PROTEIN, AMINO ACIDS AND GLUCOSE IN THE YOLK-SAC FLUIDS AND MATERNAL BLOOD SERA OF THE TAMMAR WALLABY, MACROPUS EUGENII (DESMAREST) MARILYN B. RENFREE

Department of ^oology,

Australian National University, Canberra, A.C. T. 2600, Australia

(Received 7th August 1969) Blood sera and yolk-sac fluids were collected from preg¬ tammar wallabies and analysed for free amino acids, proteins and

Summary. nant

glucose.

The total and individual concentrations of free amino acids at all stages of pregnancy were higher in yolk-sac fluids than in sera, with the excep¬ tion of glutamic acid, which was at a higher concentration in serum. Protein concentrations were lower in yolk-sac fluids than in the serum but the number of protein components in yolk-sac fluid, as determined by acrylamide gel electrophoresis, increased after implantation. Glucose concentration in yolk-sac fluids also increased with the age of the embryo. These values were lower than those in sera before implanta¬ tion, but higher than in sera after implantation. These results indicate that the transfer and accumulation of nutrient materials in the yolk-sac is not only by simple diffusion from maternal serum, but also by selective transfer through the yolk-sac placenta. INTRODUCTION

The embryo of the tammar wallaby (Macropus eugenii Desmarest), in common with that of other marsupials, develops a chorio-vitelline placenta after im¬ plantation. This is in contrast to most eutherian mammals which form a chorio-allantoic placenta. In the rabbit, both types of placenta exist concurrently, and, although transfer of most substances takes place across the chorio-allantoic placenta, it has been shown by Brambell and his co-workers (e.g. Brambell, 1958, 1961; Brambell, Hemmings, Oakley & Porter, 1960) that the chorio-allantoic placenta of the rabbit, and probably that of the rat, is impermeable to some proteins, and immunoglobulins are transferred by way of the vitelline circula¬ tion. Extensive determinations on the content of rabbit blastocyst fluid and uterine secretions with reference to various ions, bicarbonate, glucose, lactic acid and other substances have been carried out by Lutwak-Mann and coworkers (e.g. Lutwak-Mann, Boursnell & Bennett, 1960; Lutwak-Mann, 1962). 483

484

Marilyn B. Renfree & Wilson (1966), Deren & Wilson (1964) and Deren, ( 1966) have shown that the absorptive cells of the visceral

Padykula, Deren Padykula & Wilson yolk-sac endoderm in the rat possess ultrastructural features characteristic of protein-absorbing cells; they give evidence that this placental membrane may be highly specialized for the transfer of macromolecules. In the wallaby, the blastocyst lies free in the uterus during early development. Since the volume of fluid in the vesicle increases before attachment, the fluid must be derived from uterine secretions. As development proceeds, the chorion comes to lie in close apposition to the maternal epithelium, forming the chorio-vitelline placenta, so that the maternal circulation is probably able to supply most of the nourishment of the embryo (Sharman, 1961). At this point, attachment is analogous to implantation and will be referred to as such in the text. In eutherian mammals, the yolk sac is either vestigial or is inverted after early development. In M. eugenii, however, it becomes a large, fluid-filled sac. Since the membranes of this sac form the placenta, an analysis of these fluids may give an indication of the type of nutriment being provided to the embryo, and whether its contents are due to a transport across the placenta or due to a synthesis by the embryo. Beier (1968) has shown that the composition of rabbit blastocoelic fluid is similar to that of uterine fluid from the 6th day of preg¬ nancy, while McCarthy & Kekwick (1949) showed that rabbit blastocyst fluid on Days 7 and 8 was closely similar in protein composition to diluted maternal

plasma.

The present study was undertaken to compare some of the components of marsupial yolk-sac fluid with the homologous maternal blood sera at various

stages of pregnancy.

MATERIALS AND METHODS

Tammar wallabies were obtained from Kangaroo Island, South Australia, and were accommodated in open, grassy yards, the diet being supplemented with lucerne-hay and oats. Female tammars carry a dormant blastocyst during lactation and during the anoestrous period (Berger, 1966). This blastocyst normally resumes develop¬ ment after loss of pouch young or at the onset of the next breeding season, but development can be induced by injections of progesterone (Clark, 1968; Smith & Sharman, 1969). In this study, development was induced in anoestrous tammars by daily injection of 10 mg progesterone (Schering), in oil, intra¬ muscularly for 10 days. In lactating animals, development was induced by removal of the pouch young. It was noted that progesterone injections decreased the gestation period, so all embryos were compared by reference to the appropriate stage as shown in Text-fig. 1. Stages 4, 5 and 6 of tammar development are approximately equivalent to Stages 24, 32, and 34 of opossum development as described by

McCrady (1938).

Before surgery, animals were anaesthetized with sodium pentobarbital (Nembutal) injected into the ear vein. Blood was collected by cardiac puncture

Protein, amino acids

and glucose in the

wallaby

485

from each animal and serum was separated in the usual way. A mid-line incision through the pouch exposed the reproductive system and uteri were removed. In animals in which the embryo was still unattached, a small incision was made in the wall of the uterus and the endometrium carefully torn away to allow the vesicle to extrude itself. It was then placed in a sterile container and

200 180

160

?

|

140

0)

8 "120 cj

100 80 60 40 20

Embryonic stage Text-fig. 1. Comparison of the concentrations of glucose in yolk-sac fluid (diagonal shading) and the corresponding maternal serum (cross-hatching) collected at the three different stages of development which are represented by the accompanying drawings. The bar shows the range of the values observed. Stage-4 embryos have an average length of 6-0 mm, Stage-5 embryos have an average length of 9-5 mm and Stage-6 embryos a length of 13-2 mm. ys—yolk sac, am—amnion, al—allantois, ao—area opaca.

collected by rupturing the membranes. In advanced pregnancy, the yolk-sac fluid was collected with a 26-gauge needle inserted directly into the yolk sac through the uterine wall. All samples were stored at —20° C until used for

analysis. Aliquots of sera and yolk-sac fluids were subjected to the following analyses: qualitative determination of proteins was performed using acrylamide gel electrophoresis (Davis, 1964; Ornstein, 1964) and estimates of the relative

486

Marilyn B. Renfree density of the bands in gels were made with a Joyce Loebl Mark III C automatic double beam recording microdensitometer. Serum protein levels were determined using the technique of Lowry, Rosebrough, Farr & Randall (1951) but determinations on yolk-sac fluids by this technique were found to be inaccurate because of the high free-tyrosine content of the samples. An estimate of protein in the yolk-sac fluid was achieved using acrylamide gel electrophoresis, running known amounts of bovine serum albumin

as

the standard.

identification and quantitative measurement of the free amino carried out by the method of Spackman, Stein & Moore (1958), using a modified Technicon Automatic amino acid analyser fitted with a 0-9-cm Dowex-50 ion exchange column. The apparatus was calibrated with a synthetic mixture of amino acids, and on duplicate runs, the analyser gave reproducible results (error 3 to 4%). Serine, asparagine and glutamine were separated by paper chromatography using the method of Smith & Moses (1960). Precipitation of protein in the serum samples required for amino acid analysis was accomplished by the addition of trichloroacetic acid to give a final con¬ centration of 5%. The supernatant was passed through a Dowex-50 (hydrogen form) 3-cm column, using ammonia as the eluting solution. The eluant was dried in a rotary evaporator, re-dissolved in de-ionized water and dried three times to remove all traces of ammonia. The final solution was in dissolved sodium citrate buffer (pH 2-2). Analysis for glucose was performed following the colorimetrie method of Bergmeyer & Bernt (1965), using glucose oxidase and peroxidase.

Separation,

acids

were

RESULTS

A. Protein determination Quantitative. A serum protein concentration in mg/ml of 66-1 +S.E. 1-6 (n 13) was found by the Lowry technique. These concentrations lie within the normal range of various eutherian mammals (Spector, 1956) and did not change with the stage of pregnancy. Acrylamide gel electrophoresis of the yolksac fluids shows that they contain considerably less protein than serum (average 2 mg/ml as compared with 66 mg/ml) (see Plate 1 ). Protein content of yolk-sac fluid during the early stages of gestation (Stage 4) is lower than in the later =

stages.

Qualitative. The disc electrophoretic data

are presented in Plate 1. The band of the later fluids and the serum proteins are similar patterns stage yolk-sac but the relative intensities are not the same. Electrophoresis of yolk-sac fluid at Stage 4 shows one main band. It is dense with an electrophoretic mobility similar to that of albumin. Other indistinct bands are present which do not appear on a microdensitometer trace. Some material can be seen at the origin, and since it is at this position that the y-globulins are found, it is possible that this band represents these proteins. By Stage 5, ten bands are clearly visible. The relative intensity of the two fastest moving bands has changed and they are now of approximately equal

PLATE 1

Acrylamide gel electrophoretic patterns and microdensitometer tracings for each embryonic stage examined. Sample volumes were as follows: Stage 4, 0-30 ml; Stage 5, 0-22 ml; Stage 6, 0-20 ml; and serum, 1-25 µ\. The vertical scale is the same on all densitometer tracings. (Facing p. 486)

487 Protein, amino acids and glucose in the wallaby concentrations. Benzidine staining of the gel shows that there is a haptoglobin band present and a fast moving pre-albumin-like band. Amido black does not stain this fast moving pre-albumin. There is an increase of material in the y-globulin region of the gel at this stage. It is about the same concentration as at Stage 6. Table 1 free amino acids and amino compounds in the yolk-sac fluids of the wallaby

Macropus eugenii (µ-MOLE/ML) Embryonic stage Amino acid

Taurine Urea Unknown 1 Aspartic acid Unknown 2 Threonine Serine

3-82 3-36 005 006 012 207

303 8-68 016 0-16 002 1-37

1-54 3-73 0-15 005 009 0-99

0-99 2-48 0-07 002 004 4-80

0-44 5-92 009 003 008 0-56

Asparagine

918

8-23

802

303

Glycine

3-62 012 0-46 318

3-96 010 0-43 2-31 3-69 1-20 008 0-96 0-95 0-93 0-60 0-53 0-46

Glutamine Proline Glutamic acid Citrulline

1-15

0-63

5-15 007 0-90 2-81 7-14 8-50 0-07 0-46 1-15 110 0-41 104 0-21 0-01 001 0-28 1-76 0-64 3-99 015 2-48 0-02 008

56-14

49-63

53-01

Alanine Valine

4-69 1-86

Cystine

010 110 1-43 1-38 0-83 0-81 0-34

Methionine Isoleucine Leucine

Tyrosine Phenylalanine /?-Alanine Hydroxylysine GABA Orni thine Ethanolamine Ammonia

Lysine 1 -Methylhistidine

Histidine

3-Methylhistidine Arginine Total

trace

002 011 1-74 1-31 6-90 011 6-22

001 008 1-20 108 4-50 008 4-20

1-33 7-44 0-42 014 1-43

1-62 6-26 008 008 0-09 1-07

1-42 9-21 007 005 009 1-42

4-22

10-22

5-80

6-48

3-20 004 0-29 1-38 3-62 1-11 002 0-67 0-71 0-48 0-27 0-90 006

6-44 009 0-77 2-27 4-01 1-81 013 0-24 1-48 101 0-32 0-68 0-31

2-62 013 0-99 3-24 803 3-36 0-02 0-70 1-82 200 0-64 108 0-22

3-89 007 0-51 1-90 4-88

3-96 007 0-54 301 7-90 3-75 0-13 0-44 1-82 2-05 0-68 0-64 0-25 008

trace

trace

trace

002 0-44 1-71 0-21 4-13 0-16 2-31 004

0-50

1-55

30-16

41-47

016 1-41 0-33 1-89 013 1-56

trace

0-23

3-52 trace

0-44 1-76 1-76 0-48 0-50 0-29

0-20 2-76 0-94 4-66 0-25 2-27

2-39

0-32 2-68 0-82 2-98 007 0-91 002 1-77

55-49

44-58

57-30

1-95 0-51 3-16 0-20

1-22 trace

2-16

All yolk-sac fluids contained trace amounts of phosphoserine, glycerophosphoethanolamine and phosphoethanolamine, but these volumes were too small to be measured accurately. These three compounds were not detected in blood sera. = None detected.

At Stage 6, there is a similar pattern to Stage 5 but again the relative con¬ centrations of the various protein bands have changed. There is less total material in the albumin region, and the fastest moving component is again the most dense. The total concentration, determined by the densitometer, is less than at Stage 5.

Marilyn B. Renfree

488

. Amino acid analysis A total of thirty-five free amino acids and amino compounds was identified in yolk-sac fluids (Table 1) while thirty were identified in blood sera (Table 2). Both qualitative and quantitative differences were found between samples of yolk-sac fluids and blood. Taurine, cystine (including cysteine), yS-alanine, and Table 2 free amino acids and amino compounds in blood serum of the wallaby

Macropus eugenii (^-mole/ml) Embryonic stage Amino acid

Taurine Urea Unknown 1 Aspartic acid Unknown 2 Threonine Serine

2-31 0-02 0-01 001 006

1-70 002 002 002 005

2-49 0-05 006 002 015

2-30 004 0-04 003 0-11

3-37 010 003 0-02 009

0-70 001 0-01 001 007

Asparagine

0-25

0-32

0-45

0-33

0-41

0-26

0-39

007 013 002 0-39 0-24 008

009 015 003 0-40 0-38 0-09

0-24 0-64 005 0-89 0-72 0-28

010 0-29 004 0-53 0-29

016 0-26 0-07 0-49 0-40 016

008 0-12 003 0-26 0-22 019

019 017 005 0-49 0-42 0-35

001 007 008 0-03 0-04

0-02 015 0-18 007 007

Glutamine Proline Glutamic acid Citrulline

Glycine

Alanine Valine

Cystine

Methionine Isoleucine Leucine

Tyrosine Phenylalanine /}-Alanine Hydroxylysine

GABA Ornithine Ethanolamine Ammonia

Lysine 1 -Methylhistidine Histidine

018

317 trace

005 002 0-15

trace

001 004 0-07 004 004

001 005 006 004 0-05

011 0-16 0-15 006 007

0-29 0-20 0-15 007 009

0-13 010 0-10 005 006

0-10 002

010 002

0-17 002

0-14 002

019 0-02

010 0-02

019 002

008

007

008 004 008

016

004

005

trace

0-26 003 007 0-01 004

007

0-05 001

0-23 002 008 001 002

trace

trace

trace

001

trace

0-02

7-12

5-65

6-42

2-42

6-38

3-Methylhistidine Arginine

trace

Total

4-05

004

3-71

All blood sera contained trace amounts of tryptophane, carnosine and creatine, but these volumes were too small to be measured accurately. These three compounds were not detected in yolk-sac fluids. = None detected.

of y-amino-butyric acid and hydroxylysine were found in yolkfluids and not in sera, while, conversely, trace amounts of tryptophane, carnosine and creatinine were detected in blood sera but not in yolk-sac fluids. Ammonia was removed from sera before amino acid determination and so the ammonia content of yolk-sac fluid was not compared with ammonia in trace amounts

sac

sera.

489 Protein, amino acids and glucose in the wallaby Urea determinations, using the automatic analyser, are not as accurate as the simultaneous determination of the free amino acids, since urea gives a

peak heights-mole of substance than the other ninhydrin-positive compounds. The urea values for the yolk-sac fluids were consistently higher than those of the corresponding maternal serum. Serum urea values lie in the range of values given for M. eugenii on a low nitrogen diet (Lintern & Barker, 1969). Most amino compounds were ten times more concentrated in yolk-sac fluid than in serum. The average total amino acid content was 48-4 µ-moles/ml in yolk-sac fluids as compared with 5-11 µ-moles/ml in blood sera samples. The only amino acid which was consistently more concentrated in the serum was glutamic acid. Serine, asparagine and glutamine are not separated by the automatic analyser under these conditions and are expressed as a combined value, so these were separated by paper chromatography. Approximately 55% of this combined value was found to be due to serine, 40% to glutamine and the rest to asparagine. smaller

C. Glucose determination Glucose content of the samples (Text-fig. 1) doubled between Stage 4 and Stage 5, after which there was little change. This is a similar pattern to protein concentrations (Plate 1), which increased up to Stage 5 before flattening out. Serum levels fluctuate because animals were not routinely starved before operation. The mean value of 54-6 mg% and range 18 to 98 mg% agrees with a similar wide range of results found by Barker (1961) for the quokka, Setonix brachyurus. Barker suggested that this wide range was due to differences between animals with respect to their resting blood sugar levels, and showed a mean of 53-0 mg% glucose for the recently feeding quokka. Despite this large range, the concentration of glucose in yolk-sac fluids after Stage 4 was invariably higher than in

serum.

DISCUSSION

The concentrations of substances in the yolk-sac fluid change as pregnancy advances, although these changes are not reflected in the respective maternal blood sera. A marked change occurs in the proteins of the yolk-sac fluids between Stages 4 and 5, which is the period of attachment to the uterine endometrium. Before Stage 5, the embryo is an unattached vesicle in the uterus and its only source of nutrient material is the secretion of the surrounding uterine epithelium and gland cells. Once the vesicle has attached, the foetal and maternal circula¬ tions are in close contiguity throughout the extensive vascular omphalopleur, although uterine secretions may still be adding to this source. The gel patterns of the Stage-4 yolk-sac fluids are unlike that of serum. As the embryo grows, this pattern becomes more like that of the serum. The qualitative similarities of the yolk-sac fluid and the sera after attachment, and the quantitative differences in relative concentrations of their components support the suggestion that the proteins enter the yolk-sac fluid by a selective

Marilyn B. Renfree transfer through the placenta, and are not merely simple transudates of serum. This would point to the chorio-vitelline placenta as the main source of nutrients for the embryo, since the period of greatest growth occurs after attachment and the formation of the placenta. Glucose levels in the yolk-sac fluids show the same trend as the protein components except that they increase to levels which are from two to four times higher than those in sera, while protein concentrations are always approxi¬ mately thirty times lower than in serum. In ruminants, the main energy sources are the volatile fatty acids, and hexose absorption is minimal. The placenta, however, synthesizes fructose (which appears unable to cross the placental barrier) from maternal glucose (Alexander, Huggett, Nixon & Widdas, 1955). Barker (1961) has shown that the marsupial, Setonix brachyurus, produces large quantities of volatile fatty acids and in this respect resembles ruminants. Glucose levels are higher in the yolk-sac fluid of the wallaby than in the maternal blood serum so that the placenta must be selecting (or synthesizing) glucose. An analysis for fructose and a comparison of fructose concentrations and transport mechanisms with those of glucose would show whether there is an analogy between ruminants and wallabies in this aspect of their physiology. In contrast to glucose and protein, amino acid values are equally high at all stages, both before and after implantation. The total free amino acid con¬

490

is almost ten times that of maternal sera, so that, if active transport is occurring, it must be across a high concentration gradient. These values do not alter after implantation. This feature is consistent with the findings of Deren et al. (1966) and Butt & Wilson (1968) in the rabbit yolk sac. They showed that the yolk sac actively transports L-valine and other neutral amino acids, while it lacks the sugar transport mechanism present in the small intestine of a maturing foetus of the same age. A similar condition may apply in the wallaby, because the high concentrations of amino acids indicate an active transport mechanism. How¬ ever, the presence of amino acids is probably not due solely to a breakdown of protein because only one main protein component is present at Stage 4, and this is in a relatively low concentration. Of course, metabolism of this protein may provide the only source of amino acids. The only amino acid which is found in higher concentrations in the blood than in the yolk-sac fluids is glutamic acid. This particular amino acid could cross the membrane into the yolk-sac fluid by simple diffusion and then be converted to many of the other amino acids (and/or important metabolites) if appropriate keto-acids are available. This mechanism of transfer would account for the presence of several amino acids, but an active transport mechan¬ ism would still be required to account for the presence in the yolk sac of the other amino acids which cannot be derived from glutamic acid. The high concentrations of urea in the yolk-sac fluids suggest that the placen¬ ta is not freely permeable to urea, or that the rate of urea synthesis by the embryonic tissues exceeds the rate of diffusion across the placenta. In man, rabbit and dog, there is a complete equilibrium of urea between foetal and maternal blood so that it does not accumulate in the embryo. In the wallaby, tent

Protein, amino acids and glucose in the wallaby however, the yolk

excretory products.

sac

is

apparently acting

as a

(perhaps transient)

491 store

of

The high concentration of material in the yolk-sac fluids may be serving at least two purposes: (1) it may affect the osmotic pressure and so help to main¬ tain the yolk sac in a turgid condition, and (2) it may be acting as a nutrient store. In the bat, although the yolk sac is completely collapsed, Stevens & Easterbrook (1969) have shown that the endodermal cells do have a storage function, and accumulate large quantities of glycogen and lipid. The possibility remains, however, that the components of the yolk-sac fluid are due to a syn¬ thesis by the embryo. Further studies are being conducted to determine the sources of these components. It has been shown that the composition of yolk-sac fluid is different from that of maternal sera with respect to the constituents analysed. These results indicate that the yolk sac is not an inert barrier but a functional membrane that controls the transfer of a wide range of materials, both before and after implantation, by possible mechanisms ranging from simple diffusion to active transport. ACKNOWLEDGMENTS

indebted to Dr C. H. Tyndale-Biscoe for his helpful advice and encourage¬ throughout this project. I would also like to thank Mr L. B. James, Department of Biochemistry, John Curtin School of Medical Research, Australian National University, who performed the amino acid analyses, and to Mr I. A. Fox for photography. This work has been supported by a grant to Dr C. H. Tyndale-Biscoe from the Commonwealth Scientific and Industrial Research Organization. I

am

ment

REFERENCES

Alexander, D. P., Huggett, A. St. G., Nixon, D. A. & Widdas, W. F. (1955) The placental transfer of sugar in the sheep: the influence of concentration gradient upon the rates of hexose formation as shown in umbilical perfusion of the placenta. J1. Physiol., Lond. 129, 367. Barker, J. M. (1961) The metabolism of carbohydrate and volatile fatty acids in the marsupial Setonix brachyurus. Q_.Jl exp. Physiol. 46, 54. Beier, H. M. (1968) Uteroglobin: a hormone-sensitive endometrial protein involved in blastocyst development. Biochim. biophys. Acta, 160, 289. Berger, P. J. (1966) Eleven month "embryonic diapause" in a marsupial. Nature, Lond. 211, 435. Bergmeyer, H. U. & Bernt, E. (1965) Determination of D-glucose with glucose oxidase andperoxidase. In: Methods of Enzymatic Analysis. Ed. H. U. Bergmeyer. Academic Press, New York and London. Brambell, F. W. R. (1958) The passive immunity of the young mammal. Biol. Rev. 33, 488. Brambell, F. W. R. (1961) The transmission of antibody from mother to foetus. Proc. R. Soc. Med. 54, 992. Brambell, F. W. R., Hemmings, W. ., Oakley, C. L. & Porter, R. R. (1960) The relative transmis¬ sion of the fractions of papain hydrolysed homologous y-globulin from the uterine cavity to the foetal circulation in the rabbit. Proc. R. Soc. B, 151, 478. Butt, J. H. & Wilson, T. (1968) Development of sugar and amino acid transport by intestine and yolk sac of the guinea pig. Am. J. Physiol. 215, 1468. Clark, M. J. (1968) Termination of embryonic diapause in the red kangaroo, Megaleia rufa, by injec¬ tion of progesterone or oestrogen. J. Reprod. Fert. 15, 347. Davis, B. J. (1964) Disc electrophoresis. II. Method and application to human serum proteins. Ann. N.T. Acad. Sci. 121, 404. Deren, J. J., Padykula, H. A. & Wilson, T. H. (1966) Development of structure and function in the mammalian yolk sacs. 2. Vitamin B12 uptake by rabbit yolk sacs. 3. Development of amino acid transport by rabbit yolk sac. Devi Biol. 13, 349.

492

Marilyn B. Renfree

Deren, J. J. & Wilson, T. H. (1964) Development of transport mechanisms in foetal intestine and yolk sac of the rabbit. Fedn Proc. Fedn Am. Socs exp. Biol. 23, 262. Lintern, S. M. & Barker, S. (1969) Renal retention of urea in the Kangaroo Island wallaby, Protemnodon eugenii (Desmarest). Aust. J. exp. Biol. med. Sci. 47, 243. Lowry, O. H., Rosebrough, N.J., Farr, A. L. & Randall, R. J. (1951) Protein measurement with the folin phenol reagent. J. biol. Chem. 193, 265. Lutwak-Mann, C. (1962) Glucose, lactic acid and bicarbonate in rabbit blastocyst fluid. Nature, Lond. 193, 653. Lutwak-Mann, C, Boursnell, J. C. & Bennett, J. D. (1960) Blastocyst-uterine relationships: uptake of radioactive ions by the early rabbit embryo and its environment. J. Reprod. Fert. 1, 169. McCarthy, E. F. & Kekwick, R. A. (1949) Electrophoretic and ultracentrifugai examination of rabbit blastocyst fluid. J. Physiol., Lond. 108, 177. McCrady, E. (1938) The embryology of the opossum. Am. ant. Mem. 16, 1. Ornstein, L. (1964) Disc electrophoresis. 1. Background and theory. Ann. N.T. Acad. Sci. 121, 321. Padykula, . ., Deren, J. J. & Wilson, . H. (1966) Development of structure and function in the mammalian yolk sac. 1. Developmental morphology and vitamin B12 uptake of the rat yolk sac. Devi Biol. 13, 311. Sharman, G. B. (1961) The embryonic membranes and placentation in five genera of diprotodont marsupials. Proc. zool. Soc. Lond. 137, 197. Smith, M. J. H. & Moses, V. (1960) Uncoupling reagents in metabolism—1. The effects of salicylate

and 2 : 4 dinitrophenol on the incorporation of I4C from labelled glucose and acetate into the soluble intermediates of isolated rat tissues. Biochem. J. 76, 579. Smith, M. J. & Sharman, G. B. (1969) Development of dormant blastocysts induced by oestrogen in the ovariectomized marsupial, Macropus eugenii. Aust. J. biol. Sci. 22, 171. Spaokman, D. H., Stein, W. H. & Moore, S. (1958) Automatic recording apparatus for use in the chromatography of amino acids. Analyt. Chem. 30, 1190. Spector, W. S. (Ed.) (1956) Handbook of biological data. Saunders, Philadelphia and London. Stevens, R. J. & Easterbrook, N. (1969) New cytoplasmic organelle related to both lipid and glycogen storage materials in the yolk sac of the bat, Tadarida brasiliensis. Am. J. Anat. 124, 47.