When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


3 Possible Origins of COVID: Lab Escapee, Evolution, or Mutator Genes?

“Virus outbreak: research says COVID-19 likely synthetic,” shouted the headline in the Taipei Times on February 23, 2020. The idea that the novel coronavirus SARS-CoV-2 arose in a virology lab in China – by accident or as a bioweapon – has sparked an undulation of accusation and explanation ever since.

The latest chapter: An “open letter” in the April 7, 2021 New York Times, calling for “a full investigation into the origins of COVID-19.” The two dozen scientists who signed the letter cite the continuing absence of a “robust process” to examine critical records and biological samples. Their argument responds to the WHO’s March 20 press event that barely considered an origin other than from a natural spillover.

But two types of new information may counter the lab escapee hypothesis: filling-in-the-blanks of mammals that may have served as “missing links” in the evolution of disease transmission, and the rapid rise of viral variants reflecting a tendency to mutate that may underlie SARS-CoV-2 seemingly bursting from out of nowhere.

So here is my view, as a geneticist, of three possible origins of SARS-CoV-2:

1. Bioweapon – an engineered pathogen or escape of a natural candidate

2. Gradual evolutionary change through intermediate animal hosts, mutating along the way and becoming more virulent

3. “Mutator” genes that trigger mutations in other genes, speeding evolution

Bioweapon Beginnings

Until recently, the idea that SARS-CoV-2 was groomed as a weapon relied on a predecessor, a coronavirus dubbed RaTG13, found in the horseshoe bat Rhinolophus affinis.

Researchers discovered RaTG13 in bat droppings in an abandoned mineshaft near a cave in Yunnan, China, in 2013, shortly after six miners fell ill and three of them died of an unspecified pneumonia. (Bats harbor many viruses without becoming sick, which I explain here.)

RaTG13 shares about 96.1% of its genome sequence with that of SARS-CoV-2. For perspective, SARS-CoV-2’s genome is only about 80% similar to that of the original SARS coronavirus from 2003.

A key difference between RaTG13 and SARS-CoV-2 is in part of the receptor binding domain where the spike protein latches onto human cells. This differing part matches the RNA sequence from viruses in the Malayan pangolin, a spiny anteater-like creature that may be an intermediate host between bats and people in the infection chain. The transfer from bat to pangolin might have happened near the mine, or at a wet market rife with raw flesh, or in many other places where humans encroach on the territories of other animals and we simply haven’t looked.

Clues to the transition from bat virus RaTG13 to human virus SARS-CoV-2 may lie within the 4% of the genome sequences that diverge. Evolutionary biologists estimate it would have taken at least 50 years for the bat virus to have mutated itself into SARS-CoV-2, considering known, natural mutation rates of viral genomes. A bioweapon, presumably, could have been created much faster, just as it’s faster to buy a new car than to fix an old one part by part. But as we’ve learned, we can’t rely on what we know about past viruses. That is, the mutation rate of the newbie could be much faster than what we’ve seen before.

A document dubbed the “Yan report” (which I wrote about here) maintains that RaTG13 never existed. Instead, Li-Meng Yan and colleagues argue that the supposed SARS-CoV-2 predecessor is an imaginary RNA sequence uploaded to the GenBank database to provide a plausible natural explanation for the origin, to deflect attention from the idea of a bioweapon. The paper (there’s an initial and updated version) questions why, if RaTG13 was discovered in 2013, it wasn’t reported until February 3, 2020, in Nature. The Yan report never made it beyond preprint (non-reviewed) status, I suspect partly because of the pervasive tone of paranoia and the funding source, an organization connected to Steve Bannon. Researchers have eviscerated it, and Wikipedia reviews the details.

A brief report that I keep going back to appeared in Nature on March 17, 2020, when the global death toll from COVID stood at just 4,373: “The proximal origin of SARS-CoV-2.”

The “proximal origin” authors (Kristian G. Andersen, Andrew Rambaut, W. Ian Lipkin, Edward C. Holmes and Robert F. Garry) compare key parts of the new pathogen to corresponding parts of other coronaviruses, concluding “(o)ur analyses clearly show that SARS-CoV-2 is not a laboratory construct or a purposefully manipulated virus.” Part of their reasoning is common sense: the virus doesn’t bind to our cells strongly enough to have been invented. It’s an imperfect weapon. Why would a new iPhone work worse than its predecessors? It’s more likely, they argue, that the new virus, with its distinctions (like a dozen extra RNA bases inserted into the area corresponding to where the two parts of the spike protein attach), arose from natural selection. The virus had a natural advantage, so it was perpetuated – not invented.

Whatever happened, the prescient “proximal origin” researchers concluded, back in March 2020, “Although no animal coronavirus has been identified that is sufficiently similar to have served as the direct progenitor of SARS-CoV-2, the diversity of coronaviruses in bats and other species is massively undersampled.”

That’s changing. Slowly.

Evolution in a Poop Soup

A leap from the RaTG13 virus found in the bat muck of the abandoned mine in 2013 to the emergence of SARS-CoV-2 in 2019 is like reading the first and last chapters of a novel: there’s not enough of a plot to reconstruct a story. But as more chapters are being revealed, it’s looking like SARS-CoV-2 arose from a poop soup of viruses – and continues to evolve.

It turns out that RaTG13 wasn’t the only stop on the evolutionary road to SARS-CoV-2. Nor was China the only home of novel coronaviruses, although they continue to be identified there. Consider recent reports:

Cambodia, January 26, 2021 Excrement and saliva from two horseshoe bats sampled in Cambodia in 2010 yielded coronaviruses that share 92.6% of their genome sequences with SARS-CoV-2, differing at one end of the gene encoding the spike protein. Concludes a preprint in bioRxiv, “The discovery of these viruses in a bat species not found in China indicates that SARS-CoV-2 related viruses have a much wider geographic distribution than previously understood, and suggests that Southeast Asia represents a key area to consider in the ongoing search for the origins of SARS-CoV-2, and in future surveillance for coronaviruses.”

Thailand, February 9, 2021 Blood from five bats in a cave in Thailand yielded coronaviruses similar to a type found in Yunnan, China, as well as antibodies against SARS-CoV-2. The antibodies were also detected in a pangolin, according to a report in Nature. Although this study didn’t reveal the progenitor, it too extends the realm of SARS-CoV-2-like viruses beyond China.

China, March 8, 2021 Another bioRxiv preprint describes genome sequences of 411 coronavirus samples from 23 bat species collected from May 2019 to November 2020, over 2700 acres in Yunnan province. The closest relative to SARS-CoV-2, dubbed RpYN06, shares 94.5% of the genome sequence. But overall genome similarity is not as important as correspondence between individual genes, which can better predict the effect of a novel virus on a human body.

RpYN06 is actually the closest relative to SARS-CoV-2 identified so far, based on key genes that provide the tools to replicate (ORF1ab), melt into our cells and latch onto our protein synthetic machinery (ORF7a and ORF8), and encode the nucleocapsid (N) proteins that protect the viral genetic material. The study found 3 other coronaviruses whose genomes are very similar and resemble those found in pangolins.

Did SARS-CoV-2 chug along happily, in various types of bats for who knows how many years, mixing with other coronaviruses and unchanging because it’s genome served it well? Only after the jump to a new host – us – did the mutations that underlie adaptation happen, spontaneously, and then persist if they conferred an advantage. Then mutations in individual genes began to accrue into the viral variants that are now sweeping the planet. The title of a recent article in PLoS Biology sums up the forces that have molded the novel coronavirus: “Natural selection in the evolution of SARS-CoV-2 in bats created a generalist virus and highly capable human pathogen.”

The Mutator Hypothesis

A third way that SARS-CoV-2 could have come into being quickly is if a gene or genes has functioned as a “mutator,” provoking other genes to mutate.

I remember this phenomenon from my training as a Drosophila geneticist. Fruit flies that have a mutant yellow eye color can have offspring that revert to the normal red color, not from a mutation in an eye color gene, but from a mutation in a gene called mutator. It wreaks havoc on other genes. And the types of mutations it brings are like those in the new variants of SARS-CoV-2.

Half of the mutations that the fruit fly mutator gene causes are deletions – missing bits of genes. A telltale sign of the viral variant B.1.1.7, first detected in the UK, is “S gene target failure.” That’s when a PCR COVID test doesn’t replicate the RNA that encodes the spike protein because two amino acids are missing, while it does replicate the other viral genes.

In fruit flies, mutator also quintuples the rate of single-base changes, called point mutations. Those happen in the new viral variants too.

I’m not suggesting that a fly gene has run amok in viruses, but might a mutator-like gene be driving the rapid diversification of SARS-CoV-2 into a suite of variants? If so, rapid mutation might explain how the virus came into being and then became a shape-shifter, without having to invoke a mad-scientist-creating-a-bioweapon scenario or a string of hapless mammals passing along a pathogen that would kill millions of people in a year.

Identifying a mutator would require unraveling gene-gene interactions, something that hasn’t been a huge priority even in analyzing human genomes. Perhaps a well-studied gene of SARS-CoV-2 has a second function, inducing mutation of another? Although more than a million SARS-CoV-2 genome sequences have been uploaded to the database GISAID, I don’t know to what extent researchers are probing how the genes interact, rather than investigating them in conceptual silos.


When people talk about the race between variants and vaccines, they aren’t being flip. For now, the vaccines stimulate a diverse enough antibody response to handle circulating viruses. But evolution never stops. If variants arise that slip into vaccinated bodies and take hold, then spread, will those vaccines weed out the older variants while creating niches for the new? That’s what is currently unsettling the experts. And me.

And so we must anticipate and stay ahead of evolution, which the vaccine manufacturers have already been doing for months. For if there’s one constant during these crazy times, it’s that SARS-CoV-2 continually surprises us. But for now, I’m relieved to ponder alternatives to the unthinkable idea that the virus was created to destroy us.

  1. There’s literally no solid support for a natural origin in this article. Citing RpYN06, etc is not evidence of a zoonotic origin. Very poor quality article.

    1. I did not intend for there to be “solid support” for any of the 3 hypotheses — which was actually my point: that science works by accruing evidence to eliminate (disprove) hypotheses. Please link to articles you have published that are high quality. I was looking more for a reasoned discussion, particularly about my mutator hypothesis, which I have not seen brought up anywhere. Insults, especially directly to the writer, are inappropriate.

  2. hi Ricki – re the creation of SARS-cov-2.. seems like gain-of-function potential pandemic pathogen program (NIH) (GOF PPP) success story for those seeking such. The funding, the researchers, the labs, the stakeholders – it’s all there for those who choose to see. Menachery et al. (2015) seems to lay out the concept and technique (chimeric clone etc). Moreover when you look at both funding at UNC (Gilead), Wuhan (NIAID), and US (DARPA/Moderna) and elsewhere the picture gets clearer. I’d think you’d want to interview Baric (UNC), Zhengli (Wuhan), Daszak (EcoAlliance), and Fauci (NIH) perhaps even Qui (Canada/Wuhan) and Foucher (Belgium) and ask what they were up to the past decade re GOFR. Selective dissemination of information can also be selective omission of information.

    As for origin theories – zoonotic, lab leak, unrestricted BW, etc. one needs to look at fitness and cost-benefit as well as the behaviors and statements of those involved. With mRNA, CRISPR-cas9 and biogenetics – we’re in a new frontier – designer viruses, discrete population targeting, etc.. IMHO there are indeed certain dark actors with strategic intent in the mix here as too there are those seeking to head this off. Exploiting the current SITREP has tremendous momentum.

    1. Well, Bob, then surely you can provide tangible facts or evidence supporting your “humble opinion” beyond just “I’m paranoid and this feels suspicious, so it MUST be true”.

  3. Statistically, what is the probability of a natural origin of such an aggressive virus to appear naturally in a city where there were already intense and dangerous manipulations in the laboratory aiming at gains in virus function?

    1. Unfortunately statistics and even modeling can’t tell us with much certainty what happened as SARS-CoV-2 came into the world. It is an issue of logic – we can’t know what we don’t know, even with machine learning and modeling. So … the changes to the SARS-CoV-2 genome as a group are highly unusual. My thinking is Occam’s razor – that an intentional, manipulated origin is more likely. But the flip side of that is that we haven’t discovered all the bat coronaviruses in China and elsewhere, and we can’t know when we’ve discovered them all. So that is why I think the origin is unknowable.

    1. Thanks for posting. The main author, Nicholas Petrovsky, has sent me and many others intriguing links to his work since the very beginning. He has been very helpful in understanding the big picture, as well as the details. This is an in silico study; predictions based on the three dimensional structures of the important entities – receptors and the targets that they bind. It is typical for a virus to inhabit several host species, so it isn’t surprising. Pangolins were brought up early on as an intermediate host. What is intriguing about the paper is that the receptor binding to our ACE2 receptors is extraordinarily strong. I suppose one could interpret this as bolstering a lab origin hypothesis. We can never know for sure what actually happened. And one likely explanation accounts for both bats in caves and further change in a laboratory setting. So no, the tight binding to human ACE2 receptors doesn’t surprise me or change my thinking.

  4. Updated better version 6/8/2022
    On February 26, 2022, George Gao of the Beijing China National CDC published the preprint,

    1) ‘Surveillance of SARS-CoV-2 in the environment and animal samples of the Huanan Seafood Market’, by George Gao, William Liu, and many others, Chinese Academy of Sciences, National Institute for Viral Disease Control and Prevention, and others https://assets.researchsquare.com/files/rs-1370392/v1_covered.pdf?c=1645813311

    which contains the results of 33 environment samples collected January 1 or 12, 2020, from the Huanan Seafood Market (HSM) in Table 1 of the article, and results of samples collected on later dates. On the same day Michael Worobey and Christian Andersen study group released their preprint article below, that contained 33 positive environment samples from the HSM in their Table S2.

    (2) ‘The Huanan market was the epicenter of SARS-CoV-2 emergence’, by https://zenodo.org/record/6299116#.Yh_ErJhMFdh, by Michael Worobey, Joshua Levy, Pekar, and others.

    Worobey and Anderson probably consulted with Dr. Gao, to get their list of environmental samples of the HSM.

    Dr. Gao’s Table 1 lists the 33 Huanan Seafood Market positive environment samples for Covid-19, and he observed.

    “live viruses were isolated from samples F13 (wall) , F54 (ground), and B5 (ground), which were the only three samples with Ct values <30 (cycle threshhold) ….. samples F13 (wall) and F54 (ground) were from the stalls with confirmed patients ……..The genome sequences of two environmental samples, F13 (wall) and F54 (ground), were found to be highly identical to the reference strain HCoV/Wuhan/IVDC-HB-01 (WH01, sequence identity of 99.993%) and completely identical to the human stain Wuhan-Hu-1 (GenBank: NC_045512) ……. Notably, samples F13 and F54 were from the stalls with confirmed patients. All the results of successful virus isolation and the Ct values of the original samples revealed the existence of live SARS-CoV-2 with high titers in the environment of HSM……Commercial products of swabs and virus preservation solution were used for the sampling (Disposable Virus Sampling Tube, V5-S-25, Shen Zhen Zi Jian Biotechnology Co., Ltd., Shenzhen, China). For environmental samples, sampling swabs were applied to smear the floors, walls or surfaces of objects and then preserved them in virus preservation solution.”

    Dr. Gao says F13 wall and F54 ground environmental samples are highly identical (99.993%) to IVDC-HB-01 (WH01) , and completely identical to Wuhan-Hu-1, NC_045512 isolate. IVDC-HB-01 (WH01) might have been isolate EPI_ISL_402119 taken of Wuhan patient on December 31, 2019, by the National Institute for Viral Disease Control (IVDC) in Beijing, according to Table 5 in the article, ‘Nosocomial Outbreak of 2019 Novel Coronavirus Pneumonia in Wuhan, China’, by Xiaorong Wang, Qiong Zhou, and others, Wuhan Union Hospital, Beijing Capital Bio Medical, and others. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7236818/

    EPI_ISL_402119 (GISAID) was 49 years old according to Table S1 in the article, ‘Supporting Information A Single and Two-Stage, Closed-Tube, Molecular Test for the 2019 Novel Coronavirus (COVID-19) at Home, Clinic, and Points of Entry’, by Mohamed El-Tholoth, and others. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8482318/bin/ac1c03016_si_001.pdf

    NC_045512 housekeeping matters —-

    Determining who NC_045512 is, is more important. That person’s Covid-19 sequence is ‘completely identical’ to the F13 wall sample according to Dr. Gao. The NCBI database entry for NC_045512.1 in part reads,

    “NC_045512 30473 bp ss-RNA linear VRL 13-JAN-2020
    TITLE A novel coronavirus associated with a respiratory disease in Wuhan
    of Hubei province, China
    JOURNAL Unpublished
    CONSRTM NCBI Genome Project
    JOURNAL Submitted (13-JAN-2020) National Center for Biotechnology
    Information, NIH, Bethesda, MD 20894, USA
    /isolate="Wuhan-Hu-1" ”

    NC_045512.1 virological sequences (VRL) filed January 13, 2020 with NCBI. Sequence has 30473 base pairs, may have been named or sequenced in the United States by the NCBI Genome Project, using Chinese bioproject data, and was isolated in China in December 2019.
    The then unpublished article was probably, ‘A new coronavirus associated with human respiratory disease in China’, by Fan Wu, Su Zhao, and others, Shanghai Public Health Center and others. Received for publication January 7, 2020, but not published online until Feb 3, 2020, bioproject number PRJNA603194. This is Shanghai Covid-19 study # 1. https://www.nature.com/articles/s41586-020-2008-3

    ‘NC’ stands for National Center.
    NCBI Genome Project is also associated with the submission of
    NC_001451.1 Avian infectious bronchitis virus
    NC_003436.1 Porcine epidemic diarrhea virus
    NC_002645.1 Human coronavirus 229E

    On the other hand, NC_045512.2 virological sequences (VRL) were filed July 18, 2020 with NCBI. That sequence has a 29,903 base pair genome. NCBI database entry in part reads,

    “LOCUS NC_045512 29903 bp ss-RNA linear VRL 18-JUL-2020 The reference sequence is identical to MN908947. On Jan 17, 2020 this sequence version replaced NC_045512.1.”

    So according to NCBI database curators, NC_045512.2 reference sequence is identical to MN908947. A NCBI software BLAST comparing the two sequences shows 100% identity and 100% Query Cover. We know that MN908947 (identical to NC_045512.2 sequence) was a 41 year old who worked at the Huanan Seafood Market (HSM).

    “Here we study a single patient who was a worker at the market and who was admitted to the Central Hospital of Wuhan on 26 December 2019………The patient studied was a 41-year-old man with no history of hepatitis, tuberculosis or diabetes.……. During admission, BALF (bronchoalveolar lavage fluid ) was collected and stored at −80 centigrade until further processing….the viral load in the BALF sample was estimated by qPCR to be 3.95 × 10 to the 8 th power copies per ml (395,000,000 copies per ml—my numeric interpolation)…. This virus strain was designated as WH-Human 1 coronavirus (WHCV) (and has also been referred to as ‘2019-nCoV’) and its whole genome sequence (29,903 nt) has been assigned GenBank accession number MN908947”. —- in the article, ‘A new coronavirus associated with human respiratory disease in China’, by Fan Wu, Su Zhao, Bin Yu, and others, Shanghai Public Health Center, Wuhan Central Hospital and others.
    Shanghai Covid-19 study # 1.

    So F13 wall sample sequence from HSM ‘completely identical’ to NC_045512.2 sequence (according to Dr. Gao), completely identical to MN908947 sequence (according to NCBI), for 41 year old admitted to Wuhan Hospital December 26, 2019, with high viral load. Verifying 41 year old co-workers Covid-19 status and hospitalization dates, is very important to understanding how F13 sample got on the Huanan Seafood Market stall wall.

    The World Health Organization January 2021 study of Covid-19 origin lists four clusters of early Covid-19 cases, two directly involving HSM employees. Cluster # 2 affirms the existence of a 40 year old HSM employee. From page 157 of the WHO Annexes report.

    “Cluster 1: Including two confirmed cases, living together as husband and wife. Both of them denied case contact history, as well as history of exposure to Huanan Market. Spouse one, 62 years old, fell ill on 15 December 2019, spouse two, 62 years old, fell ill on 26 December 2019.”

    “Cluster 2: there were 3 confirmed cases, all of whom were traders of the same stall in Huanan Market. Stall employee one, 40 years old, fell ill on 17 December 2019; stall employee two, 32 years old, fell ill on 19 December 2019; stall employee three, 57 years old, fell ill on 25 December 2019. It was a fixed stall in Huanan Market, dealing in frozen products such as pastry and soy products. Employee two was purchasing goods from the Baishazhou market and Huanan Market back and forth. Employee three was delivering goods in Huanan Market.”

    “Cluster 3: there were two confirmed cases, living together as husband and wife, and both of them denied animal contact history and history of travel. Spouse one, 61 years old, fell ill on 20 December 2019; Spouse two, 57 years old, fell ill on 25 December 2019. Spouse one had been engaged in restaurant distribution for a long time, and often stocked up in Huanan Market. Spouse two denied a history of exposure to Huanan Market or other markets.”

    “Cluster 4: There were two confirmed cases, both of whom were employees of the same stall in Huanan Market, and both of them denied contact history of poultry and animals, as well as contact history of travel. Employee one, 56 years old, fell ill on 20 December 2019; employee two, 45 years old, fell ill on 26 December 2019. It was a fixed stall in the Huanan Market, dealing in aquatic products such as catfish”, in the article ‘WHO-convened Global Study of Origins of SARS-CoV-2: China Part’, Joint WHO-China Study
    14 January-10 February 2021 Joint Report – ANNEXES — https://www.who.int/docs/default-source/coronaviruse/who-convened-global-study-of-origins-of-sars-cov-2-china-part-annexes.pdf?sfvrsn=3065bcd8_5

    It is possible Wuhan-Hu-1 NC_045512.2/ MN908947 isolate can be found in Cluster 2 of these four World Health Organization clusters of non-HSM and HSM workers. The WHO report should be read in the context of other important accounts given in January 2020.

    Second major account of early SARS-CoV-2 cases is the article, ‘A pneumonia outbreak associated with a new coronavirus of probable bat origin’, by Peng Zhou, Zhengli Shi, and many others, Wuhan Institute of Virology, and others. Extended Tables # 1 and # 2 indicate,

    ICU-04 person was 32 years old admitted to hospital on December 27.
    ICU-10 person was 56 years old admitted December 20.
    ICU-05 person age 40 admitted to hospital on December 27.

    These hospitalization cases pretty much match the World Health Organization ‘Cluster 2’ of infected HSM workers. Only one non-Huanan Seafood Market related patient. ICU-01, age 62, admitted December 27, 2019, which may match a World Health Organization ‘Cluster # 1’ person above.

    Wuhan Institute of Virology isolates can be correlated in ‘Table S5 The integrated introduction of all 61 published 2019-nCoV’, in the article ‘Nosocomial Outbreak’ referenced above.

    Third major account of early SARS-CoV-2 cases is the Beijing CDC table, ‘Information about samples taken from nine patients infected with 2019-nCoV’, in the article ‘Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding’, by Roujian Lu, Xiang Zhao, and many others. Beijing CDC identify a WH01 person whose BALF was taken December 26, 2019. That person’s isolate was filed with the Chinese National Microbiological Data Center. Beijing CDC Figure 3 phylogenic tree in this article dated January 28, 2020 shows the Beijing CDC had knowledge of the full genome of Atlanta CDC bat coronavirus BtKy72, seven days before BtKy72’s sequences were filed with NCBI on February 5, 2020. Questions as to how the Beijing CDC got that pre-knowledge, are asked are at the end of this comment.

    Fourth major account of early HSM cases is, ‘Table 1. Clinical symptoms and patient data’, in the article ‘Complete genome characterisation of a novel coronavirus associated with severe human respiratory disease in Wuhan, China’, by Fan Wu, Su Zhao, and others. That article lists clinical data for seven early SARS-CoV-2 cases. This is Shanghai Covid-19 report # 2.

    An important aspect of the F13 wall sample is that it had ‘high titers’, ‘the Ct values of the original samples revealed the existence of live SARS-CoV-2 with high titers in the environment of HSM’. Dr. Gao does not give the possible viral load for the F13 wall sample, but he does provide the Ct value. Ct values may reflect viral load size according to some scientists. An Oxford University study group equivocated a Ct value of 24.4 or less, to a viral load of greater than or equal to 10,000 RNA copies/ml or more, in 85% of their case-contact pairs.

    “SARS-CoV-2 infectivity varies by case viral load, contact event type, and age. Those with
    high viral loads are the most infectious….predict the real-world likelihood of a SARS-CoV-2 infected individual infecting someone else……85.4% (197,677/231,497) of case-contact pairs with PCR-positive contacts, i.e., plausible onward transmission, had case viral loads of greater than or equal 10,000 RNA copies/ml (i.e. Ct less than or equal to 24.4)”, in the article, ‘ SARS-CoV-2 infectivity by viral load, S gene variants and demographic factors and the utility of lateral flow devices to prevent transmission’, by Lennard YW Lee, and others, University of Oxford

    Riddell and van Doremalen make similar correlations between Ct values and ‘viral load’ RNA copies/ml in their articles referenced below. So F13 wall (23.85 Ct) sample and F54 (25.8 Ct) ground sample Ct’s counts similar to Wuhan-Human-1 Ct (23.967 – Extended Data Figure 4) count, ‘might’ reflect similar viral loads for each sample. On the other hand, Dr. Gao’s ‘high titers’ may refer to a samples infectivity (as determined by TCID50 or plaque forming units), and not necessarily RNA ‘viral load’ as determined by quantitative RT-PCR. A University of Edinburg study defined infectivity as follows.

    “Viability refers to how stable a viral particle is outside of a host cell (e.g. in air as it travels from infector to infectee) whereas infectivity (or replicability) refers to its ability to multiply within a host cell. …..modelling study, by Lelieveld et al (2020) estimated the indoor infection risk from aerosolized viruses in four different environments. The study made assumptions on viral load and infection dose as highly infectious viral load of 5 × 10 to the 8 power RNA Copies/cm3; super infectious viral load of 5 × 10 to the 9 th power (10 to the 9 th–10 to the 10 th copies/cm3); infective dose (D50) 316 RNA copies (100–1000 RNA copies); and virus lifetime in aerosol of 1.7 days” ——in the article,
    ‘Review: What is the infectious dose of SARS-CoV-2?’, Date: 30th April 2021 Version: 29-01 , University of Edinburgh, Usher Institute

    Highly infectious viral load, super infectious viral load, and infectious dose. Wuhan-Hu-1 with his BALF of 3.95 × 10 to the 8 th power copies per ml (395,000,000 copies per ml) would be close to highly infectious, if one disregards the unit of measure ‘cm3’. Saikat Basu expanded on the notion of infectious dose and plaque forming unit as follows,

    “The infectious dose is a fundamental virological measure quantifying the number of virions that can go on to start an infection; the value of which is still not conclusively known for SARS-CoV-2 . Theoretically, according to the independent action hypothesis , even a single virion can potentially establish an infection in highly susceptible systems……the (2018 ferret) study extracted the dose in terms of the plaque-forming units (PFU)—which is a measure used in virology to describe the number of virus particles capable of forming plaques per unit volume. The quantitative estimate of the infectious dose, being of the order O (10 to the 2 nd power- 100 my interpolation), agrees with the findings presented her….how effective are these droplets at carrying virions? SARS-CoV-2 belongs to a diverse family of single-stranded RNA viruses , and as noted before, virological assessments done on the sputum of hospitalized COVID-19 patients show an averaged viral load of 7.0 × 10 to the sixth power RNA copies/mL of oral fluid (7,000,000-my interpolation), with the peak load being 2.35 × 10 to the ninth power copies/mL (2,350,000,000-my interpolation).”—in the article, ‘ Computa- tional characterization of inhaled droplet transport to the nasopharynx’, by Saikat Basu, South Dakota State University

    Taken in the context of Basu’s data, Wuhan-Hu-1 with his BALF of 3.95 × 10 to the 8 th power copies per ml (395,000,000 copies per ml) was above average viral load, and heading towards peak. Just as important, Basu defined plaque forming units as “the number of virus particles capable of forming plaques per unit volume.” Plaque forming units and TCID50 are the two basic ways of determining infectious titer.

    “Viral loads are commonly measured in two distinct ways: counting viral RNA genomes by qRT-PCR and measuring the number of infectious units in tissue culture (and thirdly by counting PFUs). The (this) second approach incubates susceptible mammalian cells with dilutions of a patient sample to determine the amount of sample required to kill 50% of the cells. This value is used to back-calculate the infectious titer in the sample in units of “50% tissue culture infective dose” or TCID50 [for example, by the Reed and Muench method]. The TCID50 is analogous (and often quantitatively similar) to the plaque-forming units (PFU) assay. Here, we refer to TCID50 and PFU more generally as “infectious units.” As these two measurement modalities (RNA genome copies and infectious units) differ in reported values and interpretation—one method measuring the number of RNAs, the other measuring the number of infectious units…….In case where a large fraction of the viral RNA copies are present as “naked” RNA (not encapsulated inside viral particles), using viral RNA copies as a proxy for the number of viral particles could lead to an overestimate.
    ——in the article ‘The total number and mass of SARS-CoV-2 virions’, Ron Sender,Yinon Bar-On, and others, Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel, and others.

    ‘Naked’ RNA, versus particle encapsulated RNA , what was the predominate case for the F13 wall coronavirus sample? More to be said later regarding using TCID50 versus plaque forming units to measure infectious units. But measuring the viral load (viral RNA genomes) itself using qRT-PCR would be the first step taken. Again, Dr. Gao was probably referring to ‘infectiousness’, when he stated F13 wall sample and F54 ground sample had ‘high titers’, but he could have been referring to the viral load size itself.

    And to this date, one study has measured the ‘titers’ of the F13 wall sample, Harbin Veterinary Institute in February 2020. In their article ‘Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS–coronavirus 2’, by Jianzhong Shi, Zhiyuan Wen, and others, Harbin inoculated ferrets with the F13 wall environmental sample coronavirus provided by the Beijing CDC. Harbin’s Figure 1A shows a ‘viral load’ of about 10 to the eight power (100,000,000 my interpolation) RNA copies per gram, on day one after inoculating ferrets with F13 sample coronavirus. Harbin’s figure 1C shows about 10 to the 6 th power infectious plaque forming units per gram (1,000,000 my interpolation) in ferrets inoculated with F13 coronavirus on day one after inoculation. On day one F13 wall sample coronavirus ‘may have been’ already infectious (able to infect ferret/human cells) and environmentally transmissible. https://www.science.org/doi/10.1126/science.abb7015

    Harbin then made the same calculations per milliliter, instead of per gram. Figures 1E shows the RNA copies per milliliter for F13 (about 1,000,000) inoculated ferrets, and Figure 1G shows about 1,000 infectious plaque forming units per milliliter for F13 inoculated ferrets on day one after inoculation. 10,000 pfu on day two after inoculation. Remember, Basu defined plaque forming units as “the number of virus particles capable of forming plaques per unit volume.” No one has determined the number of virons per particle, or the particle diameter for the F13 wall sample

    In March 2020, a Los Alamos National laboratory (Alan Perelson) and San Diego University study reviewed Harbin Veterinary Institute data, and found similar infectivity for the F13 wall sample inoculated ferrets. See Figures 1-3, in the article, ‘Modeling Within-Host Dynamics of SARS-CoV-2 Infection: A Case Study in Ferrets’, by Naveen Vaidya, Angelica Bloomquist, and Alan Perelson, Los Alamos National laboratory and San Diego University laboratory. They concluded.

    “At the peak, the viral load reached about 8 log10 RNA copies per mL (100,000,000-my numeric interpolation), among which approximately 4 log10 (10,000-my interpolation) (approximately 0.01%) per milliliter are infectious viruses. The general pattern observed in SARS-CoV-2 dynamics is similar to that for influenza virus… In general, the infection rate (b), the virus production rate (p), the proportion of infectious virus (a), and the rate of transfer from eclipse phase to infectious phase (k) positively impacted total viral-peak, infectious viral-peak, and cell loss and negatively impacted time to viral-peak. The viral clearance rate (c), the infected cell death rate (d), and the IFN-induced antiviral efficacy (s) have both positive and negative effects depending on the property being examined”.

    Los Alamos National laboratory modelling Figures 1-3 clearly shows about 10,000 infectious plaque forming units day four, after inoculation of ferrets with F13 sample coronavirus. This matches Harbin’s Figure 1G day two infectivity count. This means that .01 % of the viral load (8 log10 RNA copies per mL -100,000,000) was infectious. Los Alamos merely repeated what Weizmann Institute of Science stated earlier regarding .01% infectious part of the viral load.

    “Ratios on the order of 10 to the third power, to 10 to the fourth power between viral particles and PFUs were observed in animal viruses such as poliovirus and papillomavirus. Naively, such a ratio would suggest that only 0.01% of the virions produced are actually infectious. This ratio implies that SARS-CoV-2 is not very efficient in producing infectious progeny. While we do not have a clear explanation for this seeming low efficiency, there are several possible factors that will affect this ratio..”, in the article, ‘ The total number and mass of SARS-CoV-2 virions’, Ron Sender, Yinon Bar-On, and others, Weizmann Institute of Science, Rehovot, and others .

    Regarding the total viral load itself in the ferrets inoculated with F13 sample coronavirus peak viral load of 100,000,000 RNA copies per milliliter, according to both Harbin Veterinary Institute and Los Alamos National laboratory. In contrast, the December 26, 2019 viral load for Wuhan-Hu-1 BALF sample was estimated by qPCR to be 3.95 × 10 to the 8 th power copies per ml (395,000,000 RNA copies-my numeric interpolation) by Shanghai Public Health. Both big numbers. Approximately .01% of that may have been infectious virions (Los Alamos National laboratory and Weizmann Institute), the ‘titers’.

    But as a future baseline approach to understand the contents of one plaque forming unit,

    “Comparing the early measurements of the total virus with the measurements of infectious virus, we deduced that 1 PFU (plaque forming unit) corresponds to approximately 10 to the fourth power (10,000) virions, which is in line with published estimates.”
    Los Alamos National laboratory (Alan Perelson) and University of San Diego laboratory

    So Harbin’s Figure 1G showing about 10,000 plaque forming units in the F13 inoculated ferrets, on day two after inoculation, should be read in the context of the above ambiguous statement. Harbin and Los Alamos National laboratory seem to on the same page regarding the ‘possible’ infectivity of F13 wall sample, but Harbin suggests the infectivity was there on day one. Possibly intimating F13 wall sample was infectious January 1, 2020. The question then becomes, was F13 presence on the wall via aerosol transmission (droplet, etc), or touch transmission?

    “Moreover, rather than the conventional definition of 5 micrometer, it has recently been suggested that the size distinction between aerosols and droplets should be updated to 100 micrometers, as this distinguishes between the two on the basis of their aerodynamic behavior. Specifically, 100 micrometers represents the largest particles that remain suspended in still air for greater than 5 s (from a height of 1.5 m), travel beyond 1 m from the infectious person, and can be inhaled. Although droplets produced by an infectious individual through coughing or sneezing may convey infection…..The factors affecting airborne transmission include viral load in different sized respiratory particles, the stability of the virus in aerosols,….. Large droplets tend to reach their maximum horizontal distances quickly and fall to the ground or surfaces within a few meters,whereas aerosols can remain suspended for many seconds to hours, travel long distances, and accumulate in air in poorly ventilated spaces”, in the article, ‘Airborne transmission of respiratory viruses’, Chia Wang, Kimberly Prather, and others, Sun Tatsen University, Taiwan, and others

    What was the initial velocity of the F13 sample that allowed it to overcome gravity and drag forces and adhere to the HSM wall? Neeltje van Doremalen from the NIAID in Montana used TCID50 to help determine the infectivity of SARS-CoV-2 on aerosoled fomite surfaces, as well as determined the half-life of SARS-CoV-2 on those surfaces.

    “the titer of aerosolized viable virus is expressed in 50% tissue-culture infectious dose
    (TCID50) per liter of air….The longest viability of both viruses was on stainless steel and plastic; the estimated median half-life of SARS-CoV-2 was approximately 5.6 hours on stainless steel and 6.8 hours on plastic….On cardboard, no viable SARS-CoV-2 was measured after 24 hours and no viable SARS-CoV-1 was measured after 8 hours…… We evaluated the stability of SARS-CoV-2 and SARS-CoV-1 in aerosols and on various surfaces and estimated their decay rates”. —–in the article, ‘Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1’, by Neeltje van Doremalen, and others, National Institute of Allergy and Infectious Diseases, Hamilton, Montana

    van Doremalen using TCID50 techniques, and not plaque forming assays to test aerosoled Covid-19 infectivity on fomite surfaces. What does this say about the diameter size of the particles being tested on the fomite surfaces? Could plaque forming assays have been used? They also determined that the half-life (time taken to achieve a 50% reduction in titre/viral load) for Covid-19, on different types of fomite surfaces as being a matter of hours, to a matter of days.

    More impressive was the study, ‘The effect of temperature on persistence
    of SARS‑CoV‑2 on common surfaces’, by Shane Riddell, Sarah Goldie, and others, where they determined D values (time taken to achieve a 90% reduction in titre ) for many fomite surfaces. And how long it takes Covid-19 on fomite surfaces to reach its ‘limit of detection’. That extends to about 28 days on vinyl, stainless steel etc depending on temperature and humidity in the environment, using TCID50 assay. At least ten HSM environmental ground samples (not absorbed into the ground) tested positive for Covid-19 on January 1, 2020. Its possible they could have been there since December 1, 2019 taking into account their possible D values, limit of detection not being reached, temperature in the HSM, and half life (reduction 50% in titre) , and their then current positive Ct values. People could have been foot tracking Covid-19 throughout HSM for some time. But F54 ground sample seems rather ‘fresh’ because high viral load and possible infectivity. Not like the other ground environmental samples.

    If Wuhan-Hu-1 NC-045512.2/ MN908947 got its 100% sequence identity with F13 wall coronavirus on December 26, 2019 the day his BALF was taken, or the day before. If that F13 coronavirus on December 26 is the same taken by Beijing CDC on January 1, 2020, should we not expect ‘low titers’ January 1, 2020 when Beijing CDC sampled the wall, due to viral decay on the wall, short half-life on the wall, no virus reproduction on the wall, heat, humidity, evaporation, and possible sunlight exposure while on the wall? Early HSM cases showing 100% identity and 100% query cover to Wuhan-Hu-1/ NC_045512.2 include MN996528.1 (Dec. 30 2019-WIV04) and MN996530.1 (Dec. 30, 2019-WIV06) in Jinyintan Hospital. Their coughing and sneezing settling velocity would have carried their Covid-19 particles to the HSM ground, unless strong east or west winds carried it to the wall. But most likely the F13 ‘stuff’ was in the HSM infecting before December 29, 2019, and the fresher stuff arrived December 30/31 a day before the Beijing CDC arrived, smashed or thrown against the wall , and not yet then subject to half-life decay. (The other early December 30, 2019 HSM genomes show 99.99% identity to NC_045512.2, and hence F13 according to BLAST.). Fresh ‘Covid-19 stuff’, same sequence identity and Ct count on January 1, 2020 as on December 26, 2019, time of first hospitalization. TCID50 was not used to test F13 and F54 for infectivity by Harbin Veterinary Center, could it have been? And if it couldn’t have been used, does that mean that F13 wall and F54 ground samples were not created by airborne or droplet transmission (particles less than 100 μm in diameter)? Does it matter? F13 wall sample particles too large to be droplet?

    Perhaps Ralph Baric should be consulted. He moved from using TCID50 techniques to determine the infectivity of mice inoculated with deadly MA15 coronavirus in 2007, to using plaque forming units to determine the infectivity of Venezuelan equinine encephalitis virus loaded with an S gene from the MERS coronavirus (a real chimera) inoculated into the humanized mice in 2016. Why this move from using TCID50 to determine infectivity in 2007, to using plaque forming units in 2016 to determine infectivity? Is it because the HSM wall sample, I mean the mice were inoculated with plaque forming unit chimera coronavirus from the beginning? —in the article, ‘A mouse model for MERS coronavirus-induced acute respiratory distress syndrome’, by Ralph Baric, Boyd Yount, and others, Nature Microbiology, (2016)

    If F13 wall sample had been created from a plaque forming unit based coronavirus, would TCID50 assay have worked here for Harbin’s analysis? But does F13 wall sample and F54 ground sample because of fresh infectious ‘high titers’ from some then living entity on December 31, 2019, transmitted about in ‘possibly’ an non-aerosoled manner (many stall workers, the sneezers and coughfers, were already hospitalized), perhaps something ‘slung’ or ‘smashed’ against the F13 wall overcoming all gravitational forces attempting to bring it to the ground, in conjunction with the ‘other’ 30 positive ‘waning’ samples which may indicate a prior 30 day presence of Covid-19 HSM, represent a possible ongoing reproduceable transmission event of the zoonotic kind for lineage B Covid-19 on January 1, 2020. This does not rule out the possibility that the possible zoonotic source had not been prior inoculated with Covid-19 as we know it, or that nebulization had occurred.

    For those who say laboratory spillover, then where did the Chinese laboratory get its copy of BtKy72 to integrate into the Covid-19 genome, the Atlanta CDC? The receptor binding motif (RBM) in the spike gene came from pangolin 2019, according to many scientists. Why did the highly infectious spillover person with lineage B only infect the HSM? Is BtKy72 a laboratory construct fiction created for a plandemic?

    The 2020 interest in F13 Huanan Seafood Market wall sample extended beyond the Los Alamos National laboratory (Alan Perelson) study group. In the March 2020, Cornell University article, ‘Phylogenetic Analysis and Structural Modeling of SARS-CoV-2 Spike Protein Reveals an Evolutionary Distinct and Proteolytically Sensitive Activation Loop’, by Gary Whittaker, and others, Figure 4 shows their extremely high interest in the F13 environmental samples. The Beijing CDC had taken three different F13 wall samples in January 2020. Cornell University study group compared all three F13 wall sample isolates S genes, to the S genes of pangolin 2019 and RaTG13.

    BetaCoV/Wuhan/IVDC-HB-envF13/2020 |EPI_ISL_408511,
    BetaCoV/Wuhan/ I V D C – H B – envF13– 2 0 /2020|EPI_ISL_408514, and
    BetaCoV/Wuhan/IVDC-HBe n v F 1 3 – 2 1 / 2 0 2 0 | E P I _ I S L _ 4 0 8 5 1 5.

    The three F13 wall samples have the same amino acids S gene sequences as pangolin 2019 and RaTG13. But only the three F13 environmental samples above had the RRAR amino acid furin cleavage sites. Not the 2019 pangolin nor RaTG13. Examine figure 4 in that article closely, and you will note the QTQTNS amino acids before the furin cleavage sites, which are being ‘lost’ in some cell passage experiments, exist in Covid-19, RaTG13, pangolin 19, and no other isolate.
    Gary Whittaker (Cornell University) has written at least ten articles trying to determine the origin of the Covid-19 furin cleavage site, and is very interested in F13 wall sample having the furin cleavage site. Both Whittaker and Los Alamos National laboratory (Alan Perelson) express supreme interest in F13 wall sample.
    Dr. Gao references all three F13 wall sample isolates in Figure 2B phylogenic tree of his February 2022 article. The last two of the F13 wall samples were taken by the Beijing CDC in the 3rd or 4th weeks in January 2020.
    Jonathan Pekar would not rule out the possibility of different sources for lineages A and B Covid-19.
    “the pandemic most likely began with at least two separate zoonotic transmissions starting no earlier than November 2019, with a lineage A virus jumping into humans after the introduction of a lineage B virus.”—in the article, ‘SARS-CoV-2 emergence very likely resulted from at least two zoonotic events’, by Jonathan Pekar, Andrew Magee, Michael Worobey, and many others

    Sudhir Kumar sees one progenitor coronavirus for lineages A and B, but lineage B lacked lineage A variants as ancestors.

    “These facts support the inference that coronaviruses lacking ‘A’ variants were the
    ancestors of Wuhan-1 and other genomes sampled in December 2019 in China. Therefore, we conclude that Wuhan-1 was not the direct ancestor of all the early coronavirus infections globally.”—–in the article, ‘An Evolutionary Portrait of the Progenitor SARS-CoV-2 and Its Dominant Offshoots in COVID-19 Pandemic’, Sudhir Kumar , Qiqing Tao, and others

    Kumar would say one progenitor source coronavirus for lineages A and B is likely, but lineage A coronaviruses were not ancestors to Wuhan-person-1 (lineage B). Why not just do a molecular analysis of the F13 wall sample, and make public the photographs taken of it, and the origin of the B lineage of Covid-19 may get solved. And standardize the need to determine the size (the diameter in micrometers) of the particles that contain the virons for any future outbreaks, and publish photos of the samples in their environmental as soon as possible.

    William Gallaher retired microbiologist at LSU created a PDF breakpoint map for NC_045512 genome which may be downloaded at, https://virological.org/t/tackling-rumors-of-a-suspicious-origin-of-ncov2019/384/19 at the bottom of that page. Shows all breakpoints for that NC_045512 genome that facilitate recombination events. .

    “the two most unique peptide sequences of SARS-CoV-2 related to its ability to infect human beings and spread rapaciously, the RBD and furin cleavage site, are unified by being preceded by the CAGAC/CAGAT motif (breakpoints)……CAGAC disrupts the processivity of the viral RNA polymerase complex down its template, and facilitates, albeit rarely, copy-choice errors capable of creating potentially dangerous recombinants to humankind….This breakpoint, 23577CAGAC in MERS and 23300CAGAC in SARS-CoV-2 (double-underlined in the pdf) is identical in both viruses at the same relative position in the S1 protein sequence, where it defines an identical dipeptide, QT.”

    Also see Gallagher’s article, ‘A palindromic RNA sequence as a common breakpoint contributor to copy‑choice recombination in SARS‑COV‑2’.
    Setting aside Gallagher’s theory that important location breakpoints in NC_045512.2 genome (and there appears to be at least one hundred) shows it clearly originated in nature, the reverse can be argued because restriction site enzymes have found at important locations in Covid-19 genome indicating possible laboratory origin. Gallaher’s coming out of retirement taking supreme interest in NC_045512.2 (Wuhan-Hu-1) regarding the origin of Covid-19 at such an early date March 2020 is amazing.

    A software BLAST run on NCBI (on three different dates) website shows NC_045512.2 is 100% nucleotide identical to later date isolates,

    OK372407.1 collected July 20, 2020 Tennessee
    MZ722702.1 collected 12/21/2020 Los Angeles
    MW562722.1 collected 12/31/2020 Utah
    Is this identicalness due a founder effect of a progenitor coronavirus, or does this identicalness mean that the source coronavirus for lineage B is still active and not restricted to HSM or its locale? Note this is a question, and not a conclusion.

    “a founder effect: a genetic bottleneck that occurs when, … a new type is established from a small isolated number of infections.”——- ‘Genetic Study Three Variants of SARS-Cov-2’ — sci-news.com

    “In the context of SARS-CoV-2, and other viral pathogens, founder effects result from a chance colonization event allowing a new population of viral lineages to emerge. When a chance colonization event occurs, the growth of a new population of viral lineages can give the impression that one lineage has a growth advantage over others (Rambaut et al., 2004; Ruan et al., 2021)”, in the article, “A repeat pattern of founder events for SARS-CoV-2 variants in Alaska”, by Tracie Haan, Lisa . Smith, and many others.

    Even though Dr Gao doesn’t reference strain, BetaCoV/Wuhan/IPBCAMS-WH-01/2019 EPI_ISL_402123 (GISAID) MT019529 (NCBI) age 65 years old, sample taken December 23, 2019, earliest known human sample. Pekar and Worobey re-analyzed that sequence as being identical to Wuhan-human-1.
    “reanalysis of sequence data from the earliest sampled viruses found that three previously reported mutations in IPBCAMS-WH-01 were spurious, and the genome was, in fact, identical to the Hu-1 reference genome” .”—in the article, ‘SARS-CoV-2 emergence very likely resulted from at least two zoonotic events’, by Jonathan Pekar, Andrew Magee, Michael Worobey, and many others

    IPBCAMS = Institute of Pathogen Biology, Chinese Academy of Medical Sciences in Beijing

    1. Thank you. Can you explain briefly what the ways are that SARS-CoV-2 could have gotten to the market? Before being in the animals butchered there?

  5. Not too much more to say. Purpose of comment is to bring general public knowledge of F13 wall sample importance, and surrounding concepts. Best science will come out later this year or next year, showing whether Covid-19 was created in a laboratory. And then how it moved from there, possibly to an animal. I sent an e-mail to plosbiology and biology_editors, attn: you, that has copy of January 29, 2020 Beijing article with its Figure 3 referencing BtKy72 bat coronavirus. Same for its attached supplement. Please don’t reply. Thanks for the two posts.

Leave a Reply

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

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Related Posts
Back to top