Calcium Channel Stability Measured by Gradual Loss of Excitability in Pawn Mutants of Paramecium Aurelia

The nature of the events responsible for the action potential in excitable membranes has been investigated primarily with electrophysiological techniques. Recently, Kung and his colleagues (Kung, 1971 ; Chang et al. 1974) and I (Schein, 1976) have also applied genetic methods to the investigation of membrane excitability in Paramecium aurelia. Wild-type paramecia can reverse ciliary beat (and thus swimming direction) in response to various stimuli - tactile, chemical and thermal. The reversal mechanism, which has been shown to be sensitive to calcium concentration (Naitoh & Kaneko, 1972) is activated by the rapid influx of calcium which is responsible for the rapid rise of the membrane potential during the (graded) action potential. Behavioural mutants that are unable to reverse swimming direction have been isolated and are known as pawns. They have been shown to have a defective calcium conductance mechanism (Schein, Bennett & Katz, 1976). The mutants have been classified genetically into three genes, two of which (pwA and pwB) are represented by many alleles and are the subject of this report.

To investigate further the nature and stability of the Ca-channel and the relationship of the pawn gene-products to it, the gradual expression of the pawn phenotype has been followed (behaviourally and electrophysiologically) in cells converted from heterozygous (WT/pw - phenotypically wild type) to homozygous pawn by autogamy. A delay in the full-fledged phenotypic expression of the genotype, called phenotypic lag, was clearly described in bacteria by Davis (1948). His study, using penicillin selection of auxotrophs, led him to suggest that the wild-type enzyme had to be diluted out by growth before the auxotrophic genotype could be expressed. In Paramecium aureliaSonneborn & Lynch (1934) demonstrated a similar phenomenon which he called ‘cytoplasmic lag’ Two genetically marked individuals of opposite mating type can be mated; each conjugant produces two identical haploid germ nuclei, one of which is exchanged with the partner. The result is two genetically identical heterozygous F₁ ex-conjugants (Fig. 1). Some divisions later, the F₁ can be induced to undergo autogamy, a process similar to conjugation except that the two identical haploid germ nuclei fuse and the F₂ product becomes homozygous at all loci. There are thus two stages at which a sudden change in genotype (comparable to bacterial mutation) can occur.

Sonneborn coined the phrase ‘cytoplasmic lag’ to describe the phenotypic transformation of a pair of genetically identical ex-conjugant F₁ cultures arising from a mating of two parent strains differing in size and rate of growth. Each F₁ culture resembled for several divisions the parent from which it arose until, eventually, the two cultures became identical.

This paper describes the behaviour and electrical excitability of the homozygous ex-autogamous (F₂) progeny of a culture of F₁’s heterozygous for a pawn mutation. Firstly, phenotypic lag is quantitatively demonstrated for the pawn character, and secondly, that phenomenon is exploited to give an estimate of calcium channel half-life.

Cells are grown and mated as in Sonneborn (1970) with one exception. The growth medium was made from cerophyl (5 g/1) rye powder and buffered with 5 mm-NaH₂PO₄, 5 mm-Na₂HPO₄. For more details, see Schein (1976).

The two pawn mutants used in this study (pwA(419) and pwB(100)) have been described previously (Schein, 1976). Both mutants are extreme pawns. They show no reversal behaviour in the test situations described below, and electrophysiological results demonstrate the nearly complete absence of inward (calcium) current (Schein et al. 1976).

The spinner (sp) trait is used as a marker. The wild-type reversal response changes the direction of the ciliary power stroke by 180° (see Tamm, Sonneborn & Dippell (1975) for a more detailed description). The sp mutant acts as though the change is only 90° instead of swimming backward, it spins in place. The sp trait is easily elicited by exposure to test solution S (see below), which causes long, vigorous, frequent reversal responses. The sp trait is unrelated to the pawn trait. A sp cell responds to behavioural stimuli as vigorously as the wild type, and its electrical properties are also no different from the wild type (Schein, unpublished data).

Each of the homozygous pawns (pwA(419)/pwA(419) and pwB(100)/pwB(100)) were mated with a normally excitable marked strain (sp/sp). The ex-conjugants were grown separately through 12 divisions in 2 ml tube cultures and then tested. The products of a successful mating (the F₁) were then easily identified as phenotypically wild type (genotypically pwA(419)/ +, + /sp or pwB(100)/ +, + /sp). Several clones of each F₁ were made; these were grown through 14 more divisions in 5 ml tube cultures and then allowed to starve for 3 days, starvation inducing autogamy.

288 individual starved F₁ were cloned in three 96-well plates (Tissue Culture Plate, Microtest II, Falcon no. 3040, and Lids, Falcon no. 3041). Refeeding initiated growth following autogamy ; each slot had sufficient medium (0·2 ml) to support 10 generations of growth. At various times (18 h, 22 h, 24 h,…) the contents of wells were checked for the number of cells. These cells were also tested at the same time.

1·5 ml of starved F₁ were added to 1·4 ml medium; this allowed a single division following autogamy. Two days later and every other day thereafter, a small amount of medium was added (1·5, 1·5, 2, 2, 2, 2, 4, 4 and 4 ml).

1·5 ml of starved F₁ were added to 1·5 ml medium. Two days later the cells were placed in a refrigerator (12 ± 1 ° C) but otherwise treated in the same manner as the room temperature culture.

Not only does Ba²⁺ pass through the Ca-channel (Naitoh & Eckert, 1968), but it is also thought to block delayed rectification K-channels as well and affect repolarization (Grundfest, 1961).

Test solution S (for stimulation : 1 mm-NaH₂PO₄,1mm-Na₂HPO₄,2 mm-Na citrate, 1° 5 mm-CaCl₂, 2° 0 mm-BaCl₂) induces frequent reversals in the wild type.

Within 15 s of transfer to test solution P (for paralysing: 1 mm-NaH₂PO₄, 1 HIM-Na₂HPO₄, 2 mm-Na-citrate, 0·1 mm-CaCl₂, 10 mm-BaCl₂, 15 mm-NaCl), wild-type cells are paralysed.

Clones were tested by transferring, with a hand-drawn micropipette, one or several individuals from growth medium to the test solution. Their behaviour following transfer was observed in a stereomicroscope at approximately 20 ×.

Electrophysiological methods are described in detail in Schein et al. (1976). The cells were impaled with a current-passing and a voltage-monitoring micropipette and then stimulated with rectangular 60 ms long pulses of current. The oscilloscope records show from top to bottom, current, voltage and the derivative of voltage. Active inward current is a measure of the current flowing through the voltage-dependent calcium (early) and potassium (late) conductances, and it is computed using the curves in the three oscilloscope traces and the cell’s capacity.

The effects of the two test solutions (S and P) were used to classify the ability to exhibit reversal behaviour into four (arbitrary) types, from wild type to extreme pawn (Table 1). Test solution P was developed to demonstrate slight impairment of the normal reversal response. Wild-type cells are reproducibly paralysed 15 s after transfer from growth medium to solution P. In contrast, a pawn mutant, pwC(320), whose active inward ionic current is about 50 % of the wild type, continues swimming, slowly, for more than 10 min, although it is clearly affected. Extreme pawns are not affected by solution P. In the terms of Table 1, solution P is used to classify cells as either wild type (type 1) or not wild type (types 2, 3 or 4).

Test solution S causes long, vigorous reversal behaviour in the wild type; even mutant cells with a low level of excitability (measured electrophysiologically) reverse when exposed to solution S. However, extreme pawns, like the two used in this study, do not give any response. In the terms of Table 1, cells which give a vigorous reversal response are either type 1 (wild type) or type 2 (slightly reduced active calcium conductance). Test solution S was originally used to distinguish cells carrying the spinner (sp) trait from those not carrying the trait. Cells which do give a response, but one which is insufficient for a sp versus non-sp determination, are type 3 (excitability considerably reduced); cells which give no response at all are type 4 (extremely inexcitable, like the extreme pwA(419) and pwB(100)).

Table 2 shows the combined results from two experiments which followed the expression of the pwB(100) phenotype. At each time listed, from 16 to 32 fresh clones were tested. Autogamy should produce a random segregation of the unlinked pw and sp genes. Of the clones tested, the number of sp clones and the total number of pw clones were both about one-half, as expected.

After just a single division (cells at the ‘2-cell stage’ − 18–25 h post-autogamy) the behaviour of the pawn paramecia (type 2, and see Table 2, fourth footnote) could be distinguished from their non-pawn sibs by the use of solution P. However, the 2-cell stage pawn paramecia were clearly very excitable; their swimming was markedly slowed by solution P and they responded vigorously with reversals in solution S. As stated above, this may. be compared with paralysis of the wild type within 15 s and no effect on the fully expressed pwB(100).

With the cells exhibiting type 2 behaviour, it was also observed that the 4-, 8- and 16-cell stage paramecia showed increasing resistance to solution P, although this is not indicated in the table.

When there were 16 cells per clone (4 divisions post-autogamy) the reversal response elicited by solution S was sufficiently weak so that it was impossible to distinguish spinner from non-spinner pawns (type 3). By the 32-cell stage, the cells had become extreme pawns (type 4). They gave no reversal response at all in solution S.

A similar experiment was performed using pwA(419). In this case, the same 23 pawn clones were re-tested at each time-point. The results, shown in Table 3, were nearly identical to the results of Table 2. Although synchrony was not as good, it is clear that one could detect the pawn genotype at the 2-cell stage. Also, the cells reached type 3 by the 16-cell stage and type 4 (extreme pawns) by the 32-cell stage.

To correlate the phenomenon demonstrated by the behavioural experiments with membrane excitability, electrophysiological measurements were made on the F₁(pwB(100)/+) and post-autogamous pwB(100) homozygotes at 30 h (4-cell stage) and at 40 h (8-cell stage). These measurements may be compared both with those from the homozygous wild type and the pwB(100)/pwB(100).

The most informative trace in each oscilloscope record (Fig. 2) is the lowest trace, dV/dt. The discussion of the computation of active inward current in Schein et al. (1976) makes it clear that comparison of the height of the first peak (due to passive charging of the membrane capacitance) with the second peak gives an easily visualized, nearly quantitative estimate of inward calcium current.

The oscilloscope records of Fig. 2A–C, particularly the derivative traces, show the progressive decrease in active inward current as one proceeds from the F₁ (Fig. 2 A) to pawns (identified by their behavioural response) at the 4- and 8-cell stages (Fig. 2B and 2C) to the fully expressed pawn phenotype (Fig. 9 of Schein et al. 1976).

These records also show that the wild type behaviour of the F₁-heterozygote is reflected in its electrical properties, which are fully wild type. Records of the wild type are shown in Fig. 4 of Schein et al. (1976).

The data in Fig. 2 can be used to generate a measure of active calcium conductance, called active inward current (Schein et al. 1976). Fig. 3 shows representative tracings of active inward current from each of the recording sessions from which the pictures in Fig. 2 were taken; the ordinate in the 4- and 8-cell stage parts of the figure is expanded fourfold. The active inward current approaches a maximum (as described in Schein et al. (1976)) at 2·5 nA in the F₁ (approximately equal to 2·8 nA in a wildtype cell), 0·67 nA by the 4-cell stage (0·50 nA in another 4-cell stage individual), and 0·37 nA by the 8-cell stage. The fully-expressed mutant has essentially zero calcium conductance. These results demonstrate nearly quantitative agreement with a simple halving of active calcium conductance with each division.

Additionally, Fig. 9 of Schein et al. (1976) shows results from the homozygous pw(B100)/pwB(100) strain. It has normal passive electrical properties as well as normal delayed rectification; however, anomalous rectification (the ‘relaxation’ of the voltage when the membrane potential is hyperpolarized to − 40 mV and beyond) is not seen until the cell is hyperpolarized to about − 70 mm. A review of Fig. 2B and 2 C shows that the effect of the mutation on anomalous rectification can be seen by the 4- and 8-cell stages.

The electrophysiological results suggest that a simple halving of the number of wild-type channels occurs with each division. If there were rapid decay of channels, then a much more rapid decrease in excitability would have occurred. For example, if the half-life of a channel were 8 h (1 normal division time), then excitability would decrease by a factor of four with each division, and cells at the 4- and 8-cell stage would have and the wild-type active inward current. Since the results in the above section indicate that loss of channels agrees with a model incorporating only dilution, it was hoped that by allowing the same number of divisions in a much longer period, the effect of decay could be observed. Cells were therefore grown as described in Slow growth in Materials and Methods.

Cells were tested behaviourally (Table 4) at 11 days (2–3 divisions) and at 20 days (3–4 divisions). Normal segregation of the spinner and pawn genes demonstrate normal autogamy. The progress of the expression of the behavioural phenotype from type 2 to type 4 is similar for both genes. For example, at the 8-cell stage, after 11 days at room temperature, slowly grown pwA(419) and pwB(100) both behave as type 3 cells. The data presented in Tables 2 and 3 show that the phenotype of the normally grown pwA(419) and pwB(100) mutants appears as type 3 between the 16- and 32-cell stage. If after 11 days, an 8-cell stage paramecium behaves like a 16- to 32-cell stage paramecium grown rapidly, then the effect of channel loss was to decrease excitability additionally by to J in those 11 days. The lifetime (T) is therefore between 11 and 5·5 days. Similar limits are implied by each of the rows of Table 4 and these limits have been used to estimate the lifetime (T). The range of T is from 5·5 to 8 days in the pwA(419) experiment and from 5·5 to 9 days in the pwB(100) experiments.

Autogamy in a heterozygous paramecium leads to a sudden change in the genotype of the organism, a change comparable to mutation. Mutation in bacteria may not be reflected in phenotype until the concentration of the affected macromolecule is diluted out by cell division. The steady loss of the reversal response of homozygous pawns in the divisions which follow autogamy of both pwA/ + and pwB/ + heterozygotes has been demonstrated using two behavioural tests (Tables 2 and 3). As early as the 2-cell stage (1 division post-autogamy) the pawn homozygotes could be distinguished from their wild-type sibs by exposure to solution P. By the 16-cell stage (4 divisions) the pawns still responded to the very stimulating solution S, but the response was short in duration and infrequent. By the 32-cell stage (5 divisions) the pawns no longer showed any reversal, even with the most sensitive test stimulation.

The behaviour of F₁ the heterozygote appears wild type. The electrophysiological measurements confirmed that impression and demonstrated the lack of a gene-dosage effect in the heterozygote.

The gradual loss of excitability following autogamy, indicated by behavioural tests, is paralleled by a loss of membrane excitability measured electrophysiologically (Figs. 2 and 3). The decrease in active inward ionic current, one-quarter at the 4-cell stage and one-eighth at the 8-cell stage, agrees with a simple dilution hypothesis and demonstrates the stable nature of the calcium channel structure.

In addition, the electrophysiological data presented here, using rapidly growing cultures, reveal that channel half-life is certainly more than 40 h. The behavioural data, using slowly grown cultures, permit an estimate of from 5 to 8 days for channel half life. The model which is presented in Schein et al. (1976), based on electrophysiological data, and in Schein (1976), based on genetic and behavioural data, implies that the pawn genes’ products directly affect or are themselves structural components of the calcium channel. The experiments presented in this paper are entirely consistent with such a scheme.

The detection of non-zero active inward current in even the extreme pawn mutants, coupled with the knowledge that they were produced by nitrosoguanidine mutagenesis, suggests that the mutations are point mutations (Schein et al. (1976)) and that mutant channels are indeed synthesized. The data here suggest in addition that the newly synthesized mutant channels do not poison the old wild-type channels, and imply that the channel is monomeric or stably polymeric.

Further support for the correlation between behavioural classification and excitability, and thus for the dilution hypothesis and the measurement of half-life, is found in Schein et al. (1976). pwC(320), which this report would classify as showing State 2 behaviour, appropriate for 2-cell stage paramecia, has about (40 %) the wild-type active calcium conductance. pwA(214), which this report would classify as showing State 3 behaviour, appropriate for 16- to 32-cell stage paramecia, has 5 % the wild-type active calcium conductance.

Berger (1976) has also studied phenotypic lag in the expression of thepio.4 phenotype ; our results are in close agreement. Following autogamy and 4 days of starvation, growth through a median of 3·3 divisions was required before he observed the pwA phenotype. His threshold of detection of pawn behaviour corresponds approximately with what I have called State 3 behaviour. In my study, four divisions were required. He found that a single (+ /pwA) macronuclear fragment was sufficient to confer weak reversal behaviour (corresponding again to what I have called State 3 behaviour) on the homozygous (pwA/pwA) exautogamous cell. Under the conditions of his experiment, the single fragment would be expected to increase its original DNA content by 1-5 (Berger, personal communication). The ‘old genes’ are therefore 1·5 × (1/35) or 5 % of the total. My results show that state 3 behaviour is observed when the active inward calcium currents are 5 % of the wild type.

Berger (1974) has also shown that the old macronuclear genes are not immediately destroyed. Instead, the old macronucleus is transformed into about 33 macronuclear fragments, which are gradually destroyed during starvation following the induction of autogamy. The longer the starvation, the more fragments are destroyed for the purpose of recycling DNA precursors. The ‘rapidly grown cells’ used in this study were allowed to starve for 3 days which, according to Fig. 2 of Berger’s report (1974), leaves about 9 of the original 35 fragments. In addition, only half of the macronuclear fragment genes from the heterozygous F₁ ancestor are wild type at the pawn loci. Thus, prior to the first division of the new homozygous pawn F₂, only or of the old wild-type genes are present. A separate report by Berger (1973) demonstrates that the rate of RNA synthesis per unit of DNA is the same in old macronuclear fragments as in the new macronucleus. Thus, the contribution of ‘old wild-type genes’, which should be of the total at the 2-cell stage and halving with each division, to active calcium conductance, which is at the 2-cell stage and halving with each division, appears to be insignificant.

It could be argued that the ‘old F_t genes’ are disproportionately active. In that case, the electrophysiological results still permit use of the behavioural tests as a quantitative assay for excitability. The application of the quantitative behavioural test to the slowly growing cells remains valid, as does the 5- to 8-day half-life. Furthermore, the ‘slowly grown cells’ were subjected to the same 3 days of starvation, which was then followed by an additional 11 days of underfeeding and starvation. Berger’s (1974) data only extend to 8 days of starvation, at which time recycling pressure has left an average i’5 (of the original 35) macronuclear fragments, left to be distributed among the 4, 8, 16,… descendants. These data strengthen the conclusion that ‘oldF₁ genes’ play little or no role in this example of phenotypic lag.

I would like to thank Dr Charles David and Dr M. V. L. Bennett of the Albert Einstein College of Medicine for advice and encouragement during this work. I would also like to thank Dr George M. Katz and Mr Sidney Steinberg of the College of Physicians and Surgeons, Columbia University, for the use of the current-clamp and for giving generously of their time and energy with problems of an electrical nature. Also, I would like to thank Dr James Berger, who was kind enough to read the manuscript critically.