Evaluation of Rift Valley fever vaccine candidates in pregnant rodent models

Abstract

Rift Valley fever virus (RVFV) causes significant disease in humans and livestock. Immunogenicity of candidate vaccines rMP-12 and RVax-1 show promise, but their placental tropism and potential effects on fetal outcomes remain incompletely understood, particularly across different animal models. Understanding species-specific placental replication is essential to optimize vaccine safety in pregnant populations. To evaluate the placental tropism and fetal outcomes of RVFV candidate vaccines rMP-12 and RVax-1 in pregnant Sprague–Dawley (SD) rats and C57BL/6 mice. Pregnant SD rats and C57BL/6 mice were vaccinated intramuscularly at embryonic day 14 (E14) with 1×10⁵ or 1×10⁶ PFU of rMP-12 or RVax-1. Viral replication in maternal, placental, and fetal tissues was assessed at E18 in rats and at E17 or E19–21 in mice using viral RNA quantification and antigen detection. Fetal outcomes, including litter size, placental histopathology, and fetal demise, were recorded. In rats, both vaccines showed minimal replication in placental and fetal tissues, indicating limited vertical transmission. In contrast, mice were more susceptible: viral RNA and antigens were detected in maternal livers, placentas, and fetal compartments. rMP-12–vaccinated mice showed reduced litter sizes and autolyzed placental tissues, whereas RVax-1–vaccinated mice exhibited fetal demise, with viral antigens detected in spongiotrophoblasts, syncytiotrophoblasts, and trophoblast giant cells of the junctional and labyrinth zones. RVFV vaccines rMP-12 and RVax-1 exhibit residual placental tropism in mice but minimal replication in rats, highlighting species-specific differences. Mouse models may be useful for studying placental tropism, and these findings inform future optimization of vaccine safety during pregnancy.

Rift Valley fever (RVF) is a mosquito-borne zoonotic viral disease endemic to sub-Saharan Africa, Egypt, Madagascar, the Comoros, Saudi Arabia, and Yemen, affecting both humans and livestock, including sheep, cattle, goats, and camels [1]. During RVF outbreaks, infected pregnant sheep, cattle, and goats exhibit high rates of abortion, fetal demise, and congenital malformations, which often precede the onset of human cases. Additionally, newborn animals show markedly elevated mortality rates. Human RVF cases most commonly present as a self-limiting febrile illness; however, a subset of patients develops severe manifestations, including viral retinitis, encephalitis, or hemorrhagic fever [2]. Rift Valley fever virus (RVFV) belongs to the genus Phlebovirus within the family Phenuiviridae and is a negative-strand RNA virus with a tripartite genome composed of Large (L), Medium (M), and Small (S) segments [3]. The L segment encodes the viral RNA-dependent RNA polymerase, the M segment encodes a glycoprotein precursor co-translationally cleaved into Gn and Gc, along with the 78-kDa protein and nonstructural M (NSm) protein, and the S segment encodes the nucleoprotein (N) and nonstructural S (NSs) protein. RVFV is designated as a Category A Priority Pathogen by the U.S. National Institute of Allergy and Infectious Diseases (NIAID) and an overlapped select agent by the U.S. Department of Health and Human Services (HHS) and the U.S. Department of Agriculture (USDA) [4, 5]. In addition, RVFV is designated as a prototype pathogen by the World Health Organization (WHO) R&D Blueprint, serving as a model for epidemic preparedness, and is a notifiable disease to the World Organisation for Animal Health (OIE) [6, 7].

Veterinary vaccines for RVF have been available in several endemic countries, including the live-attenuated Smithburn and Clone 13 vaccines. However, RVF outbreaks are often unpredictable, and routine vaccination of livestock has not been widely implemented due to vaccination costs and the relatively infrequent occurrence of outbreaks compared to livestock turnover. Consequently, local RVF outbreaks in livestock can rapidly expand across large geographic areas. Thus, implementing routine vaccination of susceptible livestock in endemic regions, preferably with vaccines that confer long-lasting immunity, is critical for minimizing the impact of future RVF outbreaks. Although human cases typically begin as sporadic infections, the number of patients can increase substantially during an RVF epidemic, potentially resulting in a severe public health impact. Although no RVF vaccines are currently licensed for human use, several candidates, including RVFV-4s (four-segmented recombinant RVFV), DDVax (recombinant ZH501 strain encoding double deletions of NSs and NSm genes), and ChAdOx1-GnGc (a modified replication-deficient chimpanzee adenovirus vector encoding RVFV GnGc), are in development and have advanced to Phase 1 or Phase 2 clinical trials [8, 9]. In the United States, a formalin-inactivated RVF vaccine (TSI-GSD-200) has been evaluated in Phase 1/2 clinical trials [10–12]. However, the vaccine has limited immunogenicity, requiring multiple primary doses and periodic boosters to achieve and maintain protective immunity. In addition, its production requires high-containment manufacturing facilities, which underscores the need for investment in alternative vaccine platforms. The live-attenuated mutagenized passage (MP)-12 vaccine has been extensively characterized in both animals and humans since the 1980s, demonstrating a strong safety profile in healthy non-pregnant subjects and robust immunogenicity, with a single dose eliciting long-term protective immunity [13, 14]. The MP-12 vaccine was conditionally licensed for veterinary use in 2013 [15], and the vaccine master seed is maintained for emergency use [16]. The MP-12 strain was generated through serial passages of the pathogenic Zagazig Hospital (ZH)548 strain in human diploid fetal lung fibroblast (MRC-5) cells in the presence of 5-fluorouracil and encodes 23 mutations distributed across the genome: 4 in the S segment, 9 in the M segment, and 10 in the L segment. The attenuation phenotype of MP-12 is primarily attributed to two amino acid substitutions in the M segment (Gn-Y259H and Gc-R1182G) and two in the L segment (L-V172A and L-M1244I) [17].

Although MP-12 exhibits reduced replication compared to the pathogenic ZH501 strain in sheep placental explants, including the chorionic membrane, allantoic membrane, and villi [18], a pregnant ewe vaccinated with MP-12 at 30–50 days of gestation had an autolyzed fetus at 18–20 weeks of gestation, with the fetal brain testing weakly positive for RVFV RNA [19]. Similarly, 17 pregnant ewes in early gestation vaccinated with the arMP-12ΔNSm21/384 vaccine, which encodes deletions of the 78kD and NSm genes, gave birth to three malformed newborn lambs [20]. Since neither MP-12 nor arMP-12ΔNSm21/384 caused detectable adverse effects in pregnant ewes vaccinated during late gestation, the observed adverse effect of MP-12 may be limited to vaccination during early gestation. Similarly, ewes vaccinated with the Clone 13 vaccine exhibited fetal malformations and stillbirths, and viral RNA was detected in the newborn lambs [21].

RVax-1 was designed as a next-generation live-attenuated MP-12–based vaccine generated using reverse genetics. This vaccine incorporates deletions of the 78kD and NSm genes, along with 566 silent mutations in the MP-12 background [22]. These modifications were intended to restrict viral dissemination in mosquitoes and to incorporate a genetic signature into the vaccine strain, enabling monitoring of potential spillover events during field use. RVax-1 demonstrated protective efficacy in mice comparable to that of rMP-12, while its dissemination in mosquitoes was minimal [22]. However, it is important to characterize the attenuation phenotype of RVax-1 in pregnant animals relative to the parental rMP-12 strain.

Studies in pregnant ewes provide the most relevant information of vaccine attenuation but are highly time and cost intensive. Given these constraints, it is advantageous to develop a simpler and more rapid experimental approach for screening the attenuation level of the RVax-1 vaccine, ideally in comparison with previous data from MP-12. Previous work demonstrated that pregnant Sprague Dawley (SD) rats at embryonic day (E) 14 were highly susceptible to the pathogenic ZH501 strain, resulting in more than 50% maternal mortality, stillbirths, developmental abnormalities in pups, and widespread viral replication in maternal and fetal tissues [23]. SD rats vaccinated at E14 with the live-attenuated DDVax vaccine showed no clinical signs and delivered healthy litters [24], indicating that rodent models such as SD rats can be used to evaluate vaccine attenuation during pregnancy. Meanwhile, although the sheep model is considered the gold standard for RVF vaccine safety evaluation, studies using pregnant rodent models may provide complementary information to support safety assessments.

In this study, we aimed to assess the placental tropism of RVax-1 and the parental rMP-12 strain in SD rats and C57BL/6 mice. Evaluation of viral replication and tissue distribution in placental and fetal compartments is critical for vaccine safety, as vertical transmission or placental infection could pose significant risks during pregnancy. Given their relatively high susceptibility to RVFV, we hypothesized that pregnant C57BL/6 mice may serve as a sensitive model to reveal residual virulence or tissue tropism of vaccine strains that could remain undetected in less permissive animal models.

Materials and Methods

Media, cells, and viruses

MRC-5 cells (human diploid fetal lung fibroblast cells, ATCC CCL-171) and Vero cells (African green monkey kidney epithelial cells from Chlorocebus sp., ATCC CCL-81) were cultured at 37 °C with 5% CO₂ in Dulbecco’s modified minimum essential medium containing 10% fetal bovine serum (FBS), penicillin (100 U/ml), and streptomycin (100 μg/ml). The recombinant MP-12 strain (rMP-12) and RVax-1 strain were rescued from Vero cells via reverse genetics, as described previously [22, 25]. Both viruses underwent two passages in MRC-5 cells before being used for vaccination. The vaccine inoculum was prepared by diluting with phosphate-buffered saline (PBS) and subsequently back-titrated using a plaque assay on Vero cells.

Plaque assay for virus titration

The titration of virus stocks was performed using a plaque assay. Briefly, Vero cells in 6-well plates were infected with 400 μl of 10-fold serial dilutions of the virus and incubated at 37°C for 1 hour. After removal of the inoculum, cells were overlaid with 2 ml of a mixture containing 0.3% tragacanth gum, 1× minimum essential medium, 5% FBS, 5% triphosphate broth, streptomycin, and penicillin, and incubated for an additional 96 hours. Cells were then fixed with 25% formalin and 5% ethanol containing crystal violet at room temperature for 25 minutes, followed by washing with water. Infectious virus titers (plaque-forming units, PFU) were calculated using the formula: plaque count × dilution factor × 1000 ÷ 400 μl.

Animals

Timed-pregnant Sprague Dawley (SD) rats (n = 9, Table 1) and C57BL/6NCrl mice (n = 14, Table 2 and Table 3) were obtained from Charles River Laboratories International, Inc. (Wilmington, MA, USA). C57BL/6NCrl mice were used in this study to maintain consistency with previous evaluations of vaccine immunogenicity and efficacy [22]. The age of the timed-pregnant animals was not provided by the vendor at the time of purchase. This study was designed to assess the occurrence of any overt or abnormal findings; any notable phenotypes were evaluated on an individual basis to inform future, quantitatively powered studies. Because not all purchased animals were pregnant, group sizes were adjusted to prioritize allocation to the RVax-1 groups.

Table 1.

Vaccination and Sampling of Pregnant Sprague–Dawley Rats

Animal ID	Vaccine	PFU dose (Back-titrated)	Status of fetus/pup	Collection of tissues for RNA
#1	rMP-12	1×105 (7.8×104)	10 live fetuses	Dam (Liv, Sp. Br, Pla) 10 fetuses (Liv, Br)
#2	rMP-12	1×105 (7.8×105)	9 live fetuses 1 autolyzed	Dam (Liv, Sp. Br, Pla) 9 live fetuses (Liv, Br)
#3	rMP-12	1×106 (9.3×105)	14 live fetuses 1 autolyzed	Dam (Liv, Sp. Br, Pla) 14 live fetuses (Liv, Br)
#4	rMP-12	1×106 (9.3×105)	11 live fetuses	Dam (Liv, Sp. Br, Pla) 11 fetuses (Liv, Br)
#5	RVax-1	1×105 (7.8×104)	10 live fetuses	Dam (Liv, Sp. Br, Pla) 10 fetuses (Liv, Br)
#6	RVax-1	1×105 (7.8×104)	5 live fetuses	Dam (Liv, Sp. Br, Pla) 5 fetuses (Liv, Br)
#7	RVax-1	1×106 (1.0×106)	14 live fetuses 1 autolyzed	Dam (Liv, Sp. Br, Pla) 13 live fetuses (Liv, Br)
#8	RVax-1	1×106 (1.0×106)	16 live fetuses 1 autolyzed	Dam (Liv, Sp. Br, Pla) 15 live fetuses (Liv, Br)
#9	PBS	n/a	11 live pups	Dam (Liv, Sp. Br) 3 pups (Liv, Br)

Liv, liver; Sp, spleen; Br, brain; Pla, placenta; n/a, not applicable.

Table 2.

Vaccination and Sampling of Pregnant C57BL/6 Mice

Animal ID	Vaccine	PFU dose (Back-titrated)	Pup delivery	Collection of tissues for RNA
#1	rMP-12	1×105 (1.0×105)	3 live pups	Dam (Liv, Sp. Br, Ute) All 3 pups (Liv, Br)
#2	rMP-12	1×105 (1.0×105)	2 live pups (cannibalized)	Dam (Liv, Sp. Br, Ute) 1 pup carcass (Br)
#3	rMP-12	1×105 (7.8×105)	6 live pups 1 stillborn pup	Dam (Liv, Sp. Br, Ute) All 7 pups (Liv, Br)
#4	RVax-1	1×105 (8.8×104)	7 live pups	Dam (Liv, Sp. Br, Ute) 7 pups (Liv, Br)
#5	RVax-1	1×105 (7.8×104)	8 live pups	Dam (Liv, Sp. Br, Ute) 8 pups (Liv, Br)
#6	RVax-1	1×105 (8.8×104)	7 stillborn pups	Dam (Liv, Sp. Br, Ute) 7 pups (Liv, Br)
#7	RVax-1	1×105 (8.8×104)	9 stillborn pups	Dam (Liv, Sp. Br, Ute) 3 pups (Liv, Br)
#8	PBS	n/a	9 live pups	Dam (Liv, Sp. Br, Ute) 3 pups (Liv, Br)
#9	PBS	n/a	7 live pups 1 stillborn pup	Not sampled

Liv, liver; Sp, spleen; Br, brain; Ute, uterus; Pla, placenta; n/a, not applicable.

Table 3.

PCR-Positive C57BL/6 Mice upon Delivery

Animal ID	Vaccine	Organ	Viral RNA (copies/mg tissue)
#1 (dam)	rMP-12	Spleen	1167.6
		Uterus	1665.8
#2 (dam)	rMP-12	Liver	29.0
		Spleen	180.8
		Uterus	507.4
#3 (dam)	rMP-12	Spleen	3508.7
		Uterus	4673.4
#3 (pup #2)		Liver	457.8
#4 (dam)	RVax-1	Spleen	134.1
		Uterus	814.3
#5 (dam)	RVax-1	Uterus	595.5
#5 (pup #1)		Liver	141.3
#6 (dam)	RVax-1	Liver	37000.0
		Spleen	10309.2
		Uterus	1793.4
		Brain	299.3
#7 (dam)	RVax-1	Spleen	35.7
		Uterus	994.2

All other tissues were below the limit of detection in qPCR.

Immunization and Sampling Procedures in Pregnant Rats and Mice

Timed-pregnant Sprague Dawley rats were vaccinated at E14 via the intramuscular (i.m.) route (total volume up to 100 μl; 50 μl per injection site) with PBS (n = 1), rMP-12 at 1 × 10⁵ PFU (n = 2) or 1 × 10⁶ PFU (n = 2), and RVax-1 at 1 × 10⁵ PFU (n = 2) or 1 × 10⁶ PFU (n = 2). Health status and body weight were monitored in the Animal Biosafety Level 2 (ABSL-2) laboratory at the University of Texas Medical Branch at Galveston (UTMB). Dams and fetuses were humanely euthanized at E18 for subsequent analyses. Adult rats were euthanized by isoflurane overdose, followed by exsanguination. Pups aged 10 days were euthanized by isoflurane overdose, followed by decapitation. Fetuses within the placenta and neonates younger than 5 days, which are resistant to inhalant anesthetics were euthanized by decapitation using surgical scissors.

Timed-pregnant C57BL/6 mice were vaccinated i.m. at E14 (total volume up to 100 μl; 50 μl per injection site) with PBS (n = 3), rMP-12 at 1 × 10⁵ PFU (n = 5), or RVax-1 at 1 × 10⁵ PFU (n = 6). Health status and body weight were monitored at ABSL-2 at UTMB. Dams and fetuses were euthanized at E17 (n = 1 for rMP-12, n = 1 for RVax-1), or allowed to deliver pups and euthanized upon delivery (n = 2 for PBS, n = 3 for rMP-12, n = 4 for RVax-1), or at 10 days post-delivery (n = 1 per group: PBS, rMP-12, RVax-1). To minimize stress to the dams, body weight measurements were not performed at E19 or later. Newborn pups were counted upon delivery but remained with their mothers for several hours of nursing prior to euthanasia. Adult mice were euthanized by isoflurane overdose, followed by cervical dislocation. Pups aged 10 days were euthanized by isoflurane overdose, followed by decapitation. Fetuses within the placenta and neonates younger than 5 days, which are resistant to inhalant anesthetics were euthanized by decapitation using surgical scissors.

Tissues collected from rat or mouse dams included liver, spleen, brain, uterus, and placenta (when present). These tissues, along with fetal liver and brain, were preserved in TRIzol reagent for RT-qPCR and in 10% buffered formalin for histopathological analysis.

RT-qPCR Detection of Viral RNA

RT-qPCR was performed to measure viral RNA loads in tissues from dams, fetuses, and pups. Total RNA was extracted from collected tissue homogenates in Trizol using the Direct-zol RNA Miniprep Kit (Zymo Research, Irvine, CA, USA). Sample mass and total RNA concentration were documented to enable later normalization. First-strand cDNA was synthesized from total RNA using iScript Reverse Transcriptase (Bio-Rad Laboratories, Hercules, CA, USA), followed by PCR amplification with SsoAdvanced Universal Probes Supermix (Bio-Rad Laboratories) on a Mic qPCR Cycler (4 channels), as previously described [22]. Taqman PCR targeting the 5’ untranslated region of M RNA using the forward primer (RV-MUTR-F), the reverse primer (RV-MUTR-R), was described previously [22]. For standard curve validation, 10-fold serial dilutions of M-segment RNA (generated by in vitro transcription from the pProT7-vM(+) plasmid) were used for first-strand cDNA synthesis with iScript Reverse Transcriptase. RNA concentrations (copies/μl) were quantified by droplet digital PCR (ddPCR) using the same primers and probe set for the RVFV M-segment 5′ UTR on the QX100 droplet generator and reader (Bio-Rad Laboratories). The resulting cDNA from serial dilutions with known copy numbers was used to generate a standard curve for validating RNA copy numbers by qPCR using a Mic qPCR Cycler. Based on the standard curve, the analytical limit of detection was defined as 5 copies/μl. PCR-positive samples were normalized according to input tissue mass (mg). Owing to variability in tissue input among samples in PCR reactions, no statistical comparisons were performed for these data.

Histopathological Examination

Mouse tissues were fixed in 10% neutral buffered formalin and embedded in paraffin blocks. The purpose of the histopathological examination was to confirm the presence of viral antigens in the corresponding tissues. Accordingly, we primarily examined PCR-positive mouse tissues and did not include rat tissues in this study. Tissue sections were stained with hematoxylin and eosin staining at the Research Histology Laboratory at UTMB. For immunohistochemistry, tissue sections were treated with proteinase K antigen retrieval solution from Abcam, followed by blocking with Animal-Free Blocker from Vector Laboratories. Sections were then incubated with a rabbit polyclonal antibody against RVFV N protein [26], followed by a biotinylated goat anti-rabbit IgG secondary antibody from Vector Laboratories. Sections were subsequently incubated with streptavidin alkaline phosphatase and developed using the ImmPACT Vector Red substrate, both from Vector Laboratories, and counterstained with hematoxylin to visualize tissue architecture. Images were acquired using cellSens software with a DP74 camera mounted on an Olympus IX73 microscope. Positive IHC signals were analyzed under brightfield (magenta) or TRITC fluorescent filter (red). Merged images of brightfield and TRITC fluorescent filters were made by cellSens software.

Ethics statement

All experiments involving rMP-12 and RVax-1 were conducted with the approval of the Institutional Biosafety Committee at UTMB, as specified in the Notification of Use (#2021017). Animal studies were conducted in the UTMB Building 17 ABSL2 laboratory, accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). Animal protocol #1912097 was approved by the UTMB Institutional Animal Care and Use Committee (IACUC).

Results

Minimal Placental Tropism of rMP-12 and RVax-1 in Pregnant Sprague Dawley Rats

To assess the placental tropism of rMP-12 and RVax-1 in SD rats (2 dams per group), pregnant animals were vaccinated i.m. at E14 with PBS, either 1×10⁵ PFU (dams #1 and #2) or 1×10⁶ PFU (dams #3 and #4) of rMP-12, or with 1×10⁵ PFU (dams #5 and #6) or 1×10⁶ PFU (dams #7 and #8) of RVax-1 (Table 1). A dose of 1 × 10⁵ PFU has been frequently used in MP-12 vaccination in previous studies, whereas 1 × 10⁶ PFU is considered a high dose [13]. Body weight gains of those animals are shown in Fig. 1a. Dams vaccinated with rMP-12 or RVax-1 showed a reduction in body weight gain starting at E17 (3 dpv) relative to the mock-vaccinated control. Since pathogenic RVFV has been reported to spread into placental tissues as early as 2 days post-infection [27], dams and fetuses were euthanized at E18 (4 days post-vaccination), and viral RNA loads were quantified by RT-qPCR using a standard curve generated from in vitro–synthesized RVFV M-segment RNA (Fig. 1b). The analytical limit of detection (LOD) was set at 5 copies/μL, representing the lowest concentration distinguishable from 0 copies/μL.

Figure 1.

Detection of vaccine virus RNA in placental and fetal tissues of vaccinated Sprague–Dawley rat dams at E18. Pregnant Sprague–Dawley rats were mock-vaccinated or vaccinated intramuscularly at embryonic day 14 (E14) with 1 × 10⁵ or 1 × 10⁶ PFU of rMP-12 or RVax-1. (a) Changes in body weight of rat dams following vaccination with rMP-12 (dams #1–#4), RVax-1 (dams #5–#8), or PBS (dam #9). Data are presented as mean values, with error bars representing SEM for groups. The PBS group (n = 1) is shown as a single line without error bars. (b) The standard curve for RT-qPCR was generated using serial dilutions of in vitro–transcribed RVFV M-segment RNA, quantified in triplicate. The X-axis represents RNA copies/μL as determined by droplet digital PCR (ddPCR), and the Y-axis represents the corresponding Cq values measured by the Mic qPCR Cycler. This standard curve was used to convert Cq values of experimental samples into RNA copies/μL. (c) Viral RNA loads (copies/μl PCR reaction) are shown for maternal tissues from dam #9 (liver, spleen, and brain) and for pup liver and brain from mock (PBS)-vaccinated animals at delivery. (d-g) Viral RNA loads (copies/μl PCR reaction) are shown for maternal tissues (liver, spleen, brain, and placenta) and fetal tissues (liver and brain) from rMP-12-vaccinated animals (dams #1 and #2 [d] and #3 and #4 [e]) or RVax-1-vaccinated animals (dams #5 and #6 [f] and #7 and #8 [g]). qPCR results are expressed as copies/μL after conversion based on the standard curve and are presented as the mean. The dotted line indicates the analytical limit of detection (LOD = 5 copies/reaction). Samples with qPCR signals below the LOD are considered undetectable. All samples used in this assay are summarized in Table 1.

The mock-vaccinated dam #9 delivered 11 live pups at E22. Viral RNA was assessed in 3 of the 11 pups, and all samples were below the LOD (Fig. 1c, Table 1). At E18, dams vaccinated with rMP-12 at 1 × 10⁵ PFU yielded 10 live fetuses in dam #1 and 9 live fetuses plus 1 autolyzed fetus in dam #2. In dam #1, qPCR detected rMP-12 RNA in the liver and spleen (Fig. 1d), corresponding to viral loads of 73.3 and 26.3 copies/mg tissue, respectively, after normalization to input tissue mass. In contrast, viral RNA was undetectable in the brain, and all placental and fetal samples. In dam #2, rMP-12 RNA was not detected in any maternal, placental, or tested fetal tissues. At a dose of 1 × 10⁶ PFU, rMP-12–vaccinated dams had 14 live fetuses and 1 autolyzed fetus in dam #3, and 11 live fetuses in dam #4. Viral RNA was not detected in any tested tissues from either dam in qPCR reactions (Fig. 1e). Rats vaccinated with RVax-1 at 1 × 10⁵ PFU produced 10 live fetuses in dam #5 and 5 live fetuses in dam #6. At the higher dose (1 × 10⁶ PFU), dam #7 yielded 14 live fetuses and 1 autolyzed fetus, while dam #8 yielded 16 live fetuses and 1 autolyzed fetus. In dams #5, #6, #7, and #8, RVax-1 RNA was undetectable in all maternal and placental tissues, as well as in the livers and brains of all fetuses, in qPCR reactions (Fig. 1f and 1g). Together with the rMP-12 vaccination data, these findings suggest little evidence of placental or fetal infection following vaccination, even at the higher vaccine doses.

Outcomes of Pregnant C57BL/6 Mice Vaccinated With rMP-12 or RVax-1

To further assess placental tropism of rMP-12 and RVax-1, pregnant C57BL/6 mice were vaccinated at E14 with 1×10⁵ PFU of rMP-12 (dams #1–#3) or RVax-1 (dams #4–#7), or mock-vaccinated at E14 with PBS (dams #8 and #9), and allowed to deliver their pups (Table 2). Body weight gains of those animals are shown in Fig.2a. Dams vaccinated with rMP-12 showed slightly reduced body weight gain compared with mock-vaccinated controls. Dams vaccinated with RVax-1 showed similar body weight gain to controls until 3 dpv, with a reduction observed at 4 dpv. As shown in Table 2 and Fig. 2b, mock-vaccinated mice delivered 9 live pups or 7 live pups plus 1 stillborn pup between E19 and E20. Mice vaccinated with rMP-12 delivered 3 live pups (dam #1), 2 live pups (dam #2), and 6 live pups plus 1 stillborn pup (dam #3). Mice vaccinated with RVax-1 delivered 7 live pups (dam #4), 8 live pups (dam #5), 7 stillborn pups at E19 (dam #6), or 7 stillborn pups with 2 stillborn pups in utero at E21 (dam #7).

Figure 2.

Live pup counts and viral RNA levels in pregnant C57BL/6 mice vaccinated with rMP-12 or RVax-1. Pregnant C57BL/6 mice were vaccinated intramuscularly at embryonic day 14 (E14) with 1 × 10⁵ PFU of rMP-12 or RVax-1. (a) Maternal body weight changes following vaccination. The rMP-12 group included dams #1–#3, the RVax-1 group included dams #4, #5, and #7, and the PBS group included dams #8 and #9. Non-pregnant C57BL/6 mice mock-vaccinated with PBS (n = 2) were included for comparison. (b) Number of live pups per dam at delivery. (c) Viral RNA levels in maternal and fetal tissues. Total RNA was extracted at the time of pup delivery, and RVFV M-segment RNA was quantified by RT-qPCR. Viral RNA loads (copies/μL PCR reaction) are shown for the livers, spleens, brains, and uteri of dams #1–#3 (rMP-12), dams #4–#7 (RVax-1), and dam #8 (PBS), as well as for pups (livers and brains) from these dams. qPCR results are expressed as copies/μL after conversion based on the standard curve and are presented as mean ± SEM for groups with n ≥ 2; no SEM is shown for the PBS group (n = 1). The dotted line indicates the analytical limit of detection (LOD = 5 copies/reaction), and samples with signals below the LOD are considered undetectable. All samples used in this assay, along with normalized values (copies/mg tissue), are summarized in Tables 2 and 3.

Viral RNA analysis (Table 3, Fig. 2c) revealed that in dam #1, rMP-12 RNA was detected in the spleen (1,167.6 copies/mg), and the uterus (1,665.8 copies/mg). Viral RNA was undetectable in the liver and brain of dam, as well as tested pup tissues. In dam #2, rMP-12 RNA was detected in the liver (29 copies/mg), the spleen (180.8 copies/mg), and the uterus (507.4 copies/mg). The newborn pups were cannibalized between E19 and E20, preventing assessment of viral loads in pups. In dam #3, rMP-12 RNA was detected in the spleen (3,508.7 copies/mg), uterus (4,673.4 copies/mg), and the liver of the pup #2 (457.8 copies/mg). Viral RNA was undetectable in the liver and brain of the dam or remaining fetus tissues.

In RVax-1–vaccinated dams, RNA was detected in dam #4 in the spleen (134 copies/mg) and uterus (814.3 copies/mg), while RNA was undetectable in the liver and the brain of dam, as well as in the livers and brains of seven pups. In dam #5, RNA was detected in the uterus (595.5 copies/mg), as well as in the liver of the pup #1 (141.3 copies/mg). All other examined tissues were negative for viral RNA. In dam #6, which delivered stillborn pups, viral RNA was detected in the liver (37,000 copies/mg), spleen (10,309.2 copies/mg), brain (299.3 copies/mg), and uterus (1,793.4 copies/mg). Viral RNA was undetectable in examined tissues of stillborn pups. In dam #7, which also delivered stillborn pups, RNA was detected in the spleen (35.7 copies/mg) and uterus (994.2 copies/mg), while pup livers and brains were negative.

To examine early placental tropism, pregnant mice vaccinated at E14 with rMP-12 (dam #10) or RVax-1 (dam #11) were analyzed at E17 (3 days post-vaccination) by RT-qPCR (Table 4, Fig.3a and 3b).

Table 4.

Evaluation of Placental Tissues in C57BL/6 Mice

Animal ID	Vaccine	PFU dose (Back-titrated)	Status of fetus	Collection of tissues for RNA
#10	rMP-12	1×105 (7.8×104)	9 live fetuses	Dam (Liv, Sp. Br) 9 placentas 6 fetuses (Liv, Br)
#11	RVax-1	1×105 (7.8×104)	11 live fetuses	Dam (Liv, Sp. Br) 8 placentas 6 fetuses (Liv, Br)

Liv, liver; Sp, spleen; Br, brain; Pla, placenta; n/a, not applicable.

Dam #10 carried nine fetuses, and rMP-12 RNA was detected in the liver (10,733.9 copies/mg), spleen (11,483 copies/mg), and five of nine placentas (900; 170.7; 22.9; 23.4; 1,421.0 copies/mg). No RNA was detected in the brain of the dam or in the livers and brains of six examined fetuses. Dam #11 carried eight fetuses, with RVax-1 RNA detected in the liver (53,415.5 copies/mg), spleen (5,180.6 copies/mg), and five of eight placentas (47,263.5; 2,348.6; 754; 45.3; 91.2 copies/mg). The liver of the fetus corresponded to the placenta #1, which had the highest viral load contained viral RNA (105.7 copies/mg). No RNA was detected in the brain of the dam or in the livers and brains of other examined fetuses.

Additionally, pregnant C57BL/6 mice were either mock-vaccinated at E14 with PBS (dam #12) or vaccinated at E14 with 1×10⁵ PFU of rMP-12 (dam #13) or RVax-1 (dam #14) and allowed to deliver their pups (Table 5). Viral loads were assessed in the pups at 10 days of age. Dams #12, #13, and #14 delivered 7, 5, and 7 live pups at E19–E20, respectively. However, 5 of 7 pups from dam #14 were cannibalized at E20. The remaining pups survived until day 10, and no viral RNA was detected in the brains, spleens, or livers of those pups. Meanwhile, low-level viral RNA was detected in the spleens of dam #14 (100.5 copies/mg) at 10 days post-delivery.

Table 5.

PCR Positivity in C57BL/6 Mice at E17

Animal ID	Vaccine	Organ	Viral RNA (copies/mg tissue)
#10 (dam)	rMP-12 (E14)	Liver	10733.9
		Spleen	11483.0
		Placenta #1	900.0
		Placenta #3	170.7
		Placenta #4	22.9
		Placenta #5	23.4
		Placenta #9	1421.0
#11 (dam)	RVax-1	Liver	53415.5
	(E14)	Spleen	5180.6
		Placenta #1	47263.5
		Fetus #1 liver	105.7
		Placenta #3	30.1
		Placenta #5	2348.6
		Placenta #6	754.0
		Placenta #7	45.3
		Placenta #8	91.2

All other tissues were below the limit of detection in qPCR.

Histopathological changes of vaccinated C57BL/6 dams.

Based on the viral RNA levels detected by RT-qPCR, we next assessed the presence of viral N antigen by immunohistochemistry (IHC) and examined the corresponding histopathological lesions (Table 7). IHC using an anti-N antibody detected specific signals in the livers and placentas of some PCR-positive tissues (Fig. 4a, Fig. 5), but not in mock-infected tissues (Fig. 4b). However, nonspecific signals were observed in mononuclear cells within the red pulp of the spleen in a non-infected mouse, complicating the assessment of viral antigen presence in spleens. The uterus of Mouse #1 contained detectable viral RNA by qPCR. H&E staining of uterine samples revealed autolyzed placental debris in utero, and IHC detected viral N antigen within the autolyzed tissue (Fig. 4a). Although viral RNA was not detected in the liver of this animal, histological examination revealed focal hepatocellular necrosis with detectable viral antigen in the affected hepatocytes. Dam #6 experienced a stillbirth of seven pups and exhibited high viral RNA levels in the liver and spleen, with low levels detected in the brain. Histopathological analysis revealed widespread hepatocellular necrosis with detectable viral antigen in affected hepatocytes (Figs. 5a and 5b). Although viral RNA was also detected in the brain, IHC did not reveal viral antigen in this tissue. Dam #7 underwent stillbirth of ten pups, and viral RNA was detected in the uterus. Histology showed residual placental tissue in utero, with abundant viral antigen in the junctional zone and lesser amounts in the labyrinth zone; spongiotrophoblasts, syncytiotrophoblasts, and trophoblast giant cells were antigen-positive. (Fig. 4a). Both dam #10 and #11 were euthanized at E17 at 3 dpi. Dams #10 and #11 exhibited high levels of viral RNA in the liver. Histopathological examination revealed scattered antigen-positive hepatocytes, often associated with small numbers of inflammatory cells (Figs. 5c–f). Dam #11 also exhibited high levels of viral RNA in placenta #1, and to a lesser extent in placenta #5. Both placentas showed detectable viral antigens in spongiotrophoblast cells in the junctional zone (Fig. 4a). These results confirm that hepatocytes and trophoblasts can be infected with rMP-12 and RVax-1, likely contributing to pathological outcomes in pregnant C57BL/6 mice, including tissue autolysis and fetal demise.

Table 7.

Histopathology of Selected C57BL/6 Mice Tissues

Animal ID	Tissues	H&E	IHC (anti-RVFV N)
#1 (dam) rMP-12	Liver Spleen Brain Uterus	Focal and mild hepatocellular necrosis Lesions (–) Lesions (–) Autolyzed tissues observed in utero	(+) Affected hepatocytes (+) Mononuclear cells in red pulp (–) (+) Autolyzed tissue
#2 (dam) rMP-12	Liver Spleen Brain	Lesions (–) Lesions (–) Lesions (–)	(–) (+) Mononuclear cells in red pulp (–)
#4 (dam) RVax-1	Liver Spleen Brain	Lesions (–) Lesions (–) Lesions (–)	(–) (+) Mononuclear cells in red pulp (–)
#5 (pup) RVax-1	Liver	Lesions (–)	(–)
#6 (dam) RVax-1	Liver Brain Uterus	Widespread hepatocellular necrosis Lesions (–) Autolyzed tissues observed in utero	(+) Affected hepatocytes (–) (–)
#7 (dam) RVax-1	Liver Spleen Brain Uterus	Lesions (–) Lesions (–) Lesions (–) Necrotic placenta observed in utero	(–) (+) Mononuclear cells in red pulp (–) (+) Trophoblasts in placental tissues
#10(dam) rMP-12	Liver Spleen Brain Placenta #1 Placenta #7 Placenta #9	Scattered presence of necrotic hepatocytes with mild inflammation Lesions (–) Lesions (–) Lesions (–) Lesions (–) Lesions (–)	(+) Affected hepatocytes (–) Mononuclear cells in red pulp (–) (–) (–) (–)
#11 (dam) RVax-1	Liver Spleen Brain Placenta #1 Fetus #1 liver Placenta #4 Placenta #5 Placenta #6	Scattered presence of necrotic hepatocytes with mild inflammation Lesions (–) Lesions (–) Lesions (–) Lesions (–) Lesions (–) Lesions (–) Lesions (–)	(+) Affected hepatocytes (–) Mononuclear cells in red pulp (–) (+) Trophoblasts in placental tissues (–) (–) (+) Trophoblasts in placental tissues (–)
Control*	Liver Spleen Brain Uterus	Lesions (–) Lesions (–) Lesions (–) Lesions (–)	(–) (–)** (–) (–)

^*Non-pregnant C57BL/6 was used as a control.

^**Some mononuclear cells in red pulp showed positive nonspecifically.

Figure 4.

Histopathological changes and viral antigen localization in placental tissues of C57BL/6 mice vaccinated with rMP-12 or RVax-1. (a) H&E staining and immunohistochemistry (IHC) using an anti-RVFV N antibody are shown for dam #1 at E20, dam #7 at E21, and dam #11 at E17. PL, placenta; UT, uterus; L, labyrinth zone; JZ, junctional zone; TGC, trophoblast giant cell; SpT, spongiotrophoblast; SyT, syncytiotrophoblast; GC, glycogen cell. (b) IHC using an anti-RVFV N antibody is shown for the liver of an uninfected, non-pregnant C57BL/6 mouse and for placenta #4 of Dam #11 (PCR-negative tissue). Scale bars represent 25 μm.

Figure 5.

Histopathological changes and viral antigen localization in liver tissues of C57BL/6 mice vaccinated with rMP-12 or RVax-1. H&E staining and immunohistochemistry (IHC) using an anti-RVFV N antibody are shown for and dam #6 at E19 (a, b), dam #10 at E17 (c, d), and dam #11 at E17 (e, f). P, portal vein. Arrowheads indicate necrotic foci associated with inflammatory cells. Scale bars represent 25 μm.

Discussion

This study evaluated the outcomes of i.m. vaccination with the RVFV rMP-12 or RVax-1 candidate vaccines in pregnant SD rats and C57BL/6 mice at E14. Although MP-12 and arMP-12ΔNSm21/384 candidate vaccines have been reported to cause adverse effects in pregnant ewes when administered early in gestation [19, 20], their evaluation in surrogate rodent pregnancy models has not been previously reported. The pathogenic ZH501 strain causes 100% mortality in C57BL/6 mice, whereas pregnant SD rats infected with the ZH501 strain exhibited more than 50% maternal mortality, stillbirths, developmental abnormalities in pups, and widespread viral replication in both maternal and fetal tissues [23]. In our study, there was little evidence of productive replication of rMP-12 or RVax-1 in the placentas of pregnant SD rats at E18 (4 dpv), although vaccinated dams exhibited reduced body weight gain at 3–4 dpv. No viral RNA was detected in the placentas, livers, or brains of fetuses. Our study demonstrated that rMP-12 and RVax-1 were significantly attenuated in pregnant SD rats when administered at E14, relative to a pathogenic RVFV strain [23]. These findings were based on vaccination at E14 using doses of 1 × 10⁵ or 1 × 10⁶ PFU. Further characterization of this model across additional embryonic stages and vaccine doses will be required to fully assess the overall phenotype in vaccinated SD rat dams.

Mice are more susceptible to pathogenic RVFV infection than rats and non-pregnant adult mice typically exhibit 100% mortality. Because of this extreme susceptibility, RVFV infection in pregnant mice has not been well studied. Both rMP-12 and RVax-1 are highly attenuated in adult mice, although sporadic neuroinvasion and encephalitis can occur in up to approximately 10% of vaccinated animals via i.m. vaccination [22]. We hypothesized that pregnant mice could provide quantitative information on the tropism of vaccine viruses in placental and fetal tissues. Our study showed that pregnant C57BL/6 mice were more susceptible to rMP-12 or RVax-1 vaccine strains than pregnant SD rats, allowing viral dissemination into the livers and placentas by 3 dpv. Among the rMP-12–vaccinated mice, two of three dams that were observed until delivery gave birth to a reduced number of pups at E19–E20, and autolyzed placenta-like tissues with detectable viral antigens were found in the uterus. This suggested that rMP-12 infection in the placentas may induce fetal tissue autolysis; however, we could not trace the detailed course of rMP-12–infected placentas and fetuses within the dams beyond the presence of autolyzed tissues at the time of delivery. Among the RVax-1–vaccinated mice, two of four dams exhibited fetal demise: one dam (#6) delivered seven dead pups at E19, and another dam (#7) delivered nine dead pups at E21. Dam #6 exhibited the presence of viral RNA or antigens, including the liver, spleen, brain, uterus, and placentas, along with widespread liver necrosis and evidence of viral infection of hepatocytes, suggesting that severe systemic viral infection may have affected normal delivery. In dam #7, residual placentas in the uterus had high viral loads, with viral antigens detected in spongiotrophoblasts, syncytiotrophoblasts, and trophoblast giant cells within both the junctional and labyrinth zones. The results indicated that RVax-1 exhibits a placental tropism similar to that of rMP-12 in C57BL/6 mice. It remains unclear why RVax-1 vaccination did not result in autolysis of fetal tissues before the delivery of pups. Zika virus infects multiple trophoblast populations in the mouse placenta, and maternal type I IFN signaling, which limits viral replication, can also contribute to fetal resorption; inhibition of this pathway in certain maternal–fetal genotypes can reduce fetal loss [28, 29]. A key difference between ZIKV and RVFV is that mice are highly susceptible to RVFV infection without any manipulation of type I IFN signaling. Replication of rMP-12 in the placenta may trigger host responses distinct from those elicited by RVax-1, resulting in different placental and fetal outcomes. The mechanisms underlying fetal demise induced by specific vaccine strains remain to be elucidated.

Adult non-pregnant C57BL/6 mice may occasionally succumb to rMP-12 or RVax-1 infection due to viral encephalitis [22, 30]. However, these vaccines are not known to induce widespread viral hepatitis in mice, although the MP-12 vaccine has been reported to cause multiple focal necroses in the livers of livestock animals [31, 32]. At 3 dpv, C57BL/6 mice vaccinated with rMP-12 or RVax-1 exhibited focal hepatocellular necrosis, and viral antigen was detected in hepatocytes. Notably, Dam #6 showed viral infection in both liver and brain, with more extensive liver necrosis and widespread viral antigen distribution. The result indicated that pregnant mice were more susceptible to infection with rMP-12 or RVax-1. During pregnancy in mice, progesterone levels remain high, primarily secreted by the corpus luteum in the ovary rather than the placenta [33]. Synthetic medroxyprogesterone acetate (DMPA) can inhibit TLR9-induced IFN-α production in both human and murine plasmacytoid dendritic cells [34]. Thus, suppression of innate immune responses may contribute to the increased susceptibility to rMP-12 or RVax-1. However, further studies are needed to confirm this.

From a translational perspective, pregnant rodent models are not directly comparable to pregnant ewes or humans. Rats and mice have hemochorial placentas and short gestation periods (approximately 21–23 days in rats and 19–21 days in mice), whereas sheep possess a synepitheliochorial placenta and a much longer gestation period (147–152 days). Humans also have a hemochorial placenta but a prolonged gestation of approximately 280 days. Litter sizes differ substantially among species, with rats, mice, and sheep typically producing 6–14, 6–12, and 1–3 offspring, respectively, whereas humans usually carry a single fetus. Most laboratory mice and newborn lambs are highly susceptible to RVFV infection, whereas adult sheep and humans show lower susceptibility, and susceptibility in inbred rats varies depending on genetic background, with some inbred rat strains showing high susceptibility [35]. A viral dose of 1 × 10⁵ PFU may also influence vaccine safety in pregnant animals when considered on a per–body weight basis. Given the large differences in adult body weight—approximately 20–30 g for mice, 250–500 g for rats, 50–80 kg for sheep, and 60–80 kg for humans—the effective dose per kilogram is orders of magnitude higher in mice than in larger species. Thus, a fixed PFU dose represents substantially greater biological exposure in pregnant mice, which may contribute to the increased susceptibility and adverse pregnancy outcomes observed in this model compared with rats, sheep, or humans.

Some existing live-attenuated RVF vaccines retain placental tropism in ewes, and their safety has therefore been evaluated directly in pregnant ewes. Consequently, rodent models are unlikely to provide additional safety information for these existing vaccines beyond what has already been obtained in ewe studies. In contrast, surrogate animal models may facilitate early screening of future improved live-attenuated RVF vaccines. Pregnant mice have not previously been explored for this purpose due to their high susceptibility to pathogenic RVFV. In this study, both rMP-12 and RVax-1 vaccine strains were markedly attenuated compared with pathogenic RVFV, which typically causes rapid and lethal disease in adult inbred mice within 2–4 days after infection. Importantly, both vaccine strains exhibited placental tropism, with trophoblast infection potentially contributing to fetal autolysis or stillbirth. As noted above, direct translation of these findings to pregnant ewes is not straightforward due to species-specific differences in placental structure and immune responses. Nevertheless, this mouse model may serve as a sensitive system to assess whether future RVF vaccine candidates affect placental function via viral infection of trophoblasts or host factors such as innate immune responses. For example, genetic engineering of live-attenuated strains to reduce placental tropism could further improve RVF vaccines. Ultimately, evaluation of final vaccine candidates in pregnant ewes remains essential to validate safety. In the future, non-live models such as placental explants [18, 36] or trophoblast organoids [37, 38] may help reduce animal use, provided that vaccine strain tropism in these systems reliably predicts findings from live animal models.

Figure 3.

Detection of vaccine virus RNA in placentas and fetal tissues of vaccinated C57BL/6 dams at E17. Pregnant C57BL/6 mice were vaccinated intramuscularly with 1×10⁵ PFU of rMP-12 or RVax-1 at E14. Total RNA was extracted from tissues at E17 (3 days post-vaccination), and the levels of RVFV M segment RNA were measured by RT-qPCR. Viral RNA loads (copies/μL PCR reaction) in liver, spleen, brain, placenta, fetal liver, and fetal brain are shown for dam #10 vaccinated with rMP-12 (a) and dam #11 vaccinated with RVax-1 (b). qPCR results are shown as copies/μL after conversion based on the standard curve and are presented as mean ± SEM. The dotted line indicates the analytical limit of detection (LOD = 5 copies/reaction). Samples with qPCR signals below the LOD are considered undetectable. All samples used in this assay, along with normalized values (copies/mg tissue), are summarized in Tables 4 and 5.

Table 6.

Evaluation of C57BL/6 Mice at 10 Days Post-Delivery

Animal ID	Vaccine	Pups	Collection of tissues for RNA	Viral RNA (copies/mg tissue)
#12	PBS	7	Dam (Liv, Sp. Br) 3 pups (Liv, Br)	Not detectable
#13	rMP-12 1×105 PFU	5	Dam (Liv, Sp. Br) 5 pups (Liv, Br)	Not detectable
#14	RVax-1 1×105 PFU	7 (5 cannibalized)	Dam (Liv, Sp. Br) 2 pups (Liv, Br)	100.5 (Dam spleen)

All other tissues were below the limit of detection in qPCR.

Data availability

Data and materials are available through the agreement term made via the Office of Technology Transfer at UTMB.