Background: Multiple studies have shown both induction and inhibition of autophagy during Zika virus (ZIKV) infection. While some have proposed mechanisms by which autophagic dysregulation might facilitate ZIKV vertical transmission, there is a lack of in situ data in human and non-human primate models. This is an especially pertinent question as autophagy-inhibitors, such as hydroxychloroquine, have been proposed as potential therapeutic agents aimed at preventing vertical transmission of ZIKV and other RNA viruses.
Objectives: Given the paucity of pre-clinical data in support of either autophagic enhancement or inhibition of placental ZIKV viral infection, we sought to assess cellular, spatial, and temporal associations between placental ZIKV infection and measures of autophagy in human primary cell culture and congenital infection cases, as well as an experimental non-human primate (marmoset, Callithrix jacchus) model.
Study design: Primary trophoblast cells were isolated from human placentae (n = 10) and infected in vitro with ZIKV. Autophagy-associated gene expression (ULK-1, BECN1, ATG5, ATG7, ATG12, ATG16L1, MAP1LC3A, MAP1LC3B, p62/SQSTM1) was then determined by TaqMan qPCR to determine fold-change with ZIKV-infection. In in vivo validation experiments, autophagy genes LC3B and p62/SQSTM1 were probed using in situ hybridization (ISH) in the placentae of human Congenital Zika Syndrome (CZS) cases (n = 3) and ZIKV-infected marmoset placenta (n = 1) and fetal tissue (n = 1). Infected and uninfected villi were compared for mean density and co-localization of autophagic protein markers.
Results: Studies of primary cultured human trophoblasts revealed decreased expression of autophagy genes ATG5 and p62/SQSTM1 in ZIKV-infected trophoblasts [ATG5 fold change (±SD) 0.734-fold (±0.722), p = 0.036; p62/SQSTM1 0.661-fold (±0.666), p = 0.029]. Histologic examination by ISH and immunohistochemistry confirmed spatial association of autophagy and ZIKV infection in human congenital infection cases, as well as marmoset placental and fetal tissue samples. When quantified by densitometric data, autophagic protein LC3B, and p62/SQSTM1 expression in marmoset placenta were significantly decreased in in situ ZIKV-infected villi compared to less-infected areas [LC3B mean 0.951 (95% CI, 0.930-0.971), p = 0.018; p62/SQSTM1 mean 0.863 (95% CI, 0.810-0.916), p = 0.024].
Conclusion: In the current study, we observed that in the non-transformed human and non-human primate placenta, disruption (specifically down-regulation) of autophagy accompanies later ZIKV replication in vitro, in vivo, and in situ. The findings collectively suggest that dysregulated autophagy spatially and temporally accompanies placental ZIKV replication, providing the first in situ evidence in relevant primate pre-clinical and clinical models for the importance of timing of human therapeutic strategies aimed at agonizing/antagonizing autophagy. These studies have likely further implications for other congenitally transmitted viruses, particularly the RNA viruses, given the ubiquitous nature of autophagic disruption and dysregulation in host responses to viral infection during pregnancy.
Keywords: Zika virus; autophagy; autophagy and viral infection; congenital Zika syndrome; placental viral infection; viral transmission.