First Molecular Profiles of Severe COVID-19 Infections
Posted on by Dr. Francis Collins
To ensure that people with coronavirus disease 2019 (COVID-19) get the care they need, it would help if a simple blood test could predict early on which patients are most likely to progress to severe and life-threatening illness—and which are more likely to recover without much need for medical intervention. Now, researchers have provided some of the first evidence that such a test might be possible.
This tantalizing possibility comes from a study reported recently in the journal Cell. In this study, researchers took blood samples from people with mild to severe COVID-19 and analyzed them for nearly 2,000 proteins and metabolites [1]. Their detailed analyses turned up hundreds of molecular changes in blood that differentiated milder COVID-19 symptoms from more severe illness. What’s more, they found that they could train a computer to use the most informative of the proteins and predict the disease severity with a high degree of accuracy.
The findings come from the lab of Tiannan Guo, Westlake University, Zhejiang Province, China. His team recognized that, while we’ve learned a lot about the clinical symptoms of COVID-19 and the spread of the illness around the world, much less is known about the condition’s underlying molecular features. It also remains mysterious what distinguishes the 80 percent of symptomatic infected people who recover with little to no need for medical care from the other 20 percent, who suffer from much more serious illness, including respiratory distress requiring oxygen or even more significant medical interventions.
In search of clues, Guo and colleagues analyzed hundreds of molecular changes in blood samples collected from 53 healthy people and 46 people with COVID-19, including 21 with severe disease involving respiratory distress and decreased blood-oxygen levels. Their studies turned up more than 470 proteins and metabolites that differed in people with COVID-19 compared to healthy people. Of those, levels of about 300 were associated with disease severity.
Further analysis revealed that the majority of proteins and metabolites on the list are associated with the suppression or dysregulation of one of three biological processes. Two processes are related to the immune system, including early immune responses and the function of particular scavenging immune cells called macrophages. The third relates to the function of platelets, which are sticky, disc-shaped cell fragments that play an essential role in blood clotting. Such biological insights might help pave the way for potentially effective new ways to treat COVID-19 down the road.
Next, the researchers turned to “machine learning” to explore the possibility that such molecular changes also might be used to predict mild versus severe COVID-19. Machine learning involves the use of computers to discern patterns, or molecular signatures, in large data sets that a human being couldn’t readily pick out. In this case, the question was whether the computer could “learn” to tell the difference between mild and severe COVID-19 based on molecular data alone.
Their analyses showed that a computer, once trained, could differentiate mild and severe COVID-19 based on just 22 proteins and 7 metabolites. Their model correctly classified all but one person in the original training set, for an accuracy of about 94 percent. And importantly, in further prospective validation tests, they confirmed that this model accurately identified mild versus severe COVID-19 in most cases.
While these findings are certainly encouraging, there’s much more work to do. It will be important to explore these molecular signatures in many more people. It also will be critical to find out how early in the course of the disease such telltale signatures arise. While we await those answers, I find encouragement in all that we’re learning—and will continue to learn—about COVID-19 each day.
Reference:
[1] Proteomic and metabolomic characterization of COVID-19 patient sera. Shen B et al. Cell. 28 May 2020. [Epub ahead of publication]
Links:
Coronavirus (COVID-19) (NIH)
Blood Tests (National Heart, Lung, and Blood Institute/NIH)
Tiannan Guo Lab (Westlake University, Zhejiang Province, China)
Very interesting. Cant wait for the sample size to be increase so we can say with a certain degree of assurance that those machines can predict and that the proteins themselves are telling us what they tell now.
I am reminded of the old saying–“Can’t see the forest for the trees”. The correlation of severity is entirely based on sequential ordering by blood type and the level of Vitamin D in their blood cells!!!
The unfavorable death rate in blacks is due to both the lack of sun in Northern cities in Fall, Winter and Spring compounded by the shielding skin tone that prevents maximum vitamin D creation!!! All the other studies mentioned are focused on EFFECTS not CAUSES. Only the Chinese could be so obtuse and incorrectly targeted to miss this at the molecular level.
Study the blood type and vitamin D levels in the one test that fell outside proven parameters for either unique blood type or high or low vitamin D levels to explain the anomaly.
It’s good to hear that such progress is made.
This is an amazing scientific advance considering this disease was unknown in humans seven months ago. The results are interesting, narrowing it down to 22 proteins and 7 metabolites that could presumably be tested easily. But before this becomes useful as a screening tool predicting who advances to severe disease, there have to effective treatments that are actionable results of such testing. Otherwise, why screen for a disease that you can’t do anything about. Perhaps more important than its immediate usefulness as a screening tool, these results provide a wealth of knowledge about which molecules go awry in Covid-19. Hopefully, some of these will become drug and treatment targets which get to the heart of how this disease harms the body.
The advantage of knowing if a person was at risk for developing a severe case of Covid 19 would be that they could be very self protective and shelter in place. Others could shop for them until we have herd immunity and it dies down. High risk people could be tested first. People with lower risk could get back to work.
Another approach would be to use the person’s stool and determine the differences in the gut Microbiome. Diet and lifestyle are related factors both of which would effect blood proteins. This approach could potentially lead to the elimination of future pandemics.
It sounds like the prospective testing failed to reach high percentages, so it is probably pointless to read it. The inverse test set would be much more interesting. It would be much better to test for proteins and other chemicals, particularly lipids, which predict non-infection, and then seek for means to enhance predictors.
Actually, I can already tell you. Hugs and kisses, the proteins enhanced by hugs and the 735 lipid oddballs taken in by kissing will likely be rare among those dying of COVID-19. Pheromones, caused to be secreted by conditions experienced by populations, regulate immunity and thereby fertility to accord population numbers to available resources. This means that providing patients by mouth less than 1 gram of healthy adult male facial skin surface lipid pheromone (which alleviates symptoms of ADHD, ADD, OCD, PTSD, CD, Borderline Personality Disorder, Generalized Anxiety Disorder, opioid addiction, methamphetamine addiction, sexual perversions, racism, antisemitism, homophobia, criminal behavior, delinquency, Alzheimer’s Disease, Tourette’s Syndrome, Hashimoto’s Thyroiditis, other autoimmune disease and cancer) cleansed of aversive sentinel sub-pheromone volatiles (which cause hatred, revulsion, etc.) will diminish hospitalizations, intubations, and deaths due to COVID-19.
How would you rule out whether the proteins and lipids will be present if there is latent H. simplex, or a small eruption on the way, to meet the criteria for healthy male lip? Do the molecules exist outside the kiss bio-film, or does the one gram dose, for practical reasons, include generalized oral cavity slobber?
If a man kisses his baby or teenage son on closed lips, will the correct substance transfer to do any prevention?
This is an impressive use of state-of-the-art technology such as computer learning to search for and uncover differences between two important sample sets (severe disease and death versus mild to no disease). However, a blind search will miss significant differences if it is not guided by a working hypothesis that looks at function. A molecule or metabolite may be present in similar amounts in both sample sets but in forms that function differently in each set to yield completely different outcomes in disease severity. A hint on functionality usually emerges from the study of natural experiments that lead to the selection and global distribution of the associated variants. This was true for falciparum malaria and innate resistance traits such as the sickle hemoglobin. In the case of COVID-19 infections, the correlation between the geo-location of the Duffy blood group phenotypes – Fy(a+b+), Fy(a-b+), Fy(a+b-) and Fy(a-b-) and severity of infection and death is striking and cannot be ignored. For example, the epicenter for COVID-19 deaths (Western Europe) is also the epicenter for the FY*B and the epicenter of Fy (a-b-), sub-Saharan Africa, remarkably shows the least amount of severe disease and death. This makes the Duffy blood group antigens prime putative candidates for the elucidation of the pathogenesis of COVID -19 infections and urgently needs further investigation. A combination of these powerful computer-based molecular tools with a simple working hypothesis may point the way to the new tests and treatment strategies that are so desperately needed.
A scientifically tantalizing Covid-19 molecular-signature innovative research study in specific well-defined genetically distinct population subset with appreciable subgroup-stratification of negative vs mild vs severe positive Covid-19 patient-subsets; however, the relatively low study statistical power could lead to skewed data-trends and statistically significant outputs in the mild vs severe subsets. Cases and age-/ethnicity-matched control sample-numbers should be larger in future multicentric studies with pooled clinical samples from populations of varying genetic landscapes for eventually drawing meaningful patient-friendly immunotherapeutically significant end-points in terms of cost-effective clinical management of mild vs severe Covid-19 positive patients worldwide. Further, the sophisticated computer-based medical-information-technology based learning protocol with protein molecular regulatory signatures appear pharmacologically attractive for future design of Covid-19 molecular-signature genetic-screens/pharmacological scaffolds with ethnicity and gender-specific stratification of Covid-19 susceptible “at-risk” population(s) on a global platform.
With my expertise in multidisciplinary medical research spanning diverse clinically-significant specialities as demonstrated in my competitive publication-track record till date, I certainly gained critical research insights in the overwhelming and ever-expanding Covid-19 disease pathogenesis area, and wish to congratulate my American colleagues and contemporaries for propeling the Covid-19 research area for eventual patient-friendly, safe, cost-effective outcomes with public health impact.
As a scientifically logical research strategy, long-term ethnicity-specific genetic-association studies with Single Nucleotide Polymorphisms with pharmacogenetics/genomics, case-control robust study-designs should be encouraged at NIH, USA, and collaborating sites globally for developing predictive and prognostic biomarkers in Covid-19 clinical management!
Glad to see positive progress is being made!
yes, the gut is the harbinger of ACE2 receptors.. whereever can be found epithelial lining, one can find ACE-2-receptors.
the infection involves ACE-2 .. but the infection is not the issue, everybody gets the flu,, a viral infection.. does not by itself make you sick.. Instead it is your bodies own response to the infection that distributes the flu systems throughout your body.. unfortunately certain genes are located in your epithelia lining of your various organs which respond to virus transfected cells.
It is here in these responses, that things are happening which lead to death. Try RASV12, it is a gene located in the short arm of chromosome II from the base bp 522,241 to 525549 and other better known genes in different chromosomes like maybe ROS, SRC, and Erb2.
The problem of the virus is not a medical problem it is a molecular biology problem that is why medicine has had such a difficult time with it.
Re “Proteomic and Metabolomic Characterization of COVID-19 Patient Sera”, the authors mention that “Corticosteroid treatment was effective in suppressing MERS-CoV and SARS-CoV but showed negligible effect on COVID-19 patients and may even have induced lung injury”.
Quite an interesting difference, I thought. However I followed the citation re MERS (Arabi et al), and found this, instead:
“Corticosteroid therapy in patients with MERS was not associated with a difference in mortality after adjustment for time-varying confounders but was associated with delayed MERS coronavirus RNA clearance.”
Further reading of citations, the authors cite Russel et al (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7134694/), that corticosteroids “showed negligible effect on COVID-19 patients and may even have induced lung injury.” Actually, that paper surveys the past use of corticosteroids in SARS and other viral respiratory infections, and concludes that a benefit is not at all likely and harm is possible. But it includes no clinical data re COVID-19.
That paper, “Clinical evidence does not support corticosteroid treatment for 2019-nCoV lung injury” perhaps should have been titled “PAST Clinical evidence does not support corticosteroid treatment for 2019-nCoV lung injury.”
We should pay more attention to asymptomatic infected people who recover with little or no need for medical care. I wonder is there any interference between covıd 19 and other viruses? Also, what were the CD4-cell levels among the infected patients?
Hey, really amazing article, this to so informative, I’m loving it …
so would platelets imply a coagulopathy issue (like suggested by low tech clinical observation?)?
My money is on hydrogen sulfide (not detected as a metabolite), an inhibitor of the NLRP3 inflammasome, whose depletion will promote a cytokine storm. Note the Greek study correlating lower serum H2S levels with adverse outcomes.
Can we please do a research on AAT protein count comparing severe, mild and non symptomatic people. Why?? It keeps neutrophil elastase in check and is a general protease inhibitor that can potentially inhibit TMPRSS-2.
AAT is produced in the liver and lymphocyte and monocyte (two of the three suppressions mentioned in this research).
Just look at AAT and severe CVOID, relation is striking!
It’s said that AAT is rare, maybe true, but most people don’t have symptoms so how much is really known?? Look at the numbers of severe CVOD-19 cases and you might just hypothesize… but a study should be done on AAT relation to Covid-19 …
It was a good experience reading your blog. Thanks for sharing with us.