INTRODUCTION
Congenital cytomegalovirus (HCMV) infection is the most common nongenetic cause of fetal malformation in developed countries, with a mean global incidence of 0.64% (
1). Congenital HCMV (cCMV) may case neonatal death, prematurity, and chronic illness from lung, liver, and neurological disease. Approximately 10% of HCMV-infected babies will be symptomatic at birth, presenting with chronic conditions such as sensorineural hearing loss, vision loss, prematurity, intrauterine growth restriction, microcephaly, and motor defects, with ∼10% among these dying of multiorgan dysfunction. A significant proportion (∼15%) of initially asymptomatic HCMV-infected babies additionally develop disease between birth and 5 years of age.
Maternal primary infection, reinfection, or reactivation of latent virus during pregnancy can lead to HCMV infecting and replicating within the placenta, crossing the materno-fetal interface, and subsequently infecting the fetus—known as mother-to-child transmission (MTCT). Fetal injury results from direct viral cytopathic damage, although placental infection may also cause fetal injury via HCMV-induced immunomodulation, dysregulation of placental development, and placental dysfunction, as we and others have shown (
2–7). The placenta is the life support system for the developing fetus, providing for exchange of oxygen, nutrients, antibodies, hormonal compounds, and excretory products between mother and fetus. An underdeveloped or abnormal placenta can lead to placental insufficiency and subsequent fetal injury. The placenta therefore represents the main organ of cCMV pathogenesis and an important target for therapy to prevent fetal damage.
Despite the clinical and social importance of cCMV, there are no licensed therapeutics available for use during pregnancy to treat HCMV placental infection and prevent subsequent transplacental transmission and congenital disease (
8,
9). This is partly due to systemic drug toxicity, potential teratogenicity, or limited evidence for efficacy of the established and licensed HCMV antiviral drugs. The licensed HCMV-specific direct-acting antivirals (DAAs) of the type nucleoside/nucleotide/pyrophosphate analogues, such as ganciclovir (GCV), cidofovir (CDV), and foscarnet (FOS), have significant toxicity issues, which make them unlikely to be used during pregnancy.
There are four antivirals of potential interest in the prevention and treatment of cCMV during pregnancy: valaciclovir (VACV), letermovir (LMV), maribavir (MBV), and brincidofovir (BCV). VACV, the prodrug of the nucleoside analogue aciclovir (ACV), is the only HCMV antiviral to be investigated in pregnant women to date (
10–12), due to its established high safety profile. ACV is first converted by viral thymidine kinases to ACV monophosphate; however, HCMV lacks this kinase and so the exact mechanism of action against HCMV has not been deciphered in detail, but appears to be dependent on the nucleoside-phosphorylating activity of the HCMV protein kinase pUL97 (
13,
14). LMV is a 3,4-dihydro-quinazoline and targets the HCMV terminase complex, interfering with HCMV DNA concatemer cleavage and packaging of HCMV DNA into capsids. LMV was recently approved by the FDA for prophylaxis to prevent HCMV infection and disease in hematopoietic stem cell transplant (HSCT) recipients. MBV is a riboside benzimidazole that binds to the HCMV-encoded kinase pUL97, inhibiting viral nuclear egress and virus production efficiency. MBV has recently been granted Orphan Drug Designation by the European Commission and Breakthrough Therapy Designation by the FDA as a treatment for HCMV in transplant patients. BCV is an alkoxyalkyl ester analogue of the nucleoside analogue CDV, designed to release CDV intracellularly, allowing for higher intracellular concentrations and lower toxicities. A recent phase III trial failed to meet the primary endpoint of reduction in clinically significant HCMV infection in HCT recipients (
15).
The efficacy and toxicity of these HCMV antivirals has yet to be investigated in detail and in parallel in model systems of pregnancy. Given some very promising findings in other settings of antivirals such as LMV (
16), MBV (
17), CDV/BCV (
18), and VACV (
11), we investigated the efficacy and toxicity of these antivirals in first-trimester extravillous trophoblast cells and third-trimester
ex vivo placental explant histocultures.
DISCUSSION
The antiviral efficacy and toxicity of established and more recent antivirals was studied in human placental cell models and ex vivo placental tissue relevant to assessing their future use during pregnancies with congenital HCMV infection. Recently developed HCMV antivirals (MBV and LMV), compared with the reference antiviral drug GCV, displayed high efficacy and low placental cell toxicity profiles in pregnancy model systems.
The antiviral EC
50 values obtained from placental trophoblasts infected with the genetically intact Merlin strain of HCMV showed LMV displayed the most potent antiviral efficacy, followed by MBV, CDV, GCV, and lastly ACV. The relative EC
50 values obtained are consistent with data from human fibroblast cell cultures (
19,
21–27). The higher EC
50 and EC
90 values observed for ACV treatment is unsurprising given that HCMV lacks the thymidine kinase enzyme to convert ACV to ACV monophosphate and as such is not a specific antiviral for HCMV. Paradoxically, a nonrandomized, single group assignment phase IV clinical trial (
12) and a recent prospective, randomized, double-blind, placebo-controlled phase II/III trial (
11) showed that VACV treatment resulted in a reduction in the number of symptomatic children at birth as well as the number of terminations of pregnancy for fetal anomalies, and a reduction in the rate of fetal infection after maternal primary infection acquired early in pregnancy, respectively. There is some evidence that the HCMV pUL97 kinase phosphorylates ACV in the absence of the thymidine kinase, albeit with lesser efficiency (
13,
14). This may be a possible explanation for the efficacy observed in our study at high antiviral concentrations and the efficacy observed
in vivo in clinical trials.
The inability of LMV to inhibit viral replication by more than 80% relative to untreated cells at the high concentration of 10 μM (77.3% reduction) and the results obtained being similar to treatment at 0.0015 μM (76.3% reduction) was surprising given the known high potency of LMV in fibroblast cell culture models using plaque reduction and fluorescence reduction assays (
19,
23,
24). As these assays measure encapsulated, enveloped, replication-competent viral particles, whereas RT-qPCR measures viral DNA, the accumulation of replication-incompetent HCMV concatemers from LMV treatment could be a possible explanation for this discrepancy. As DNase digests any unencapsulated, unenveloped DNA, but is unable to penetrate the viral capsid, we treated supernatants with DNase prior to nucleic acid extraction to eliminate any unencapsulated, unenveloped HCMV concatemers from the RT-qPCR measurement of viral load. Consistent with our hypothesis, DNase treatment of LMV-treated cell culture supernatants resulted in significantly lower estimates of viral loads than observed in DNase treatment of GCV-treated cell culture supernatants. This novel method can be of use clinically, where viral loads may be artificially inflated by RT-qPCR detection of free floating concatemers in patients undergoing LMV treatment. Using this method, the viral loads will more accurately reflect both the true result and clinical evaluation of LMV efficacy.
Antiviral treatment at 10 μM, which is much higher than the EC
50 values (all <0.42 μM) for all the HCMV-specific antivirals, did not show any cytotoxic effects. This is consistent with limited animal model studies that showed both LMV and MBV have low toxicity and no teratogenic effect or maternal toxicity (
28–30). Several studies in animal guinea pig models also showed CDV and BCV have potential benefits in preventing HCMV transmission during pregnancy without toxicity (
31–33). Clearly, some of these antivirals can have significant toxicity issues
in vivo (e.g., GCV and CDV) which do not manifest in our cell proliferation or LDH assays. While these assays do not fully demonstrate potential safety of antiviral treatment during pregnancy, they do suggest these antivirals are nontoxic to the placental organ. Further animal studies and phase II safety studies on antiviral toxicity and potential teratogenicity would provide additional safety data for the potential use of these therapeutics during pregnancy.
The viral kinetics in HCMV-infected placental explants in response to treatment with LMV, MBV, and CDV showed viral replication was inhibited over the 14-day time course, whereas there was a consistent increase in viral DNA in untreated placental explants. A recent study investigating the placental transfer of LMV and MBV using a placental perfusion model showed both antivirals crossed the placenta at a low to moderate rate, with the mean concentration in the fetal compartment being superior to the EC
50 for both molecules (
30). Furthermore, they showed some accumulation of the antivirals in the placental tissue where HCMV replicates, causes placental damage, and where MTCT occurs. These treatments could therefore limit HCMV replication within the placenta, which can be an indirect mechanism of HCMV causing adverse pregnancy outcomes through placental dysfunction, and potentially prevent HCMV from entering the fetal circulation and causing direct cytopathic damage to the developing fetal organs.
These data from ex vivo experiments show recently licensed and studied antivirals, particularly LMV and MBV, display high efficacy and low toxicity profiles in first-trimester placental trophoblast cell cultures and third-trimester placental explant histocultures. Further investigations are warranted to characterize these antivirals more profoundly in terms of applicability for use in extended regimens of antiviral treatment. It will be particularly exciting to learn more about their potential for use specifically during pregnancy, or in other applications, to support the control of congenital HCMV disease.
MATERIALS AND METHODS
Antiviral compounds.
Analytical grade aciclovir was obtained from Sigma-Aldrich. Letermovir and maribavir were obtained from MedChemExpress. Analytical grade cidofovir and ganciclovir was obtained from Sigma-Aldrich. Stock aliquots were prepared in dimethyl sulfoxide (DMSO) and stored at –80°C.
Cultured cells and viruses.
Human first-trimester TEV-1 extravillous trophoblast cells (
34,
35) were maintained in Ham’s F10 nutrient mix (Life Technologies) supplemented with 10% fetal bovine serum (FBS) and 100 U/ml penicillin G, 100 μg/ml streptomycin, and 29.2 μg/ml
l-glutamine (1XPSG), (Life Technologies) at 37°C with 5% CO
2. Genetically intact HCMV strain Merlin (UL128+, RL132−) was derived from a Merlin-BAC recombinant, pAL1120, kindly provided by Richard Stanton (University of Cardiff, United Kingdom) (
36) and propagated in RPE-1 cells kindly provided by Barry Slobedman. Titer of virus stocks was determined using standard plaque assays.
TEV-1 culture assays.
TEV-1 cells were seeded in 24-well plates and inoculated with virus in triplicate at a multiplicity of infection (MOI) of 2 PFU/cell. Mock-infected cultures were established concurrently. Plates were centrifuged at 770 × g for 30 min followed by 2 h of incubation at 37°C with 5% CO2. Supernatant was removed and replaced with fresh medium with or without antivirals and incubated at 37°C with 5% CO2.
Clinical placentae and placental villous explant histocultures.
Term placentae were collected with consent from women undergoing elective Caesarean section delivery who had had a healthy pregnancy and were not in labor, under ethics approval SESIAHS HREC 09/126. Placental villous explant histocultures were established as previously described (
37). Briefly, placental explants were inoculated with 1 × 10
7 PFU of HCMV strain Merlin and incubated for 5 days at 37°C supplemented with 5% CO
2. At day five postinfection, explants were washed in 1× PBS and transferred to fresh plates with fresh medium. The explants were then treated with 25 μM antivirals and incubated for a further 7 days. At day 12 postinfection (7 days postinhibitor treatment), medium was again replaced with fresh medium and fresh compounds and histocultures were incubated until explant harvest at 19 days postinfection (14 days postinhibitor treatment).
Nucleic acid extraction and quantitative real-time PCR.
Total nucleic acid from the TEV-1 trophoblast cell culture supernatants and the placental explant tissues were extracted using MagnaPure LC Total Nucleic Acid kit according to the manufacturer’s protocol (Roche) as previously described (
37). Quantitative real-time PCR was performed using a Roche 480 LightCycler with Kapa Sybr Fast qPCR master mix (Merck). The number of cell-associated HCMV major immediate early (MIE) DNA copies was normalized against cellular albumin copies as previously described (
3,
21), using previously published oligonucleotide primers (
3). Reactions were carried out under the following conditions: denaturation at 95°C for 5 min, followed by 45 cycles of denaturation at 94°C for 15 s, annealing at 60°C for 20 s, and elongation at 72°C for 15 s. Product specificity was determined using melt peak analysis.
HCMV genome detection.
The TEV-1 trophoblasts were infected with HCMV Merlin strain (2 PFU/cell) or mock infected and either treated with 10 μM letermovir or left untreated. Total nucleic acid from TEV-1 trophoblast cell lysates was extracted using DNeasy blood and tissue kit according to the manufacturer’s protocol (Qiagen). Concatemer HCMV DNA was amplified between nucleotide 232,815 (forward primer: 5′-GCACGTCCCAAACTGGCTTGAGGAG-3′) within the TRS region of one genome and nucleotide 2,136 (reverse primer: 5′-GGAAAGAGCGTGTGTGATCTGGCCGAG-3′) within the RL1 region of the next genome, yielding a 4,967-bp product spanning the HCMV genomes (
Fig. 7). PCR amplification of the 4,967-bp region would not be able to occur on cleaved genomes, as cleavage occurs in the middle of the amplified region between the “a” sequences of two concatemer genomes, as shown in
Fig. 7. PCR was performed using the Expand High Fidelity PCR system as previously described (
38). PCR products were detected by electrophoresis on 1% agarose gel in 0.5× Tris-borate-EDTA (TBE) and visualized with Sybr Safe. Specificity of product was determined by Sanger sequencing.
DNase treatment.
Merlin-infected TEV-1 trophoblast cell culture supernatants treated with letermovir, ganciclovir, or left untreated were incubated with amplification-grade DNase I (Merck) or H2O at room temperature for 15 min. Stop solution was added and samples heated at 70°C for 10 min to denature the DNase enzyme, and then placed on ice for qRT-PCR analysis.
Western blot analysis.
Western blot analysis was performed by standard procedures as described previously (
39). Immunostaining was performed with the antibodies mouse MAb-β-actin (Ac-15, Sigma), mouse MAb-IE1p72/pUL44 (IE/E; clones DDG9 and CCH2, Dako), mouse MAb-pp65 (Abcam), mouse MAb-pp28 (Abcam), and horseradish peroxidase (HRP)-conjugated anti-mouse secondary antibody (Pierce). Densitometry of immunostaining was performed using ImageJ software. The mean densitometry values for control-infected cells normalized against β-actin were assumed to be 100% and this value was used to calculate relative protein expression.
MTT assay.
TEV-1 cells were treated with 10 μM antivirals at 40 to 50% confluence and incubated for 24 h at 37°C supplemented with 5% CO2. Medium was removed and replaced with medium containing 1 mg/ml MTT (3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyl-2H-tetrazolium bromide) and cells were incubated for 2 h. Cells were washed with 100% isopropanol for 10 min to dissolve the formazan and plates were read at an optical density of 570 nm (OD570).
Cytotoxicity assay.
Cell damage was measured using LDH release assays performed using the CytoTox 96 nonradioactive cytotoxicity assay (Promega), with the supernatant of placental cells cultured for 1, 4, and 7 days in the presence of inhibitors at 10 μM according to the manufacturer’s protocols.
Statistical analysis.
The EC50 and EC90 values for antivirals were determined using nonlinear regression and dose-response inhibition. For comparison between two groups, a Student’s t test was performed. For comparisons between three groups or more, the nonparametric Kruskal-Wallis test was initially performed to identify the presence of a difference between treated and nontreated groups. Where a significant difference was detected (P < 0.05), a one-way ANOVA was performed with post hoc Bonferroni correction applied for multiple comparisons. Statistical analysis was performed using GraphPad Prism v7.0.
ACKNOWLEDGMENTS
The study was funded by an Early Career Award from the Thrasher Research Fund (grant RG181876-Hamilton) and was supported by grants from the Australian National Health and Medical Research Council (grant APP1127717-Hamilton), the Australia-Germany Joint Research Cooperation Scheme (grants 2017-18/RG162050, 2020-21/RG192195), the DAAD-Go8 (grants 2017-18/MM-WDR-STH-JM, 2020-21/MM-WDR), the Wilhelm Sander-Stiftung (grant 2018.121.1/MM-SBT), and the Interdisciplinary Center for Clinical Research (IZKF) of the Faculty of Medicine of FAU (grant A88/MM-HS). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
The authors are grateful to Richard Stanton (University of Cardiff, Wales, UK) and Barry Slobedman (University of Sydney, Australia) for the supply of valuable materials. We thank Joanna Youngson, Maria Jimenez, and Rebecca Anderson (Sydney Cord Blood Bank, Sydney Children’s Hospital, Sydney, Australia) for arranging consent for woman to donate placental tissue. We thank Megan Lenardon and Lynn Tran for designing HCMV concatemer primers.
W.D.R. and M.M supervised the project; S.T.H., M.M., and W.D.R. designed the research; S.T.H. performed experiments; S.T.H. collected and analyzed data; S.T.H., M.M., and W.D.R. wrote the manuscript.