The Main Protease of SARS-CoV2: Finding drug candidates with X-rays

“With the ending –ase you can always assume that it is a really bad protein that breaks something else.”

My cell biology professor in the first semester of my biology degree

As the study of biology progresses, one naturally learns that the above statement – quoted freely from my memory – is not always entirely true. But in fact, most proteins that end in –ase are those that can break down something else. What they split is usually indicated by the syllable in front of it. A protease thus is a protein that can cleave (other) proteins. So does the Coronavirus have a protease to break down the proteins in our cells? No, because at least the main protease of the coronavirus. main protease, also known as MPro for short, cleaves the viruses own proteins. And why this is so important for the coronavirus that MPro is being intensively researched as a potential target for drugs against Covid-19, is what I would like to explain in this article.

The genome of SARS-CoV2

Viruses are usually classified according to the type of their genome. There are DNA and RNA viruses, with coronaviruses belonging to RNA viruses. More precisely, they belong to the single-stranded RNA viruses with positive polarity, +ssRNA viruses for short. So let’s take a look at what is written on this single RNA strand of the coronavirus:

Reproduced from Alanagreh, L.; Alzoughool, F.; Atoum, M. The Human Coronavirus Disease COVID-19: Its Origin, Characteristics, and Insights into Potential Drugs and Its Mechanisms. Pathogens 2020, 9, 331,, under Creative Commons Attribution (CC BY) license.

All proteins which are needed for the coronavirus to multiply in host cells are encoded on this RNA. The proteins that contribute to the structure of the virus are encoded in the back part of the virus’ RNA. This includes, for example, the famous spike protein (S), with which the virus docks to the host cells (to ACE2 to be exact), and whose blueprint is given to us with the RNA vaccines from Pfizer/Biontech or Moderna.

In the front part the so-called non-structural proteins (NSP) are encoded, which are simply numbered from 1 to 16. These proteins are no longer present in the mature, circulating virus. But they are fundamentally important to get the host cell to produce new viruses. This article will deal with the fifth non-structural protein, NSP5, the main protease (MPro), which is sometimes also called 3CL protease.

Of slippery RNA and stumbling Ribosomes

While researching this article, I came across a mindblowing mechanism that I hadn’t heard of before. Most of my readers already know that proteins are built by the ribosome. Ribosomes scan along the RNA and add a specific amino acid into the growing protein for each specific triplet on the RNA. Thus, the DNA and RNA that encodes for a protein must actually always consist of at least 3 times as many nucleotides (A, C, T, G) as the number of amino acids that make up the protein.

Reproduced from Bhatt, Pramod R.; Scaiola, Alain; Loughran, Gary; Leibundgut, Marc; Kratzel, Annika; McMillan, Angus et al. (2020): Structural basis of ribosomal frameshifting during translation of the SARS-CoV-2 RNA genome. In: bioRxiv, 2020.10.26.355099, DOI: 10.1101/2020.10.26.355099, under Creative Commons Attribution (CC BY) license.


However, there is great evolutionary pressure on viruses to keep their genome as small as possible. The coronavirus genome therefore uses a trick that can also be found in some other viruses. At one point, shortly after NSP10, the following RNA sequence segment is found: UUU UUA AAC. The spaces indicate the reading frame used up to that point. For the ribosome this sequence section is quite “slippery”. Here, it often gets tangled up and stumbles into a new reading frame. This process is also called ribosomal frameshifting. From this point on there are two variants of amino acid chains, thus two different proteins. One version of the protein (where the frame continues with GGG = glycine, UUU = phenylalanine and so on) will give rise to NSP11. The other version (continuing with CGG = arginine, GUU = valine and so one) will give rise to NSP12-16.

But this only as a side note and without further details, because the main protease, NSP5, is in fact encoded before this section.

The function of the main protease

First of all, I always find it remarkable that there are proteins that arise as part of a long amino acid chain and then fold within this chain in such a way that they can become biologically active in order to cut themselves out of this chain. MPro is exactly such a “self-liberating” protein, which you can see schematically in the cool video at the very end of this article. But that’s not all: after MPro has freed itself from the amino acid chain, it cuts the chain at 9 other sites (see the little blue triangles in the picture below). Only then can the NSPs 6-16 really start their work, namely the multiplication of virus components. If the main protease could be disrupted in its function, one should put a stop to virus replication. A substance that inhibits MPro could therefore be a much-needed drug for the treatment of acute Covid-19 infections.

Searching for drug candidates using X-rays

In order to identify inhibitors for proteins, one should first know exactly the structure of the target protein. X-rays have been used quite successfully for decades to determine the structure of proteins. To do this, the purified proteins must first be crystallized in large quantities. This crystal is now exposed to very high-energy X-ray light, the wavelength of which roughly corresponds to the atomic distances in the crystal. This then acts as a diffraction grating. The pattern of X-rays captured after the scattering on a detector allows the structure of the protein to be reconstructed. The structures of all proteins of the coronavirus have already been described in this way, even if these structures are constantly being refined and improved with great care.

Reproduced from Lubin, Joseph H.; Zardecki, Christine; Dolan, Elliott M.; Lu, Changpeng; Shen, Zhuofan; Dutta, Shuchismita et al. (2020): Evolution of the SARS-CoV-2 proteome in three dimensions (3D) during the first six months of the COVID-19 pandemic. In: bioRxiv, 2020.12.01.406637. DOI: 10.1101/2020.12.01.406637. With kind permission from Stephen Burley.

But where does this strong X-ray light come from? Nowadays this is mostly generated using synchrotron light sources.

PETRA III, a Synchrotron light source at DESY

The German Electron Synchrotron (DESY) is located in the west of Hamburg since the early 1960s. A number of elementary particles were discovered here in the 1970s, including my favorite particle, the gluon. But this is another story.

Since there was less and less to add to the particle zoo over the past few decades, the PETRA ring accelerator has been operated as a synchrotron light source. Here, groups of electrons are accelerated on a so called storage ring to almost the speed of light with the help of electric fields in cavity resonators. Large magnets constantly send them along their circular path: at PETRA they circulate 130.000 times per second around the 2.3km long accelerator. ALong their path the electorns are send through so-called undulators, where magnetic fields in alternating directions make them oscillate. During this, they emit the much sought-after intensive X-ray light, which is tangetially led from the ring to recording stations. We are currently in the third technical generation of PETRA: PETRA III. Over the next few years, however, PETRA III is to be further developed into PETRA IV, and thus the most powerful synchrotron light source in the world.

Left: An aerial photo of the west of Hamburg shows where the mostly underground accelerators of DESY and the European XFEL are located. © DESY / R. Schaaf und D. Schröder. Right: In the beamlines, the X-ray light is guided tangentially from the PETRA III ring to the measuring stations. © DESY / illustrato.

Along the PETRA III ring lie the so-called beamlines and hutches, on which the researchers can take their measurements. And it was in three of those beamlines, P11, P13 and P14 to be precise, that an international research team led by Alke Meents at DESY held its samples in the X-ray beam. However, these samples were not just crystals of the main protease alone. Instead, MPro was crystallized together in almost 6000 different reaction vessels, each with a potentially active substance. For this purpose, so-called chemical libraries were used, which only contain substances that have already been approved or are currently tested as medicinal substances for other diseases. In the event of one or more hits, this would significantly accelerate the approval of such a substance for the treatment of Covid-19, since all of these substances have already been tested for their basic tolerability.

Calpeptin and Pelitinib as potential drugs against Covid-19

In this way, 37 active substances were identified that can bind to the main protease. In a next step, the Bernhard Nocht Institute for Tropical Medicine in Hamburg tested whether these active ingredients could slow down virus replication in cultivated cells. After all, this was the case for 7 substances. Two of those, calpeptin and pelitinib, showed such a strong inhibitory effect that they are now being investigated in further preclinical studies.

Upper left: DESY researcher Alke Meents during the drug screenings at PETRA III beamline P11. Upper right: The sample holder (left) at measuring station P11 rotates the sample in the X-ray light from the PETRA III accelerator. Meanwhile, a stream of cold nitrogen cools the sample. With the help of many scattering images from different directions, the molecular structure of the protein-drug complex can be calculated. Lower left:
In the control hutch of the PETRA III beamline P11, DESY researcher Wiebke Ewert shows on a so-called electron density map where a drug candidate (green) binds to the main protease of the corona virus (blue). Lower right: Surface representation of the corona main protease Mpro with bound active substance calpeptin (yellow). © 2021 DESY/C. Schmid

These promising results, for which researchers from many other research institutions, such as the University of Hamburg, the EMBL Hamburg and the European XFEL, also contributed, were recently published in Science.

By the way, DESY explained the research behind their article also in this cool video. Have fun with it!

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *

This website is using cookies to improve the user-friendliness. You agree by using the website further. Privacy policy