Virus mutations of COVID-19 pose a challenge as latest wave of coronavirus infection in India highlights need to ramp up genome sequencing efforts

The rising wave of coronavirus infection, which began in Maharashtra, and is spreading to hitherto unaffected areas, underscores the importance of greatly stepping up the genome sequencing efforts and looking specifically for mutations of concern.

Undergraduate students, who have tested positive for COVID-19, appearing for final year examination at Maharani College of Arts & Science in Bengaluru on March 19

AFTER touching the lowest number of confirmed COVID-19 cases at around 9,100 on February 9, India is witnessing a continuous and worrisome increase (Fig 1a & b). The number of new cases on March 18 was 39,726 with the seven-day moving average on that day being around 29,355, which is of the same order as what it was around December 11, 2020. While, given the graph trend, it is reasonable to ask if we are heading towards a second wave of infection spread, which interestingly the country has not seen unlike many countries of the West, it may be a little too early to say.

Manindra Agrawal, Director of Indian Institute of Technology-Kanpur, one of the authors of the National ‘Super Model’ Initiative of the Department of Science and Technology, was quoted by The Hindu in the last week of February as saying that, on the basis of the model, the rising wave of infections, which began in Maharashtra, was unlikely to last beyond three weeks and that the number of cases would at most rise to 11.3-11.4 million cases by April 2021.

This, according to Agrawal, was because, according to Super Model, 60 per cent of India’s population had already been exposed to the infection and the country was close to achieving herd immunity if it had not already. While this is contrary to the 21 per cent seropositivity rate found by the Third National Serological Survey conducted by the Indian Council of Medical Research (ICMR) between December 17 and January 8, Agrawal argued that since the serosurvey was based on detection of antibodies in people it did not reveal the actual levels of immunity. “We know that antibody levels can wane but still be enough to confer immunity. Given the [recent] relaxations in movements… only 20 per cent exposure cannot explain the confirmed symptomatic infections, active cases and recoveries that we now see,” he said.

We are in mid-March and the number of cases (March 18) has already touched 11.5 million. With many underlying assumptions, models can go wrong; there is nothing surprising in that. But what is important to note is that the increase in March has been steeper, with Maharashtra continuing to contribute the highest share of nearly 63.2 per cent to the number of active cases as of March 18 (Fig. 2), according to the Ministry of Health and Family Welfare (MoHFW).

The Manaus cases

The resurgence in Manaus, Brazil, in January is a case in point. Infections in the city peaked in April 2020 and then decreased dramatically and by October-November 2020 it had flattened out. To boot, a serosurvey based on blood donors suggested that about 76 per cent of the city’s population had antibodies and scientists opined that the city had achieved herd immunity and a second wave was impossible.

While many arguments have been advanced as to why the serosurvey based on blood donors—who probably represent a subset of the population, and the sample would be biased— may have overestimated the positivity rate, the counter view is that it could not have been so low as to witness such a resurgence. Of course, since more than eight months had passed when the second wave hit Manaus, the immunity would have waned. However, a Lancet study published on January 27 noted that waning immunity alone was unlikely to explain the scale of resurgence. It further stated that new SARS-CoV-2 variants might drive resurgence in places where they circulated if they had increased transmissibility compared with the earlier variant(s), particularly if they had “escape mutation(s)” that could evade virus neutralising antibodies.

The three recently detected SARS-CoV-2 variants—the British B.1.1.7, the South African B.1.351 and the Brazilian P.1 “are unusually divergent and each possesses a unique constellation of mutations of potential biological importance”, the study said. Of these, two (B.1.1.7 and P.1) are known to be circulating in Brazil and P.1 was isolated in Manaus on January 12. Genomic analysis has shown that the P.1 variant has accumulated 10 unique spike protein mutations. These include N501Y (common to all the three variants), which is believed to be much more transmissible, as well as E484K, which is supposed to be immunity evading. There is also in vitro evidence suggesting that the presence of the E484K mutation reduces neutralisation by polyclonal antibodies.

One case of SARS-CoV-2 reinfection in Manaus has been associated with the P.1 variant. Further, according to Lancet , a new variant called P.2 has been identified in Manaus, which too carries the escape mutation E484K. Two other cases of reinfection have been reported in Manaus with the P.2 variant. These empirical associations suggest a possible linkage between the new variants and resurgence. Of course, this can be unambiguously proved only through genomic sequencing of isolates from new infections at a much higher rate than during the earlier wave and see if the new variants are indeed causing community transmission.

Role of mutations

During a pandemic, as the virus passes through many hosts, thousands of mutations occur and accumulate in a virus due to random errors during its replication cycles. Many just disappear. Only those that render some evolutionary advantage to the virus for its continued replication remain. Some of these could turn out to be more infective and virulent, which calls for constant vigil.

We had seen in May 2020 ( Frontline , May 22, 2020) the emergence of a variant, characterised by the defining mutation D614G, replacing the original Wuhan strain almost globally. The mutation, D614 G, codes for the amino acid G (glycine) in place of the original D (aspartic acid) at a specific genomic site 614 on the spike (S) protein, the critical viral protein that enables the virus’ entry into human cells. Glycine, being less bulky than aspartic acid, affords more flexibility to the virus, enabling it to bind more efficiently, and endows it with increased infectivity and transmissibility to make it the dominant strain in the world.

Fortunately, D614G lies outside the region responsible for producing neutralising antibodies and, therefore, has not been a cause for concern for severity of the disease and effectiveness of vaccines and therapeutics. But it underscores the importance of not underestimating or underplaying the role of mutations. Continued investigation and monitoring of these is necessary to understand sudden localised infection spreads and to control them.

Genome sequencing

Until December 2020, for example, India saw over 7,000 mutations in the 6,400 genomes submitted in the open Indian database. On December 30, the government launched the Indian SARS-CoV-2 Genomic Consortia (INSACOG) comprising 10 laboratories. The consortium is to monitor the genomic variations in SARS-CoV-2 on a regular basis through the multi-laboratory network.

So far, over 6,000 genomes from isolates are believed to have been sequenced by this consortium. This is of the same order of isolates from March to December 2020. But INSACOG has apparently not made public data of these genomes, particularly about the kind of variants that have shown up in different populations and their rates among the new infections following the surge.

In fact, researchers outside the network have privately voiced their frustration in not being able to access the consortium data and analyse the new genomes because these have not been made accessible to other labs. However, R.K. Mishra, Director of the Centre for Cellular and Molecular Biology (CCMB), Hyderabad, which is part of the consortium, told this correspondent that this was essentially due to the early hiccups in creating appropriate servers and data sharing platforms, but since then things have improved and genome data have become accessible. However, other researchers are yet to confirm this.

According to a Parliament question on March 18, the Minister of State for Health, Ashwini Choubey, said 400 cases with the United Kingdom, South African and Brazilian variants had been detected in the country so far since these variants spread to India in January through travellers from overseas. Of these, 158 had been detected in the two weeks after March 4, the Minister said. This should be of concern. However, there was no information about how many of each variant had been detected.

On February 23, the ICMR issued a press release, which had a stand-alone statement: “There is no direct relation between the recent surge in COVID-19 cases in Maharashtra and some other States with the strains containing the mutations N440K and E484Q.” Why only these mutations were picked out and not the others, particularly the U.K., South African and Brazilian strains, especially in view of the statement in Parliament, is not clear. In fact, while E484K has been widely discussed in the context of South African and Brazilian variants, particularly because of its immune defence evading ability, this was the first time one was hearing of mutation E484Q in the Indian context.

The ICMR further said that while these mutations were not specific to India, they had been found earlier in some States. Apparently, the E484Q mutation had been seen as early as March and July 2020 in four samples from Maharashtra, and N440K had been reported on 13 different occasions between May and September 2020 in Telangana, Andhra Pradesh and Assam. E484Q was not discussed at all then probably because it was found in very few samples and also because it was not seen to persist.

However, according to a recent work by a group of scientists from Seattle, Washington State, which was posted in the online preprint database bioRxiv, the site E484 in the receptor binding domain (RBD) of the S protein appears significant and the replacement of E (glutamic acid) to three other amino acids K (lysine), Q (glutamine) and P (proline) found in different genomes decreased the virus neutralising ability by over 10 per cent and even 35-60 per cent in some cases.

It is perhaps for this reason that the ICMR statement had flagged E484Q rather than E484K. The latter has so far shown up only in three samples in Maharashtra.

In the wake of the resurgence, a three-member expert team constituted by the MoHFW visited some key towns (Nagpur, Amravati, Yavatmal, Pune, Mumbai and Thane) in Maharashtra on March 1-2 to assess COVID management in the State. The team’s report seems to be based on its brief interactions with some officials, field visits and visits to some hospitals. It did not seem to find it necessary to include inputs from epidemiological and biological investigations by research institutes. The report was cursory and descriptive and sadly lacked numbers and data.

The report’s basic conclusion was: “While the exact causes of surge are not known—since laxity in COVID behaviour is not specific to the State—the possible factors are COVID inappropriate behaviour due to lack of fear of disease, pandemic fatigue; miss outs and super spreaders; and enhanced aggregations due to recent gram panchayat elections, marriage season and opening of schools, crowded public transport, etc.” It also noted that the virus was spreading to hitherto unaffected areas and most cases were asymptomatic, that there was a sense that the current wave is less virulent, and that the health machinery may have become lax after cases came down after September.

With regard to Nagpur, the report noted that serosurveillance there had shown a high positivity rate in the population (average 50 per cent). This, it said, may be leading to herd immunity and to the infection spreading to second-tier towns such as Amravati and even to rural areas, where people had hitherto believed that COVID was a disease of urban areas and so did not follow protocols. For Pune, it made this general statement without any substantiation, especially when the ICMR had only a week before the release of the report said otherwise: “Possibility of indigenous virus mutation making it more infectious and less virulent.”

One of the chief goals of the INSACOG is to sequence 5 per cent of all the positive cases in the country. Adding up the 6,400 of the earlier database and the 6000+ of INSACOG, only around 12,500 genomes are available for analyses, which is just 0.1 per cent way below the INSACOG target. As a recent review in Journal of Biosciences by scientists of the CCMB observed: “India has so far not been sequencing SARS-CoV-2 isolates to full capacity…. Exploiting advances in genomic epidemiology by monitoring and increasing sequencing efforts following local spikes will go a long way in staying on top of mutations of concern…”

Significantly, the review noted: “The variant landscape is mostly concordant across States in India. However, there appear to be instances of high representation of specific variants in selected States”. However, there is a caveat to this statement as the review pointed out. The over-representation in particular States is also a function of the States sequencing larger number of samples and submitting most genomes to the Indian and international databases. The review called for increased sequencing particularly from under-represented States to get a clearer picture of country-wide spread of COVID.

The mutation N440K, which finds mention in the ICMR statement of February 23, was first identified in Andhra Pradesh in late June last year and has been present in about 42 per cent of samples collected from Andhra Pradesh, which constitutes 6 per cent of all samples collected in India, since then, according to the review. However, according to reliable sources, this mutation is now showing up in samples collected from Kerala during the recent surge. This will be besides what the review itself has noted about Kerala. A study on 200 samples from Kerala up to December 2020 apparently has revealed 89 variants, of which four are novel, which were exclusive to Kerala and not present in other parts of the country. This study is now being scaled up by the State government to gain insights into the transmission. Such an effort, the review said, needed to be mounted across all States.

According to the review, the top genes where most mutations have been seen in India are, besides the S-protein, ORF1a and ORF1b (Fig.3). Some of the recently identified Spike mutations that could be of concern include N439K, N440K, Q493K and E484K. The other mutations, as of now, are absent in Indian isolates. The review has identified key Spike mutations of global concern which should be prioritised for surveillance in the Indian context as well (Table 1). The mutations A222V and S477N have also been seen in Indian samples.

The mutation N501Y, which is characteristic of the U.K., South African and the Brazilian variants and has been reported to have increased transmissibility (by as much as 70 per cent in the U.K.), has, however, not been seen in India, at least in genomes from the public database that have been analysed so far. According to the review, the mutation P681H, which has been found in association with N501Y in the U.K. strain, has been seen 54 Indian isolates, which includes 4 per cent of samples from Maharashtra.

The site of the mutation P681H being close to the cleavage site on the S-protein (which facilitates the virus entry) could be of concern, the review observed. Though there seems to be no evidence of community transmission of the U.K. variant in India, strangely enough the ICMR statement makes no mention of this mutation. With regard to double mutations of concern (as has been seen in the overseas variants), the review observed: “Such multiple mutations can drive higher transmission and/or immune escape and are strong contenders for enabling reinfections and lowering vaccine efficacy.”

Variants of concern

On March 17, The Telegraph (India) and The New Indian Express reported, on the basis of inputs from anonymous but reliable sources at the National Centre for Disease Control (NCDC), a member of INSACOG, that 20 per cent of 200 samples analysed from Maharashtra recently carried a double mutation which includes the mutations of concern E484Q (which, as mentioned earlier, has been associated with higher immune evading ability) and L452R. The latter mutation is a relatively unknown one even globally. However, on March 9 scientists from the University of California, San Francisco, posted a paper on the web repository medRxiv, which has discussed the finding of a new variant from the whole genome sequencing of 2,172 samples in the State of California. The variant carries three Spike mutations, which significantly includes the mutation L452R in the S-protein’s RBD region. These are new mutations not found in the other variants of concern (VOCs) from the U.K., South Africa and Brazil. This novel variant is named B.1.427/B.1.429 and has also been tagged as a VOC.

According to researchers, the variant emerged around May 2020 and increased from an insignificant number to over 50 per cent of the cases between September 1, 2020, and January 29, 2021. This variant is said to cause 19-24 per cent increased transmissibility, a twofold increased viral shedding by patients, increased infection in pseudo-virus cell cultures and lung organoids (cell cultures of specific organs derived from stem cells) caused specifically by L452R mutation and 4 to 6.7-fold and twofold reduced susceptibility to neutralising antibodies respectively in convalescent patients and vaccine recipients.

In the light of this finding, detecting a significant percentage of samples carrying the same mutation (along with another mutation of concern E484Q) in Maharashtra should raise the antennae of the government teams and other researchers engaged in finding the cause of the new surge. The need of the hour is to step up the whole genome sequencing of samples drawn from the newly infected populations of regions where there is evidence of resurgence. “NCDC may have flagged this double mutation because 20 per cent is not insignificant,” said Mishra of the CCMB. “But it is still too early to say whether this indeed is the cause for the resurgence. We have to closely watch the situation for the next couple of weeks, and analyse more samples, before we can say anything conclusively,” he added.

This again underscores the importance of greatly stepping up the sequencing efforts and looking specifically for the mutations highlighted as mutations of concern by the CCMB review (Table 1) and mutations E484Q and L452R that have most recently been flagged by the NCDC, rather than issue dismissive statements as the ICMR has done on the basis of small data that may finally turn out to be premature.