In May 2021, SCIREQ had the pleasure of hosting a virtual panel discussion with leading preclinical infectious disease researchers focusing on Preclinical Models of COVID-19.
Since the beginning of 2020, critical research efforts have been aimed at understanding the transmission, clinical manifestation, and pathogenesis of SARS-CoV-2 infections. There has been much progress in identifying relevant in vitro and in vivo modelling for this disease for vaccine development and therapeutic intervention.
Here are some highlights of what we learned during the panel discussion, as the experts below delve into the advantages and disadvantages of various COVID-19 animal models.
RECAPITULATING COVID-19 FEATURES IN PRECLINICAL MODELS: DIFFICULTIES AND CHALLENGES
While animal models have been very useful for recapitulating some of the pathophysiology of COVID-19, many clinical features seen in patients have been hard to reproduce in these preclinical models.
Coagulopathy is a big component of human clinical disease and one of the main features of COVID-19 that has been difficult to reproduce in animal models. It is unclear how well any of the animal models are tackling and mimicking that aspect of disease and it is possible that some of the lung dysfunction is secondary to the coagulopathy rather then the other way around. Viral hemorrhagic fever is one of Dr. Bailey’s primary research focuses and very little is currently understood about how that process works. The coagulopathy seen in COVID-19 is different then in other diseases like Ebola for example, and an additional problem is that the mouse has some changes in its coagulation cascade which makes it extremely resistant to developing coagulopathy. As such, studying it in mice becomes difficult and is one of the reasons why we have such a poor understanding of it.
Risk factors and comorbidities are another big challenge, as the characteristics of a human that develops severe COVID-19 are very complicated to reproduce in an animal model: how do we recapitulate decades of unhealthy lifestyle and aging in a mouse model?
Until it is understood why older people are so much more susceptible to severe disease, it makes it hard to study. The need for the mouse model is to be able to use it to anticipate the possibility that some mutation of the virus will make it more susceptible in younger people, which would have serious ramifications for the human population and for the studies in the mice. Understanding this mechanism of why it is happening in humans will allow researchers to develop a mouse model for that and bypass the challenge of replicating 40 years of risk factors.
Everything indicates that the mechanisms behind COVID-19 are multifactorial, which makes it hard to recapitulate in preclinical models. Hence the importance of understanding the underlying mechanisms. In the end, scientists should also always keep in mind that it may be the wrong risk factors that they are trying to model.
AGING MICE: THERE IS NO TIME MACHINE TO QUICKLY AGE MICE
Older patients are among the high-risk population for severe COVID-19 and studies show a clear age and sex dependent factor in disease severity and mortality with the mouse adapted virus and the humanized ACE2 k18 mice. However, obtaining aged mice from vendors for aging research is not an easy task: not many provide them and when they are available, they are costly. In example, 3–4-month-old mice are much easier to obtain then 6–8-month-old. If researchers choose to purchase relatively younger strain, there is no time machine to age these mice, and the only option is to wait.
Researchers believe there should be increased government funding for aging research. There needs to be more investment from the federal government to invest in breeding up stocks of aged mice. There are currently available stocks at NIH but these are difficult to obtain. Aged mice around 1,5 years old can also be purchased from the Jackson Laboratories but in “aging research”, these are still considered late-middle age.
FOCUSING ON THE HOST: WHAT SPECIFIC QUESTIONS ARE TRYING TO BE ADRESSED
Studying the pathophysiology of COVID-19 in animal models is certainly not an easy task. Although challenges are present, preclinical models still have utility and a lot of potential in terms of preclinical testing of counter measures like vaccines and anti-virals.
When diving into the physiology, researchers have to think carefully about the host itself and what are the questions they are going to ask, to make sure that the model is a suitable system to even address these. Questions and system have to be tailored to ensure that studies are meaningful. Many of the most important questions still remain to be addressed despite the various animal models currently available:
- Monkeys and hamsters are off the table for studying severe COVID-19
- The k18 mice that do get severe COVID-19 have a distribution of ACE2 receptors that is completely non-physiologic.
In the end, the utility of animal models depends on what questions are being asked and what the plan is. In the setting of vaccines and anti-virals in models that have accelerated progressive disease, even including brain infection which is not typical of humans, if you can use an anti viral or vaccine that will stop that infection in an experiment that takes a week, it remains a good model for addressing a set of very specific questions. Animal models can teach us some things at a reasonable timeframe even though they do not fully replicate exactly what happens in the human disease.
» No model is perfect, but they keep improving! «
Models are not perfect, and none are a photocopy of human disease. The current best mild disease model does not recapitulate the fatal progressive disease seen in the worst COVID-19 patients, but these are going through a period of evolution and refinement.
There is so much of the disease that appears to be unique to humans, including many patient-reported symptoms that are relatively rare and also not easy to measure in animal models, such as brain fog and COVID toes. In this case, clinical studies remain highly valuable and will help address questions that animal models cannot.
CHANGING THE VIRUS, NOT THE HOST
Mouse adapted viruses are now available thanks to the work of several groups, making it possible to probe genetically modified animal models to answer specific questions. With the mouse adapted virus, mice do no get brain infection but there is a very accelerated disease course: the virus replication peaks in the lung at day 1 to 2 and by day 7 the animals are moribund. This is not seen in humans and is an example of the built-in limitations of these particular mouse models.
However, mouse adapted viruses hold a lot of potential and as they become more popular, they will be useful to further study comorbidities. It is easier from a practical perspective to take a diabetic knock out mouse and infect it with a mouse adapted virus, than to try to age a k18 mouse or make it diabetic/obese. Having these mouse adapted viruses is going to help study the role of particular genes, as researchers can infect knock out mice strains without having to cross them with k18 mice.
It is also important to remember that while these current models are not perfect, they are already much better when compared to what was available last year. Wether studying models of obesity or other chronic inflammatory diseases, scientists have to acknowledge the limitations and the specific applications the models are best suited for.
VARIANTS: THEY ALSO HOLD POTENTIAL AND OPPORTUNITIES
New variants following mutations of the virus have certainly been a matter of attention and fear since the emergence of COVID-19. However, with vaccination rolling out around the world, some argue that there will be no patients left to recruit for clinical trials or other studies. Researchers are afraid we will see a repetition of previous events like MERS or SARV-CoV, where research slowly died as the virus was not a matter of public concern anymore. However, according to scientists, these variants will help keep the research alive: evidence indicates that the emergence of these variants at some point will escape vaccine efficacy and protection, and we may end up having new vaccine roll outs every 3 or 4 years.
COVID-19 RESEARCH GRANT SUBMISSION: WHAT REVIEWERS ARE LOOKING FOR
At the beginning of last year, preliminary data was not a must yet, and grant reviewers mostly looked at the ideas behind submissions. Things have now evolved and reviewers are not forgiving of people who have no COVID-19 preliminary data. Below is some advice from the panelists for anyone looking into joining COVID-19 research:
- Preliminary Data: If no preliminary data is available, researchers should show that they can have access to these models.
- Study Relevance: Researchers must also prove that the study is relevant and will actually address an important question, rather then simply characterize a strange phenomenon observed in the animal model. It is important to make sure that the model recapitulates the aspect of disease that is being addressed and that it can be characterized in a relevant way. Making a convincing case that the model proposed is able to meet the demands of the questions being asked. That is where some preliminary data and experience with the model will pay off.
- Pathophysiological Readouts: Cytokine panels, RNAseq, weight loss and histopathology are not sufficient, and these are not readouts that can be measured in humans. Further, histopathology is not particularly predictive of human disease. Pathophysiological readouts of lung injury become crucial and add a lot of value, such as measuring preclinical lung function decline in SARS-CoV-2 Models with the flexiVent. These are highly clinically translatable and bridge the gap/disconnect there is between mouse and human data.
- BSL3 Considerations: Researchers mut be aware of the constraints of working in a BSL3 facility. A lot of experiments that can be performed in a BSL2 are simply not feasible inside the BSL3 lab. Many facilities do not have sophisticated equipment that can enter and live in the BSL2 and some samples obtained in a BSL3 cannot leave the lab.
The final advice and general consensus is: if you do not have access to a BSL3, find a collaborator who does!
LEARN MORE ABOUT THE PANELISTS:
PAUL MCCRAY – UNIVERSITY OF IOWA
Dr. McCray’s research interests include airway epithelial cell biology ad COPD pathogenesis and treatment, host-pathogen interactions at the airway surface, and responses to bacterial and viral respiratory pathogens. Learn more about Dr. McCray’s research here.
WAYNE MITZNER – JOHN’S HOPKINS UNIVERSITY
Dr. Mitzner’s primary research interests include the structural basis of physiologic lung function and how pathologic situations and environmental exposures can change lung function and structure. Learn more about the Mitzner lab here.
ADAM BAILEY – WASHINGTON UNIVERSITY IN ST. LOUIS
Dr. Bailey’s lab focuses on furthering infectious disease research by developing animal models to study viral disease pathophysiology and evaluate the effectiveness of vaccines and new therapeutics. Learn more about Dr. Bailey here.
IAN DAVIS – OHIO STATE UNIVERSITY
Dr. Davis’s research focuses on the pathophysiologic effects of pulmonary viral infections. More specifically, this group studies the impact of influenza A infections on alveolar epithelial cell function in a mouse model of influenza-induced ARDS. Learn more about the Davis lab here.