ABSTRACT
The degree of abnormality found in amphibian interspecies hybrids varies from one species pair to another, and may also differ between reciprocal matings (Moore, 1955). Although some general ideas exist about the mechanism of arrest and the course of abnormal development, our knowledge is far from complete. One major aspect of development which has received attention only at the gross level is the nature and diversity of those proteins which compose the greater part of the cytoplasmic sap, the ‘soluble’ proteins in the sense of this paper (see below). This paper reports the results of studies on the development of normal (Ranapipiens) and gastrula-arrested hybrid (R. pipiens ♀×R. sylvatica ♁) amphibian embryos.
The present work was undertaken (i) to provide a detailed survey of the saline-soluble proteins in R. pipiens during the embryonic period, and (ii) to apply the same techniques to the gastrula-arrested hybrid with R. sylvatica. The findings do not indicate a major rôle in morphogenesis for any of the components studied, but do provide some insight into degenerative changes in the hybrid embryos. The principal technique was zone electrophoresis in a synthetic gel, polyacrylamide, followed by protein staining.
Embryos of the two parent species are quite similar (Pollister & Moore, 1937; Shumway, 1940), but the hybrid embryos differ from both parents in several ways. Morphologically they are arrested during gastrulation. They gain functional cilia, swell, burst, and finally, about 3 days after arrest, they cytolize (Moore, 1946). The inductive and differentiative capacities of hybrid tissues are subnormal (Moore, 1947, 1948). Metabolic abnormalities of the hybrid embryos include a depressed respiratory rate (Barth, 1946) and a depressed respiratory potential (Gregg, 1960). Glycogen degradation apparently begins earlier but continues at a very subnormal rate (Gregg, 1948). Glycolytic capacity is subnormal (Gregg, 1962). Creatine phosphate under aerobiosis is present at subnormal concentrations (Harrison, 1963). Acid phosphatase activity, at least in the dorsal region, is subnormal (Mezger-Freed, 1953). DNA of whole embryos increases normally to almost the end of maternal gastrulation, then stops completely (Gregg & Lovtrup, 1960). During the period of arrest individual nuclei may increase their DNA content to an amount ‘greater than necessary for a subsequent mitosis but cell division has already ceased’ (Moore, 1957). Cytoplasmic RNA also becomes quite subnormal and in the absence of R. pipiens chromosomes the R. pipiens cytoplasm inhibits nucleolus formation in R. sylvatica nuclei (Moore, 1957). An intact nucleolus appears to be required for the normal production of ribosomes when this begins in tailbud-stage embryos, but the lack of nucleoli in mutant Xenopus embryos does not prevent the synthesis of messenger RNA (Brown & Gurdon, 1964).
The pattern of fast-green staining, believed to reflect changes in localized histones, changes at the onset of gastrular movements in a variety of amphibian embryos, but the hybrids of this paper cannot be distinguished from normal embryos by this criterion (Horn, 1961).
Studies of amphibian development are approaching the molecular level in descriptions of fine-structural changes during development (Karasaki, 1959a, b, 1963), and in descriptions of specific structural units such as yolk platelets (Wallace, 1963a, b). Various methods reveal developmental changes in the proteins of whole embryos: see, for example, reviews by Ranzi (1962) and by Tyler (1957).
The present knowledge of mechanisms in protein synthesis suggests the importance of genetic factors, and their unambiguous and controlled translation, in producing the normally occurring spectrum of proteins. One might expect hybrid lethality to involve interference by the foreign genome in the pattern of protein synthesis, either eliciting the production of some chemically abnormal protein, or altering quantitatively or temporally the production of some normal protein. Alterations in respiratory and DNA metabolism which may express such an interference occur in a variety of hybrid embryos: see, for example, Chen & Baltzer (1964). In the case of the hybrid R. pipiens ♀ ×R. sylvatica ♁ some kind of interference seems indicated since neither gynogenetic nor androgenetic haploid embryos of R. pipiens are arrested at gastrulation nor do they show the remainder of this arrest syndrome (Moore, 1955).
From this introduction it should be clear that we do not know the mechanism of arrest. All of the factors studied (with the possible exception of glycolytic metabolism) are normal until after one can already tell that hybrid embryos are arrested because of their delay in forming a blastopore lip. There is as yet no description of the course of development of the major cytoplasmic proteins, and there is reason to suspect a causal abnormality in some aspect of protein synthesis.
MATERIALS AND METHODS
Biological
Adult frogs of both sexes of the species Rana pipiens and adult males of the species Rana sylvatica were obtained from C. H. Mumley, Alburg, Vermont.
Normal (Rana pipiens) and hybrid (R. pipiens ♀×R. sylvatica ♁) embryos were prepared by fertilizing ova stripped from pituitary-stimulated females (Rugh, 1934; Hamburger, 1942). A sperm suspension of two testes macerated in 40 ml. of a standard medium, bicarbonate-free 20 per cent. Holtfreter’s solution (Holtfreter, 1931), was made for each species. Ova from a single female were stripped alternately into two dry dishes. Sperm suspension was added by pipette. After 30 min. the dishes were rinsed and flooded with medium. After the jelly swelled, the embryos were distributed in groups of about fifty among several shallow dishes. Six different clutches were prepared in this manner. All were reared at 18°C. in standard medium which was changed at 2-day intervals. Fertilization exceeded 95 per cent.; normal development of R. pipiens embryos to swimming tadpole stages exceeded 90 per cent.
In this paper embryonic stages for R. pipiens embryos are those of Shumway (1940). Hybrid embryos were staged by the morphology of half-siblings. Developmental stages of hybrid embryos are designated by prefixing ‘H’ to the stage designation which would have applied to R. pipiens embryos which had been reared identically. That is, hybrid embryonic stages were numbered by reference to the morphology of a group of half-sibling R. pipiens embryos which had been handled identically and reared simultaneously (Gregg, 1957).
From four clutches, groups of normal and hybrid embryos were harvested at specified times after fertilization (6-, 12- or 24-hr. intervals) as shown in Text-figs. 2 and 3. The jelly coats were removed with forceps. The developmental stage was noted. Groups of embryos, twenty-five each, were placed in 0·4 ml. polyethylene vials and rinsed in fresh rearing medium. The excess fluid was withdrawn by a capillary pipette. During all subsequent operations the embryonic material was chilled and remained in these capped, marked vials.
Development of proteins in normal embryos. Ordinate: embryonic stage and equivalent age at 18°C. Abscissa: mobility in 5 per cent, gel expressed in mm. per (V./mm.) per hr., cathode at right. Intensity of staining is shown by the degree of darkening by stippling; horizontal widths represent actual width of stained region in gel.
Development of proteins in normal embryos. Ordinate: embryonic stage and equivalent age at 18°C. Abscissa: mobility in 5 per cent, gel expressed in mm. per (V./mm.) per hr., cathode at right. Intensity of staining is shown by the degree of darkening by stippling; horizontal widths represent actual width of stained region in gel.
Development of proteins in arrested hybrid embryos. Represented as in Text-fig. 1.
Partial fractionation of egg proteins: Crude histones. Abscissa: mobility as in Text-fig. 1. WO: whole extract (stage 1 of Text-fig. 1). CH: crude histone extract of coelomic oocytes.
Embryos were homogenized without additional fluid by thrusting a platinum wire into the storage vial and spinning the wire at 800-1000 r.p.m. for 30-60 sec. Centrifugation at 5000-5500 g. for 30 min. at 0°C. produced a pellet and a clear supernatant fraction of 20-50 μl. The latter was removed and stored in a clean vial at -20°C. The supernatant fraction so prepared is a whole-embryo extract of the saline-soluble proteins, i.e., the proteins soluble in a small addition of standard rearing medium (20 per cent. Holtfreter’s). Samples from two clutches were frozen prior to homogenization (pre-frozen); all other samples were frozen only after transfer of the clear embryo extract to a clean storage vial. The whole embryo extract was separated by zone electrophoresis without prior dialysis.
Two clutches of R. pipiens embryos were reared for comparison of dorsal and ventral fragments. Dorsal and ventral fragments were prepared from stage 18 embryos by cutting them between two tungsten needles. Ventral fragments consisted of the yolk-laden endoderm and the overlying belly epidermis. The remainder of the embryo, in front of the heart region and dorsal to the nephric prominence, was included in the dorsal fragment. Samples consisted of the dorsal or ventral fragments of fifty embryos. Subsequent homogenization, centrifugation and electrophoresis were accomplished in the manner described for whole-embryo extracts.
Electrophoretic techniques
The method of electrophoresis was derived from methods for starch (Smithies, 1959) and for polyacrylamide gel (Raymond & Wang, 1960). Gel slabs were wrapped in a vapor-barrier film (Saran-Wrap, T. M. Reg. U.S. Pat. Off., Dow Chemical Corp.) and laid horizontally between brine-cooled plates so that their ends dipped directly into buffer reservoirs.
Acrylamide, A,N’,-methylenebisacrylamide, and N,N,N’,N’,-tetramethylethylenediamine (Davis, 1962), all from Distillation Products, Rochester, were polymerized in 6-mm. thick sheets by ammonium persulfate. The gel was made in tris-borate buffer, 10 mM. tris (hydroxymethyl) aminomethane (Sigma, Saint Louis) and 10 mM. boric acid. The reservoir buffer was the same tris-borate but with 3 mM. ammonium persulfate added, to avoid conductivity differences. The pH in the gel was determined with a Beckman Zeromatic pH meter with temperature compensation. Both temperature and voltage gradient were monitored in the gel. During electrophoresis a mercury thermometer in contact with the gel registered temperature; voltage gradient was monitored with a voltmeter independent of the electrophoretic power source, measuring via platinum electrodes across 20 cm. of gel. The conditions for electrophoresis which gave the most satisfactory resolution were: 10 V./cm., pH 9·0, 3·0°C., 5 per cent, polyacrylamide gel (0-25 per cent, bisacrylamide), for 60 min.
At least three aliquots of the undialyzed embryo extract of each sample of embryonic material were subjected to electrophoresis. Tabs of Whatman paper 41H (4x6 mm.) containing 5 μl. of extract were placed along the end of a slab of gel and a second slab abutted against it. Such a set contained both hybrid and normal material from two or more stages or two or more clutches. Three such sets were placed in electrical series between the electrode reservoirs. There was no evidence of regional effects on mobility with this arrangement.
Immediately after electrophoresis the gels were stained in amido black (Matheson-Coleman and Bell, Cincinnati) 0 · 5 per cent, in glacial acetic acid : methanol : water : 1:5:4 (Smithies, 1959). Excess dye was removed with changes of 10 per cent, acetic acid. Electrophoretic destaining was found to produce distortions in the banding pattern and therefore was not used. Except as already noted chemicals were reagent grade from Mallinckrodt, New York. Crude histone extracted from coelomic oocytes of Ranapipiens was prepared and kindly provided by Professor E. C. Horn, Department of Zoology, Duke University.
Protein concentration in the embryo extract was estimated according to Lowry et al. (1951), using bovine serum albumin as a standard. For both normal and hybrid embryos to stage 12 (or H 12) protein concentration was about 40 mg./ml. in the whole embryo extract, a value comparable to the value for extractable protein calculated from the data of Gregg & Ballentine (1946). Because of changes in the volume of the perivitelline fluid (normal) and blastocoelic fluid (hybrid), protein concentration in the embryo extract decreases during neurula and tailbud stages (14 to 18). Older embryos were removed from their vitelline membranes even though this involved rupture and loss of blastocoele fluid for the hybrid embryos. The volume of embryo extract was the same for each electrophoresis; changes in total protein concentration were taken into account in the interpretation of banding patterns.
Samples of embryo extract from several developmental stages or several clutches were separated side by side in the same gel slab. The appearance of novel bands in a few of the samples could not be blamed therefore on mobility artifacts arising in the gel.
The destained gels were compared and enlarged scale drawings were made. The prominent features described below were quite consistently seen, especially from one sample to another within the same gel. Where differences are described they are supported by at least one simultaneous separation of the two samples contrasted.
RESULTS
Nomenclature
The stained pattern resulting from electrophoresis was arbitrarily divided into zones, as in Text-fig. 1, on the basis of band width, staining intensity and resolution. These zones were named from the anodic side (a) to the cathodic side of the pattern (d). Each band was designated by a subscript numeral to the zone name, thus c1, c2 and c3 were the bands in order from anodic side to cathodic side of the c zone. Bands which were apparent only during a brief developmental interval were designated by adding + or - to the name of the nearest band, to indicate a position on the anodic or cathodic side of that band, respectively. Capital letters refer to hybrid material.
Major protein zones
Zone a, the most rapidly moving anodic zone, consisted of a pair of broad bands, a1 and a2, which stained with a moderate intensity. The two were well defined from each other and from the slower zone b by regions without appreciable staining (i.e., there was no background).
Zone b stained with the least intensity and showed the poorest resolution of any of the zones. There appeared to be from three to five resolvable bands. The b bands were stained only slightly more intensely than the background.
There was no sharp drop in the intensity of background stain between this zone and the next slower anodic zone, c.
Zone c showed the most intense staining and the best resolution of any of the zones. It usually included two or three very intensely stained, very sharply defined and very thin bands. The background stain within the zone was moderately intense (usually much greater than in zone b), but between the origin and the slowest of the sharp bends, c3, the background was negligible.
All of the cathodic bands were included in zone d. They were all moderately broad. Two stained with moderate intensity; the remainder were usually faint.
Changes during embryogenesis
Text-fig. 1 presents a graphic summary of the electrophoretically resolvable protein bands during early development. It is a plot of electrophoretic mobility (abscissa) versus developmental age (ordinate) in which staining intensity is indicated by the degree of darkening by the stippling. Horizontal width represents the actual width of the stained region in the gel. The general picture is one of slight change.
Zone a. Over the entire developmental period studied there was a slow increase in the staining intensity of a1 and a slight decrease in its width resulting in an increase in the degree of resolution of a1 There was no demonstrable change in the band designated a2 during early development.
Zone b. There appeared to be at least one change in zone b during development. The region containing b1 and b3 began to increase in staining intensity toward the end of gastrulation, and then the resolution increased to reveal three discrete bands during the tailbud stages. The discreteness of the slowest band of this zone, b5, appeared to decrease during gastrulation and again toward the end of embryonic life. During these same periods the intensity of the background staining increased, nearly obscuring b5.
Zone c. The changes in this slow anodic zone were fairly clear-cut. The first major band (cj was double during a short period of time; the extra band (c1 +) was demonstrated only at the end of gastrulation. When present, Cl + stained with as great an intensity and was as sharply defined as c1. The two bands were very close together. The slowest band in this zone (c3) also appeared to be double during part of development. Text-fig. 2 shows first a faster (c3+) and then a slower (c3—) band close to c3. The former appeared during the end of blastulation and persisted through most of gastrulation, while the latter appeared in the neural tube stage and persisted at least to the establishment of circulation. Only c3 — was as sharply defined or stained as intensely as the broader c3.
Zone d. In the cathodic zone two additional bands, d3 and d5, appeared during gastrula and early neural stages. Then d5 reappeared in tailbud stages. Both were always very diffuse and stained only faintly.
Comparison of normal and hybrid proteins
Proteins of the hybrid are summarized similarly in Text-fig. 2. The nybrid material was found to differ slightly from the normal, and only in zones C and D.
Zone C. There was no evidence of a split in C1 at any time. Such a split might have been partly obscured by the intensity of the background, but if C1 + did occur, it was with much less intensity than c1 + (in normal). The staining intensity of C2 appeared to be increasing toward the end of the period studied, whereas c2 remained constant throughout the developmental period studied. The staining intensity of C3 and its subsidiaries, C3+ and C3-, decreased progressively in hybrids after stage H12, whereas c3 +, c3, and c3remained constant in the normal. The hybrid bands faded into the low intensity background in stage H20.
Zone D. The most dramatic difference between hybrid and normal protein development was in the pattern of the cathodic components. The intermediate minor band, D3, persisted from its appearance during the formation of the blastoporal lip, instead of fading as d3 did after neural closure. In intensity, width and resolution it did not distinguishably differ from a maximally intense dy The fastest cathodic band in the normal material, D5, did not fade in the hybrids, D5. From its appearance during gastrulation, D5 became more intense until it approached in intensity and width both D4 and d4. An additional band, D6, appeared late in the life of hybrid embryos. It migrated faster than any cathodic band demonstrated in the normal embryos. D6 appeared at about the time normal embryos became motile. It was broad, diffuse and of low intensity, resembling the slower D4.
Comparison of whole extract with crude histones of oocytes
Text-fig. 3 shows a direct comparison of the bands resolved from the saline whole-embryo extract and from the crude histone fraction of Rana pipiens oocytes, as revealed by simultaneous migration under the conditions already described. Where a component of the histone extract appeared to have the same mobility as a component of whole-embryo extract, the same designation was applied to both. The previously described anodic component (Horn, 1961) appears to be multiple in the gel. Two slow-moving components are seen in the gel separation which were also apparently obscured by the initial boundary during free electrophoresis. One was anodic and one cathodic at pH 9.
Comparison of dorsal and ventral fragments of tailbud embryos
Electrophoresis of the dilute-saline extract of both dorsal and ventral embryonic fragments resulted in patterns such as those illustrated in Text-fig. 2. There was an extremely great similarity between patterns derived from dorsal fragment extracts and those from ventral fragment extracts. Within the limits of the technique there appears to be no significant difference between dorsal and ventral fragments. In an attempt to provide further resolution of the proteins of these two regions, some samples were subject to electrophoresis for twice the time previously used. The relative position of the various bands still did not differ from dorsal to ventral extracts.
Effects of freezing and thawing
For both normal and hybrid embryos the intensity of cathodic bands was greater in the pre-frozen extracts than in post-frozen extracts of the same development stage. The changes during development described above for normal and for hybrid embryos were most clearly seen in the pre-frozen material, but were also evident in post-frozen material. In no case did a d§ band appear in pre-frozen material.
DISCUSSION
Migration in gel
Gel electrophoresis involves a sorting both by molecular size and by molecular net charge (Smithies, 1959; Raymond & Nakamichi, 1964). Although no rigorous theory of the sorting processes has been widely accepted, approaches to such an understanding are discussed by Kunkel & Trautman (1959), with proposals by Smithies (1959), Raymond & Nakamichi (1962), Ornstein (1962) and Boyack & Giddings (1963).
Comparison with other work
Spiegel (1960) analyzed a similar protein extract of Rana pipiens embryos at several stages up to stage 21. His densitometrie tracings of bromphenol-blue stained protein after paper electrophoresis at pH 8-6 are very similar to what would be expected from the present study. A pair of well separated anodic bands was followed by a ‘plateau’ region containing a number of bands. He distinguished only two cathodic components, possibly d2 and d4 of the present work. A large amount of material remained at the origin, possibly an irreversible adsorption associated with the use of paper (Block et al., 1958). Little if any change was found in the embryo.
Denis (1961) analyzed a similar protein extract of Pleurodeles waltlii over the entire embryonic period. His electrophoretic studies on cellulose acetate film also reveal no major changes in number or intensity of bands until about the time of blood circulation. His patterns are very different from those of the present study. No correlation is evident.
Reports that the quantity or variety of soluble protein increases rapidly once blood circulation is established have been interpreted as evidence of yolk mobilization (Denis, 1961; Gregg & Ballentine, 1946). Direct evidence of yolk platelet degradation (Karasaki, 1963) is consistent with such a view. However, the present work does not reveal any changes in stage 19 through 21 that could be interpreted as yolk mobilization. This difference may result from differences in sensitivity or operational definition of ‘soluble’.
The proteins of zone d apparently correspond to most of the components of a crude histone fraction previously described (Horn, 1961). Since the histone fraction was prepared by extraction of washed yolk platelets with 0 ·1 N HC1, the d bands represent the basic histone-like proteins of yolk platelets. The fact that these proteins appear in extracts of all stages need not mean that this material would be free in the cytoplasm under conditions which would prevent yolk platelet damage. The enhancement of these d bands by a freeze-thaw cycle prior to homogenization is consistent with a localization of these basic proteins in the amorphous zone of yolk platelets (Karasaki, 1963). There is no direct evidence of other locations for these proteins. The present study does not illuminate the reported abrupt shift in fast-green affinity of yolk platelets and nuclei (Horn, 1961), since the pattern of basic proteins in pre- and post-gastrular stages changes only slightly and very slowly.
Discussion of the molecular size, shape or heterogeneity of components in these or other bands is reserved until more direct evidence can be offered.
Comparison of normal and hybrid proteins
There are two zones showing differences between electrophoretic patterns of normal and hybrid embryos. In zone C of hybrid material C1 + never appears, C2 becomes abnormally intense after stage Hl7, and C3 and C3 — become abnormally faint after stage Hl4. In zone D, D3 remains visible after Hl3, D5 intensifies after Hl7, and D6 appears.
Since the C components are yet to be characterized or their position localized intracellularly, a discussion of their rôle must be speculative. Furthermore, since these differences occur most markedly well after morphological arrest, they cannot represent primary defects. However, basic proteins, particularly histones, can act as stabilizers of nucleic acid and may thus be involved in genetic regulation and in control of protein synthesis (Irvin et al., 1963).
One interpretation of the appearance of the D bands is that subcytolytic changes occur for some time prior to gross cytolysis and that these changes allow the proteins represented in D bands to be more readily extracted. Karasaki (1963) has clearly shown the presence of a limiting membrane surrounding yolk platelets prior to degradation and elaborate concentric membranes during their normal degradation. If pre-cytolytic degeneration involves a decrease in the stability of cellular membranes in general, one could expect the amorphous zone of yolk platelets to be more readily extractable and more readily liberated by a freeze-thaw cycle. Those basic proteins found in yolk platelets are believed to lie in the amorphous layer (Horn, 1961).
The intensity of the C2 band could be similarly interpreted if some identity between C2 protein and either of the major proteins of yolk platelet central region (Wallace, 1963a, b) were established. At present we have only the suggestive evidence that the low mobility and high resolution of this band are consistent with the large molecular size of yolk platelet proteins.
Comparison of dorsal and ventral fragments of tailbud embryos
The failure to demonstrate differences between dorsal and ventral fragments of tailbud-stage embryos is surprising. The dorsal fragments represent tissues engaged in relatively rapid differentiation in the sense of histogenesis and organogenesis while the ventral fragments remain morphologically conservative. Other studies (see, for example, review by Boell, 1959) indicate the existence of marked physiological differences between dorsal and ventral fragments. Two factors may account for the similarity found in the present study. First, enzyme changes are more readily detectable than are changes in non-enzyme proteins. Therefore the enzymatic activity of an extract may in fact change drastically without changes in the quantity of major constituent proteins. Second, adsorption to particulate material, for which frog embryos are notorious, could remove significant components from the embryo extract during a centrifugal separation.
SUMMARY
Polyacrylamide gel electrophoresis of dilute saline extracts of embryonic Rana pipiens resolved an array of six cathodic and at least seven anodic bands.
During embryonic development changes in electrophoretic pattern were restricted to intensities of a few of the bands.
The major differences between normally developing (R. pipiens) and gastrula arrested hybrid embryos (R. pipiens ♀ x A. sylvatica ♁) were the increased intensity of cathodic bands (probably histones) and the existence of one extra cathodic band in the hybrid material.
No differences between extracts of dorsal and ventral fragments of tailbud stage normal embryos were detected.
RÉSUMÉ
Analyse par électrophorèse des protéines du suc cellulaire d’embryons de grenouilles normaux et hybrides
L’électrophorèse, sur gel de polyacrylamide, d’extraits salins dilués d’embryons de Rana pipiens a révélé la présence de six bandes cathodiques et d’au moins sept bandes anodiques.
Au cours du développement embryonnaire, les modifications du type électrophorétique ont été restreintes à des variations d’intensité d’un petit nombre de bandes.
Les principales différences entre les embryons à développement normal (R. pipiens) et les gastrulas hybrides bloquées (R. pipiens ? xÆ. sylvatica £) consistaient en un accroissement d’intensité des bandes cathodiques (probablement des histones) et en l’existence d’une bande cathodique supplémentaire dans le matériel hybride.
On n’a pas décelé de différences entre les extraits de fragments dorsaux et de fragments ventraux d’embryons normaux au stade du bourgeon caudal.
ACKNOWLEDGEMENTS
It is a pleasure to acknowledge the guidance and helpful criticism of Professors J. R. Gregg and E. C. Horn during the course of this research, and also the critical readings given this manuscript by Professors Florence Moog and D. J. Fluke. This work was partially supported by grants from the U.S. Public Health Service (GPM-16, 483) and from the American Cancer Society (IN-63-C9). Portions of this work were done in partial fulfillment of requirements for Ph.D. degree at Duke University.