Severe COVID-19 patients have impaired plasmacytoid dendritic cell-mediated control of SARS-CoV-2

Type I and III interferons (IFN-I/λ) are important antiviral mediators against SARS-CoV-2 infection. Here, we demonstrate that plasmacytoid dendritic cells (pDC) are the predominant IFN-I/λ source following their sensing of SARS-CoV-2-infected cells. Mechanistically, this short-range sensing by pDCs requires sustained integrin-mediated cell adhesion with infected cells. In turn, pDCs restrict viral spread by an IFN-I/λ response directed toward SARS-CoV-2-infected cells. This specialized function enables pDCs to efficiently turn-off viral replication, likely via a local response at the contact site with infected cells. By exploring the pDC response in SARS-CoV-2 patients, we further demonstrate that pDC responsiveness inversely correlates with the severity of the disease. The pDC response is particularly impaired in severe COVID-19 patients. Overall, we propose that pDC activation is essential to control SARS-CoV-2-infection. Failure to develop this response could be important to understand severe cases of COVID-19.


March 2021
Data Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A description of any restrictions on data availability -For clinical datasets or third party data, please ensure that the statement adheres to our policy Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
Sample size

Data exclusions
Replication Randomization Blinding Authors can confirm that all relevant data are included in the article and/or its supplementary information files. we have included a Data availability section to mention that 'Authors confirm that all relevant data are included in the article and/or its supplementary information files, i.e. datasets used in the study along with appropriately accessible links/accession-codes. Source data are provided with this paper.' For the patient cohort analysis, our outcome variable is a frequency circulating in a beta-distribution. Since it falls within the standard unit interval ] 0; 1 [, we, in a preliminary study, logit-transformed the frequency and performed an Anova power test on some randomly selected biomarkers in order to assess the minimum sample size required in each group. With a power test (probability 1-") = 0.80, alpha risk = 0.05 and Effect Size = 0.94 (ES = the difference between the largest and smallest means divided by the square root of the mean square error). Our result showed that we need at least 5 people in each patient group. The minimum observed in our data = 8, for the Rea group. Thus, we are confident that no significant sample size problem has occurred. For in vitro experiments, the number of used samples was sufficient to reach required statistical power for all of our statistical tests. Our rational is that when there are not enough samples, the test will not reject our null hypothesis (H0) regardless of the data, regardless H0 is true or false. Thus, before proceeding with the statistical analysis, we ensure by mathematical simulation that H0 can be rejected using our sample sizes when H0 is actually false. To this aim, we generate highly contrasted dummy data having the same shape as our observed data (i.e., same number of samples and same number of experimental conditions). We test the dummy data, if the test cannot reject H0 despite maximum contrasts, this means that the sample size is not sufficient for the choosen test, then the analysis is stopped and no result is presented.
In addition, there is not assumption on the shape distribution (i.e., small sample size) and thus we used non-parametric test.
Data exclusion of flow cytometry analysis when the total count of cells were too low to be accurate when analyzed in the gated population..
The reproducibility of the results was addressed by performing at least three independent experiments and using for each one cells collected from distinct donors. The numbers of independent experiments are indicated in the figure legend section. The results show the means of the parameters measured in the independent experiments, with error-bars representing the standard deviations. When appropriate, each individual experiment is represented by a single dot or symbol, as indicated in the legends.
The constitution of the cohort was done by the collaboration of the Hospices Civils de Lyon (HCL). This cohort consists of patient groups recognized by clinicians as: i) patients admitted in intensive care units for severe disease at hospital admission (i.e., acute respiratory distress syndrome or severe pneumonia requiring mechanical ventilation, sepsis and septic shock) are referred to as Severe group ii) patients with mild symptoms (i.e., low-grade fever, cough, malaise, rhinorrhea, sore throat) are referred to as Mild early group when collected in the first two weeks and Mild late group for later time points.
For in vitro test, assigning samples in different experimental conditions was completely random because: i/ our data were homogeneous across factors that can substantially affect our measurements and ii/, our study had an unidimential effect to test. In addition, the collected cells were carefully divided and put in each well without preconceptions before their subjection to the different experimental conditions. Then cells underwent treatments in the same environmental circumstances, i.e., same time, laboratory bench, instruments for the analysis, experimenter etc… to avoid introducing any known technical bias. Thus, statistical analysis valides that the treatment is solely responsible for the biological response and test for the treatment effect without correcting for potential interfering technical bias.
The processing of samples for flow cytometry analysis was performed with coded samples, thus investigators were blinded to group allocation during data collection.
For other types of experiments the blinding was not possible and/or not relevant (e.g., analysis with automatic measurement for ELISA, RT-qPCR etc..).
All cell line were test negative for Mycoplasma contamination None