NEW YORK (GenomeWeb) – Researchers from the US, Malawi, and the UK recently debuted a single-cell sequencing strategy aimed at characterizing the diversity across individual malaria parasites — an approach that's expected to prove particularly useful for understanding features of so-called multiple genotype or multiple clone infections.
"Multiple clone infections are when we have more than one parasite strain infecting a single patient," the Texas Biomedical Research Institute's Ian Cheeseman told In Sequence. "These are incredibly common, especially where malaria is common."
In a study published in the advance, online edition of Genome Research earlier this month, Cheeseman led a team that described the single-cell sequencing method and its development. The paper also included findings from proof-of-principle experiments done using hundreds of individual Plasmodium falciparum or P. vivax cells obtained from either cell cultures or patient blood samples.
As it stands, the single-cell sequencing method reliably produces sequences spanning around 50 percent of each Plasmodium genome, revealing variant and more far-reaching haplotype profiles that coincide well with those generated on Illumina's VeraCode GoldenGate custom array that targets 96 SNPs.
"This is really the first chance we've had to have a look at the extent of the genetic diversity within infections with that type of haplotype information," Cheeseman said.
"One of the major benefits is that [with] increased resolution, we're starting to see things that we couldn't see before," he added. "We're starting to see that there are particular drug mutations that are co-inherited within the same parasite. So there are some parasites that bear more drug resistance mutations within a single infection than others within the same infection."
Along with genotyping and haplotype information on each cell, those involved in the study ultimately hope to generate complete genome sequences for each parasite, which Cheeseman called the "ultimate level of resolution for understanding differences between parasites."
Past research suggests such multiple-genotype infections may contribute to virulence and drug resistance related to mixing between parasites residing in the same vicinity. Such infections occur throughout regions where malaria is endemic, though Cheeseman noted that they are especially common in Africa, where the majority of malaria cases and related fatalities occur.
But while deep sequencing and genotyping experiments on Plasmodium isolates from individual cases can provide information on the overall set of variants present in that population of parasites, they don't allow for a haplotype-level look at the genetics of each cell contributing those variants.
"People have looked largely in bulk at infections: they take blood out of a patient, take DNA from all of the parasites simultaneously, and [are] not able to distinguish which mutations go with which parasite," Cheeseman said.
Prior efforts to study the genetics of single Plasmodium cells have involved growing individual parasites into clones in the lab, Cheeseman explained. But that approach is laborious and not feasible for the tricky-to-culture P. vivax parasites.
In an effort to perform high-throughput genomic studies of single Plasmodium cells, Cheeseman and his team considered methods used for sequencing DNA from individual cells in other clinical and research contexts, including efforts to assess genetic diversity and progression using sequences from individual tumor cells.
But there were Plasmodium-specific complications to consider, too. For instance, the malaria parasite genomes are notoriously repetitive and rich in adenine and guanine bases, Cheeseman noted, making them difficult to deal with in the sequencing setting.
Each parasite is present in a haploid form during the red blood cell infection stage of its life cycle, meaning just one copy of the genome is available for sequencing unless the parasite happens to be replicating when the sample is collected.
Moreover, the parasite's DNA is tucked away inside not only its own cell membranes, but also the membrane that surrounds each host red blood cell. That additional membrane presents a challenge when trying to extract parasite DNA without damaging it.
"It's tough to get in to that DNA safely, without damaging the DNA," Cheeseman said. "Once the DNA is damaged, it's essentially lost from the experiment."
He and his colleagues came across a solution to that problem somewhat serendipitously when they stored some samples in a -80C freezer overnight. As it turned out, freezing and thawing the samples helped in cracking open cell membranes and dramatically improved results from the team's sequencing experiments.
The freeze-thaw step has since become a routine part of the single-cell Plasmodium sequencing protocol, which begins with DNA dye- and flow cytometry-based techniques for identifying and isolating infected red blood cells.
After looking at existing single cell kits and protocols — including more than a dozen different combinations of DNA dye, amplification, sequencing, and analysis approaches — the team settled on a single-cell sequencing workflow that begins with parasite DNA staining by the Vybrant DyeCycle Green dye.
Since human red blood cells do not contain nuclei, Cheeseman noted, any nuclear DNA-containing cells from patient blood samples that are the appropriate size and shape are considered likely to contain Plasmodium parasites.
Once parasite DNA is extracted from each cell, it gets amplified by whole-genome amplification and fed into an Illumina library preparation protocol that uses the Kappa HiFi enzyme rather than the standard Illumina PCR enzyme to bump up coverage across the AT-rich parasite genomes.
To avoid swamping out some sequences in the genome, Cheeseman noted that the researchers typically try to limit the rounds of whole genome amplification as much as possible while still producing enough DNA for sequencing.
As they reported in Genome Research, the researchers attempted single-cell sequencing on hundreds of P. falciparum and P. vivax cells while developing, tweaking, and validating the method. These included cells from patient samples, cultured parasites, and combinations of genetically characterized Plasmodium strains mixed together in the lab.
The latter laboratory lines "were really used during the optimization to make absolutely sure that when we took a single cell, there was no contamination from other genotypes present in the infection," Cheeseman said.
After demonstrating that it could use single-cell sequencing to obtain genotypes that matched one strain or the other in a mixed-parasite sample, the group became more confident in applying the same approach to more uncharacterized, patient samples, he explained.
For example, when the team took a crack at sequencing a subset of cells from 95 single-species (either P. falciparum or P. vivax) infections that had already been genotyped with the VeraCode array, it found near perfect agreement between the variant patterns obtained with each approach.
In those experiments, the sequencing-based scheme appeared to accurately pick up on the distinct haplotype patterns present in various parasites in each sample, highlighting its potential usefulness for those interested in understanding the details of parasite diversity.
The researchers also looked at how well the sequencing method fares when attempting to produce whole-genome sequences from individual Plasmodium parasites. Using Illumina's HiSeq 2000 instrument, they did single-cell sequencing on nearly two-dozen P. vivax or P. falciparum cells obtained from patient infections.
Reads generated for each of the 11 sequenced P. vivax cells were subsequently compared to the reference genome for that species, covering almost one-third of it to a depth of 10-fold or higher.
Similar experiments on cells from P. falciparum-infected individuals or from lab samples of that species generated slightly higher coverage of the corresponding reference genome, with more than half of the genome represented to a depth of 10-fold or more.
The reason for the discrepancy in coverage for each species is still unclear, Cheeseman said, noting that it may diminish as more and more single cells are sequenced from each species.
Nevertheless, the variant patterns identified in just a handful of the sequenced P. vivax or P. falciparum cells proved useful for beginning to unravel haplotype blocks, the researchers reported, revealing relatedness and shared ancestry amongst the parasites from each species. Similar experiments pointed to the presence of varied drug resistance allele combinations within individual parasite cells.
From these and other findings so far, the study's authors argued that "our single cell genomics approach can be used to generate parasite genome sequence directly from patient blood to unravel the complexity of P. vivax and P. falciparum infections."
"These methods open the door for large-scale analysis of within-host variation of malaria infections, and reveal information on relatedness and drug resistance haplotypes that is inaccessible by conventional sequencing of infections."
The team is considering ways of improving the approach, including possible tweaks to the oligos used to amplify parasite DNA. It noted that additional resolution would also be required to reliably call non-SNP forms of variation, such as copy number variants, in the Plasmodium genomes.
The researchers used Illumina instruments for their proof-of-principle experiments, though Cheeseman said there is "absolutely no reason why this [can't be] compatible more broadly with other platforms."
He and his colleagues plan to begin applying the sequencing strategy to better understand malaria parasite biology — from ways in which the parasites are transmitted to information on how genetic diversity is represented within individual infections from malaria-prone locations, starting in Africa.
"Our next steps are very much focused on looking into the within-host population structure of infections in Africa, where we expect to see the greatest diversity within single hosts," he said.