SARS-CoV-2 infection during pregnancy results in increased mortality and adverse pregnancy outcomes. A study conducted in Washington State estimated that 10% of SARS-CoV-2-infected pregnant women required hospitalization, whereas for non-pregnant adults this number was only 2.8% (19). In addition, pregnant women diagnosed with COVID-19 are at higher risk for preeclampsia/eclampsia, severe infections, intensive care unit admission, maternal mortality, preterm birth, still birth, and severe neonatal and perinatal morbidity index, compared to uninfected pregnant women (3, 20). Current data indicate that mother-to-child transmission (MTCT) rarely occurs. Notably, in a recent study, SARS-CoV-2 RNA was only detected in the umbilical cord plasma and placental tissue of 2 out of 31 infants born to SARS-CoV-2 positive mothers (21), indicating limited replication capacity of the virus in the in-utero environment. This phenomenon may be explained by the unfavorable localization and scarcity of ACE2 and TMPRSS2 receptors in the placenta (22, 23). Similarly, only a few studies have reported the detection of SARS-CoV-2 RNA in the breast milk of infected mothers (24, 25). Zhu et al, reported an association between the presence of SARS-CoV-2 RNA in the breast milk and mild symptoms in infants (25), but this association was not observed in other studies (26). Importantly, the detection of viral RNA does not necessarily indicate the presence of replication competent SARS-CoV-2. In fact, to date, infectious SARS-CoV-2 particles have not been isolated from breast milk (27), highlighting the need for further investigations to confirm if postnatal MTCT of the virus via breast milk can occur.

Several studies have demonstrated that SARS-CoV-2-specific antibodies generated in the infected women during pregnancy are transferred to the neonates (28, 29). Flannery et al, reported that sera of 72 out of 83 infants born to SARS-CoV-2 IgG and/or IgM positive pregnant women had detectable IgG responses, and SARS-CoV-2 IgG concentrations in maternal sera correlated with those of cord blood (29). The timing of maternal infection may critically impact the efficiency of placental antibody transfer, as pregnant women with recent or ongoing infection have been reported to have overall reduced transfer of SARS-CoV-2-specific IgG against the spike, receptor binding domain, and nucleocapsid antigens compared to SARS-CoV-2 recovered individuals (30). This suggests that the timing of maternal vaccination could profoundly influence the ability to achieve neonatal protection (31, 32). SARS-CoV-2-specific maternal antibodies could also be transferred to babies through breast milk. In a recent study, 84% of SARS-CoV-2 positive and COVID-recovered mothers had IgG, IgA or IgM against the receptor binding domain in breast milk, with 53% being positive for all three isotypes (24). Conti et al, reported significantly higher levels of SARS-CoV-2 spike-specific IgA in the saliva of infants born to COVID positive women who were exclusively breast-fed compared to those that were formula-fed (28). Notably, both IgG and IgA responses in breast milk exhibited neutralization activity against live SARS-CoV-2 virus (33), indicating that immunization of pregnant women could induce infant protection via the transfer maternally acquired antibodies.

After initial hesitation to recommend SARS-CoV-2 immunization in pregnant women due to lack of safety data, both the CDC and the WHO now recommend the use of the COVID-19 vaccine during pregnancy, with several studies reporting its safety (34, 35). Similar to infected women, women who received SARS CoV-2 vaccine during pregnancy transferred virus-specific antibodies to their babies (36, 37). In a recent study, Low et al. demonstrated that two doses of a SARS-CoV-2 mRNA vaccine in lactating mother was able to elicit virus-specific IgG and IgA antibodies in the breastmilk (38). Antibody titers (36, 38) as well as durability (37) of the maternally transferred antibodies are typically higher in infants born to vaccinated mothers compared to infants born to unvaccinated infected mothers, suggesting that humoral responses transferred from mothers following prenatal vaccination could provide a better protection of infants than those obtained from natural infection, although larger clinical studies will be needed to verify such a hypothesis. Moreover, prenatal vaccination has the advantage of protecting the mother from severe disease consequences (39), thereby preventing development of adverse maternal-fetal consequences. For instance, fully vaccinated pregnant women infected with SARS-CoV-2 during the Omicron wave had lower disease severity, compared to unvaccinated women (40). Importantly, since the timing of maternal vaccination may impact the efficiency of IgG transfer, the prenatal immunization schedule must be carefully tailored to ensure optimal protection of both mother and infant (41). In addition, to design a protective prenatal immunization regimen, it will be important to understand the kinetics and durability of maternally transferred antibodies and their possible interference with the developing immune system of infants.

A variety of animal models including humanized mice, hamsters, ferrets, pigs and non-human primates (NHP) are used to study SARS-CoV-2 infection, pathogenesis and immune responses. While all of these species express ACE-2 receptors and can be infected by the virus (42), only a few of these animal models are able to recapitulate human disease. Golden Syrian hamsters are considered as valuable models for therapeutic research due to their ease of handling, low cost and similar lung pathological phenotypes post-SARS-CoV-2 infection (42, 43). Vaccinology research however has shown particular interest in using rhesus macaques (RMs), owing to their similarity to human anatomy, physiology and immunology and ability to generate neutralizing antibody responses (44). Moreover, like humans, SARS-CoV-2 replicates in the respiratory tract of RMs and viral infection results in pulmonary infiltrates and a shedding pattern similar to mild to moderate human infections (45, 46). To date, there is no published animal model of SARS-CoV-2 infection during gestation or preclinical animal studies testing the safety and efficacy of SARS-CoV-2 vaccines during pregnancy. The development of such animal models will be critical to design and evaluate prenatal immunization regimens that will confer maximum protection to the mother and their neonates against SARS-CoV-2 infection.

Pregnant women are highly vulnerable to influenza virus infection and virus-associated severe disease outcomes, resulting in increased rates of mortality (50). During the 2009 H1N1 pandemic, pregnant women accounted for 5% of deaths, while representing less than 1% population (51). While the risk of mother-to-child transplacental transmission of influenza is extremely rare (52), maternal infection during pregnancy is associated with adverse maternal-fetal outcomes such as preterm delivery, spontaneous abortion, low birth weight and fetal death (50). In addition, maternal influenza virus infection disrupts the cytokine and hormonal signaling pathways responding to respiratory pathogens, compromising lung and placental architecture, and impacting the health of both mothers and their offspring (53). Maternal influenza infection may also have an impact on the neonatal immune response. Notably, offspring of mice infected with influenza during pregnancy demonstrated significant challenges in clearing secondary respiratory infections compared to offspring of healthy dams and it has been hypothesized that reduced levels of natural killer and B cells in the pups of influenza-infected dams might contribute to the increased vulnerability to infection (54). Whether infants of women who experience influenza infection during pregnancy demonstrate a similar increased susceptibility to respiratory infections remains unclear.

To prevent maternal influenza infection and reduce the risk of complications during pregnancy, the CDC recommends that all pregnant women receive the inactivated influenza vaccine (55). The safety of maternal vaccination has been widely demonstrated (56). Vaccination during pregnancy reduced the risk of influenza-associated acute respiratory infection by one-half, lowering the risk of being hospitalized with virus-associated disease by 40 percent (57).

Maternal Influenza vaccination also provides passive immune protection to the infants, with virus-specific IgG transferred through the placenta and virus-specific IgA transferred through breastfeeding (58, 59). These antibodies are particularly important for protecting infants during their first few months of life, as influenza vaccination during the first 6 months of life is not advised. In fact, maternal vaccination during pregnancy reduced infant hospitalization due to influenza infection by up to 91% (60–63). In addition, cord blood antigen-specific hemagglutination inhibition (HAI) titer was found to correlate with protection from IBV during the first 2 months of life and higher titers were associated with reduction in the risk of infections (64). In fact, trans-placentally transferred IAV-specific IgG antibodies in cord blood neutralized IAV in vitro (65). Moreover, anti-influenza IgA antibodies transferred to infants from their vaccinated mothers during breastfeeding significantly decreased the number of respiratory illnesses during early life (58).

Unfortunately, to date the efficacy of seasonal influenza vaccines is fairly limited, usually around 40–60%, even dropping to as low as 20% in certain seasons (66). Thus, investigations of novel vaccination platforms and universal influenza immunization strategies that will confer protection against multiple influenza strains are warranted. Potential vaccination strategies will need to demonstrate efficacy, immunogenicity and safety in pre-clinical animal models, before proceeding to human clinical trials. To date, several animal models have been used for influenza vaccinology research, including mice and ferrets. Although NHPs are not natural hosts for influenza viruses, they can be experimentally infected with human isolates and exhibit disease symptoms, as observed in humans (67). To specifically study mother-to-child transfer of antibodies post-immunization, mice models have been mostly employed (68). While primary findings from rodent models can provide the proof of concept that maternal antibodies can confer protection to their offspring, the differences in placentation and mechanisms of antibody transfer between rodents and humans, must be taken into consideration and there remains a need for more physiologically relevant models.

DENV is a positive-sense single stranded RNA virus belonging to the Flaviviridae family (14, 69). Transmission of the four similar but antigenically distinct DENV serotypes (DENV-1, DENV-2, DENV-3, DENV-4) occur primarily through the bite of the female Aedes mosquito in tropical and subtropical countries, but can also occur via blood transfusions and vertically from mother to their infants (70–75). Each year ~400 million people are infected with DENV, of which 40,000 die from severe DENV disease. All four serotypes of DENV can co-circulate during an outbreak (76). In addition, co-infection between DENV serotypes and other flaviviruses such as ZIKV has also been reported (77). DENV infection is of high concern for pregnant women due to virus-associated severe maternal and fetal outcomes (78). Pregnant women are approximately 3.5 times more susceptible to development of severe DENV-disease and the mortality rate of DENV-infected pregnant women is also significantly higher, compared to non-pregnant individuals (79, 80).

DENV is endemic in over 100 countries including central and south America, the middle east, Asia, and the Pacific Islands (81). DENV infection and re-infection was ranked the second most severe vector born disease worldwide (82) as it can cause varying levels of disease, ranging from asymptomatic infection to dengue hemorrhagic fever, dengue shock syndrome and death (81, 83). Viral infection during pregnancy is associated with premature birth, fetal death, low birth weight, acute fetal distress, and hemorrhage (71, 73, 84, 85). DENV transmission from mothers to their infants have been reported. In a prospective study, DENV transmission to their infants was reported in 22.7% women infected during the 2012-2013 epidemic in French Guiana (74). However, the pathogenesis of DENV infection in neonates and infants is poorly understood. DENV transmission from mother to child often occurs during delivery, although, in-utero transmission of DENV through the placenta has been observed (74). Interestingly, Vats et al., demonstrated that umbilical cord blood cells are highly permissive to DENV infection, and hence were thought of a site of viral amplification before vertical transmission (86). To date, although DENV virus has been found in breastmilk (73), the virus has been rarely documented to be transferred postnatally during breastfeeding. Hence, CDC still encourages breastfeeding for all mothers even if those from DENV endemic areas.

Maternal DENV-specific IgG antibodies are efficiently transferred across the placenta to the fetus and persists in the infants during the first year of their life (87). Presence of maternally acquired DENV-specific antibodies has been associated with low prevalence of symptomatic DENV infection in infants <3-4 months of age (88). Presence of sub-neutralizing levels of antibodies generated during primary DENV infection in adults has been suggested to cause antibody-dependent enhancement (ADE) during a subsequent DENV exposure, leading to severe disease outcomes (89, 90). In neonates and infants, ADE is especially relevant due to presence of trans-placentally transferred maternal antibodies in the first year of their life. Although the pathogenesis of DENV infection in infants is poorly understood, the presence of maternally derived sub neutralizing levels of anti-DENV antibodies were postulated to be a critical risk factor for severe DENV disease outcomes in infants (91). A study conducted with 75 Vietnamese infants with primary DENV infection indicated that infants experienced dengue hemorrhagic fever when maternally derived neutralizing antibody titers have declined (92). Additionally, Ribeiro et al. reported DENV in fetal tissues of pregnant women diagnosed with DENV, when a spectrum of clinical presentations ranging from asymptomatic to dengue hemorrhagic fever and dengue shock syndrome was observed in the newborn. However, whether sub-neutralization levels of maternally derived anti-DENV serotypes enhanced the risk of newborn disease was not defined (93). This finding highlights that maternal immunization against DENV, followed by boosting during the first year of life might protect mothers, their fetus and their infants from severe consequences of DENV-associated diseases but also urges for caution as sub-optimal vaccination may promote ADE. To date, a live attenuated tetravalent vaccine (Dengvaxia^®) is licensed for use in dengue endemic regions. However, the use of this vaccine is contraindicated during pregnancy, although accidental administration in pregnant women was not associated with adverse outcomes (94). Hence, mother-infant pre-clinical models to study DENV pathogenesis, maternal antibody associated ADE and protective efficacy of trans-placentally acquired antibodies will be instrumental for development of safe and protective vaccine strategies targeted to this population.

Development of a suitable DENV pathogenesis model has been challenging due to the poor replication competence of clinical DENV strains in animals. To study peripheral replication of DENV, IFNα/β and γ deficient mice (AG129) have been used (95). Furthermore, mouse-adapted DENV strains were used to identify antiviral therapeutics for DENV replication. To mimic human dengue like clinical symptoms, humanized mice were generated by engrafting human hematopoietic progenitors in immunodeficient mice (96). Efficacy of DENV vaccine candidates has been tested in both immunocompetent and immunocompromised mice models (95). NHPs such as RMs intravenously infected with DENV clinical isolates resulted in hemorrhagic manifestations as observed in humans (97). Even though NHPs do not exactly recapitulate the clinical manifestations of human DENV infection, they develop comparable anti-DENV antibody kinetics to those observed in humans (98). More recently, Tree shrews infected with DENV clinical isolates resembled DENV disease outcomes as observed in humans. Tree Shrew infection model also resulted in development of neutralizing antibodies (99). While maternal immunization for DENV infection is crucial, testing the safety and efficacy of such vaccines are challenging due to lack of reliable animal models that can results in high viremia and mimic DENV clinical manifestation as observed in humans. Hence, there is an urgent need to develop such pregnant DENV transmission animal models, perhaps by utilizing the pregnant infection models already developed for other flavivirus infections, such as ZIKV, described below. Importantly, the co-circulation of different DENV strains and co-infection with other flaviviruses must be considered while developing these models.

EBOV and MARV associated febrile hemorrhagic disease symptoms begin 3-21 days after exposure and typically increase in severity. Filovirus-induced diarrhea can be quite serious, reaching 10 liters per day and resulting in severe dehydration, contributing to the pathogenesis of these viruses (104). Hemorrhage presents as petechiae, bleeding from gums, venipuncture sites, subconjunctival hemorrhage, as well as blood in vomit and stool (104). Patients typically experience multiorgan failure and hypovolemic shock leading to death (104).

Pregnant women and infants face an elevated risk of Ebola hemorrhagic fever and increased risk of mortality (105). Previous outbreaks have recorded extremely high case fatality rates for pregnant women and newborn infants, reaching 90% and 100% respectively (106). Filovirus infection during pregnancy often leads to fetal loss (107), but MTCT is not well defined, due to the sporadic and often rural nature of outbreaks as well as the potential of the viruses to penetrate multiple tissue compartments. The detection of EBOV RNA in the amniotic fluid well after viral clearance in the mother, suggests that transplacental viral transmission might be feasible (108). In addition, EBOV RNA was recovered from breastmilk (109, 110) and multiple studies suggest possible MTCT of EBOV during breastfeeding (111, 112).

To date, there is one FDA approved vaccine (Ervebo®) for EBOV infection for adults >18 years of age. Ervebo® is a replication competent recombinant vesicular stomatitis virus (VSV) expressing EBOV glycoprotein. This vaccine elicits potent humoral responses generating high titers of IgG (geometric mean titer of 969.9-1307.3, day 28 post vaccination) as well as high titers of neutralizing antibody (geometric mean titer of 122.8-211.7, day 28 post vaccination) (113). The efficacy of Ervebo® was evaluated during an outbreak in Guinea and was found to be 100% at preventing EBOV infection (113). While there are no universally established immune correlates of protection for the EBOV vaccine, the role of humoral responses in EBOV-associated disease outcomes has been implicated (114).

Currently, there are no vaccines tested prenatally for the prevention of EBOV or MARV. Clinical trials for development of the Ervebo® vaccine did not explicitly involve pregnant women. In fact, this population was considered an exclusion criterion. However, follow up studies with women who became pregnant during the course of the clinical trial demonstrated that pregnancy loss amongst the immunized group was not statistically different than that of unvaccinated group (115). Based on these safety data, WHO now recommends Ervebo® during pregnancy under certain circumstances, when the benefits outweigh risks. Very recently, Johnson and Johnson have set up a phase 3 clinical trial to investigate the safety and immunogenicity of this vaccine in pregnant women (116). There are currently no vaccine options for MARV. However, preclinical studies of replication-competent VSV-based vaccines demonstrate protective efficacy in NHPs (117).

Mice, guineapigs, hamsters and NHPs have been widely used for filovirus research (118). As expected, NHPs such as, marmosets have similar MARV disease pathology to humans (118, 119), and different filoviral vaccine platforms have been studied in nonpregnant NHP models (120). However, to date, animal models of pregnant filoviral infections are lacking and data for vaccine efficacy during pregnancy remains limited. Thus, preclinical animal models that were utilized for development of the FDA-approved. EBOV vaccine for non-pregnant individuals could be employed to investigate the safety and efficacy of maternal vaccination against Filoviruses.

Good clinical management such as early antiretroviral therapy (ART) initiation and adherence to treatment can reduce the risk of MTCT of HIV to <1-2% (125, 126), but implementation challenges such as suboptimal ART coverage of pregnant and breastfeeding women, poor access or adherence to ART, and late presentation for prenatal care prevent the complete elimination of vertical HIV transmission through ART alone. In addition, ~50% of the new infant HIV infections occur when a newly infected mother, who is not on therapy unknowingly transfers the virus to their offspring (127). Infants and children who are infected with HIV early in life from their mothers have to rely on lifelong ART to keep viral replication suppressed, and as a consequence, they are predisposed to ART-associated metabolic complications (128). Moreover, some ART regimens during pregnancy are associated with fetal complications such as neural tube disorders (129), preterm birth (130), hematological abnormalities (131), and mitochondrial dysfunction, later in life (129, 131). Congenital defects and neonatal abnormalities of HIV infection include developmental delays, neurological dysfunction, susceptibility to bacterial, viral and yeast infections, high persistent fever and weight loss (132).

Studies from the pre-ART era have indicated that in the absence of any intervention, ~45% of infected pregnant women would transmit the virus to their babies (133). The fact that more than 50% of exposed infants remain uninfected, suggests that maternal genetic factors such as receptor and HLA subtypes, as discussed in (134) or immune factors could play a crucial role in providing partial protection against infant infections. Several studies have attempted to identify the maternal immune correlates of protection against vertical transmission. Notably, maternal antibodies against conserved regions of the HIV Env protein responses have been associated with reduced MTCT risk in multiple studies. For instance, the magnitude of maternal IgG antibodies specific for Env third variable loop (V3), neutralization of tier 1 viruses and the magnitude of CD4- binding site blocking antibodies were associated with reduced risk of MTCT in subtype B HIV-infected maternal cohort (135). The specificity of maternal V3-specific IgG responses that were associated with reduced MTCT risk was further mapped to the C-terminal region of the V3 loop (136). Another study linked maternal Env-specific Abs targeting the membrane-proximal external region (MPER) of gp41 with reduced risk of vertical HIV transmission among HIV subtype C-infected ART-experienced mothers (137). In contrast, maternal antibodies against certain HIV Env protein regions have also been associated with increased risk of MTCT. For instance, gp41 ectodomain targeting Abs have been associated with increased MTCT risk in clade C HIV-infected, ART-naïve, breastfeeding mother-infant pairs (138). Additionally, in a cohort of clade C HIV-infected, breastfeeding, ART-experienced women, IgG against the variable loops 1 and 2 (V1V2) and anti-CD4 binding-site Abs were associated with increased risk of MTCT of HIV, and no association was observed with maternal V3-specific IgG binding nor tier 1 virus-neutralizing responses (139). Thus, these findings suggest that maternal immune correlates of protection against MTCT might differ based on infecting virus clade, transmission mode and ART status.

Maternal autologous virus-specific neutralizing antibody responses have also been studied in the setting on MTCT of HIV. Some of these studies have indicated that viruses transmitted to infants are generally resistant to neutralization by maternal plasma suggesting that viral escape to neutralization may be important during vertical transmission (140–143). Additionally, maternal broadly neutralizing antibody responses may select for neutralization resistant T/F variants (144). Nevertheless, other studies have reported higher levels of neutralizing antibodies in women who transmitted the virus to their babies as compared to non-transmitting infected mothers (145). Thus, our understanding of the role of neutralizing antibodies during vertical transmission and the potential impact of transmission mode remains incomplete. Moreover, the role of maternal non-neutralizing antibody responses and MTCT risk also remains poorly defined. Breast milk Env-specific IgG responses with antibody-dependent cellular cytotoxicity (ADCC) activity have been associated with reduced risk of MTCT in some studies (146) but not in others (147). Interestingly, some studies have reported that ADCC activity of passively acquired antibody is associated with improved survival in HIV-infected infants, whereas neutralization activity is not (148, 149).

Boosting maternal immunity has been explored as a possible strategy to reduce vertical transmission. HIV Phase I vaccine regimens have demonstrated high levels of immunogenicity and safety in nonpregnant adult populations. Few therapeutic vaccine trials in pregnant women also demonstrated good safety profile (150). However, these limited sample size trials were not designed to evaluate efficacy. A recent analysis of samples from 11 HIV infected women who received Env-subunit vaccines during pregnancy reported an increase in binding antibody responses following vaccination. However, autologous neutralization potency in this small group of vaccinated women were comparable to that of 6 unvaccinated HIV infected pregnant women (151). As novel HIV vaccine immunogens are under development, further exploration of maternal immunization strategies to limit infant infection especially in the setting of postnatal breast milk transmission should be pursued.

A globally effective HIV vaccine strategy must cope with the broad genetic diversity of HIV viruses, its ability to escape neutralizing antibodies, as well as different transmission routes, including MTCT. Preclinical NHP models, such as RMs infected with Simian/Simian-Human Immunodeficiency Virus (SIV/SHIV) recapitulates most virological and immunological hallmarks of HIV infection in humans and have been widely used in testing HIV vaccine immunogens. To model viral transmission during pregnancy, female RMs have been experimentally infected with SIV during various stages of gestation (152–154). Additionally, breastmilk transmission of HIV to infants has been mimicked by orally infecting infant NHPs (152, 154, 155). Using such an infant RM model of oral SHIV infection, the ability of human breastmilk-derived polyfunctional antibodies, with weak neutralization and antibody-dependent cellular cytotoxicity functions, to reduce HIV transmission was demonstrated (156). In addition, infant RM models have been utilized to study HIV in early life and to evaluate candidate pediatric vaccine regimens (157, 158) and immunotherapies (159–161).

ZIKV is primarily transmitted by Aedes mosquito bites, however, sexual transmission and transmission via blood transfusion has been reported (167). In the 2015-2016 outbreak, ZIKV was also identified to be transmitted vertically from mothers to their infants during pregnancy (163). In healthy individuals including pregnant mothers, ZIKV infection is typically asymptomatic or might cause mild febrile disease with rashes, conjunctivitis and arthralgia (168). However, vertical transmission results in neonatal birth defects. The congenital birth defects and abnormalities caused by ZIKV infection during pregnancy, also known as congenital zika syndrome (CZS), results in neurodevelopmental disabilities, severe microcephaly, visual impairments, motor delay and increased morbidity in the newborn (169, 170). ZIKV is transmitted from mother to child in 1 out of 10 ZIKV-infected pregnancies (169–171). Interestingly, although vertical transmission of the virus can occur at any time during gestation, first- and second-trimester infections have the highest risk of developing CZS in the newborn (172). Whether ZIKV can be transmitted during breastfeeding is currently unknown. Although ZIKV RNA and infectious virus has been detected in the breastmilk of ZIKV-infected mothers, viral transmission during breastfeeding has not been confirmed (173), and future studies with bigger cohorts will be needed to completely rule out the possibility of MTCT of ZIKV via breast milk

To date, there is no approved vaccine to prevent ZIKV infection. Diverse vaccine platforms including DNA, mRNA, subunit, live-attenuated and purified inactivated vaccines have been tested in non-pregnant animal models of ZIKV infection, and several of them have already entered Phase I/II clinical trials. In addition to being protective against ZIKV infection, an ideal vaccine candidate should be able to prevent vertical transmission of the virus and associated fetal complications. Studies have indicated that humoral immunity, specifically neutralizing antibodies play an important role in achieving protection against ZIKV infection (174, 175). Furthermore, ZIKV-specific IgG antibodies were found to efficiently cross the placenta (176). Hence developing a prenatal vaccination regimen able to protect mothers from infection and prevent vertical transmission seems to be an attractive strategy. Since ZIKV infection early during gestation has been associated with increased fetal disease severity (172), it is likely that vaccination prior to conception or early during pregnancy would be necessary to prevent transmission and improve fetal outcomes. Furthermore, delineating the correlates of fetal protection and most importantly, levels of antibodies necessary to achieve protection against in-utero transmission will be necessary to design a protective vaccination regimen. The fact that ZIKV is only one serotype makes vaccine design slightly less challenging as neutralizing antibodies developed against one strain would potentially neutralize others. However, a key concern for ZIKV vaccine design is elicitation of cross-reactive antibody responses with co-endemic DENV serotypes, which are antigenically similar flaviviruses (177). Cross reactive antibodies that can target conserved epitopes of ZIKV and DENV serotypes, may contribute to ADE of infection against subsequent viral infections (178, 179).

To determine the mechanisms of severe CZS in the fetus and ability of maternal vaccination to prevent vertical ZIKV transmission, relevant animal models to study maternal immune response and fetal ZIKV pathogenesis will be necessary. Mice models are often utilized to study ZIKV pathogenesis. However, for efficient replication of the virus in mouse models, transgenic type I interferon knockout mice or wild type dams treated with type I IFN receptor specific blocking antibody are preferred (180). More recently, several groups have also successfully recapitulated vertical ZIKV transmission and adverse fetal outcomes in wild type mice (181, 182). In the context of pregnancy, while the short breeding times and large litter size of mice is definitely advantageous, this model cannot accurately mimic human pregnancy primarily owing to the differences in the placental organization (183). Other animal models such as rats and pigs have been used to study vertical ZIKV transmission (180). The most relevant and widely used model for studying vertical ZIKV transmission during pregnancy and CZS-associated pathogenesis are NHPs such as RMs. In fact, ZIKV was originally isolated from a RM in 1947 from the Ziika forest and hence, RMs serve as natural ZIKV hosts (184). Coffey et al. established an intravenous and intraamniotic ZIKV-infected pregnant RM model, in which ZIKV pathogenesis and ZIKV-specific immune response comparable to humans were observed (185). In addition, ZIKV-infected RM model was also able to mimic the higher rates of congenital defects in fetus of mothers infected during first trimester vs. those infected during third trimester (186, 187). These models are currently used to assess the efficacy of prenatal ZIKV vaccination to prevent fetal transmission and associated abnormalities. Notably, Van Rompay et al. assessed a DNA vaccine regimen in pregnant RM model of ZIKV infection, where RMs were immunized before conception. While this vaccine was not able to completely prevent vertical transmission of the virus, maternal viral loads and fetal disease severity was reduced by vaccination (188), highlighting the need for further research in this area.

Among healthy individuals including pregnant women, CMV infection, is typically asymptomatic or mild with flu-like symptoms (197, 198). Congenitally infected infants develops symptoms such as rash, jaundice, microcephaly, hepatosplenomegaly and vision loss, (199–201). In 20% of cases, congenital CMV is associated with hearing loss and developmental delays (194, 202, 203). In addition to in-utero transmission, vertical transmission of CMV can also occur through viruses secreted in the birth canal, saliva, and urine (204, 205). Moreover, the virus can be transmitted postnatally via breastfeeding causing symptomatic disease in preterm (<32 weeks) and very low weight infants (205–211), while, breastmilk transmission in full-term infants is usually asymptomatic (212). The risk for in-utero congenital transmission following primary maternal infection increases with gestational age with 40-70% transmission risk at the third trimester of pregnancy (213–216). However, the risk of an infant developing defects and sequelae is greatest when transmission occurs early in gestation (217).

CMV-specific antibodies transferred from mothers to infants in utero and during breastfeeding confers partial protection from CMV infection to infants (218). Interestingly, Saccoccio et al., determined no association between plasma levels of CMV virion-specific IgG and postnatal CMV, yet identified a weak association between infant glycoprotein B (gB)-specific IgG levels and protection against mucosal virus acquisition (219). Although CMV-specific antibodies in breastmilk may confer some protection in infants, there are instances of CMV reactivation in the mother, resulting in viral shedding in the breast milk (218) that can result in infection of the infant. To avoid this mode of transmission to a vulnerable preterm infant population, in France, breast milk pasteurization is recommended for premature infants (<32 weeks gestational age, <1500g birth weight) (220). Currently, there is no licensed CMV vaccine, both due to lack of defined immune correlates of protection and the vast immune evasion techniques of the virus (221). Of the 12 vaccines under clinical development since 2017, the gB/MF59 subunit vaccine has been one of the most successful to date, conferring approximately 50% efficacy in prevention of CMV acquisition in two studies. The elicited antibodies exhibited limited virus neutralization, highlighting the potential importance of non-neutralizing antibody functions (221–223). Moreover, it has been hypothesized that T-cell responses and both neutralizing and non-neutralizing antibody functions will likely be needed for a protective CMV vaccine (224). An mRNA-based vaccine (mRNA-1647) comprised of mRNAs of CMV surface proteins glycoprotein B and pentameric complex was tested for safety and immunogenicity in a phase I trial by Moderna (225). The advantage of this vaccine over a gB vaccine was inclusion of pentameric complex (PC), since antibodies to PC can neutralize infection of epithelial cells, endothelial cells and myeloid cells (228, 229), as compared to gB, that can only neutralize infection of fibroblasts (230, 231). In addition, the majority of the neutralizing antibodies in CMV-seropositive individuals are targeted against CMV PC (228). Based on the Phase I data for this vaccine, a phase II safety and efficacy trial involving healthy seronegative and seropositive individuals is underway (226) and very recently, the company has launched a phase III trial to test the safety and efficacy for this vaccine in women of childbearing age (227).

Mice, rat, and guineapig models using species-specific CMV are the primary small animal models used to study viral dissemination (232). The guinea pig model can also be used to study vertical transmission at a much lower resource cost compared to NHPs, but it may be less advantageous and clinically relevant because of differences in the virus homology, placental anatomy, and pathogenesis in guineapigs vs. human (233–237). Species-specific CMV infecting RMs (RhCMV) and those infecting humans (HCMV) are similar in their genomic and amino acid composition, viral pathogenesis, fetal transmission, fetal sequelae (238). HCMV and RhCMV share 60% identity in their gB protein which is critical to further study the current vaccine platforms (237, 239). Notably, Bialas et al. developed a RM model of congenital CMV infection and demonstrated that maternal CD4+ T cell immunity during primary rhCMV infection is important for controlling maternal viremia and inducing protective immune responses that prevent severe CMV-associated-fetal disease (240). Furthermore, this RM model have also been utilized to study the protective roles of passively infused maternal antibodies (241) and hence are instrumental for pre-clinical studies to guide development of a protective vaccine for prevention of congenital CMV infections (238, 242).

RV mainly infects cells in the villi of intestines and virions are shed in high titers in stool (>10¹⁰-10¹¹ RV particles/gram of stool) and vomitus (248). Vertical transmission of RV perinatally or during breastfeeding has not been reported to date. In adults, RV infection can be asymptomatic, or associated with nausea, malaise, headache, abdominal cramping, mild diarrhea and fever (249, 250). RV disease symptoms in infants and children typically begin within 48 hours of infections and include diarrhea, intestinal pain and vomiting (245). Children over 3 months of age and immunocompromised infants usually experience prolonged disease which can increase risk of complications such as dehydration, electrolyte imbalance and metabolic acidosis (251).

To date, there are 4 licensed vaccines for prevention of RV-associated gastroenteritis in infants, of which 2 are primarily recommended by FDA and WHO (Rotarix and Rotateq) (251). Currently, CDC recommends infant RV immunizations in 2-3 doses with in the first 6 months of life, but these pediatric vaccines do not provide protection during the first months of life, highlighting that a maternal immunization approach will be needed to ensure protection of newborns. Payne et al, recently demonstrated the presence of maternal RV-specific IgG in infants born to naturally RV-infected US mothers (252), highlighting the feasibility of maternal vaccine-induced antibody transfer in infants. However, to achieve maximum protection in the first months of life, the timing of maternal RV immunization in the setting of current infant vaccination schedule needs to be evaluated. Interference from maternally acquired antibodies is considered as a major correlate for reduced infant vaccine efficacy, particularly those from the low-and-middle income countries (LMICs) (253). For instance, a clinical trial of Rotavac in Indian infants determined an inverse relationship between maternal IgG levels and infant vaccine response (254). Similarly, anti-RV antibodies in breastmilk have been associated with decreased vaccine efficacy in infants (255). To circumvent this maternal antibody interference, vaccinating the pregnant women and delaying the timing of infant immunization until the maternal antibody wanes could be beneficial. This strategy would passively protect the infants from RV-associated diseases for the first few months of life, followed by boosting of immune response with infant vaccination. Innate susceptibility factors such as polymorphism in specific genes (alpha 1,2 fructosyltransferase, FUT2) can impact development of RV and norovirus (NV)-specific IgG, IgA and neutralizing antibody titers both in adults and infants (256). Hence, while developing RV or NV maternal vaccines, the roles of such innate susceptibility factors need to be taken into consideration.

Several pre-clinical animal models of RV infection and pathogenesis are available that has the potential to be developed as tools to investigate prenatal vaccination strategies and protective efficacy of maternally derived antibodies in infants. RV disease and lethality are studied in wild type and immunodeficient mouse models, although mouse-specific RV strains are critical to infect wild type mice, due to lack of lethality and limited symptoms of human strains in such models (257). Mice also lack emetic response which makes studying vomiting, a hallmark symptom of RV infection, in the mouse model challenging (250). Very recently, Langel et al. has highlighted the protective efficacy of maternally-transferred dimeric IgA on suckling mouse pups (258). Guinea pigs are often used in infectious disease and vaccine development studies as they more closely model disease pathology to humans than mice (259, 260). A recent study utilized guinea pigs in a pre-clinical trivalent RV vaccine study to determine immunogenicity characteristics (261). Neonatal and gnotobiotic piglets have been used to study breastmilk-induced protection against rotavirus C (RVC), where lower levels of milk IgG and IgA in nursing pigs resulted in higher levels of clinical symptoms in their offspring (262). RV disease pathogenesis has also been tested in NHP models such as neonatal RMs (263). While maternal RV vaccination to attain protection in infants are being studied using mouse models (264), development of a RM models that can mimic clinical symptoms of RV-induced diseases, such as vomiting, to study maternal immunization and transplacental transfer of maternal antibodies will be crucial.

Signs and symptoms of NV-associated disease include diarrhea, nausea, vomiting, stomach pains, fever and body aches. This virus has a brief incubation of only 12-48 hours post exposure, and complications from infection arise mainly from dehydration and electrolyte imbalance. Children under the age of 5 are more likely to develop dehydration with NV acute gastroenteritis (269), however neonates often experience asymptomatic infection with NV (269, 270), suggesting that maternally acquired virus-specific antibodies might play a crucial role in protection against NV during early life. Although, genetic factors such as presence of histo blood group antigens (HBGA) may also play crucial role (271, 272). NV evolution results in novel dominant pandemic strains every 2-5 years (273), and owing to the fluctuations in prevalence and a lack of reliable culturing system, there are currently no maternal or infant licensed vaccines for NVs (274). MTCT of NV is uncommon.

The protective role of maternal breast milk antibodies against NV has been highlighted in recent studies, where maternal anti-NV-IgA was associated with lower prevalence and less severity of infant NV diarrheal disease (270, 275). Interestingly, while there was no difference in disease severity with breastfeeding in older infants (276), other studies have found that high titers of NV-specific IgA in breastmilk was protective against symptomatic NV infection in infants (275), potentially suggesting that virus-specific IgA are important immune correlate of NV-associated severe disease outcome. Hence, maternal vaccines that can elicit robust NV-specific mucosal IgA responses, that eventually gets transferred to infants via breastfeeding needs to be explored as a strategy to mitigate infant NV, RV or other gastrovirus-associateddiarrheal disease.

NV pathogenesis, host immune response and vaccination approaches are studied in multiple animal models such as transgenic mouse, gnotobiotic pigs, calves, pigtailed macaques, tamarins and RMs. (277, 278). However, a challenge of employing some of these models such as mice and calves to study maternal vaccination is their dissimilarities in placentation mechanisms compared to humans and lack of transplacental IgG transfer. Therefore, further research in developing NHP models such as RMs which have long gestation period, and a hemochorial placenta like humans (279) to study NV pathogenesis and maternal antibody transfer remains critical.

RSV infection in healthy adults typically resolve in a week or two, however the virus can cause severe disease manifestations, resulting in hospitalization in infants and children. Severe RSV diseases is associated with breathing difficulties, cough, respiratory distress syndrome, pneumonia and respiratory failure (282). RSV infection during pregnancy also results in acute respiratory illness, which may be clinically severe and lead to hospitalization (283, 284). While evidence of MTCT of this virus is rare, maternal RSV infection can potentially alter infant immunity, resulting in lower airway lymphocyte and cytokine profiles that are significantly different from that of an RSV-naïve host during a first postnatal RSV lower respiratory tract infection (285).

The protective role of RSV-specific antibodies that are trans-placentally transferred from mother to infants have been recently studied in a cohort of 587 mother-infant pairs who were followed from the time of birth to 6 months of age (286). While a strong correlation was observed between maternal and infant RSV antibody titers, it was not very clear why certain RSV-infected infants suffer from severe disease symptoms even in the presence of high levels of RSV-specific antibodies from their mothers. In order to define immune correlates of protection against RSV-infected in infants, Jans et al, compared the RSV-specific antibody responses between infected hospitalized infants to those of uninfected infants. While infected infants demonstrated lower RSV-specific avidity as compared to uninfected infants, there was no association between neutralization potency or antibody epitope-specificity and disease severity (287). Identifying the functional response of antibodies associated with RSV-protection is critical to facilitate the development of maternal vaccination approaches that can elicit protective antibody response to prevent RSV infection severity in infants.

There are currently no licensed vaccines against RSV, although several platforms are under investigation in both non pregnant and pregnant groups. For instance, a RSVpreF3 vaccine was found to be well tolerated and immunogenic in a cohort of non-pregnant women (288). In addition, a single intramuscular dose of RSV fusion protein-based nanoparticle vaccine in pregnant women showed protection against RSV infections in infants, with 1.5% of infants in the vaccine group getting infected compared to 2.4% in the placebo group. This immunization regimen was also able to successfully reduce the percent of hospitalization from 3.7% in the placebo group to 2.1% in the vaccine group (289). This immunization regimen did not result in any adverse side effects in pregnant women. However, this study was not powered to estimate the vaccine efficacy, and hence, better powered, and extensive studies of maternal immunization and protection from RSV infection during infancy are warranted.

To study RSV replication and pathogenesis, animal models such as mice, cotton rats, sheep and most commonly NHPs have been infected with human RSV (290). Chimpanzees have been previously used in vaccinology research to evaluate the protective efficacy of RSV vaccine candidates (291). However, due to their restricted use in research, NHP models such as African Green Monkeys (AGM) and macaques are often used. AGM is only semi-permissive to the virus, whereas RMs do not mimic the clinical symptoms observed in humans (290). In a recent study, three different NHP models, baboons, AGM and RMs were experimentally infected with RSV and the kinetics and titers of virus-specific antibody development was evaluated (292). Of these NHP models, AGMs demonstrated high levels of RSV-specific antibody titers and serum IgG levels of the mother correlated with IgG levels of the infants that remained detectable 4-5 months post birth. Hence this model was selected as most suitable for studying prenatal vaccination strategies. This study indicated that, immunization of pregnant mothers with RSV prefusion F protein resulted in virus-specific antibodies in the infants that were both higher in magnitude and durability compared to those of unvaccinated controls. Further studies to evaluate protective efficacy of these antibodies against infant RSV challenge will be crucial.

In adults, HSV is primarily a sexually transmitted virus associated with physical symptoms like cold sores and genital lesions. HSV can be largely asymptomatic as the virus persist in a latent state throughout the lifespan of the infected individual, with occasional reactivation (296, 297). Infant’s symptoms can vary and may include ulcerations of eye or mucous membranes, lethargy, seizures, encephalitis, respiratory failure, and mortality (298). Up to 85% of MTCT of HSV occurs in the birth canal during the time of delivery from exposure to HSV in the vaginal mucous membranes, whereas 10% of transmissions occur postnatally, when HSV is primarily transmitted to the infant through microlabial or cutaneous lesions (298–300). In-utero transmission is rare, occurring in 1-5% of cases (298). In 2017, HSV-1 and HSV-2 were estimated to cause 14,000 annual infections globally among neonates (301).

HSV-specific antibodies with potent neutralizing functions have been detected from cord blood and neonatal sera of infants born to HSV-1 and HSV-2-infected mothers, highlighting the efficient transplacental transfer of HSV-specific antibodies (302, 303). Using a mouse model, successful maternal transfer of HSV-specific IgG to the neonatal trigeminal ganglion (304) and maternal breast milk was also demonstrated (305). These findings indicate that a maternal vaccine for HSV infection might be able to protect neonates via placentally-transferred HSV-specific IgG during their first few months of life. However, to date, there is no licensed HSV vaccine available for non-pregnant or pregnant populations (306, 307). HSV’s immune evasion techniques including cell to cell spread, Fc receptor analog production, inhibition of complement-mediated lysis, and the establishment of latency makes vaccine design especially difficult (308–314). Furthermore, a randomized, double blind efficacy trial conducted in 2012 with over 8,000 women found that even though an HSV-2 gD subunit vaccine elicited neutralizing antibodies, vaccine efficacy was only 20% (223, 315), further underscoring the need for investigating the potential protective role of additional antibody and other immune functions, in order to guide development of an efficacious vaccine.

Pregnant and neonatal mice models have been used to study HSV infection (316, 317). A mouse model has been used to study the trivalent HSV-2 subunit, replication defective live attenuated d15-29 and HSV-1 0ΔNLS vaccine candidates (307). Cotton rats have also been used for HSV vaccinology research ( (316).The guinea pig model of genital herpes best mimics human disease and has been currently employed for HSV vaccine research. This model allows for assessment of acute and recurrent HSV disease and viral shedding (306, 316). RM model of HSV-2 infection demonstrates histologic evidence of acute pathology, latency establishment in the ganglia, and viral reactivation. However, further modification will be necessary to conduct studies of maternal infection and placental antibody transfer in the RM model, as HSV-2-specific antibodies were only detected in 30% of non-pregnant animals after vaginal inoculation of HSV-2 strains (318).

As a leading cause of liver fibrosis, hepatocellular carcinoma, and other related comorbidities, the effect of HBV infection on health is profound (325, 326). There are two major courses of HBV pathogenesis; acute hepatitis that resolves completely over the course of weeks to months, and chronic hepatitis that results in a carrier state in the infected individual (322). While chronic hepatitis is largely asymptomatic, there is ongoing viral replication that over the course of years often progresses to fibrosis and is associated with high risk of hepatocellular carcinoma, (327). Chronic HBV is also associated with extrahepatic diseases including serum-like sickness, vasculitis (polyarteritis nodosa), and kidney diseases (membranous glomerulonephritis, and membranoproliferative glomerulonephritis) (328). Individuals infected as neonates and children have the highest burden of HBV disease due to the length of chronicity. Additionally, the likelihood of acute hepatitis progressing to chronic one in an individual is largely dependent on their age at the time of infection. Adults infected with HBV have a 1-5% chance of becoming carriers, children have a 20-30% chance, and neonates have a 90-95% chance of chronic infection (329–333).

Because of the risk associated with perinatal HBV transmission due to primary or chronic maternal infection, pregnant women are routinely screened for HBV surface antigen. If the pregnant women is HBV surface antigen negative and at high risk of acquiring HBV, maternal vaccination is recommended to achieve protection against HBV infection. Alternatively, if the mother screens positive for current infection at delivery, the neonate receives hepatitis B immunoglobulin (HBIg) prophylaxis in addition to the standard first dose of the HBV vaccination series within 12-24 hours of birth. Subsequent boosters for the vaccination regimen occur at 1 month and 6 months of age (334). The virus usually takes more than 24 hours to establish chronicity and hence HBIg treatment is capable of reducing infant infcetion by more than 90% (335). The major barriers currently faced in this effort is the lack of maternal screening in endemic areas and failure of implementation of this prophylactic regimen at birth. Moreover, even in areas of widespread adoption of this practice, there are still 1-9% of cases of breakthrough infection, mainly associated with delayed or incomplete immunoprophylaxis (336). Notably, cord blood HBV antibodies correlated strongly with maternal HBV antibody levels post maternal vaccination, highlighting prenatal vaccination could be an additional approach to protect the newborn against HBV infection (337). Importantly, a three dose HBV vaccine regimen has been found to be safe when administered during pregnancy (338).

HBV’s complex life cycle and narrow host range has significantly limited the availability of animal models that can recapitulate HBV pathogenesis in humans. Consequently, animal models to study HBV prenatal vaccinations are lacking. Chimpanzee is the only NHP model which closely mimics human HBV disease pathogenesis and has been previously used for HBV vaccine and therapeutic research (339). However, chimpanzee models are no longer used for research purposes due to limited animal availability and ethical constraints. The other available models to study HBV pathogenesis are mouse models than can be manipulated to support infection, and mammals and bird species that support hepatitis viruses with a similar life cycle to HBV (340, 341). However, mouse models require genetic manipulation, humanization, and immune cell depletion to support infection and mammalian and bird models such as tree shrew, woodchuck, and duck do not closely recapitulate HBV MTCT. Although there is a lack of appropriate animal models for HBV vaccine research, models for MTCT are limited as effective vaccines and immunoprophylaxis have already been developed and are widely and successfully utilized.