EVALUATION OF CARDIOVASCULAR FLOW CHARACTERISTICS IN THE

129SV MOUSE FETUS USING COLOR-DOPPLER-GUIDED SPECTRAL

DOPPLER ULTRASOUND

WILFRIED MAI, DVM, PHD, JOHANN LE FLOC’H, PHD, DIDIER VRAY, PHD, JACQUES SAMARUT, PHD,PAUL BARTHEZ, DVM, PHD, MARC JANIER, MD, PHD

The purpose of this study was to evaluate color- and spectral Doppler ultrasound in the establishment of normal

functional cardiovascular development features in the mouse fetus. Mouse fetuses (129Sv strain) were studied in

utero between embryonic day (EDs) 9.5 and 19.5. Time–velocity curves were derived from Doppler interro-

gation of the aorta and umbilical artery. The sample volume was accurately placed on the vessels of interest

based on color-Doppler images. From these curves, the following parameters were obtained: heart rate (HR),

acceleration time (AT), and deceleration time (DT). HR increased between EDs 9.5 and 19.5 from 102.9 to

303.2 b.p.m. For the other parameters, the most significant change observed was the increase of DT in the

umbilical artery at the end of pregnancy, corresponding to the appearance of a diastolic flow. We report the use

of a commercially available, clinical, ultrasound unit to obtain quantitative data on the cardiovascular devel-

opment in the mouse fetus. These results may be useful for the recognition of in utero cardiovascular dysfunction

in transgenic or knock-out fetus. Veterinary Radiology & Ultrasound, Vol. 45, No. 6, 2004, pp 568–573.

Key words: circulation, color-Doppler, echocardiography, fetus, imaging, mouse.

Introduction

MURINE GENETIC MODELS associated with structural

and functional cardiovascular defects are now avail-

able. Some of these defects may be identified in utero and

there is a need for noninvasive techniques, such as Doppler

ultrasound, to evaluate murine fetal heart in utero. Dop-

pler measurements of blood velocity at the atrioventricular

cushions and outflow tract have already been performed in

mouse fetuses.1 But these experiments were invasive be-

cause measurements were made after laparotomy and hys-

terotomy. Furthermore, they were conducted in a limited

period, from embryonic day (ED) 10.5 to ED 14.5.1 In

another invasive study umbilical arterial blood flow pattern

was reported in mouse embryos between EDs 10.5 and

16.5.2 Noninvasive methods have been described allowing

Doppler examination of heart function in a mouse fetus

using 7.5-MHz transducers.3–6 Nevertheless, the low reso-

lution provided by such transducers was not adequate for

fine evaluation of specific fetal vessels. More recently, other

studies have described the assessment of cardiovascular

anatomy and function in the mouse fetus using very high

resolution ultrasound equipment, the so-called Ultrasound

Backscatter Microscopy, or Ultrasound Bio-Microscope

(UBM).�7–9 In these reports, a dedicated 40-MHz ultra-

sound imaging system was used via a trans-abdominal ap-

proach. These studies provided data characterizing

circulatory hemodynamics in the early developing mouse.

Doppler values obtained through noninvasive techniques

differed quantitatively from results obtained in invasive

studies. Nevertheless, these studies were also conducted

over a limited period of the whole pregnancy, i.e., between

EDs 9.5 and 14.5.

So far, all Doppler studies with the UBM have been

made placing the sample volume based on gray-scale two-

dimensional (2D) images. The vessels of interest, mainly

the dorsal aorta and the umbilical vessels, were identified

based on the hyperechoic streaming patterns within the

lumen.7,9 In previous studies using lower frequency com-

mercially available units, only flow within the cardiac cav-

ities could be registered because of a lack of spatial

resolution.3–6 In addition, fetal blood is not that echogenic

at lower frequencies, making it difficult to identify blood

flow without color-Doppler.

Commercial ultrasound units operating at a reasonably

high frequency (i.e., improved spatial resolution) are now

available. Although standard gray-scale 2D images do

not have the degree of resolution achieved with the UBM,

Address correspondence and reprint requests to Wilfried Mai, DVM,MSc, Radiology, School of Veterinary Medicine, University of Pennsyl-vania, 3900 Delancey Street, Philadelphia, PA 19104-6010.E-mail: wmai@vet.upenn.edu

Received October 13, 2003; accepted for publication February 27, 2004.doi: 10.1111/j.1740-8261.2004.04098.x

From the Small Laboratory Animal Imaging Platform ANIMAGE, 59Boulevard Pinel, 69003 Lyon, France (Mai, Janier), INSA, CREATIS,69621, Villeurbanne cedex, France (Le Floc’h, Vray), Ecole Normale Su-perieure de Lyon, Biologie Cellulaire et Moleculaire, 69364, Lyon, France(Samarut), Universite Claude Bernard, 69622, Villeurbanne cedex, France(Samarut), Ecole Nationale Veterinaire de Lyon, Radiologie, 69280,Marcy l’Etoile, France (Mai, Barthez, Janier).

�Visual Sonico, Toronto, Canada.

568

color-Doppler capabilities may enable their use in rodent

cardiovascular research. Clinical ultrasound units are wide-

ly available and easily accessible to researchers, as opposed

to specifically research-dedicated units like the UBM. The

aim of this study was to establish normal functional car-

diovascular development features in 129Sv mouse fetuses.

We evaluated the feasibility of accurately recording Dop-

pler parameters from major fetal vessels with a commer-

cially available ultrasound unit, using color-Doppler

guidance to position the sample volumes. In addition, we

conducted the study over an extended period of time (from

ED 9.5 to ED 19.5), to provide additional information

about the cardiovascular physiologic events that may occur

at the end of pregnancy.

Materials and Methods

Animal Conditioning

Animal-care procedures were conducted in accordance

with the guidelines set by the European Community Coun-

cil Directives.

Twenty-four timed pregnant mice (129Sv strain) were

included in this study. We chose to conduct our experi-

ments on the 129Sv strain because this strain is widely used

as a model of cardiovascular diseases, and it happens to be

the strain used in our laboratory to produce knock-out

mice lacking receptors to thyroid hormone (largely impli-

cated in the regulation of cardiovascular function).

Embryos were staged in days of gestation, with ED 0.5

being defined as noon of the day a vaginal plug was detected

after overnight mating. Mice were studied between EDs 9.5

and 19.5 (term is reached at ED 20.5 for this strain).

A commercially available ultrasound machine with a

linear transducer (7–15MHz) was used.� Studies were per-

formed with mice under general anesthesia (isoflurane 1–

1.5%) using specific equipment.w This system, developed

for small rodents, allows delivering of anesthetic gas

through a mask. The animal lies on a heating pad and

the delivered gas is heated to prevent hypothermia. The

degree of heating is controlled by a feedback system de-

pending on the actual rectal temperature of the mouse

measured with a rectal thermistor. This way, the body

temperature could be maintained at 37� 11C through-

out the experiments.10

The hair on the abdominal wall was clipped prior to

experimentation and acoustic gel was applied to provide

good coupling between the probe and the skin.

Ultrasound Acquisition

A midline or parasagittal transabdominal approach was

used to examine the uterine horns. The abdominal cavity

was virtually divided into four quadrants: left and right

cranial and left and right caudal. When more than four

fetuses were present, measurements were performed on

only one fetus per abdominal quadrant to ensure that no

fetus would be studied twice during the experiment.

The ultrasonic probe was operated at the highest fre-

quency range (‘‘resolution mode,’’ 15MHz; axial resolu-

tion measured at �20dB: 600mm; lateral resolution

measured at �20dB: 700mm (data provided by the man-

ufacturer)). Fetuses were located using gray-scale 2D im-

ages, and nearby maternal anatomic landmarks (kidneys,

urinary bladder) allowed for consistent identification of the

fetus at each measurement step.

A color-Doppler volume was superimposed on the gray-

scale image to locate blood flow within the heart, the dorsal

aorta, and the umbilical cord (Figs. 1 and 2).

Fig. 1. Color-Doppler ultrasound image in a mouse fetus at embryonicday 13.5. Signal from the aorta and the heart is visible. The spinal cordappears in the far field as a hyperechoic line.

Fig. 2. Color-Doppler ultrasound image of the umbilical cord in a mousefetus at embryonic day 17.5. The umbilical artery and the umbilical veinare visible. The vertebral column appears in the far field as multiple hyper-echoic dots.

�HDI 5000t; Philips Medical Systems, Bothell, WA.wEquipement Veterinaire Minervet, Esternay, France.

569Evaluation of Cardiovascular Flow Characteristics in 129SVMouse FetusVol. 45, No. 6

Following identification of the vessels of interest,

spectral Doppler interrogations were performed, placing

the sample volume successively on the dorsal aorta and

the umbilical artery. The Doppler settings were as

follows: sample-volume size: 0.5mm; wall filter: medium

level; gain level: 65–80%; and angle of insonation: less than

601. The pulse repetition frequency was set in order to

record the maximum velocities without aliasing (3500–

4000Hz).

From these time–velocity curves, the following nonan-

gle-dependent parameters were evaluated in dorsal aorta

and umbilical artery: heart rate (HR) derived from the

cardiac cycle (CC) duration measurement, acceleration

time (AT), and deceleration time (DT) (Fig. 3). The ma-

ternal HR was also recorded at the beginning and at the

end of the procedure from Doppler interrogation of the

abdominal aorta.

AT and DT were automatically measured and averaged

over five consecutive CCs using the specific automatic in-

tegrated software available on the ultrasound machine. HR

was manually measured and averaged over three consec-

utive CCs. For each vessel, two consecutive measurements

were made and values were then averaged. AT and DT

were normalized to the CC duration.

Data were acquired from embryos and fetuses by

scanning the pregnant mice daily between EDs 9.5 and

19.5. The total number of data for each recorded param-

eter was 186 acquired in 75 fetuses (fetuses were generally

not studied every day from ED 9.5 to ED 19.5, but rather

underwent one to five consecutive examinations). The

number of fetuses studied for each gestational stage was as

follows: NED9.5¼ 7; NED10.5¼ 7; NED11.5¼ 18; NED12.5¼19; NED13.5¼ 25; NED14.5¼ 27; NED15.5¼ 20; NED16.5¼ 15;

NED17.5¼ 13; NED18.5¼ 19; NED19.5¼ 16. AT and DT were

not recorded at EDs 9.5 and 10.5 because of difficulties

and lack of consistency in identifying the dorsal aorta and

umbilical artery at these stages.

Statistical Analysis

Results are presented as mean � standard deviation.

Normality and equal variance tests were performed on all

recorded parameters (HR, AT, DT). A one-way ANOVA

was used on normally distributed data. A Tukey test was

then performed to extract the groups that differed from the

others. If the normality test or equal variance test failed,

data were submitted to a Kruskal–Wallis one-way ANO-

VA on ranks and then pair-wise multiple comparison pro-

cedures were performed according to Dunn’s method. A

value of Po0.05 was considered significant.

Results

Color-Doppler Images

Good delineation of fetal vasculature was provided by

color-Doppler images as early as ED 11.5, and major fetal

vascular structures such as the umbilical artery and vein,

the aorta, the ductus venosus, and the caudal vena cava

could be identified accurately (Figs. 1, 2 and 4).

HR

HR increased from ED 9.5 to ED 19.5 (Fig. 5). The

differences in mean values among the gestational stages

were statistically significant for fetal HR (Po0.001),

whereas they were not for the maternal HR (P¼ 0.051).

The fetal HR remained lower than the maternal HR at the

end of pregnancy (Fig. 5).

Aortic and Umbilical Artery Doppler Parameters

For all parameters, there was a statistically significant

difference between gestational groups (Po0.05) (Fig. 6).

Fig. 3. Doppler parameters recorded from an umbilical arterytime–velocity profile in an embryonic day 13.5 fetus. Notice that the time–velocity curve contains two opposite flows: one positive and continuousflow from the umbilical vein, and one negative and systolic flow fromthe umbilical artery (CC, cardiac cycle; AT, acceleration time; DT, decel-eration time).

Fig. 4. Color-Doppler ultrasound image from an embryonic day 17.5 fe-tus. Note the ductus venosus (shunting blood from the umbilical vein to thecaudal vena cava and thus bypassing the liver), the dorsal aorta, the caudalvena cava (CVC), and the heart. The fetal spine is at the bottom of theimage, and the vertebral column can be seen as hyperechoic dots dorsallylocated with respect to the aorta.

570 Mai et al. 2004

A significant decrease of AT in the aorta was observed

from ED 16.5 to ED 19.5, associated with a significant

increase of DT at EDs 18.5 and 19.5. For the umbilical

artery a significant increase of AT was observed at EDs

18.5 and 19.5. A significant increase of DT in the umbilical

artery was observed at EDs 18.5 and 19.5 as compared

with the ED 9.5–16.5 period. This corresponded to the

progressive installation of a diastolic flow in this artery. At

the end of pregnancy, the flow within the umbilical artery

was often a continuous systolo-diastolic flow (37% of fe-

tuses at ED 18.5 (n¼ 19) and 44% of fetuses at ED 19.5

(n¼ 16)).

Discussion

This study provides a data set of normal values of non-

angle-dependent Doppler parameters in mouse embryos

and fetuses from the umbilical artery and the dorsal aorta,

in the specific 129Sv strain.

Data were acquired using a commercially available ul-

trasound machine, demonstrating the feasibility of such

analysis. In addition, this is the first study where color-

Doppler imaging was used to assess the cardiovascular

system in mouse fetuses from the onset of cardiac activity

until the end of pregnancy. Although clinical units have a

lower spatial resolution than the UBM, color-Doppler ca-

pabilities greatly facilitate the recognition of the vessels of

interest. Other functional studies were performed using the

UBM only for research purposes.7–9,11,12 This unit has a

high spatial resolution (axial resolution: 40mm; lateral res-

olution: 57–104mm), but the depth of exploration is limited

at the highest frequencies, and high blood velocities such as

in the heart and large arteries in late pregnancy are off-

scale with the UBM, which is not the case with commercial

machines.12

Although one study focusing on the heart chambers re-

ported measurements of fetal HR at EDs 17.5 and 19.5,4

previous studies reporting Doppler parameters from the

dorsal aorta and umbilical artery were limited to a shorter

period (ED 10.5–14.5 or ED 10.5–16.5).2,7,9 The rationale

for limiting the time interval of study was that most of the

cardiac changes take place before ED 14.5.13,14 However,

heart and major vessel development continues after ED

14.5: organogenesis is achieved, the final circulation is

established, and the placenta undergoes some modifica-

tions. Moreover, cardiovascular function is not only

determined by the heart shape and form, but also by sev-

eral humoral and vascular changes that may vary in

the late stages of pregnancy and thus influence cardiovas-

Fig. 5. Mean fetal and maternal heart rate (b.p.m.) as a function of em-bryonic day. The fetal curve best fits with a third-order polynomial functionwith a correlation coefficient R of 0.779 (Po0.001). Errors bars, standarddeviation.

Fig. 6. Mean acceleration time (AT) and deceleration time (DT) in aorta (left figures) and umbilical artery (right figures) as a function of embryonic day(ED) (AT and DT are normalized to the cardiac cycle duration, and therefore without unit). AT and DT did not vary in dorsal aorta throughout pregnancy,except at the end of pregnancy, where an increase of DT associated with a decrease of AT were observed. In umbilical artery there was an increase of AT at EDs18.5 and 19.5. An important and statistically significant increase of DT was observed from ED 18.5, corresponding to the appearance of a diastolic flow withinthis artery. Error bars, standard deviation.

571Evaluation of Cardiovascular Flow Characteristics in 129SVMouse FetusVol. 45, No. 6

cular function. A previous study conducted between EDs

9.5 and 14.5 reported that forward diastolic flow was ab-

sent in the umbilical artery. A diastolic flow has been de-

scribed in the human fetus as early as week 13 of

pregnancy.15 Our study, extended over a longer period of

the pregnancy, demonstrated that a diastolic flow is often

present within the umbilical artery in the mouse fetus after

ED 17.5. This consequently implies that the mouse

fetus may represent a potential model for human placen-

tal morphological and vascular abnormalities that can be

evaluated using Doppler examination of the umbilical ar-

tery flow.

The question of anesthesia is critical when performing

fetal physiologic studies. Indeed, embryos and fetuses are

particularly sensitive to homeostasis variations. Previous

studies used similar doses of pentobarbital delivered intra-

peritoneally.1,2,4,7,8 However, pentobarbital is known to

decrease cardiac output and arterial pressure and may not

be suitable when performing studies of physiologic param-

eters, especially on embryos or fetuses.16 We chose to in-

duce and maintain anesthesia with isoflurane. Even though

the cardiovascular effects of volatile anesthetics in murine

prenatal hearts have not been well investigated, a recent

study reported a comparison of the effects of halothane

and isoflurane at clinically relevant concentrations on chick

embryos, measuring the dorsal aortic blood velocity by

Doppler techniques.17 In that study halothane, but not

isoflurane, caused a significant decrease in cardiac stroke

volume and maximum acceleration of blood, which is an

index of cardiac performance. Also the embryonic HR was

not affected by either drug. Isoflurane appears to be an

anesthetic agent of choice to perform studies on the cardio-

vascular function in embryos and fetuses, despite the lack

of specific data on the mouse. Another advantage of vol-

atile vs. fixed anesthesia is that gases are rapidly eliminated

by the lungs, allowing more precise control of anesthesia

and faster recovery. This is important when performing re-

peated anesthesia on the same pregnant mice on a day-to-

day basis.

Our measurements of HR differ from those reported in

other studies. Phoon et al. found higher values of HR than

previously reported, and ascribed this observation to the

noninvasive nature of their assessment, and especially to

the stringent thermoregulation used.1,2,4,7 Despite the fact

that we also used a noninvasive method and a stringent

thermoregulation technique, we did not find such high

values of HR.7 Rather, our results seem to be closer to two

other studies: (1) a noninvasive technique and thermoreg-

ulation using warming pad and radiant lamps and (2) an

invasive technique and thermoregulation in a warmed

bath.2,4 Therefore, the type of anesthesia is probably not

the cause of these differences and it can be hypothesized

that they are because of the genetic particularities of the

strains used: CD1 mice,8 vs. ICR mice,2 vs. Swiss-Webster

mice,7,9 vs. 129Sv mice (our study). Each strain may be

characterized by individual variations of physiologic pa-

rameters such as HR, and this could explain the differences

that we, and others, observed.18

Our results confirmed previous descriptions of the evo-

lution of the murine fetal HR during pregnancy, where an

increase of HR during embryonic and fetal development

was reported.1,2,4,7 We provided additional information on

what happens at the end of pregnancy. Interestingly, the

fetal HR at the end of pregnancy remained below the ma-

ternal HR, as opposed to the human fetus, in which HR is

higher than the maternal HR.

We limited our measurements to a few spectral Doppler

variables, and did not measure other parameters reported

in previous studies, like, for instance, the peak velocity.7,9

This parameter is highly dependent on accurate alignment

with flow, and we preferred to measure parameters that are

not influenced by the accuracy of alignment. Also, we

chose not to measure parameters such as the nonejection

time (NET) or the ejection time (ET), reported in other

studies, since these parameters can be derived from the

knowledge of HR, AT, and DT. Hence, ET and NET

would have been redundant.

AT and DT measurements can be influenced by the lo-

cation of the Doppler sample-volume within a given vessel.

This could account in part for the variability in our results.

In the aorta, we paid attention in always placing the sample

volume right distally to the aortic arch, which was easily

recognized on color-Doppler images. Placing the sample

volume consistently at the same place in the umbilical ar-

tery was more challenging, and this is probably a limitation

of the study.

In our experiments, we could not consistently identify

blood flow within the aorta and umbilical artery before ED

11.5. This is because at the beginning of spontaneous car-

diac activity, the unseptated heart is capable of maintaining

some circulation but blood flow is probably too weak to be

identified with the ultrasound unit we used. Similarly, we

were not able to identify heartbeats and Doppler signal

before ED 9.5, and in two mice cardiac activity was not

visible at the time we began the measurements, at ED 9.5.

In a recent study of mouse embryos using the UBM,

rhythmic cardiac activity appeared a little before ED 8, and

a Doppler signal was identified as soon as ED 8.5.19 By ED

9.5, 100% of embryos had both rhythmic cardiac activity

and vascular Doppler signals. Our different observation

may be explained by a lack of spatial resolution and Dop-

pler sensitivity of the ultrasound unit we used, which op-

erates at a lower frequency than the UBM. In addition, in

the mentioned study using the UBM, embryos were imaged

after laparotomy and exteriorizing the uterus, which

increases the sensitivity by decreasing both the distance

and tissue interposition between the ultrasonic crystal and

the embryos.

572 Mai et al. 2004

Conclusion

With new high-frequency clinical ultrasound units now

widely available, gray-scale, color- and spectral Doppler

ultrasound can be used to monitor cardiovascular function

in mouse fetuses and embryos from ED 9.5 until the end of

pregnancy. Although the sensitivity/specificity of this

technique to detect pathologic changes is still to be deter-

mined, it is likely that values reported herein in the 129Sv

strain might be used as reference values to assess cardio-

vascular function in genetically engineered mice of this

specific strain.

REFERENCES

1. Keller B, Keller B, MacLennan MJ, Tinney JP, Yoshigi M. In vivoassessment of embryonic cardiovascular dimensions and function in day-10.5to -14.5 mouse embryos. Circ Res 1996;79:247–255.

2. MacLennan MJ, Keller BB. Umbilical arterial blood flow in themouse embryo during development and following acutely increased heartrate. Ultrasound Med Biol 1999;25:361–370.

3. Huang GY, Linask KK. Doppler echocardiographic analysis ofeffects of tribromoethanol anesthesia on cardiac function in the mouseembryo: a comparison with pentobarbital. Lab Anim Sci 1998;48:206–209.

4. Gui YH, Linask KK, Khowsathit P, Huhta JC. Doppler echocar-diography of normal and abnormal embryonic mouse heart. Pediatr Res1996;40:633–642.

5. Linask KK, Gui YH, Huhta JC, Khowsathit P. Evaluation of heartdefects in early TS 16 mouse embryos using echocardiography. Mol Biol Cell1993;4:144a.

6. Linask K, Huhta J. Use of Doppler echocardiography to monitorembryonic mouse heart function. Methods Mol Biol 2000;135:245–252.

7. Phoon CK, Aristizabal O, Turnbull DH. 40 MHz Doppler charac-terization of umbilical and dorsal aortic blood flow in the early mouse em-bryo. Ultrasound Med Biol 2000;26:1275–1283.

8. Srinivasan S, Baldwin HS, Aristizabal O, et al. Noninvasive, in uteroimaging of mouse embryonic heart development with 40-MHz echocardi-ography. Circulation 1998;98:912–918.

9. Aristizabal O, Christopher DA, Foster FS, Turnbull DH. MHzechocardiography scanner for cardiovascular assessment of mouse embryos.Ultrasound Med Biol 1998;24:1407–1417.

10. Suckow M, Danneman P, Brayton C. Important biological features.Lab Mouse 2001;1–12.

11. Foster FS, Zhang MY, Zhou YQ, et al. A new ultrasound in-strument for in vivo microimaging of mice. Ultrasound Med Biol 2002;28:1165–1172.

12. Zhou YQ, Foster FS, Qu DW, Zhang M, Harasiewicz KA, Adam-son SL. Applications for multifrequency ultrasound biomicroscopyin mice from implantation to adulthood. Physiol Genom 2002;10:113–126.

13. Vuillemin M, Pexieder T. Normal stages of cardiac organogenesis inthe mouse: II. Development of the internal relief of the heart. Am J Anat1989;184:114–128.

14. Vuillemin M, Pexieder T. Normal stages of cardiac organogenesis inthe mouse: I. Development of the external shape of the heart. Am J Anat1989;184:101–113.

15. Wladimiroff JW, Huisman TW, Stewart PA, Stijnen T. Normal fetalDoppler inferior vena cava, transtricuspid, and umbilical artery flow velocitywaveforms between 11 and 16 weeks’ gestation. Am J Obstet Gynecol 1992;166:921–924.

16. Daniel E, Fulton J, Middlesjon M, Martin W, Foulks J. An analysisof the mechanism of barbiturate induces cardio-vascular depression. ArchInt Pharmacodyn 1956;58:457–472.

17. Wojtczak JA. The hemodynamic effects of halothane and isofluranein chick embryo. Anesth Analg 2000;90:1331–1335.

18. Blizard DA, Welty R. Cardiac activity in the mouse: strain differ-ences. J Comp Physiol Psychol 1971;77:337–344.

19. Ji RP, Phoon CK, Aristizabal O, McGrath KE, Palis J, TurnbullDH. Onset of cardiac function during early mouse embryogenesis coincideswith entry of primitive erythroblasts into the embryo proper. Circ Res 2003;92:133–135.

573Evaluation of Cardiovascular Flow Characteristics in 129SVMouse FetusVol. 45, No. 6