2.1.

OCT System

A phase-stabilized swept source OCT system (PhS-SSOCT) was used for angiographic imaging. In summary, the system composed of a broadband swept source laser source of central wavelength 1310 nm, scan rage of 150 nm, scan rate of 30 kHz, output power of 39 mW, axial resolution of $11\mu \mathrm{m}$ in air, and lateral resolution of $16\mu \mathrm{m}$. A complete description of the system can be found in our previous publications.^47,51

2.2.

Animal Manipulation and Dosing

Detailed animal manipulation procedures are described in our previous work.⁴⁹ In summary, pregnant CD-1 IGS mice (Crl: CD1{ICR}, Charles River Laboratories, Inc. Wilmington, MA) at gestational day 14.5 were anesthetized and placed on a heated surgical platform to maintain body temperature throughout the imaging procedure. Abdominal fur was removed, and an incision was made on the abdomen exposing the uterine horn for imaging. After stabilization, initial OCT images of the fetal brain were recorded. The mother was then administered the corresponding dose of ethanol via intragastric gavage. Subsequent OCT measurements were taken for a period of 45 min at 5-min intervals. The uterus was hydrated with $1\times$ phosphate-buffered saline 1 min before every measurement.

Our first study evaluating the effects of ethanol on the fetal brain used ethanol at a dose of $3\mathrm{g}/\mathrm{kg}$⁴⁷ to simulate binge drinking in humans. At this dose, severe vasoconstriction was observed in the fetal brain within 45 min of exposure. Hence, in this study, two lower doses of ethanol, 1.5 and $0.75\mathrm{g}/\mathrm{kg}$, were used to evaluate the effects of dose on the murine fetal brain vasculature. Sham (control) experiments were performed utilizing distilled water because the ethanol was diluted in distilled water.

There were a total number of six samples in the sham group, three samples in the $3\text{-}\mathrm{g}/\mathrm{kg}$ ethanol group, four samples in the $1.5\mathrm{g}/\mathrm{kg}$ of ethanol group, and five samples in the $0.75\mathrm{g}/\mathrm{kg}$ of ethanol group. The number of the samples refers to the number of separate fetuses from separate pregnancies. Additionally, three mice were used to study the effect of a dose of $3\mathrm{g}/\mathrm{kg}$ on the maternal peripheral and fetal brain blood vessels simultaneously.

2.3.

Mother and Fetus Comparisons

Since there is a paucity of research on the acute changes in the developing fetal brain vasculature after maternal ethanol consumption, we simultaneously assessed the effects of maternal alcohol consumption on both the mother and the fetus. Ethanol has different effects on blood vessels based on the location of the vessel in the body, so vessels on the skin of the hind limb of the mother were imaged with the mother in a supine position. This location was chosen because the peripheral vessels on the forearm in humans are known to dilate when exposed to alcohol.⁵² Since the forelimb of the mother was relatively difficult to image due to the surgical procedures on the abdominal cavity, the peripheral vessels on the hindlimb were imaged instead.

A dose of $3\mathrm{g}/\mathrm{kg}$ of ethanol was tested on three mice. Instead of the aforementioned procedure where the fetal brain was imaged every 5 min for 45 min after maternal ethanol consumption, imaging was performed after 30 min to ensure proper imaging of the peripheral blood vessels of the mother and fetal brain. Results from this group were compared to the sham group, where the mother was administered lactated Ringer’s instead of ethanol.

2.4.

Imaging, Quantifications, and Statistics

The 3D OCT images consisted of B-scans with 600 A-scans each, and a total of 600 B-scans per volume. Five B-scans were recorded at each spatial position to perform angiographic imaging. A total area of $\sim 6\mathrm{mm}\times 6.2\mathrm{mm}$ of the fetal brain was imaged. The total acquisition time for each dataset was 84 s, including scanning mirror flyback time. A correlation mapping cm-OCA algorithm was used to obtain the vasculature maps from the acquired OCT images.⁵⁰ A discrete Fourier transform-based sub-pixel registration technique was used to correct the axial shift caused due to bulk motion between each pair of 5 B-scans that were recorded at the same spatial position.⁵³ The temporal correlation for each frame with the reference frame was calculated and averaged. The threshold was then determined as the average of the mean correlation values minus the standard deviations of the mean correlations. The SNR-dependent artifacts were corrected using the temporal variance of the background noise as a function of imaging depth.⁵⁰ The 3D vasculature maps were obtained by mapping the temporal correlation coefficients of the entire 3D image. A maximum intensity projection (MIP) was calculated to obtain the en face images of the dorsal surface arterial blood vessels on the fetal brain. A frequency rejection filter⁵⁴ was applied to the 2D MIPs to remove bulk motion artifacts due to maternal respiration and heartbeat. Amira software (EFI Co., Portland, Oregon) was used for denoising and to form final MIPs.

All quantifications were performed on the MIP images, and vessel diameter (VD) was used to quantify the changes in vasculature. Amira was used to perform the quantifications. All VD quantifications were performed on the main branch of the vessel. The results in this study also include images and quantifications from fetuses that were exposed to ethanol at a dose of $3\mathrm{g}/\mathrm{kg}$ from our previously published data.⁴⁷ However, since this study uses a different angiography algorithm, the data were reprocessed using the current cm-OCA algorithm.

For the dose-response study, two nonparametric Kruskal–Wallis ANOVAs were performed to assess the effects of different doses and time on the vasculature. This was followed by two-sided Mann–Whitney $U$ tests that were performed to test for statistically significant changes between every ethanol group and the sham group (three pairs) and between the three ethanol groups (three pairs). Thus, there were a total of six pairwise Mann–Whitney tests that were performed. Bonferroni correction was included in this case for multiple pair-wise tests.

For the mother and fetus comparisons, Mann–Whitney $U$ tests were performed to test for statistically significant changes between the ethanol and sham groups of the mother and between the ethanol and sham groups of the fetuses.

3.1.

Dose-Response Analysis

Vasculature images of one typical example from each of the groups are shown here. Figures 1Fig. 2Fig. 3–4 show MIPs of 3D cm-OCA images before and 45 min after gavage with distilled water (sham), and ethanol at doses of 3, 1.5, and $0.75\mathrm{g}/\mathrm{kg}$, respectively.

Fig. 1

MIPs of cm-OCA images of murine fetal brain vasculature (a) before and (b) 45 min after maternal exposure to distilled water. The dashed rectangle shows the main branch of the vessel on which the quantifications were performed.

Fig. 2

MIPs of cm-OCA images of murine fetal brain vasculature (a) before and (b) 45 min after maternal exposure to ethanol at a dose of $3\mathrm{g}/\mathrm{kg}$. The dashed rectangle shows the main branch of the vessel on which the quantifications were performed.

Fig. 3

MIPs of cm-OCA images of murine fetal brain vasculature (a) before and (b) 45 min after maternal exposure to ethanol at a dose of $1.5\mathrm{g}/\mathrm{kg}$. The dashed rectangle shows the main branch of the vessel on which the quantifications were performed.

Fig. 4

MIPs of cm-OCA images of murine fetal brain vasculature (a) before and (b) 45 min after maternal exposure to ethanol at a dose of $0.75\mathrm{g}/\mathrm{kg}$. The dashed rectangle shows the main branch of the vessel on which the quantifications were performed.

All three doses of ethanol showed dose-dependent vasoconstriction. Figure 5 shows the percentage change in VD for a period of 45 min at 5 min intervals where the pre-gavage measurement was used for comparison for that given dose. All samples from all four groups were used for these calculations. The plot depicts the inter-sample means and standard deviations. While the sham group and ethanol at a dose of $0.75\mathrm{g}/\mathrm{kg}$ showed similar trends, the greater doses of 1.5 and $3\mathrm{g}/\mathrm{kg}$, ethanol exposure resulted in dose-dependent vasoconstriction.

Fig. 5

Percentage change in VD after exposure, every 5 min for 45 min. The error bars represent the standard deviation.

The Kruskal–Wallis ANOVAs showed that the dose and time had statistically significant effects on the VD. The results of the Kruskal–Wallis ANOVAs are given in Table 1.

Table 1

Results of the Kruskal–Wallis ANOVAs. DF represents the degrees of freedom.

Figure 6 shows the dose-dependent changes in the fetal brain VD 45 min after maternal exposure. The inter-sample averages and standard deviations are plotted. The results from the ethanol groups were compared to results from distilled water (sham) and to each other, and the statistically significant results ($P<0.008$ after Bonferonni correction) are shown by the asterisks.

Fig. 6

Percentage change in VD 45 min after maternal exposure to water (sham), and 16.6% ethanol at doses 0.75, 1.5, and $3\mathrm{g}/\mathrm{kg}$. The asterisk indicates $P<0.008$ using a two-sided Mann–Whitney $U$ test.

Six two-sided Mann–Whitney $U$ tests were performed between each of the groups for pair-wise testing. A statistically significant difference was found between the sham group and the $3\mathrm{g}/\mathrm{kg}$ ethanol group, between the sham group and the $1.5\mathrm{g}/\mathrm{kg}$ ethanol group, and between the 0.75 and $3\mathrm{g}/\mathrm{kg}$ ethanol groups. Table 2 shows the results from the six pair-wise Mann–Whitney $U$ tests. $P$ values in bold indicate statistical significance after Bonferroni correction for multiple tests ($P<0.008$).

Table 2

Summary of the Mann–Whitney U tests. P values in bold indicate statistical significance (P<0.008 after Bonferroni correction for multiple tests). n1 and n2 are the number of samples in group 1 and 2 being compared. U is the test statistic, and P<0.008 indicates the criterion for statistical significance.

3.2.

Comparison of Acute Vasculature Changes in the Mother and the Fetus

Figures 7(a) and 7(b) show MIPs of 3D cm-OCA images of the fetal brain before and 30 min after exposure to lactated Ringer’s, respectively. Figures 8(a) and 8(b) show MIPs of 3D cm-OCA images of the mother’s skin on her hindlimb before and 30 min after exposure to lactated Ringer’s, respectively. No significant change in vasculature was seen in the results from the fetus and the mother.

Fig. 7

MIP of cm-OCA images of fetal brain vasculature (a) before and (b) 30 min after maternal exposure to lactated Ringer’s at a dose of $3\mathrm{g}/\mathrm{kg}$.

Fig. 8

MIP of cm-OCA images of the hindlimb of the mother (a) before and (b) 45 min after maternal exposure to lactated Ringer’s at a dose of $3\mathrm{g}/\mathrm{kg}$.

Figures 9(a) and 9(b) show MIPs of 3D cm-OCA images of the fetal brain before and 30 min after exposure to ethanol, respectively. Similar to results shown in Figs. 2 and 3, a drastic vasoconstriction was seen in the fetal brain within 30 min of ethanol exposure.

Fig. 9

MIP of cm-OCA images of fetal brain vasculature (a) before and (b) 30 min after maternal exposure to ethanol.

Figures 10(a) and 10(b) show MIPs of 3D cm-OCA images of the mother’s skin on her hindlimb before and 30 min after ethanol exposure, respectively. Drastic vasodilation was observed, which is in contrast to the vasoconstriction seen in the fetal brain. These results are similar to previous studies in humans where vasodilation was seen in forearm skin, whereas vasoconstriction was noticed in the underlying muscle.⁵²

Fig. 10

MIP of cm-OCA images of the hindlimb of the mother (a) before and (b) 45 min after maternal exposure to ethanol.

Figure 11 plots the comparison of percentage changes in VD at 45 min after exposure. The results from the ethanol groups were compared to results from the sham group from the mother and fetus, respectively. The inter-sample averages and standard deviations of percentage change in VD were plotted.

Fig. 11

Comparison between percentage change in VD 30 min after exposure to ethanol and lactated Ringer’s. The asterisk indicates $P<0.001$ using a two-sided Mann–Whitney $U$ test.

Results from Mann–Whitney $U$ tests performed to compare results from the ethanol and sham samples in the mother and fetus groups, respectively, are shown in Table 3.

Table 3

Summary of the Mann–Whitney U tests. P values in bold indicate statistical significance. n1 and n2 are the number of samples in the ethanol and sham groups. U is the test statistic and P<0.05 indicates the criterion for statistical significance.