A Chinese gain-of-function study with 100% lethality?

Insert obligatory fear mongering taglines.

rl;dr: a new Chinese study looking at a mutant pangolin coronavirus has garnered a lot of attention with it’s pronouncement of “100% lethality”. The claim comes with serious flaws and is heavily confounded. Although this study itself warrants critiques, there are still questions over the widespread utility of gain-of-function research.

---

“Urgent!” “Breaking!”, and “whatever words are will garner your attention and make you click in fear!”

A preprint study published in early January has made an absolutely scary finding- that a lab-mutated coronavirus derived from pangolins appear to kill all of the transgenic mice, possibly leading to the death of all of us!

I mean, if the Daily Mail has headlines such as these then surely we are all at risk of this new virus:

From Daily Mail

But is this study really all that it’s hyped up to be, or is it a sensationalist, clickbaity coverage intended to garner attention and views?

I came across this study earlier this week and didn’t plan to cover it, but as I saw more and more articles around this study it’s worth taking a gander and pointing out some of the issues and what this sort of study may mean in the grand scheme.

First, the preprint was released early January from Wei, et al.¹ and reads more like an editorial rather than a study. This doesn’t mean anything in particular, but it’s weird how the study is presented.

The study itself builds off of the researchers’ prior work² where they examined this so-called new variant of a pangolin coronavirus.

A strange title, and we’ll take a look at what the authors were intending with this prior study.

The prior study looked an isolated pangolin coronavirus named GX_P2V and its changes after serial passage in Vero cells:

The pangolin coronavirus GX_P2V, originally cultured from the lung-intestine mixed samples of a pangolin captured in anti-smuggling operations in 2017, was passaged in Vero cells [Citation6].

In total, this pangolin coronavirus was passed through 8 Vero cell lines, and afterwards was analyzed for mutations, with the assumption that the mutations would be associated with tropism and increased viral replication when paired with these Vero cells. The isolate after serial passage was given the label GX_P2V(short_3UTR):

We next compared the sequence differences between the first passage of the GX_P2V sample and the eighth passage of the GX_P2V isolate. Compared to the single cell-passaged sample, the multiply passaged GX_P2V isolate has complete 5'/3'-terminus sequences and two mutations: a nonsynonymous mutation in the final amino acid codon of the nucleoprotein gene (GCU→GUU) and a 104-nucleotide deletion in the hypervariable region (HVR) of the 3′-terminus untranslated region (3′-UTR) (Figure 1). Thus, the multiply passaged GX_P2V isolate is a cell culture adapted variant.

As noted above, the two main mutations was a swap of one base within the nucleoprotein gene as well as a large deletion within the 3’-UTR. Interestingly, these mutations didn’t appear in early passages, but nearly all of the later serial passaged GX_P2V showed these mutations, suggesting that the passage through the Vero cell lines acted as a selective pressure for these mutations.

From Lu, et al. Figure 1. A comparison between SARS-COV2’s genome, the originally obtained pangolin coronavirus (GX_P2V sample), and the isolate found after serial passage (GX_P2V isolate). The two main mutations found within the isolate are the nonsynonymous mutation boxed in green above, a well as the large deletion within the 3-UTR denoted by the dashed lines which are intended to keep the sequence in alignment. Dots denote homological regions (i.e. pangolin coronavirus bases that are similar to SARS-COV2’s).

So far, this sounds like what would be grounds for possible concern if the mutations are associated with increased cytotoxicity and lethality.

However, when the researchers tested this isolate among cell lines, golden hamsters, and young/old BALB mice there didn’t appear to be strong evidence of severe infection and cytopathy.

For instance, in cell lines cell damage appeared highly attenuated relative to SARS-COV2:

Consistent with the observed slow rate of cell damages, GX_P2V(short_3UTR) produced small plaques on Vero and BGM cells at five days postinfection (Figures 2(C–D)). In comparison, SARS-CoV-2 produced substantial cell damages and large plaques in Vero cells at two days postinfection [Citation4, Citation21]. Notably, GX_P2V(short_3UTR) showed no visible cytopathic effects or plaques in Calu-3 cells (Figure 2(A–C)). Compared to reported SARS-CoV-2 data, GX_P2V(short_3UTR) is highly attenuated in causing cell damage and plaque formation.

And in golden hamsters the results were similar, where there were no noticeable changes when they were challenged with GX_P2V(short_3UTR):

The infection outcomes are clearly contrast to those described in the SARS-CoV-2 infected hamsters. Hamsters infected with GX_P2V(short_3UTR) had similar body weight increases to the mock infected hamsters and no apparent clinical symptoms in the two week observational period (Figure 4(B)). Consistent with the absence of clinical manifestations, there was only a slight hyperemia noted on the surface of lungs following infection with GX_P2V(short_3UTR) at 2 and 5 d.p.i. (Figure 4(C)). Moreover, no signs of bronchopneumonia were found in the tracheas of infected hamsters (Figure 4(C)).

Further evidence points to lack of pathological features, although this finding would merely suggest that this serial-passaged variant may not be a proper match for hamster models:

Compared to the mock group, no pathological lesions were found in all the tissues of all GX_P2V(short_3UTR)-infected hamsters over the entire observation period (Figure (5C and S2)). Thus, GX_P2V(short_3UTR) did not induce potentially damaging inflammatory changes in any tissues of either infected groups. The lack of detectable pathology and short duration of viral replication in infected hamsters suggests that GX_P2V(short_3UTR) has a replicative deficiency in hamsters.

And this seems to be the case for BALB mice as well, in which case the the GX_P2V(short_3UTR) isolate did not appear to be replication competent when given to BALB mice, with very few pathologies noted as well as quick viral clearance:

BALB/c mice were intranasally infected with 1 × 10⁵ TCID₅₀ of GX_P2V(short_3UTR) and euthanized to collect samples at specific days postinfection (Figure 7(A)). Following inoculation, no apparent clinical signs were observed in any inoculated mice. Notably, the body weights of young BALB/c mice continued to increase (Figure 7(B)) and no gross pathological changes on the lung surfaces of inoculated mice were detected (Figure 7(C)). The viral RNAs in the lung, turbinate, and trachea tissues of young mice were rapidly cleared, and aged mice had an even faster rate of viral clearance (Figures 7(D and S3)). This is in contrast to the slower rate of viral clearance in aged mice inoculated with SARS-CoV-2 [Citation27].

Now, all of these studies don’t provide any details for what may happen if humans are presented with this variant, however for all intents and purposes these findings seem to point towards reduced virulence and replication capabilities when presented to the rodent models above, and this appears to be findings that the authors were hoping for.

That’s because these pangolin coronaviruses share large homology with SARS-COV2’s spike, and because of this this newly discovered GX_P2V(short_3UTR) isolate may serve as a model for cross-reactive immunity against SARS-COV2. Or put another way, if this new isolate is less virulent and replication competent then it may serve as a model for creation of a attenuated vaccine against SARS-COV2, as noted in the authors’ concluding remarks:

Live attenuated vaccines are attractive because they can stimulate a strong and durable immune response with both humoral and cellular immunity. The challenge of constructing live attenuated vaccines is the difficulty of adopting acceptable attenuation strategies. Currently, two live attenuated SARS-CoV-2 vaccine candidates generated by using genome recoding have been reported [Citation35, Citation36]. The original GX_P2V virus is not a human pathogen, which is advantageous of safety concerns. Our characterization of GX_P2V(short_3UTR) suggests that this variant has a live vaccine potential against SARS-CoV-2 or can be developed as a vaccine vector for the expression of protective antigens like SARS-CoV-2 RBDs. Our data also suggest that shortening the 3′-UTR is a novel strategy for making a live attenuated SARS-CoV-2 variant that warrants further investigations.

In this case, the findings here appeared to suggest that this possibly weakened virus may make for a suitable vaccine candidate worthy of future research (at least in the words of the researchers). It doesn’t necessarily mean that I would agree with those conclusions, but the point being made here is that the pangolin coronavirus bearing these mutations appear to be weak, and so much so that they may work as vaccine candidates, and it’s likely where the interest in this mutant strain came from, and likely where further research was conducted.

But how does that compare to the current preprint in question?

In contrast, the recent preprint looked at transgenic mice bearing human ACEII receptors, in which the mice were challenged with the GX_P2V(short_3UTR) variant. Unlike the prior study, the mice in this study showed clear pathologies related to the variant, so much so that all of the mice infected died:

We initially assessed whether GX_P2V C7 could cause disease in hACE2 mice by monitoring daily weight and clinical symptoms. A total of four 6 to 8-week-old hACE2 mice were intranasally infected with a dosage of 5×10⁵ plaque-forming units (pfu) of the virus. Four mice inoculated with inactivated virus and four mock-infected mice were used as controls. Surprisingly, all the mice that were infected with the live virus succumbed to the infection within 7-8 days post-inoculation, rendering a mortality rate of 100% (Figure 1B). The mice began to exhibit a decrease in body weight starting from day 5 post-infection, reaching a 10% decrease from the initial weight by day 6 (Figure 1C). By the seventh day following infection, the mice displayed symptoms such as piloerection, hunched posture, and sluggish movements, and their eyes turned white.

100% mortality rate?

Well, that would be scary on its own, although one should also notice that such a small sample size would likely cause any presentation of mortality to appear far scarier than they are. In this case, because only 4 mice were inoculated any loss of life would be a 25% increase in mortality. Lose 2 mice and you’re already at 50% mortality. Strangely, the Daily Mail article seems to have misreported on the “8 rodents” and taken that from the tropism portion of the study, although note that mice were sacrificed across the groups at certain days post-infection and therefore isn’t an actual assessment of mortality.

This is what happened when another previous study was widely reported, in which case an Omicron/Wildtype hybrid virus appeared to show 80% mortality…among the 10 mice that were challenged with this newfangled Omni-S:

With such a low sample size it becomes easy to see how these numbers could be misinterpreted. Heck, the Omni-S study itself suggested that 100% of Wuhan infected mice died/were euthanized, and it’s obvious that this did not happen (most of us are still here, aren’t we?). It’s findings such as these that should warrant some reservation.

It’s also funny to see that a lot of the fear around this study mimics that of this other “gain-of-function” study.

As to the supposed brain damage in these mice, note that the authors are not sure if the outcomes are related to the transgenic nature of the mice, to the extent that transgenic mice may express ACEII within the brain, although this comment seems to be unsubstantiated and instead is extracted from the findings within this study and serves more as a rationalization for the pathological results:

However, it is important to note that our hACE2 mouse model may be relatively unique. The company has not yet published a paper on this hACE2 mouse model, but our results suggest that hACE2 may be highly expressed in the mouse brain. Additionally, according to the data provided by the company, these hACE2 mice have abnormal physiology, as indicated by relatively reduced serum triglyceride, cholesterol, and lipase levels, compared to those of wild-type C57BL/6J mice.

Note that these studies don’t provide us any information on how humans would be affected, and although we should hesitate to allow for human studies in the case of possible crossover and escape of a virus, at the same time this an example of fearing something that we have been exposed to already.

That is to say, the prior study and evidence of strong homology between this pangolin virus and SARS-COV2 also suggests that, in the case of this pangolin coronavirus variant, that we’d probably be nowhere near a high mortality rate, because we would likely already have immunity against this GX_P2V(short_3UTR) strain.

Note here that the authors themselves didn’t compare live GX_P2V infections, and so their results are compared to groups which would already show relatively reduced pathology. Granted, prior evidence does seem to suggest that infection of transgenic mice with GX_P2V don’t appear to show strong evidence of harm. Nonetheless, the fact that the authors used inactivated and mock viruses means that this study itself uses no original GX_P2V comparator.

By missing out on key information, I don’t find the assertions about how lethal this virus is to be something substantial, and so it’s curious why so many have taken to strike fear with these “100% lethality” remarks. It’s quite clear that there’s more to the lethality story than is being reported, and given the limited information it’s quite clear that we should be hesitant with these numbers.

---

Now, the main question that this study brings up is whether gain-of-function research of any sort should be conducted.

Weirdly, in the case of this mutation it doesn’t appear that this variant actually came about due to serial passage alone, but rather is possibly part of the pangolin coronavirus stew that gets selected for through passage in Vero cells.

At least, that is what one of the coauthors (Lihua Song) states based upon alleged misinterpretations of her study (the following coming from the Comment section of the preprint):

1. The GX_P2V virus has been published in Nature in 2020 (doi: 10.1038/s41586-020-2169-0). It is not a brand-new virus.

2. The GX_P2V(short_3UTR) mutant was published in Emerging Microbes & Infections in 2022 (doi: 10.1080/22221751.2022.2151383). This cell-adapted mutant is the actual isolate published in the Nature paper. So, the original GX_P2V virus was not isolated. Clearly the original GX_P2V virus in the pangolin sample has severe growth deficiency in Vero cells.

This again seems to be a comment that has not been substantiated, but at least makes the assumption that this mutant pangolin coronavirus may exist in general but at very low levels, only to be selected for through Vero cell line passage. Again, an unsubstantiated claim.

But does this mean any gain-of-function research should not be conducted?

The problem is figuring out what necessarily constitutes an alleged “gain-of-function” study. But it also requires a balance of understanding why such research is conducted and what can necessarily be gained from such research.

One growing issue among skeptics is to label any bit of research with the “gain-of-function” moniker to outright dismiss the claims. It works even more when these studies appear to come from Chinese labs, even though these sorts of studies are occurring all the time. I wouldn’t be surprised if outlets such as the Daily Mail are constructed in such a way to allude to the Chinese lab leak hypothesis for SARS-COV2.

Serial passage is itself a mainstay for virology research, which itself comes with rightful concerns given the fact that serial passage is known to lead to various mutations, although this can operate to either increase or decrease virulence.

This creates a precarious situation in which we must contend with the idea of “just because you can, doesn’t mean you should”. How much of this research should be allowed, and what should be the intent of such research?

And it also comes with the notion that we can argue against generalizations made by studies while also noting that such studies may be ethically concerning.

In that case, it’s clear that we should be careful with these “100% lethality” remarks since the data raises serious contextual issues, while we can also raise concerns over the need and utility of gain-of-function studies, especially if the findings may provide us with evidence of something that would otherwise likely not be a real threat to us humans but may still lead to the emergence of such a threat.

---

If you enjoyed this post and other works please consider supporting me through a paid Substack subscription or through my Ko-fi. Any bit helps, and it encourages independent creators and journalists such as myself to provide work outside of the mainstream narrative.

Click on the image above or visit the website at https://ko-fi.com/moderndiscontent

1. Lethal Infection of Human ACE2-Transgenic Mice Caused by SARS-CoV-2-related Pangolin Coronavirus GX_P2V(short_3UTR)

Lai Wei, Shuiqing Liu, Shanshan Lu, Shengdong Luo, Xiaoping An, Huahao Fan, Weiwei Chen, Erguang Li, Yigang Tong, Lihua Song

bioRxiv 2024.01.03.574008; doi: https://doi.org/10.1101/2024.01.03.574008

2. Shanshan Lu, Shengdong Luo, Bingke Bai, Zhenping Fan, Bixia Hong, Wen Xu, Hongbin Ma, Weiwei Chen, Huahao Fan, Yigang Tong & Lihua Song. (2023) Omicron breakthrough infections in wild‐type SARS‐CoV‐2 vaccinees elicit high levels of neutralizing antibodies against pangolin coronavirus GX_P2V. Journal of Medical Virology 95:8.
Crossref