Bioluminescent imaging of ABCG2 efflux activity at the blood-placenta barrier

Abstract

Physiologic barriers such as the blood placenta barrier (BPB) and the blood brain barrier protect the underlying parenchyma from pathogens and toxins. ATP-binding cassette (ABC) transporters are transmembrane proteins found at these barriers, and function to efflux xenobiotics and maintain chemical homeostasis. Despite the plethora of ex vivo and in vitro data showing the function and expression of ABC transporters, no imaging modality exists to study ABC transporter activity in vivo at the BPB. In the present study, we show that in vitro models of the placenta possess ABCG2 activity and can specifically transport D-luciferin, the endogenous substrate of firefly luciferase. To test ABCG2 transport activity at the BPB, we devised a breeding strategy to generate a bioluminescent pregnant mouse model to demonstrate transporter function in vivo. We found that coadministering the ABCG2 inhibitors Ko143 and gefitinib with D-luciferin increased bioluminescent signal from fetuses and placentae, whereas the control P-gp inhibitor DCPQ had no effect. We believe that our bioluminescent pregnant mouse model will facilitate greater understanding of the BPB and ABCG2 activity in health and disease.

Figure 1. Choriocarcinoma and HVT cell lines express ABCG2 but not P-gp or MRP1.

(A) Western blot for ABC transporters P-gp, ABCG2, and MRP1 in BeWo, Jeg-3, and HVT cell lysates and positive control HEK cells expressing ABCG2, P-gp, or MRP1. (B) Flow cytometry antibody staining of ABCG2. (C) Flow cytometry antibody staining of Pgp. Data normalized to fluorescence measured in the antibody stained condition for each cell line from three experiments ± SD (*p < 0.05, **p < 0.01 by Student’s t test).

Figure 2. Accumulation of fluorescent substrates of ABCG2, P-gp, and MRP1, measured by flow cytometry.

(A) Choriocarcinoma cells and positive control ABCG2-expressing cell line H460-MX20 were incubated with ABCG2 substrates mitoxantrone (10 μM) or purpurin 18 (15 μM) with and without Ko143 (100 nM, 5 μM). (B) Choriocarcinoma cell lines incubated with P-gp substrate Rhodamine 123 (2 μM) with and without P-gp inhibitor tariquidar (100 nM, 5 μM). (C) Choriocarcinoma cell lines incubated with MRP1 substrate Calcein-AM (1 μM) with and without MRP1 inhibitor MK571 (10 μM, 25 μM). All accumulation values are normalized to accumulation of the maximally inhibited condition. Data represent means ± SD of three experiments (**p < 0.01, ***p < 0.001, ****p < 0.0001 by Student’s t test).

Figure 3. ABCG2 inhibition increases D-luciferin accumulation and bioluminescent signal in choriocarcinoma cells.

(A) D-luciferin (2 mM) accumulation in BeWo cell line, with and without ABCG2 inhibitors Ko143 (100 nM, 5 μM) and fumitrogen C (FTC; 100 nM, 5 μM) measured by flow cytometry. (B) D-luciferin (2 mM) accumulation in Jeg-3. (C) D-luciferin (2 mM) accumulation in H460-MX20. (D) To measure the effect of ABCG2 on bioluminescence, all three cell lines were transiently transduced with a baculovirus (Bacmam) containing firefly luciferase. Bioluminescence with and without 5 μM Ko143 is reported. All values are normalized to the 5 μM Ko143 condition. Data represent means ± SD of three experiments (*p < 0.05 **p < 0.01, ***p < 0.001, ****p < 0.0001 by Student’s t test).

Figure 4. In vivo bioluminescent pregnant mouse model.

(A) Illustration of the breeding strategy to generate pregnant mice carrying placentae and fetuses expressing fLuc. (B,C) Diagrammatic representation of the BPB barrier with the hypothesized distribution of d-luciferin (B) before and (C) after ABCG2 inhibition. (D) Bioluminescence images (left to right) of a male CAG-fLuc mouse, non-pregnant female FVB mouse, pregnant FVB mouse, and pregnant FVB mouse imaged immediately post euthanasia with peritoneum opened to visualize uterus, all after being administered 5 mg/kg D-luciferin i.p. for 20 min. (E) Bioluminescence images from (left to right) of intact uterus demonstrating BLI signal from fetuses and placentae, and female mouse after uterus was removed with no signal seen, after 5 mg/kg IP d-luciferin for 20 minutes.

Figure 5. Validation of in vivo mouse model.

(A) Dose escalation study of pregnant mice, carrying fLuc-expressing fetuses and placentae, receiving increasing dosages of D-luciferin (arrow) over 80 minutes. Data normalized to the maximum signal of each mouse and represent mean ± SD of at least three different animals. (B) Maximum bioluminescence signals measured every two days between embryonic day 5.5 and 16.5. Data are normalized to the maximum BLI signal measured at the last time point, and are represented as mean ± SD of at least three different animals. (C) Bioluminescence signal intensities represented in photons per sec. (D) Data presented in (C) normalized to the 20-minute time point. Data represent mean ± SEM of five different animals. Number of mice, (A) n = 3, (B–D) n = 4, except for E5.5–6.5 (n = 2).

Figure 6. ABCG2 inhibition increases D-luciferin-mediated bioluminescence signal in vivo.

(A) Injection strategy utilized to study the effect of inhibitors. A baseline signal is gathered for the first 20 minutes, when the signal plateaus. At this point mice are removed from the imager and injected with vehicle, or inhibitors, after which they are returned for an additional 25 minutes of imaging. (B) Time activity curves of total flux (photons per sec) of pregnant mice, carrying fetuses and placentae expressing fLuc, injected with 5 mg/kg D-luciferin i.p., plus vehicle (negative control), Ko143, Gefitinib, or DCPQ (negative control). All inhibitors were injected at a dosage of 16 mg/kg. Data normalized to the 20-minute time point of each animal (baseline signal). (C) AUC acquired from the time activity curves for vehicle control and all three inhibitors. (D) Maximum BLI signals acquired from the post-injection phase from all four treatment injection conditions. Data represent mean ± SEM of at least four different animals (*p < 0.05, **p < 0.001 by Student’s t test). Number of mice, (B–D) Vehicle n = 5, Ko143 n = 5, DCPQ n = 6, gefitinib n = 4.