A newin vitro culture technique for rat embryos

A New In Vifro Culture Technique for Rat Embryos' MAURICE A. ROBKIN, THOMAS H. SHEPARD AND TAKASHI TANIMURA Department of Nuclear Engineering a n d Central Laboratory f o r Human Embryology, Department of Pediatrics, Uniuerszty of Washington, Seattle, Washington 98195

ABSTRACT A new technique for culturing embryos, which permits detailed observation and manipulation, is described. The method has been applied to rat embryos on day 10, 11, or 12 of gestation, Twenty-four-hour culture of day-11 rat embryos resulted in nearly normal increases in somite number, and embryo length and protein increment of about half that occurring in vivo. After 24 h of culture the embryos generally were translucent and showed no apparent areas of gross tissue necrosis. Sections from representative embryos are presented. The apparatus is extremely compact and requires only about 9 ml of medium which is sufficient to support three embryos. The small volume allows the measurement of metabolites consumed or liberated with reasonable accuracy. The method permits a detailed and continuous record to be made of the heart rate and the addition of drugs to and sampling of the culture medium. The apparatus, named the PLASMOM, was used to measure the effect of changes in the ambient temperature on the heart rate of day-11 rat embryos. The results showed a linear dependence of heart rate on temperature between 30 and 40" C.

Teratologists and developmental biologists have long recognized the advantages of being able directly to observe the developing embryo. Nature permits this for oviparous species in a relatively straightforward way. In the case of viviparous species nature is not so cooperative. From the early days of this century workers have attacked the problem of culturing growing mammalian embryos. Their conclusions (see e.g., Waddington and Waterman, '33; Nicholas and Rudnick, '34, '38; Nicholas, '38) tend to be fairly consistent. Although it is possible to grow presomite and early-somite embryos with tissue culture techniques in static medium, embryos at stages beyond about 30 somites require a circulating medium to survive for more than a very short period of time. Nicholas' ('38) techniques were not widely adopted and further advances in the culture of mammalian embryos were not made for 30 years. The work of New (New, '67; New and Daniel, '69) showed that, in fact, using a circulating culture medium would enable rat embryos to grow and develop in vitro past the 30-somite stage. This was true for embryos up to about 50 somites as measured by increases in somite number, TERATOLOGY, 5: 367-376.

crown-rump length, and protein content. This work put the in vitro culture of rat embryos on a solid foundation and provided a dependable preparation for experiments in development, both normal and abnormal. New's ('67) apparatus, however, requires that the perfusing gas drive the circulation which results in fluid pulsations that cause the embryos to be agitated on their attachments to their rafts. The gas bubbles cause frothing when human serum is used as the culture medium which in some cases requires the use of an antifoamant as an adulterant. In addition the circulator must be maintained in a thermostatically controlled enclosure which inhibits access, observation, and manipulation. Tamarim and Jones ('69) have proposed a culture device, but we have had no experience with it. A desirable culture system should have the following attributes: (1) visual accessibility, (2) physical accessibility for instrumentation and obtaining specimens of medium, (3) small internal volume, and 1 Supported by the NIH grants HD02392, HD00180, and HD00836, and by the Graduate School Research Fund, University of Washington. 2 Present Address: Department of Anatomy, Kyoto University, Kyoto, Japan.

367

368

M. A. ROBKIN, T. H. SHEPARD AND T. TANIMURA

(4) simple construction and repair. Such

a device has been constructed. Twenty-four-hour culture experiments with day-11 rat embryos (24-26 somites) have given good growth results as determined by the increases in somite number, crown-rump length, and protein content. The device has been named the PLASMOM and is described in detail below. In addition an electronic method of measuring and recording the heart rates of these embryos is described and illustrative data from its use are given. APPARATUS, MATERIALS, AND PROCEDURES

A schematic of the PLASMOM system is shown in figure 1. Taking each part in order of circulation, the various components are described as follows: Pump. The circulating medium is driven by a small-displacement roller pump with a torque control monitor which permits the turning rate to be accurately set (Cole-Parmer Ultra-Masterflex with Laser Beam modulated by heart beat

Laser Beam Loser

-

Embryo

Photo Multiplying Tube

A

w

Strip Chart Recorder

L,1 Filter

Water Bath

Fig. 1 Diagram of the components of the PLASMOM embryo culture device.

Model 7014 pumphead). Depending on the tubing used the pumping rate tends to change with time as the tubing fatigues. Silicone rubber tubing has the best fatigue vs. time characteristics, but is sufficiently permeable to oxygen to make i t impossible completely to saturate the nutrient medium. Although Tygon tubing has an inferior fatigue rate it was preferable because it is relatively impermeable to oxygen. Using this tubing the medium may be kept near saturation for the duration of the experiments so far attempted. Using a larger pumphead with larger diameter pump tubing permits higher flow rates with lower RPM and so prolongs tubing life. (This improvement must be balanced against an increase in the agitation of the embryos with the increased surging of the medium due to the increased volume displaced per pulse.) Filter. The filter is a simple glasswool-filled tube which serves to trap particulates circulating with the medium. The glass wool may be autoclaved, and the system is assembled with the filter in the circuit. Bubble trap and thermistor port. The bubble trap and thermistor port is a cruciform glass construct which serves to provide a preferred diversion for any incidental bubbles flowing with the medium. The thermistor port is covered with a serum stopper, and needle-mounted thermistors are inserted to monitor the temperature on both sides of the embryo chamber. One of the thermistors may be used to control the heating circuit, if desired. Oxygen electrode port. The oxygen saturation level is monitored by a polarographic-type Clark oxygen electrode (Yellow Springs Instrument Co., Yellow Springs, Ohio) held in a glass bubble with small inlet and outlet lumens which give a rapid Aow of medium across the face o f the electrode membrane. The oxygen electrode is usually placed just upstream from the embryo chamber and monitors the relative oxygen saturation of the medium as it is delivered to the embryos. The oxygen level is calibrated at the start of each experiment at the appropriate temperature by adjusting the readout value to 21% before the oxygen flow is started. Presumably the medium is in equilibrium with the atmosphere until the gas flow is

I N V l T R O CULTURE OF R A T EMBRYOS

started. The flask is usually left open to the atmosphere so no hyperbaria develops, but by strengthening the glassware and closing the openings to the atmosphere the system could be operated hyperbarically with relatively little modification. E m b r y o c h a m b e r . The embryo chamber is a high-quality glass tube of square cross section made from a cuvette for a spectrophotometer with round glass tubing fused on the ends to accommodate the silicone rubber connectors. The optical properties are excellent as the opposite sides are parallel and carefully polished. The internal dimension of the chamber used in our apparatus is about 7 mm. (Using a two-cell flashlight, we have obtained color motion pictures of the embryos at high enough magnification to see the flow ofindividual red blood cells in the yolk sac circulation.) L u n g . The lung of the PLASMOM is composed of a bundle of silicone rubber tubes immersed in a water-filled flask. The perfusing gas of the desired composition is bubbled into the water and passes by diffusion through the tubes into the culture medium which is flowing inside the tubes. The silicone rubber is extremely permeable to oxygen, carbon dioxide, and water vapor, with transfer rates proportional to the difference in partial pressure across the tube walls. The diffusive transfer eliminates the bubbles in the medium. Since there is no direct contact there is no frothing of the serum and no danger of serum protein denaturation. Thus, there is no need to add an antifoaming agent to the medium. For a total volume of 10 ml flowing at about 10 ml/min the 02 saturation level increases from 21 to 80% in about 5 min when 95% 02 is used as the perfusing gas and the lung has a total of 50 linear feet of 0.020 X 0.037-inch tubing made up as a bundle of 25 %foot lengths. W a t e r b a t h . The flask in which the tube bundle is immersed is supported in a heating mantle controlled by either one of the circuit thermistors or a third thermistor in the water bath (if desired). The heated water warms the circulating medium at the same time that gas exchange occurs. Once the desired temperature is reached long-term thermostasis is excel-

369

lent and there are practically no shortterm fluctuations. The water bath does not interfere in any way with observability of the embryos since they are held well above the flask. A s s e m b l y . The various parts of the circuit are connected by rubber tubing, and assembly and disassembly are very simple. The assembled system and the components described above are shown in figure 2. Heart-rate determination. Automatic measurement of the embryonic heart rate is made by a method of transillumination using a laser and photomultiplying tube. The beam from a red (He-Ne) low-power laser (Coleman Instruments, Maywood, Ill.) is passed through the embryo. A lens focuses an image of the beating heart into the face of a photomultiplying tube, and the flicker produced by the beating heart is detected and recorded either by an oscilloscope or strip chart recorder. The signal can also be passed into the input of a cardiotachometer, and the heart rate obtained directly. The rate also may be obtained from the oscilloscope or strip chart recorder, so that a simultaneous record of both the heartbeat pattern and rate is obtained. Embryos have been exposed to the laser for up to an hour with no apparent ill effect. Figure 3 shows a typical strip chart recording of individual heart beats of a day-11 rat embryo. Sampling of the m e d i u m . A PLASMOM circuit with a single embryo chamber (as many chambers as desired may be used simultaneously) with room for three embryos has an internal volume of about 9 ml. This is a small enough volume so that the amounts of metabolites consumed or liberated may be measured with reasonable accuracy. Since the entire circuit of the PLASMOM is accessible small aliquots of medium may be withdrawn from time to time for analysis. M e d i u m . Based on his experiments New ('67) concluded that the best results were obtained by culturing rat embryos in homologous serum. Tanimura and Shepard ('70) found no statistically significant differences between cultures using homologous serum and human serum. The best results with human serum, however, were obtained with freshly drawn and prepared blood.

370

M. A. ROBKIN, T. H. SHEPARD A N D T. TANIMURA

Fig. 2

Assembled PLASMOM system

Time i n seconds Fig. 3 Strip chart recording (as recorded) of heartbeats of day-11% rat embryo. Time scale of 1 secldivision.

The circulator is usually filled with about 9 ml of human serum to which 667 IU penicillin and 12 Fg kanamycin have been added, In the explant procedure (Shepard et al., '71a) the embryos are placed in the chamber, which is filled with

about 2 ml of Hanks' solution, so that upon insertion into the PLASMOM the final volume is brought to about 11 ml and the serum slightly diluted. For initial chemical analysis a n additional 3 - 4 ml of serum is equilibrated with the circulating

I N VITRO CULTURE OF RAT EMBRYOS

371

imately three embryos. Figure 8 shows the somite counts from the same experiments. Heart rate and ambient temperature. To illustrate the use of the PLASMOM system further, the relation between the embryonic heart rate and the ambient temperature was determined. A length of rubber tubing was submerged in the water bath and connected to a separate pump drawing from an ice-water bath. This cooling coil provided for rapid cooling of the water bath and hence the circulating serum. The embryos were explanted into the PLASMOM. The temperature of the medium at the inlet and outlet of the embryo chamber was monitored by needle-mounted thermistors and the heart rate monitored by the laser-transillumination technique described above. The results of the experiment are presented in two graphs (figs. 9, 10). In the first, pooled data for the heart rate of several embryos are plotted against ambient temperature between 30 and 40" C. The experiments were done by first stabilizing the heart rate at a nominal base line temperature of 37-39' C. However, since the initial base line values varied somewhat each experiment was analyzed in terms of the change in temperature from the base line. Since the heart rates tended to become erratic above 39" C only experiments done below 39" C were analyzed. A null hypothesis of a linear response RESULTS was made and the data were regressionEmbryonic growth. The PLASMOM fitted to a straight lineof the formRate = circulator has been used in a series of 18- of a - b (AT - AT), where AT is the to 24-h growth experiments in which mean of the temperature changes over the Sprague-Dawley rat embryos were ex- pooled data. The results showed a temperplanted on day 11 of gestation. Figure 4 ature coefficient of 7.2 & 0.1% (value shows three embryos cultured for 20 h SE) change in heart rate per degree cenfrom explantation at day 11. These em- tigrade change in temperature. Student's bryos were translucent and showed no t value for the fit is t = 77 with 16 deapparent areas of gross tissue necrosis. grees of freedom which yields a probabilMitotic figures were present in histologic ity of rejection of the null hypothesis of sections of the embryos. There were some linearity of P < 0.000001 (Abromowitz necrotic areas in the center of the allan- and Stegun, '64). toic placenta. Figures 5 and 6 show secIn the second graph (fig. 10) the temtions of a day-11 embryo cultured €or 22 h. perature and heart rate are plotted against Figure 7 shows the protein content of the time for a single embryo to illustrate the experimental embryos and littermate con- close correlation. The gaps in the figure trols from each of seven experiments. Each near 40" C are due to the irregularity data point represents the mean of approx- observed in the heartbeat at these higher

medium using two hypodermic syringes as injection and withdrawal reservoirs. The extra serum is used for analysis and the net protein dilution owing to the Hanks' medium is 85% (2 ml out of 1415 ml total). Pre- and postexperimental maintenance. After assembly with new rubber tubing and glass components which have been soap and water washed the PLASMOM is wrapped in paper and sealed with autoclave tape. After being steam sterilized, the assembly is left wrapped until used in the next experiment. New pumphead tubing is sterilized by circulating 70 % ethanol followed by distilled water rinses prior to connection into the PLASMOM circuit . After each experiment all tubing is discarded and the glassware immediately rinsed with running tap water and then washed with soap and water. The tube bundle lung is copiously flushed with distilled water to wash out as much medium as possible. After flushing the lung is boiled to denature any serum proteins adhering to the tubing and soaked overnight in a 2 mglml solution of pronase in 0.2 M Tris buffer at pH 8. The buffer solution is flushed out and the PLASMOM reassembled. Explantation and assessment of the embryos. The methods of explantation and assessment of the embryos were described in Shepard ('71b).

*

372

M. A. ROBKIN, T. H. SHEPARD AND T. TANIMURA

3 73

I N VITRO CULTURE OF RAT EMBRYOS

IIOO. b

a. I

.V

900.

\ \

s

700.

AF CA

Ll

AA

od

b

300-

Of

\

-D

G

l' 100V.

0

8 16 24 32 40 48 Age- hours after 8 am. on day 10

56

Fig. 7 Protein content versus age of cultured rat embryos explanted on day 11. Upper case values are experimental results after culture. Lower case values are littermate controls. A data pair, (e.g.. a,A), corresponds to a single experiment. In vivo values denoted by V. Each data point is the mean of approximately three embryos.

temperatures which made measurement impossible. DISCUSSION

As a laboratory apparatus the PLASMOM provides certain advantages for the culture of mammalian embryos. The ratio of the mass of the embryos with their associated tissues in a preparation to the mass of the circulating serum is sufficiently large so that the embryos are able meaFig. 4 Three embryos cultivated in vitro for 22 h after explantation on day 11 (about the 25somite stage). Fig. 5 Stained section from an embryo grown for 22 h in nitro after explantatibn on day 11. Some increase i n necrotic cells was found in the peripheral part of the extremity. Scale represents 200 p.

Fig. 6 Section through the chorioallantoic placenta of a n explant. Same conditions as embryos shown in figures 4 and 5. Arrow indicates Reichert membrane and parietal yolk sac. Visceral yolk sac is at the top. Scale represents 100 p.

surably to modify the concentrations of metabolites in the medium. The compactness, accessibility, and ease of visualization of the preparation are a benefit in laboratories that are cramped for space. Since the system is separated by solid surface from any part of the external environment including the perfusing gas it is protected to a considerable extent from contamination. The absence of direct gas contact eliminates frothing of the medium and so reduces some associated uncertainties such as the effect of adding antifoaming agents or protein denaturation due to bubbles. The PLASMOM is easy to assemble and can readily be autoclaved. Maintenance of the apparatus is relatively straightforward. Some of the difficulties with the preparation that have not yet been worked out are the following. 1. To maintain the oxygen saturation level close to 90% it has been necessary

3 74

M. A. ROBKIN, T. H. SHEPARD AND T. TANIMURA

40

30

be

ao

20 cr, Q,

.s '4

G

v.

10 8 6 4

2 0

I

0

I

I

I

I

I

16 24 32 40 48 A g e - h o u r s a f t e r 8 a.m. on d a y 10

8

I

56

Fig. 8 Somite count versus age of cultured rat embryos explanted on day 11. Upper case values are experimental results after culture. Lower case values are littermate controls. In wino values denoted by V. Each data point is the mean of approximately three embryos.

to use roller pump tubing with a low oxygen diffusibility. Unfortunately the tubing used (Tygon) has a relatively short life in service. Long-lived silicone rubber tubing is much too permeable to oxygen to maintain high 0 2 levels. 2. It is difficult to maintain the system completely closed and still record the heartbeats with the laser technic. If the bubble bypasses, which act as surge chambers, are closed with serum stoppers rather than gauze dust traps, enough motion is induced in the embryos owing to pumping to introduce serious artifacts

into the heartbeat record. However we have observed relatively little problem with microorganism contamination when only gauze covers the bubble bypass standpipes. The antibiotics used appear to be completely effective over the course of our 24-h experiments, but they represent an additional uncertainty with respect to growth and development of the embryos. 3 . Although the components of the PLASMOM itself cost relatively little a complete system including a controlled pump, instrumentation for monitoring the ambient conditions, and heart-rate

375

I N VITRO CULTURE OF RAT EMBRYOS

4. Our 24-h growth results are still generally rather inferior to what is observed in viuo over the same interval of time. It is not clear whether this is due to a factor in the PLASMOM system itself or to the lack of maternal mediation. Until it is possible to maintain normal growth and function of the explanted chorioallantoic placenta we may be unable to improve on the reported growth results. The effect of maternal metabolism on the embryo certainly must be another important element that is nearly impossible to reproduce in this in vitro system. The biological in vitro results as measured by the increment in somite number are about 89% of what is achieved in uiuo during the corresponding embryonic period. The protein and length increment however are less, namely 59 and 52% of the in viuo change. A large number of factors could contribute to these shortcomings, but the lack of regulative maternal organ systems to modulate the level of glucose, oxygen, and other substances is the most likely fault. Oxygen may not be

. D

a = -21.8f0.2(fS.E)% b = 7.2 f 0. I O h per degree 3 = 2.6 degrees

/

60

6

1

I

I

7

5

4

3

2

1

0

Drop i n temperature from baseline "C ( A T )

Fig. 9 Effect of temperature on heart rate of 25-somite (day 11) rat embryos. Linear regression fit to data averaged at each temperature and restricted to temperature changes below base line temperature. Each data point is the mean of between one and six observations.

measuring equipment represents a substantial financial investment, so that experiments requiring multiple preparations become very expensive.

Heart Rate Response of 25 Somite (Day 1 1 113) Rat Embryo to Variations of Ambient Temperature in PLASMOM Culture

- 40 I6O-

-

cr,

A

.

Heart rate Temper at u re

E

-"A*.A .A

\

8120-

-

2

P

80-

'4

-

& $60

AA -&A I

I

I

I

.&A I

I

. * A A I

-34

cb

-32

2 3

-

*A

.A A%

-30 I

I

5

Q

a

4..A

x P,

-36

2

2.A t da

-

I

\

9,

L a

&

9 P, 100-

cb

Heart rate irregular and not measurable

a

b

-383

.n

9 5 140-0

.s

2

AA

I

I

I

I

I

I

-

Elapsed time - minutes Fig. 10 Correlated heart rate and temperature for a day-I1 rat embryo. Elapsed time measured from start of temperature transient.

3. 5

3

a

3 76

M. A. ROBKIN, T. H. SHEPARD AND T. TANIMURA

a limiting factor since New and Coppola (‘70) have grown 29-somite rat embryos in 2 atm of oxygen for 40-45 h and found no increase in protein and somites over those grown in 1 atm. Tanimura and Shepard (‘70) reported high utilization of glucose by rat embryos in vitro, and we have evidence from heart-rate increases that addition of extra glucose to the system after 8-10 h is a benefit. We have also observed that the lactate in the 10 ml of medium doubles during a 20-h culture period . An exact comparison with the results obtained by New (’67) is not possible because he did not give in vivo values after day 11 and most of his measurements were made after 40-h culture periods. His best 40-h results after explanting 22- to 28somite embryos were means of 43 somites and 1152 mg of protein. Our explanted embryos, starting at 24 somites and 236 mg of protein, attained the 33.5-somite stage and 632 mg of protein in a 22-h in vitro period. A discrepant increase in somite number over protein during in vitro culture was also reported by Berry (‘68) in rats, and Klein et al. (‘64) in chicks. Our studies, in view of the lack of any marked cell death, suggest that increased embryonic protein and morphogenesis may not be completely integrated. LITERATURE CITED Abromowitz, M., and I. A. Stegun 1964 Handbook of Mathematical Functions with Formulas,

Graphs and Mathematical Tables. National Bureau of Standards, Washington, D. C. Berry, C. 1968 Comparison of in vivo and in vitro growth of the rat foetus. Nature, 219: 92-93. Klein, N. W., E. McConnell and D. J. Riquier 1964 Enhanced growth and survival of explanted chick embryos cultured under high levels of oxygen. Dev. Biol., 10: 17-44. New. D. A. T. 1967 Develoument of exulanted embryos in circulating medium. J. Embryol. Exp. Morph., 17: 513-525. New, D. A. T., and P. T. Coppola 1970 Development of explanted rat fetuses i n hyperbaric oxygen. Teratology, 3: 153-159. New, D. A. T., and J. C. Daniel 1969 Cultivation of rat embryos explanted at 7.5 and 8.5 days of gestation. Nature, 223: 515516. Nicholas, J. S. 1938 The development of rat embryos in a circulatory medium. Anat. Rec., 70: 199-210. Nicholas, J. S., and D. Rudnick 1934 Development of rat embryos in tissue culture. Proc. Nat. Acad. Sci., 20: 656-658. 1938 Development of rat embryos of egg-cylinder to head-fold stages i n plasma culture. J. Exp. Zool., 78: 205-232. Shepard, T. H., T. Tanimura and M. A. Robkin 1971a In vitro studies for analysis of teratogenic events. In: Malformations Conghitales des Mammifkes. H. Tuchmann-Duplessis, ed. Masson, Paris, pp. 51-66. 1971b Energy metabolism in early mammalian embryos. Dev. Biol. Suppl, 4: 4258. Tamann, A., and K. W. Jones 1968 A circulating medium system permitting manipulation during culture of postimplantation embryos. Acta Embryol. Morph. Exp., 10: 288-301. Tanimura, T., and T. H. Shepard 1970 Glucose metabolism by rat embryos in vitro. Proc. SOC. Exp. Biol. Med., 135: 51-54. Waddington, C. H., and A. J. Waterman 1933 The development in vitro of young rabbit embryos. J. Anat., 67: 355-370.