Dr. Kulasinghe, from QUT Centre for Genomics and Personalised Health and School of Biomedical Sciences, delivered the findings to a special virtual meeting of the American Association of Cancer Research (AACR) on “COVID-19 and Cancer” earlier this month.
He said results of the test could inform doctors those patients that were likely to develop a severe infection and require a ventilator when they first present and thus differentiate them from patients likely to experience a milder case and who could go home and self-isolate.
“This is extremely important for the triage of patients when hospitals are running near or beyond capacity,” Dr. Kulasinghe said.
“We used spatial transcriptomic profiling (a technique which enables researchers to map cell-to-cell interactions and genes) to study lung samples from COVID-19 patients who had died.
“These spatial profiling biology approaches to understand complex tissues were voted the method of the year in 2020 Nature Methods.
“We drew upon our previous experience in spatial profiling of lung cancer to study COVID-19 in the lungs.
“Using high-resolution imaging and genomic profiling, we were able to map the presence of the virus in the lungs down to the single cells present in the lung tissue.
“We discovered a handful of pro-inflammatory genes which were upregulated (higher expression) in COVID-19 cases when compared with the closest pandemic virus, swine flu or H1N1, and the lungs of healthy people.
“The pro-inflammatory genes, including one called ifi27, are involved in type 1 interferon response—an inflammatory response to defend the body from viruses and other pathogens.
“The value of measuring this biomarker, ifi27, in a nasal swab or blood sample is in triaging patients because it can tell us how severe the COVID-19 disease is as soon as the patient seeks medical help with COVID symptoms.”
Dr. Kulasinghe said the researchers had measured ifi27 in asymptomatic, mild, moderate and severe COVID-19 cases.
“We saw that ifi27 is elevated in a step-wise manner with severe cases having high ifi27 levels.”
Dr. Kulasinghe said researchers knew that ifi27 was elevated in the blood of COVID-19 patients.
“But there had not been any evidence of where the signal for the high ifi27 levels was coming from.
“By spatial profiling lung tissue of COVID-19 patients who had died we got a much deeper picture of the cellular changes driven by viral infection and that the lungs were a source of the raised ifi27.
“This technique also allowed us to identify which cells in the lungs the virus was binding to.”
This collaborative research project with University of Queensland Diamantina Institute and the Walter and Eliza Hall Institute of Medical Research was awarded the Ausbiotech Johnson & Johnson Industry Excellence Collaboration Award and Industry Choice Award in 2020.
“Spatial Profiling of Lung SARS-CoV-2 and Influenza Virus Infection Dissects Virus-Specific Host Responses and Gene Signatures” is published on the MedRxiv pre-print server.
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that emerged in late 2019 has spread globally, causing a pandemic of respiratory illness designated coronavirus disease 2019 (COVID-19). Robust blood biomarkers that reflect tissue damage are urgently needed to better stratify and triage infected patients. Here, we use spatial transcriptomics to generate an in-depth picture of the pulmonary transcriptional landscape of COVID-19 (10 patients), pandemic H1N1 (pH1N1) influenza (5) and uninfected control patients (4). Host transcriptomics showed a significant upregulation of genes associated with inflammation, type I interferon production, coagulation and angiogenesis in the lungs of COVID-19 patients compared to non-infected controls. SARS-CoV-2 was non-uniformly distributed in lungs with few areas of high viral load and these were largely only associated with an increased type I interferon response. A very limited number of genes were differentially expressed between the lungs of influenza and COVID-19 patients. Specific interferon-associated genes (including IFI27) were identified as candidate novel biomarkers for COVID-19 differentiating this COVID-19 from influenza. Collectively, these data demonstrate that spatial transcriptomics is a powerful tool to identify novel gene signatures within tissues, offering new insights into the pathogenesis of SARS-COV-2 to aid in patient triage and treatment.
Severe COVID-19 is associated with a limited, differential transcriptome when compared with severe influenza
Only a limited number of genes were identified as differentially expressed between COVID-19 and pH1N1 influenza samples (2 downregulated and 4 upregulated genes, Fig. 6). Of the six genes significantly differentially regulated in COVID-19 samples, three were associated with the type I IFN response (LY6E, IFI27 & IFI6), two were heat shock protein family members (HSPA6 & HSPA1A) and one, CT45A1, when combined with various growth factors, is associated with cell survival and tumorigenesis. Interestingly, all three of the interferon-inducible genes were also significantly upregulated in COVID-19 samples compared to those without a viral infection (Extended Data Table 7). Two heat shock protein family members (HSPA1A and HSPA) were also identified as significantly down-regulated in COVID-19 tissues. These two proteins were found to be significantly upregulated in pH1N1 tissues compared to uninfected tissues (data not shown), suggesting a potential influenza specific signature. Together, these data indicate that pH1N1 influenza virus and SARS-CoV-2 induce only subtly different host pulmonary transcriptomes but may be potentially distinguished by a core disease-associated immune signature.
Understanding the biological functions, networks, and host-pathogen interactions that impact organ and disease development requires both cellular information and a spatial context. Analyses of bulk RNA sequencing or scRNA sequencing provides a global overview of an organ’s response to a pathogen, typically identifying broad inflammatory pathways.
However, such approaches fail to identify subtle individual cellular changes that are spatially distinct. This has particular impact when considering innate and adaptive immune cells that may be crucial for understanding pathogen-specific responses, and to distinguish the profile of one pathogen from another.
In contrast, spatially resolved transcriptomes of virally-infected tissues offers the possibility of disentangling the individual infected cells, contributions of viral load, cellular responses and hence patient-to-patient variability.
The clinical spectrum of COVID-19 is highly variable. Association of viral load with disease severity would provide a mechanism for early stratification of patients for treatment options. Indeed, analyses of nasopharyngeal swabs suggests that nasopharyngeal SARS-CoV-2 RNA is independently correlated with disease severity16, similar to earlier observations with SARS-CoV-117.
However, the association between viral replication in the lung and severe disease remains more complicated. Here, we observed that 8 out of 10 patients who died because of COVID-19 had no detectable viral RNA in lung biopsies as determined by RNAscope. Transcriptomic analysis showed that areas of high viral load were associated with a pronounced type I IFN response, consistent with other preliminary spatial analysis of the lungs of COVID-19 patients 11 and assumedly due to increased viral pathogen-associated molecular patterns (PAMPs) available for type I IFN stimulation. Nevertheless, even in samples where no viral RNA was detected, a pronounced type I IFN gene signature could still be observed in the pulmonary tissue.
These observations add weight to the growing understanding of the role type I IFNs in SARS-CoV-2 infections18. Current evidence suggests that an early and robust induction of type I interferons can help control viral replication and help ensure a mild infection.
In contrast, a delayed induction of type I IFN (i.e. after the peak of viral replication) is largely irrelevant for viral control as viral titres have already declined at this point in the infection but may help perpetuate the detrimental pro-inflammatory response and lung damage 18,19. The transcriptomic data presented here thus speak to the nuanced role of type I IFNs in SARS-CoV-2 infection.
Pulmonary transcriptomic analysis is a powerful tool to delineate the pathogenesis of SARS-CoV-2 and identify how it differs compared to other respiratory pathogens such as influenza virus. Previous studies have suggested that the lungs of COVID-19 patients display an increased incidence of alveolar capillary microthrombi and thrombosis with microangiopathy compared to those of influenza patients5.
In the present study, D-dimer levels were elevated in six of ten COVID-19 patients and some thrombosis was observed. Consistent with these observations, an earlier report using bulk lung RNA analyses to identify 69 differentially-expressed angiogenesis-related genes in COVID-19 patients, but not in influenza patients11.
Strikingly, while we observed that both COVID-19 and influenza patients had an upregulation of genes associated with coagulation and angiogenesis, these genes were not differentially expressed between the two virus-infected patient groups. These contrasting data point to the complementary role of blood profiling, bulk and tissue-specific spatial analyses in defining clinical parameters for therapy.
Accurate detection, timely diagnosis and effective treatment are all essential to the management of COVID-19 and will be instrumental in curbing viral spread as the number of new infections increase daily. Several potential clinical and transcriptional biomarkers for triaging patients with COVID-19 have been explored.
Clinical biomarkers include C-reactive protein, serum amyloid A, interleukin-6, lactate dehydrogenase, neutrophil-to-lymphocyte ratio, D-dimer, lymphocytes and platelet count 20,21. However, the majority of these studies have focussed on biomarkers in patient blood. While peripheral blood samples are a convenient and highly accessible site for clinical sampling, this is not the primary site of viral infection.
Thus, information delineated from analyses of blood may not reflect tissue severity and thus may hamper biomarker discovery. Furthermore, a biomarker that could be rapidly identified in the respiratory tissue (e.g. via a nasal swab) may be more accessible in low resource settings than one requiring a blood sample. Interestingly, previous analyses of IFI27 (interferon-alpha inducible protein 27) in blood of COVID-19 patients revealed that IFI27 was upregulated in SARS-CoV-2 infection22-24. Here, we identified IFI27 as differentiated upregulated in the lungs of both COVID-19 patients vs control patients and in the highly restricted set of genes differentially expressed between COVID-19 and influenza patients.
IFI27 has previously been used as a biomarker to successfully differentiate bacterial pneumonia from influenza virus infection25.
These data raise the exciting possibility that IFI27 may not only represent a biomarker for severe COVID-19 but that it may also help differentiate this disease from other clinically similar viral infections. This becomes particularly important as the ‘second wave’ of COVID-19 in the US and Europe is set to overlap with the winter influenza season. Validation of this gene in nasal specimens, as well associations with disease severity will be required to confirm IFI27 as a gene signature that is useful in stratifying COVID-19 patients.
This study was subject to several important limitations. Firstly, all data was derived from a small sample cohort derive from a single study site, and it remains to be determined how much these data can be extrapolated to other patient populations. Furthermore, additional studies are required across a broader range of patients (i.e. those with mild and moderate disease) to determine the therapeutic value of any of the putative tissue biomarkers identified herein.
However, despite these limitations these data reveal the unprecedented power of spatial profiling combined with detailed multiparameter bioinformatic analyses to dissect the key variables that contribute to differential gene expression across highly variable patient cohorts and the heterogeneous distribution of virus and immune responsiveness within tissues.
This study also demonstrated the value of using the suite of linear-modelling tools available in limma to interrogate the complex multi-factorial experiment design in this study. In particular, limma’s ability to model complex experimental designs and to implement information borrowing amongst genes to handle the relatively small number of samples in the panel, make it a highly attractive analysis tool for spatial profiling techniques with targeted gene panels.
The present study suggests that spatial profiling would present many advantages in analysing COVID-19 samples across different patient cohorts to identify fundamental response signatures distinct from background effects. This work therefore established the foundation to evaluate larger tissue cohorts to assess the relationship between blood and tissue biomarkers, disease progression, pathology and repair.
reference link: https://www.medrxiv.org/content/10.1101/2020.11.04.20225557v1.full
More information: Arutha Kulasinghe et al. Spatial Profiling of Lung SARS-CoV-2 and Influenza Virus Infection Dissects Virus-Specific Host Responses and Gene Signatures, (2020). DOI: 10.1101/2020.11.04.20225557