Enter Flavonoids

How a novel approach to discovering and trialing plant-based antivirals is being attempted in Kenya to fight the pandemic

EMSKE Phytochem
An Idea (by Ingenious Piece)
9 min readSep 28, 2020


(Prev article #4 of 5) Note: this is the 5th of 5 articles in our series on flavonoids.

vImportant sources of key flavonoids — Citrus × sinensis (sweet orange), mangifera indica (common mango), menthe piperita (common peppermint)

As we’re preparing for engaging here in Kenya to champion a distributed clinical trial of key flavonoids against covid, I think it’s an appropriate time to discuss the Big Picture in what we are doing.

What is a flavonoid?

When I first learned the word ‘flavonoid’, I thought it must have to do with flavoring. In fact these similar-sounding names are largely a coincidence, as the prefix for flavonoid comes from ‘flavia’, or ‘blond’ / flax-like / yellowish — the color that several common flavonoids take in their isolated powderized form.

From Ali et al 2019 (CC by 4.0)

So let’s get into the nitty gritty:

When you start looking at compounds that come from plants, several classes of compounds routinely present themselves: Saponins, alkaloids, and sesquiterpenoids to name just a few. One class that very frequently comes up are the flavonoids.

Flavonoids are a class of a couple thousand identified compounds that follow this basic structure (flavones specifically shown):

flavone backbone structure [image: public domain]

In the plants that flavonoids turn up in, they typically present themselves most prevalently in the skins of fruit, in seeds, and in flower buds. In my review of phytochemical literature, it’s rare to see any plant part with more than 10% dry weight mass of flavonoids, and usually closer to 1% or less.

Why do plants produce flavonoids? Among other roles, there is broad appreciation of flavonoids’ antimicrobial activity in plants. Only over the past 30 years though has the plant biology literature begun to recognize that flavonoids form a key part of plants’ defenses against plant viruses.

Here’s the example of our primary go-to flavonoid for trialing, hesperidin:

The hesperidin molecule [image: public domain]

Note the similar structure on the bottom right matching the “framework” shown just earlier. What are the rings on the left side? Those are sugars — and crucial for augmenting the molecule’s binding action as they help wrap around the enzyme target’s structure. Two of them together like this are called a disaccharide. The particular double-sugar shown is the joining of rhamnose and glucose sugar rings, and is called rutinose.

What could *flavonoids* have to do with COVID-19?

If you’ve been following our other articles, you know that we like to simulate the binding of plant-derived compounds to the key enzyme of COVID-19 virus (SARS-CoV-2)’s that serves as its “seamstress”. Remember: the more you can bind up the seamstress’ scissors, the harder you make her job to cut new viruses to shape. But why were we motivated to look at plant compounds at all?

Our work began with this February 2020 paper in F1000Research (a research publishing house supported by Bill & Melinda Gates Foundation) from Hong Kong Polytechnic University (and whose first author, Yu Wai Chen, is now one of our closest technical advisors):

Chen’s team screened over 7000 purchasable drugs on the main protease. Which compounds turned up as the tip-top performers? The flavonoids hesperidin and diosmin.

Returning readers will also remember from our (technical!) calibration article that biomedical researchers across the world assayed flavonoids on the SARS protease between 2004 right up until 2019 before the COVID-19 (SARS-CoV-2) pandemic:

Broadly, two classes of compounds were identified to perform well on the assays in vitro. These are,

  1. flavonoids, (which come from land-based plants), and
  2. phlorotannins (which come from sea plants / brown algae — i.e. seaweed )

Purely for ease of supply chain due to our location in Africa, we focus on the land-based plant sources — — flavonoids.

Our in silico findings

Benefiting from the calibration referenced above, and shining a spotlight on nearly 300 flavonoids and other plant-based compounds, we ran our assay and came to the results that the following compounds would most likely inhibit the SARS-CoV-2’s Main protease in a cell-free in vitro assay:

projected IC50 concentration of 295 primarily plant-derived compounds (units: micromolar; source: EMSKE Phytochem)

In the above chart, the lower the IC50 value, the better. Why? Because IC50 describes the concentration of the compound required to inhibit 50% of the virus’ protease activity. The lower the concentration of drug required, these less you needs to consume to be effective, and the less likely a drug is to effect anything else in the body that you don’t want it to. Zooming in on the top 5% performing compounds from the far left side:

projected IC50 concentration — top inhibiting compounds

As you can see, hesperidin remains among the top performers. (Some top performers like hinokiflavone get ruled out for cell toxicity or sulcatone A due to insufficient supply of known source plants). Here’s what hesperidin itself looks like in its highest binding pose to the SARS-CoV-2 Main protease:

top view of hesperidin binding within the COVID-19 virus’ Main Protease— tunnel (invisible) runs left-to-right; Can you see the hesperidin structure (in yellow) matching the chemical diagram above? [product credit: PyMol]

The flavonoid sits in the tunnel, or ‘cleft’, formed by the joining between the two symmetrical sides of the protease. And by the way, the highest-binding poses look similar to the above graphic for all of the high-performing flavonoids we’ve assayed. Now, look at the same pose but from a different viewpoint:

Front view of hesperidin binding in Main Protease (tunnel seen head-on in progressive un-cutaway shots). Can you see just the edge of the bright yellow hesperidin molecule inside the tunnel? [product credit: PyMol]

(Note the tell-tale heart-shape in this projection of covid’s main protease). This view is called a ‘surface’ view. It is useful for visualizing the influence of charges throughout the Main Protease. We see a colored surface representing each surface atom’s field’s ability to influence on molecules around them — molecules such as the bright yellow hesperidin buried into that tunnel. Here we are looking head-on into the ‘tunnel’ formed by Mpro’s structure.

What is the implication of this picture? While the in silico results indicate this compound binds strongly to the protease, the 3D visualization demonstrates that it doesn’t just stick randomly to the side of the protease. Rather, it binds the protease in a very central way: the seamstress’ scissors are gummed up together and are unlikely capable of cutting anything — especially new viruses.

Where are we sourcing these compounds from?

Mango seed kernel — [credit: WalterReeves.com]

While flavonoids such as hesperidin can be found in many, many, plants throughout the world, we focus on three strategic sources for the purpose of a trial. The 1st source is mango seed kernel extract (MSKE).

[from: Abdel-Aty et al 2018]

In addition to containing 3% hesperidin by dry weight, it also contains 1% tannic acid which performed as strongly through the in silico assay.

Mangos are easily sourced in sub-tropical regions of the world. India and China are the largest exporters, and yes, sub-Saharan Africa could position itself as a major exporter as well.

The 2nd source is simply the pre-existing medication marketed (outside of the US) called Daflon which is a 90% hesperidin & 10% diosmin formulation. In the context of a clinical trial, for either of these two sources, the key would be getting these administered early in the course of a patient’s infection.

[from: Chashitsu LaSere CC BY-NC-SA 2.0]

For patients where the infection has unfortunately progressed to the lungs, we propose clinical trials of the 3rd source: peppermint extract as an inhaled therapeutic.

[from: Srok et al 2014]

Decocted peppermint extract is rich in high-performing flavonoids: eriocitrin, luteolin-7-O-rutinoside, and hesperidin. Pep extract is interesting strictly as an inhaled therapeutic — otherwise the rosmarinic acid can lead to an upset stomach.

Peppermint propagates easily and grows quickly, reaching 6 times its volume every two months. And a tiny amount of it goes a long way: Yielding out at roughly 500 mg of dry weight per nebulized dose, existing plantations cultivated for the packaged tea market can easily handle demand.

But again we’re talking about catching the lung infection during its viraemia period. If it’s already progressed to the infamous cytokine storm, then anybody’s guess still on the most appropriate cytokine storm-alleviating therapeutic.

Are there other viable sources?

Of course — for northern hemisphere markets the traditional sources of hesperidin will be citrus peels. (Orange, lime, lemon, and grapefruit).

What’s the catch?

Indeed, no panaceas here — Among the very same flavonoids we focus on the most — (hesperidin and naringin specifically, but undoubtedly others that bind strongly in silico as well) are those that are already known to bind and inhibit one of several human enzymes, OATP1A2, that the body uses to breakdown other drugs for common chronic conditions:

Source: FDA.gov

The result being that a patient taking say, Lipitor, ends up getting a much higher apparent dose than their doctor (and Pfizer) ever intended. Same too if they take an ingestible diabetes, heart arrhythmia, or blood-thinner type of drug.

Then how to go about patients that take ongoing medications for such chronic conditions? What if they ever came down with covid and needed these compounds as a therapeutic? Well in that case their primary care provider under a pharmacologist’s guidance becomes an important stakeholder for determining how best to prescribe these compounds vis-a-vis their existing medication.

What’s next?

[Licensed under CC BY-NC 2.0 from GSK]

With the support of a local plant medicinals manufacturer here in Kenya, now we’re focused on clinical trialing. Critically, we’ll be trialing using a distributed clinical trial model that spans multiple clinical sites around the world. In the days ahead (Sep 2020) we’ll be hosting our flavonoid-covid trial protocol, FLAVOCOV, on the Pan-Africa Clinical Trials Database. Update 15-Oct: Our trial registration is now live on the PACTR database at: https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=12365

Do you want to see plant medicinals trialed against covid and want to know how to help? If you know clinical decision-makers at a research hospital near you, then we absolutely want to talk with you. Reach out to us at clinicaltrials@emske-phytochem.com.


Probably one of the most pernicious questions we’ve had to continuously ask ourselves is: If biomedical researchers had assayed drugs on the SARS protease to success since 2005, why didn’t they initiated clinical trials of these compounds during the COVID-19 pandemic?

When I’ve surveyed friends in life sciences, they impress upon me that researchers’ goals in life practically speaking are just to publish and hope that another entity picks up their work (and especially, cites it). They’re not going to pursue clinical trialing, as they’re too busy fundraising for experimental work leading to their next academic paper. While it may be a difficult posture to fathom in the context of an economy-freezing pandemic, it is understandable nonetheless.

Update 2 Jan 2021

Pleased to see that a flavonoid glycoside, rhoifolin (having been reported in the graphic above) was found by a lab in Germany to inhibit the SARS-CoV-2 main protease in vitro with better than 20uM IC50 efficacy. To our knowledge this is the first time a flavonoid glycoside has been assayed in vitro on the SC2 main protease, and we’re pleased that the results align with our model’s prediction.

title images credits: citrus x sinensis [lulucmy |unsplash]; mangifera indica [Dinesh Valke | CC BY-SA 2.0]; mentha piperita [tillwe | CC BY-SA 2.0]



EMSKE Phytochem
An Idea (by Ingenious Piece)

Coming from a multidisciplinary technical background in Silicon Valley, SE Asia, & East Africa, the author builds awareness of plant medicines. Tw: @EMSKEPhyto