key: cord-0312085-jc2jqhph
authors: Freeman, Megan Culler; Wells, Alexandra I.; Ciomperlik-Patton, Jessica; Myerburg, Michael M.; Anstadt, Jennifer; Coyne, Carolyn B.
title: Respiratory and intestinal epithelial cells exhibit differential susceptibility and innate immune responses to contemporary EV-D68 isolates
date: 2021-01-05
journal: bioRxiv
DOI: 10.1101/2021.01.05.425225
sha: 652f92314fbbaec6971fdaae3814218f22dbe5f0
doc_id: 312085
cord_uid: jc2jqhph

Enterovirus D68 (EV-D68) has been implicated in outbreaks of severe respiratory illness and acute flaccid myelitis (AFM) and is detected in patient respiratory samples and from stool and wastewater, suggesting both respiratory and enteric routes of transmission. Here, we used a panel of EV-D68 isolates, including a historical isolate and multiple contemporary isolates from AFM outbreak years, to define the dynamics of viral replication and the host response to infection in primary human airway cells and stem cell-derived enteroids. We show that some recent EV-D68 isolates have decreased sensitivity to acid and temperature compared with an earlier isolate and that the respiratory, but not intestinal, epithelium induces a robust type III interferon (IFN) response that restricts infection. Our findings define the differential responses of the respiratory and intestinal epithelium to contemporary EV-D68 isolates and suggest that some isolates have the potential to target both the human airway and gastrointestinal tracts.

EV-D68 has undergone rapid evolution since the 1990s, leading to the emergence of four 74 clades, termed A-D (Du et al., 2015; Tokarz et al., 2012) . This degree of evolution has led to loss 75 of neutralization from pre-existing antibodies, highlighting the potential significance of these 76 changes (Imamura et al., 2014) . Contemporary EV-D68 isolates exhibit different biologic 77 properties than historical reference isolates, including replication in neuronal cells (Brown et al., 78 2018). EV-D68 is often detected in patient respiratory samples, however, EV-D68 has also been 79 isolated from stool specimens and wastewater, suggesting that it may also be transmitted by the 80 fecal-oral route (Bisseux et al., 2018; Pham et al., 2017; Weil et al., 2017) . The viral and host 81 determinants that influence EV-D68 tropism remain largely unknown, particularly in the respiratory 82 and gastrointestinal epithelium. Moreover, whether there are differences in the replication 83 dynamics and/or host responses to isolates circulating prior to AFM outbreaks versus 84 contemporary isolates is also unclear.

The EV-D68 reference isolate Fermon is often used as a historic isolate, due to its isolation 86 in the mid-1960s. However, this isolate has undergone decades of passage through cell lines and 87 has thus likely undergone changes that make it well-adapted for replication in cell culture and less 88 representative of its original sequence when it was isolated from a child with pneumonia (Schieble 89 et al., 1967) . These changes highlight the need to perform comparative studies using pre-90 outbreak and contemporary EV-D68 isolates in order to define the viral and host determinants of 91 infection. In this study, we performed comparative studies of replication kinetics, temperature 92 sensitivity, polarity of infection, and cellular responses to infection using a panel of EV-D68 93 isolates, including a historic isolate and multiple isolates from AFM outbreak years. To define host 94 cell-type specific differences in EV-D68 replication and/or host responses, we performed 95 comparative studies in primary human bronchial epithelial (HBE) cells grown at an air-liquid 96 interface and in primary human stem-cell derived intestinal enteroids. We found that respiratory 97 and intestinal cell lines were permissive to both historic and contemporary EV-D68 isolates, but 98 that there were isolate-specific differences in temperature sensitivity at 33 o C or 37 o C. In contrast, 99 primary HBE cells were largely resistant to EV-D68 replication, with only one isolate, 100 KY/14/18953, able to replicate. KY/14/18953 and MA/18/23089 were able to replicate in human 101 enteroids. Primary HBE, but not enteroids, mount a robust innate immune response to EV-D68 102 infection, characterized by the induction of type III interferons (IFNs) and to a lesser extent type I

IFNs. Lastly, we show that inhibition of IFN signaling enhances EV-D68 replication in primary 104 HBE, supporting a role for this signaling in the control of viral replication in the airway. Collectively, 105 these data define the differential responses of the respiratory and intestinal epithelium to historic 106 and contemporary EV-D68 isolates. Figure 1A) . Collectively, these 128 studies suggest that select EV-D68 isolates exhibit cell type-specific sensitivity to temperature in 129 cell lines (summarized in Figure 1C ).

Some contemporary EV-D68 isolates have increased acid tolerance

We found that all EV-D68 isolates were capable of replicating in gastrointestinal-derived cell lines.

However, in addition to cellular tropism, enteric viruses must be stable in acidic environments to 134 infect the GI tract. Previous work suggested that select EV-D68 isolates were destabilized 6 following exposure to low pH (pH 4-6) as a mechanism of genome release during vial entry (Liu 136 et al., 2018) . However, whether this instability might influence the enteric route of transmission is 137 unclear. In order to define the stability of EV-D68 virions in various conditions that mimic the 138 environment in the GI tract, we exposed historical and contemporary isolates of EV-D68 to 139 simulated intestinal fluids of the stomach and fed and fasted states of the small intestine over 140 short (30 min) and long (60-120 min) exposure times. These fluids reflect not only the differential 141 pH of the GI tract, but also contain bile acid and phospholipids that better recapitulate some 142 aspects of the GI luminal content. To compare the stability of EV-D68 to other members of the 143 enterovirus family that are transmitted primarily via the fecal-oral route, we performed similar 144 studies with echovirus 11 (E11) and EVA71. E11 and EVA71 were stable in both fed state small 145 intestine (FeSSIF pH 5) and fasted state small intestine (FaSSIF pH 6.5) for all exposure times 146 tested (Figure 2A, 2B) . However, whereas E11 exhibited significant reductions in titer when 147 exposed to fasted state simulated gastric fluid (FaSSGF pH 2.0), EV71 was less impacted by this 148 exposure (Figure 2A, 2B) . None of the EV-D68 isolates tested were able to withstand the most 

Next, we determined whether EV-D68 could infect GI-derived primary cells, particularly 178 given that all isolates replicated to high titers in a GI-derived adenocarcinoma cell line (Caco-2).

To do this, we used human primary stem cell-derived enteroids, which we used previously to 180 define the cellular tropism of other enteroviruses in the GI epithelium (Drummond et al., 2017;  181 Good et al., 2019) . We found that only one isolate, KY/14/18953, replicated in human enteroids, 182 which occurred in a temperature-independent manner but that there were very low levels of 183 infection by other isolates tested, although MD/09/232229 exhibited some capacity to replicate to 184 low levels ( Figure 4A-B) . A limitation of the above-described model is that enteroids grown in 

In this study, we define differences in the dynamics of EV-D68 replication and pH stability using 281 a panel of isolates from AFM peak years and a pre-outbreak isolate. In addition, utilizing two 282 primary human cell models representing common tissue sites targeted by enteroviruses in 283 humans, we define differences in epithelial responses to EV-D68 between the respiratory and GI 284 tracts. Collectively, this work details the varied responses of the respiratory and intestinal 285 epithelium to historic and contemporary EV-D68 isolates and defines the role of type III IFN 286 signaling in the control of EV-D68 infection in the respiratory, but not intestinal, epithelium.

We found that most isolates of EV-D68 efficiently replicated in both respiratory and 288 intestinal epithelial cell lines, although there were some isolate-specific differences in temperature 289 sensitivity. By comparison, primary HBE cells were less permissive to EV-D68 infection. One 

Previous work with EV-D68 before the emergence of AFM suggested that due to 309 preferences for replication at 33°C and sensitivity to acid in vitro, it was more suited to be a phospholipids, that more closely mimic the gastrointestinal environment. Our studies using 314 multiple contemporary isolates after the emergence of AFM suggest that these isolates are 315 relatively stable at 37°C and also have improved acid stability. However, none of the EV-D68 316 isolates tested were stable during even short incubations with the most acidic fluid, the simulated 317 fasted state stomach fluid, at a pH of 2. Our data also indicate that many isolates of EV-D68, even 318 those associated with AFM outbreaks, are unable to replicate efficiently in human enteroids. Table   398 1. Viruses were grown in HeLa cells at 33°C in 5% CO2 until CPE was observed, purified by 399 ultracentrifugation over a 30% sucrose cushion as previously described (Morosky et al., 2016) .

Purity of all viral stocks was confirmed by Sanger sequencing of VP1 using enterovirus-specific 401 primers, as described previously (Oberste et al., 2003) . Plaque assays were performed in HeLa 

ICAM-5/Telencephalin Is a Functional Entry Receptor for Enterovirus D68

Human enterovirus D68 in clinical and sewage 571 samples in Israel

MA/18/23089 (pink), or KY/14/18953 (blue) and incubated at 33°C (A, 638

) surfaces. Supernatants were sampled at 640 the indicated hpi from both apical (A-D) and basolateral (E-H) compartments and titers 641 determined by TCID50 assays. Titers are shown as mean ± standard deviation from three 642 independent replicates. Dotted line denotes limit of assay detection

ANOVA. *p<0.05, **p<0.005, ***p<0.0005, ****

Comparison of EV-D68 strain-specific growth characteristics in primary human 656 enteroids. Primary enteroids grown on Matrigel (A-B) or transwells (C-F) were infected with 10 6 657 PFU of the indicated EV-D68 strains

For intestinal cells grown on transwells, cells were infected 660 from either the apical (C, E) or basolateral (D, F) surfaces. Supernatants were sampled at the 661 indicated hpi from both apical (C-D) or basolateral (E-F) compartments and titers determined by 662 TCID50 assays. Titers are shown as mean ± standard deviation from three independent 663 replicates. Dotted line denotes limit of assay detection

Figure 6. IFN signaling restricts EV-D68 replication in primary human airway cells

B) Levels of IFN-b (A) or IFN-l1 (B) as determined by Luminex-based assays in medium 694 harvested from primary human bronchial epithelial (HBE) cells pretreated with the JAK1/2 inhibitor 695 ruxolitinib (5 mM, Ruxo) or DMSO control for 1 hour and then

Induction of the interferon stimulated genes (ISGs) CXCL10 or IFIT1 in control 699 (DMSO)-or Ruxo-treated HBE infected with the EV-D68 strains MD/09/23229 or KY/14/18953 as 700 assessed by RT-qPCR. Symbols represent individual biological replicates from at least two 701 unique donors

or KY/14/18953 for 48 hours at 33°C. Statistical significance was determined using a Student's t-703 test

Whole genome RNAseq-based transcriptional profiling from total RNA 712 isolated from primary human bronchial epithelial (HBE) cells grown at an air-liquid interface or 713 primary human enteroids infected with EV-D68 isolates MD/09/23229 and KY/14/18953 was 714 performed in HBE infected from the apical or basolateral domains of in human enteroids infected 715 at 33 o C or 37 o C. (A) Heatmap of vRNA FPKM (fragments per kilobase per million reads mapped) 716 values for apical and basolateral infection of HBE with the indicated isolate and of enteroid 717 infection at 33°C and 37°C with KY/14/18953. Key is at right. Purple indicates high viral reads and 718 white indicates low viral reads

and 721 between HBE and enteroids infected with KY/14/18953 (E). (F) Heatmap of select interferon 722 stimulated genes (ISGs) in primary HBE infected with the indicated isolates of EV-D68 from the 723 apical of basolateral domains or in mock-infected controls. Scale at right. Red indicates higher 724 expression and blue indicates lower expression

Heatmap of transcripts (based on log2 RPKM) associated with type I, II, or III interferons (IFNs)

HBE cells infected apically and basolaterally at 33°C with the indicated strains or in enteroids 729 infected with KY/14/18953 at 33 o C or 37 o C. Scale at right, Red indicates higher RPKM values, 730 blue represents low RPKM values

Supplemental Figure 2. IFN induction in response to EV-D68 infection is independent of 734 temperature. (A), Luminex-based multianalyte profiling of 37 cytokines and chemokines in 735 primary human bronchial (HBE) cells infected with 10 6 PFU of

Shown is a heatmap based on cytokines induced relative to mock 738 infected controls (key at right), with blue denoting significantly increased cytokines in comparison 739 to uninfected. Grey denotes little to no change (scale at top right). Data are based on three 740 independent experiments. Levels of IFN-b (B), IFN-l1 (C), or IFN-l2 (D) from HBE infected at 33 o C 741 or 37 o C are shown

Levels of IFN-l1 as determined by Luminex-based assays in Calu-3 (E) or Caco-2 (F) cells 743 infected with MD/09/23229 or KY/14/18953 at 33°C or 37°C. Symbols represent individual 744 biological replicates. Statistical significance was determined using a Student's t-test