minimum assures us that the rate at which such catenary systems can proceed depends on the component with the slowest rate of acceleration. Of course, in any given instance, a particular value of μ could be distorted for purely physical reasons or might even be an average resulting from two or more reactions so balanced that momentarily one, then the other would occupy the position of control. However, to explain on this basis the existence of large groups of processes garnered from modes of behaviour and forms of life having superficially nothing in common, but nevertheless characterised by essentially identical values of μ, requires unwarranted, unmanageable and improbable assumptions regarding the prevalence of very peculiar quantitative adjustments, both physical and chemical. Accordingly, our analysis will rest more securely on the simpler and defensible postulate that constant and recurrent values of μ, implicate constant and recurrent reactions, chemical in type and specific.

Before attempting an application to the heart-beat of vertebrate embryos, it is well to recall the problems which this type of heart presents, and to consider what results our method can achieve.

The heart in these forms begins to pulsate several days before its innervation by the vagus nerve (Lillie, 1908). This fact, so long befogged by neurogenic theorists (Martin, 1905), has found its proper place in current interpretation. Yet the chemical processes that underlie myogenic and neurogenic rhythms, remain inadequately characterised. We do not know if the physical-chemical background for neurogenic rhythm is identical with that of its myogenic predecessor. With the same temperature characteristics recurring in the most remote situations (Crozier, loc. cit.’, Glaser, 1925), and with metabolism in muscle and nerve in all likelihood essentially the same (cp. the review of recent work in this field by Forbes, 1922), it would be most striking to find values of μ specifically diagnostic of either myogenic or neurogenic rhythms. Should the physical-chemical steps behind these two types of beat differ in number or kind, such differences might of course become manifest when suitable stages of development are studied at the proper temperatures. Still, the mere discovery of different increments in the two cases would not at once entitle us to assume more than one catenary series. Indeed, it is easy to imagine quantitative adjustments in these reactions such that with the onset of vagus control, certain components of a single chain might emerge more or less readily than before. In this event a change in the frequency with which a particular temperature characteristic appears could create the impression of a structural difference in the controlling series, whereas in reality we might be dealing merely with a change in some quantity or perhaps only with a revision of certain time relations.

In the closely relevant studies on Limulus acceleration of the adult heart, by heating the ganglion alone, so far has yielded only μ = 12,200 (data from Garrey, 1921 ; Crozier, 1924, c), whereas embryos with myogenic rhythm exhibit increments of 11,500, 16,400, 20,000, and 25,500 (Crozier and Stier, 1927). Here uniformity in the adult contrasts with diversity in the embryo. However, Garrey’s procedure (loc. cit.) was not identical with that of Crozier and Stier who subjected the entire organism to the various temperatures. Also, the adult Limulus in all probability was starving, whereas the larvae had their normal food supply. Until these and perhaps other variables have been controlled, the apparent difference between myogenic and neurogenic rhythms in this form must be accepted with reservation.

With these considerations in mind, we should feel no surprise if Fundulus larvae in comparable stages of development failed to conform to the Limulus pattern as we have it at present. Such failure would not necessarily conflict with any future evidence that the two types of rhythm harbour in their chemical foundations an essential difference. On the other hand, if the anticipations, which now appear warranted, are realised, discrepancies between Limulus and Fundulus can be harmonised with the idea that analysis by means of temperature so far has given indications in one or perhaps in both forms of parts only of the actual chemical series, and that very possibly such studies are inherently capable of uncovering in neurogenic and myogenic hearts only such differences as depend on changes in the modus operandi of one and the same basic mechanism (cp. discussion in Crozier and Stier, 1927, pp. 501, 502 and 511 ; also Glaser, 1925, p. 279).

The observations on which our treatment rests agree with those available in the literature, but were made especially to fulfil the requirements set by the present type of analysis. Polimanti’s work on Gobius eggs (1911) is not available because the temperature intervals are irregular, too large, and the total range too narrow. The same and other disadvantages are inherent in Cesana’s observations on the chick (1911). Of the older records, those of Loeb and Ewald (1913) on Fundulus, Menidia, and their hybrids, and of Laurens (1914) on Amblystoma, are the most authentic. Yet, adequate as they undoubtedly are for demonstrating the R.G.T. rule between 5⁰ and 30° C., these experiments also fail to offer the materials necessary for an independent derivation of temperature characteristics.

The most convenient form available at the Marine Biological Laboratory, Woods Hole, proved to be Fundulus heteroclitus. After artificial insemination, the eggs, on reaching the desired stages of development, were first examined with reference to the visibility and regularity of the heart beat. The most favourable embryos were then transferred to individual test-tubes hung in a glass thermostat. The temperature was regulated and held constant to within 0 · 1 ° C. by means of ice, an electric heater, and a stirring device. When the system had cooled and remained constant for 30 minutes at some point between 4° and 6° C. one test-tube at a time was corked so as to enclose a large bubble of air, and placed in a horizontal position on a submerged wire tray illuminated in one region by a beam of light passing through the water from an electric bulb located under the thermo-stat. Only at the extremes of the thermal range were special precautions necessary to insure that the temperatures at the surface of the tray did not differ measurably from those to which the embryos had been exposed while the tubes were in the vertical position. By shifting the tray, rolling or tilting the tubes and adjusting both the light and the binocular microscope, it was possible in all cases to secure satisfactory visibility.

Preliminary tests, arbitrarily made at 20° C., indicated that, in due time, readings at this temperature can be duplicated after exposures to both 4·0° and 29·5° C. For the final experiments, the average lower extreme was 5·8° C. The passage from room temperature to this limit occupied about 2 hours. A run with four embryos and observations one degree apart required in the neighbourhood of 10 hours. With thermal stability established, it took approximately 5 minutes to make all the adjustments and record three readings on one individual. In certain check series only one reading was taken at each temperature and the intervals were two degrees or more. Time was measured with a stop-watch graduated to tenths of a second. For the more rapid hearts, the number of beats taken was increased with the temperature. In the lower ranges ten, at intermediate points twenty, and, in certain instances around 25° C., fifty beats were counted for each reading.

Within the thermal limits mentioned, 4-day embryos fall into two distinct groups. One of these, comprising six animals, has a prevailing increment that fluctuates about 16,500; the other, with two instances, yields an average of μ = 14,400. The clearest illustrations of these two groups are given by embryos A₅ and A₆, Fig. 1.

In both cases there is little or nothing to suggest more than one temperature characteristic for the entire range. This fact is very strikingly indicated in Table I, in which the time required for one beat at the various temperatures is calculated from the equation and the results compared with the closest observations for these temperatures. No matter what theory we may finally adopt, the agreement between biological fact and physical hypothesis is unescapable.

From embryo A₅, Fig. 1, it is very evident that at any give temperature the same heart frequently fails to yield a specific reading more than once. This failure at precise duplication cannot be explained solely in terms of experimental errors, since these remain essentially constant, whereas the discrepancies in question vary in different embryos and in one and the same individual may change abruptly at certain temperatures. Also in those numerous instances in which the readings actually do coincide, it would become necessary to deny experimental errors altogether. We accordingly adopt the suggestion of Loeb and Chamberlain (1915) and assume that the processes under analysis are themselves variable. This variability may be attributed to fluctuations in the effective concentration of some chemical, perhaps a catalyst, essential for the manifestation of the beat. The requisite amount of this substance present with each beat cannot be less than a certain minimum, nor can it at any moment exceed the capacities of the system necessary for its production and liberation.

All of which, if true, gives variation in thermal data great significance. Over a range of temperatures yielding a single value of μ, the fastest and the slowest readings (if they approximate sufficiently to the theoretical limits), should be faster or slower than their corresponding means, not necessarily by identical, but by respectively constant fractions. It is easily seen, therefore, how averages based on different animals or even on the same individual might prove deceptive particularly at temperatures where the value of μ changes and a new process, very possibly with a different latitude of variation, is supposed to assume control. Thus, constancy in the latitude of variation becomes an added criterion by which we may judge of constancy or change in the controlling links of our catenary chains. Hence the graphs are all plotted with due reference to the marginal variates and, with certain interesting exceptions, all observations remain confined within narrowly parallel lines.

The results in 4-day embryos are not always so simple. As shown in Fig. 2, 19·5° C. and especially 8’7° C. are temperatures frequently characterised by vertical shifts that may or may not involve significant changes in either p-value or variability.

If we restrict ourselves to the region where the observations are most convincing we observe that the shift about 8’7° C. does not necessarily involve any change of increment. At the temperatures in question observed and calculated times diverge, yet the steps from one datum to the next in the observed series are identical with the intervals between calculated values if these are based on a constant μ. On this supposition, we are confronted here by changes of rate with no corresponding changes of temperature characteristic. The situation is similar, perhaps to be identified with, that encountered by Crozier and Stier (1927) in the data of Murray (1926) on isolated fragments of the chick heart. In Murray’s figures the latitude of variation sometimes changes and because the heart-beat at any moment must be governed by its fastest pace-maker, Crozier and Stier suggest for displacements of this type a change in pace-maker control. If sudden and not too frequent, the transfer of control from a certain pace-maker to a faster or slower one should externalise as an abrupt increase or decrease in rhythm. Obviously, if we are at all times measuring the slowest chemical rate in the fastest pace-maker, two pacemakers differing in frequency, but themselves controlled by the same link of their respective catenary chains of chemical processes, could expose in the heart a change of frequency without any change of increment. Since these data on Fundulus give no indication of a continuous change in the latitude of variation, it is assumed that we are here dealing with a single temperature characteristic exhibited by not less than two different pace-makers. This interpretation, though perhaps not the only one possible, appears to harmonise with the data (Table I).

Conditions on the 12th day of development are quite different. As shown in Table II, the increments are of the orders 20,000, 18,000, 16,000, 14,000, and 12,000. None of the eight embryos gave a single value of μ over the entire range. At the higher temperatures the rate changes from either 20,000 or 18,000 to one of the lower orders.

The evaluation of these terminal increments is not absolutely certain. The critical points at which the prevailing increments break, cannot well fall below 16 · 8° C. or above 21·1° C. Within these limits, however, we are apt to encounter a downward shift of the data. If we again postulate a change in pace-maker control, the facts above 20 ° C., as well as the dislocations at low temperatures, might be described by the prevailing increment, e.g. D₂, Table II. This suggestion is of course not the only one. In certain instances, between 21 ° and 29 ° C., μ might be of the order 5000 and in cases B₃, C₃, A₂ and C₂, μ = 12,000 would appear to be equally adequate and might even be justified since values of this order occur so frequently in similar situations and under other circumstances may emerge very distinctly from the embryonic Fundulus heart. Yet however this may be, the values of μ given in Table II do actually reproduce the original data and hence we may conclude that 12-day hearts differ from the earlier ones, in a sharper definition of critical points and by the certain disclosure of at least two new increments, 20,800 and 18,200.

As the observations on the 4th and 12th days, respectively, precede and in all likelihood come after cardiac innervation, it is desirable to have another developmental crisis for comparison. The moment of hatching was chosen, not only for this reason, but also because at that time one can be fully certain of the vagus connections. There are of course difficulties. The fish that has left its egg-case cannot be observed adequately without restraint, but no confinement practicable would do away with the vibrations of the fins. It is necessary, therefore, to select embryos likely not to hatch before the experiment is completed, nor to delay too long afterwards. In two instances there was success. The larvae, aged 17 days, 11 hours and 45 minutes, at the beginning of the run, had apparently reached identical stages of development. They were under observation for 8 hours and 15 minutes. The first A_H was run from 5-5° to 24° C. and hatched within 12 hours of the last reading; the second, B_H, was carried from 4·9° to 22 ° C. and was free-swimming before a record at 23^{0 °} C. could be made. With half an hour between observations the last reading on B_H was taken less than 30 minutes prior to the actual moment of escape. The data for these two embryos are given in Table III and Fig. 4.

At first sight these two hearts suggest a curvilinear relationship between log. frequency and 1/T Abs. As the table indicates, however, except for the terminal readings—always the least reliable—the greater parts of both records can be described essentially by two familiar characteristics. Of these the orders 20,000 and 16,000 are well substantiated in the earlier embryos. Conceivably, μ = 23,000 is a distorted representation of 20,000 or 25,000—the latter reported for other cardiac rhythms, cp. Crozier, and Crozier and Stier (loc. cit.). Such distortions, though possibly the outcome of changes in pace-maker control, might also result from interference traceable to the frequent twitchings and the almost incessant vibratory movements of the pectoral fins characteristic of embryos about to hatch. While the data differ from those of the earlier stages, both larvae disclose one of the increments and the high frequency of critical points characteristic of the 12th day.

If the differences between 4-day larvae and those of the 12th day and later are associated with the innervation of the heart, the restoration of essentially myogenic conditions, in so far as this is possible in older embryos, should be followed by results some of which might be anticipated. As contrasted with 12-day and hatching larvae, such embryos should exhibit with greater frequency a single increment over the entire temperature range, and this increment should be either of the order 14,000 or 16,000; along with the more or less precise definition of increments inadequately shown by the earliest normal cases, such embryos, if exposing the value 20,000 at all, might be expected to restrict this chiefly to the lower ranges; and furthermore the frequency of critical points should be reduced.

The simplest and, after practice, least harmful method of converting neurogenic into myogenic hearts, depends on a necrosis limited to the medullary region of the brain and started by means of a hot needle. Properly localised, mere contact on the vitelline membrane of a 10- or 12-day embryo, without either puncture or loss of fluid, is sufficient to kill any desired portion of the nervous system. In the course of a few minutes, the affected region becomes opaque and white. The most favourable defects are those in which the opacity remains limited to a section extending roughly from the level of the pectoral fins to the mid-brain. Such animals frequently live 4 or 5 days. The heart-beat is invariably slower, its rate being apparently related inversely to the degree of injury. In the more severe cases, the volume of blood is below normal and pulsations too feeble to insure adequate circulation. Sometimes the only indication of activity in the heart is a rhythmical swaying with no sign of progressive contractile waves.

Control animals with approximately equal amounts of nervous tissue deleted in the tail region exhibit normal heart-rates, unless the necrosis approaches too near the medulla. In that event the rate is intermediate between that of normal and fully depressed hearts.

From individuals treated in this manner we can scarcely expect ideal or uniform results. Very likely ganglion cells imbedded in the heart wall remain unaffected, and in all probability there are unintended anatomical derangements and secondary effects of the primary lesion, such as breakdown products. Also, depression of the heart-rate is not to be anticipated from mere vagus destruction, and might be taken as a symptom, either of continuous excitation in certain fibres that happen to remain functional, or of some other unusual state among the tissues involved.

Data on five cauterised individuals are given in Table IV and Fig. 5. In one individual, C_c, the increment 14,200 prevails from 5 · 8° to 27 · 4° C.; in a second, B_x,μ= 14,700 from 15 ° to 27·6° C.; a third, B_c, exhibits μ = 16,700 from 5·9° to ^I5’9° C., whereas in a fourth, D_c, μ= 14,000 from 5·9° to 17·8° C. Thus the two increments most definitely characteristic of 4-day embryos emerge—in one instance over the entire range. In two cases, A_c and B_x, an increment of the order 20,000 is restricted to the lower half of the thermal range. A_c, B_e and D_c exhibit respectively 11,300, 7,800 and 6,300 in the upper ranges. However, the value ·1,300 is the only one of this group that appears to be sufficiently substantiated.

In accordance with our expectations the elimination of the vagus connections is followed by a restriction of increments of the order 20,000 to the lower temperatures and by an increase in the frequency of μ = 14,000, both at high and at low temperatures.

Setting aside the rare or possibly distorted increments, and considering as homogeneous those that deviate from the most frequent orders of magnitude by less than 10 per cent., we arrive at the following table of values for Fundulus embryos :

Two of these averages agree quantitatively not only with the characteristics reported for larval Limulus, but likewise with the increments deduced from other cardiac rhythms. The very low rates suggested here and there and omitted from the table, very possibly are identical with p, = 5000 sometimes encountered in Limulus embryos (Crozier and Stier, loc. cit.)> or even with μ = 7300 found at the same temperatures in adult Tiedemannia and Pterotrachea (Glaser, 1925). Between 4 ^· and 9 ^· C. both molluscs yield a mean of p, = 22,500. On careful scrutiny, the items on which this average rests fall into two groups (loc. cit.), one fluctuating about 20,500 and the other, less clearly, about 24,700. Both of these values (20,000 and 25,500) are reported for Limulus. The lower one (20,900) is recurrent in Fundulus embryos. Again, between 9 · 7° and 25 · 5° C., Limax yields μ = 16,300 ± ; for Tiede-mannia from 8·7° to 26·4° C., μ = 16,300 ± and at times 14,400 ±. Regardless, then, of our immediate problem the catenary chain of chemical processes underlying heart rhythm in Fundulus embryos conforms to a general type and contains certainly not less than three basic reactions common to phyla as wide apart as molluscs, arthropods and vertebrates. This statement is conservative and excludes increments which in one type or another as yet happen to be rare ; on the other hand, it does take cognisance of the heart-beat of Blatta where Fries (1926) reported a particularly clear case of μ = 14,100.

The quantitative and serial identities common to the catenary chains of Fundulus and Limulus do not involve corresponding coincidences in the developmental periods during which the several increments emerge. In fact all the increments characteristic of larval Limulus, perhaps even the one reported for the adult, sooner or later emerge either from neurogenic Fundulus or else from individuals that have been cauterised or are still normally myogenic. On this basis, therefore, the respective catenary series underlying neurogenic and myogenic rhythms cannot be considered constitutionally different, but diversified only by changes in the relative ease with which particular increments emerge under the two sets of conditions. Without implying a final commitment, this conclusion may prove to be inevitable. In general, the establishment of neural control can only mean that the nervous system has come to regulate the activities of another tissue by imposing the limits that characterise its own. If the propensity of muscle and nerve to react in certain fundamentally similar ways is really a symptom of properties common to all living cells it is clear that nervous control of heart rhythm is not apt to involve the introduction of incremental novelties, but rather to depend for its effectiveness on a revision of time-relations. In any particular instance, such revision might easily fail to show in the presence of certain equilibria or other quantitative adjustments unfavourable to incremental change ; on the other hand, if revision does show, the new time-relations should externalise as changes in the frequency with which certain increments appear ; as extensions or contractions of ‘he thermal ranges over which they apply; as increases or decreases, at certain temperatures, in the actual rate of pulsation; and finally, as changes in the frequency, localisation, or both, of critical points. Variations compatible with these suggestions are amply illustrated by the data here made available on myogenic and neurogenic hearts.