Details related to the animals used in this study are listed in Table 1.
Anatomical analyses of the femoral vein
Anatomical analyses of the femoral vein were performed using a 2-year-old male common marmoset (I 7713M) undergoing euthanasia. The femoral veins and arteries are located in the femoral triangle. The femoral triangle is formed at the boundaries between the abdominal wall and thigh muscles (Figure 1B-D). At the base of the thigh, a large vein runs through the center of the inverted triangle, and an artery runs in parallel outside of the vein. In the lower region, the veins and arteries become thinner and overlap, with the arteries positioned on top of the veins (Figure 1D).
For blood sampling, the vein should be targeted because an arterial injury can cause femoral artery hematoma, which may lead to cardiovascular shock when hemorrhage is severe13. Although blood can be drawn from any part of the vein in the triangle and its distal area, venipuncture from the proximal site of the femoral triangle is recommended because of the large size of the vein and its seemingly little overlap with the artery. Additionally, the vein in the proximal site is superficial, allowing for easy location using a needle. Because it pulsates, the artery in the femoral triangle is sometimes identified visually or by palpation, and the vein runs just medial to it. Thus, the observation of artery pulsation helps to predict the location of the vein.
Furthermore, a blue coloration indicative of the vein is usually observed under the skin at the top of the triangle (Figure 1B,C). However, lymph nodes in the triangle are often located close to the vein and show a dark blue color, so their appearance is similar. Fortunately, they can be distinguished by their mobility: the vein is stationary, and lymph nodes are movable. Thus, the pulsation of the artery and the blue color of the vein are two major clues used to locate the vein, although it may be necessary to shave the hair to visualize them.
Determination of the stage in the ovarian cycle
P4 and E2 levels were monitored in six female marmosets (ranging from 1- to 3 years old) to investigate the duration of the follicular and luteal phases. The results showed an average duration of 11.58 days (n = 6 from four marmosets) and 16.8 days (n = 5 from three marmosets) for the follicular and luteal phases, respectively (Table 2). In the following, the P4 and E2 dynamics of one (I6751F, 3-year-old) of the six marmosets is described in detail. Blood sampling and hormone measurements in this animal were performed every few days for 38 days (Table 3).
Luteal phase (days 1-10)
The start day of the measurement was set as day 1. Based on the P4 level (21 ng/mL), the animal was likely in the luteal phase. A high P4 level was observed from days 1 to 10 (≥21 ng/mL), suggesting the luteal phase. On day 12, it dropped sharply to 4 ng/mL. This significant decrease indicated the transition from the luteal to the follicular phase. A decrease in E2 level from day 10 (241 ng/mL) to 12 (189 ng/mL) also supported this transition.
Follicular phase (days 12-22)
After transitioning to the follicular phase, the P4 level remained low (4-6 ng/mL) until day 19 and then slightly increased from day 19 to 24 (day 19, 6 ng/mL; day 22, 8 ng/mL; day 24, 9 ng/mL). In contrast, the E2 level significantly increased from day 19 (94 ng/mL) to 22 (322 ng/mL) and then decreased from day 22 to 24 (158 ng/mL). Based on the increased P4 and decreased E2 levels, ovulation was predicted to have occurred between days 22 and 24, transitioning from the follicular to the luteal phase.
Luteal phase (days 24-36)
After ovulation, the P4 level remained high until day 36 and then dropped to 3 ng/mL on day 38. Thus, it is likely that the transition from the luteal to the follicular phase occurred between days 36-38. Consistent with this transition, the E2 level decreased during this period (day 36, 2517 ng/mL; day 38, 73 ng/mL).
Prediction and determination of the timing of ovulation
To examine the relationship between urine CG level and the ovulation date, seven marmosets (No. 1-7) were prepared. Cloprostenol was injected to reset the ovulatory cycle (day 0). Then, monitoring of blood P4/E2 levels and urine CG levels was performed from day 7. Urine was collected immediately after lighting. Blood sampling was essentially performed in the morning. It was assumed that the ovulation occurs when the E2 level was largely dropped compared with that in the previous day as reported17. On that day, the follicular rupture was indeed observed on the surface of the ovaries, and the presence of zygotes/oocytes in the oviducts (Table 4).
CG was first detected during days 7 to 11 (day 7, N = 1; day 8, N = 2; day 9, N = 1; day 10, N = 2; day 11, N = 1) (Table 4). The large drop in E2 levels (indication of the ovulation) was observed 0-3 days after the first CG signal. The duration from the first CG detection to the E2 decrease seemed to be large when the first CG detection was earlier (Figure 3A). For example, one animal (No. 1) showed a drop in E2 signal 3 days after the first CG detection on day 7 (Table 4). In contrast, the co-occurrence of the first CG detection and the drop of E2 signal was observed in one animal (No. 2) on day 10. Thus, although the first CG detection and ovulation were observed on the same day in one of the five animals, ovulation occurred within a few days after the first detection of CG.
The immunochromatographic test kit for marmoset CG is designed to use urine for testing. Examination of the CG level usually accompanies the determination of P4/E2 levels using blood plasma. It will be helpful if blood plasma, instead of urine, can be used for the CG test. To test this, the blood plasma that was left after the P4/E2 level measurement in the above experiments was examined. Blood plasma and urine were collected from the four animals on either of the days of the above experiments (the days of examination were indicated by double asterisks [**] in Table 4). Using urine, two of the four animals showed positive results (scores 5 and 3), and the other two showed negative results (score 1). Blood plasma showed essentially the same results as urine (Figure 3B). The stronger signal was obtained by using blood plasma. Therefore, when using blood plasma, the judgment should be made before 10 min, which is set for urine.

Figure 1: Blood collection from marmosets. (A) A restrainer is used for blood collection. (B) Marmoset upper thigh. (C) Femoral triangle and blood vessels. The femoral artery in the triangle is often visible and shows pulsation. The blue color of the femoral vein is sometimes visible in the proximal area of the triangle (indicated as the phlebotomy site). Lymph nodes also show a blue color. However, lymph nodes are movable as lymph nodes are attached to the skin. (D) Anatomical view of the thigh of the euthanized animal. The artery, vein, and phlebotomy site are indicated. The same animal is shown in B-D. (E) Blood collection using a restrainer. Holding position of the marmoset's legs and blood being drawn are shown. (F) Urine collection from a marmoset. Please click here to view a larger version of this figure.

Figure 2: Typical patterns of P4, E2, and CG levels during ovarian cycle in marmosets. The time points of FSH, hCG, and PGF2α injections are indicated. Dashed lines show the expected hormonal pattern after PGF2α injection. Abbreviations: P4 = progesterone; E2 = estradiol; CG = chorionic gonadotropin; FSH = follicle-stimulating hormone; hCG = human chorionic gonadotropin; PGF2α = prostaglandin F2α. Please click here to view a larger version of this figure.

Figure 3: Determination of CG level to predict the occurrence of ovulation. (A) The possible relationship between the first day of CG detection (score > 1) and duration until ovulation (E2 drop). (B) Blood plasma can be used for immunochromatographic CG tests. A representative result of each score (top). The score was determined 10 min after the sample loading. Score 2 represents no band within 5 min but the appearance of a band within 10 min. Blood collection was performed soon after the collection of urine. Four marmosets (No. 1, 2, 3, 5 in Table S4) were examined. Samples used here are indicated as double asterisks in Table 4. Blood plasma was diluted to 50% by dilution buffer used for E2 measurement. Please click here to view a larger version of this figure.
Table 1: Animals used in this study. Please click here to download this Table.
Table 2: Duration of follicular and luteal phases. The phase was determined based on the P4 level (follicular phase P4 ≤ 8, luteal phase P4 > 8). When phase change was observed between the measurements, the midpoint between the measuring dates was determined to be the change point. The duration between the two change points was regarded as follicular or luteal phases. To ensure the measurement of the duration of a single phase, the data with a long (≥7 days) interval or twice of ≥6 day intervals in the same phase, including when spanning the two phases, were not used for the analyses. Only when two measurements were conducted in the same phase, the data were employed.Please click here to download this Table.
Table 3: Results of P4 and E2 measurements for 38 days in one marmoset.Please click here to download this Table.
Table 4: Determination and prediction of ovulatory days. The traces of ovulation were considered to be present if bleeding sites or rupture sites were observed in ovaries. Oocyte/zygote indicates the stages of oocyte/zygote obtained from oviducts on the day of E2 drop.Please click here to download this Table.