Plasmodium malariae is the second most abundant Plasmodium species in sub-Saharan Africa causing malaria in humans (Garnham, 1966; Coatney, 1971; Sutherland, 2016; Nundu et al., 2021; Culleton et al., 2023; Miezan et al., 2023). It is detected worldwide in many tropical and subtropical regions and is often found in co-infections with other malaria parasites, mostly P. falciparum (Zhou et al., 1998; Herman et al., 2023; Nguiffo-Nguete et al., 2023). Over the past decades a worldwide decline in malaria cases and deaths has been reported by the World Health Organization (WHO) (Organization, 2023). However, this decrease in case numbers has not been observed for P. malariae. On the contrary, there are many endemic countries reporting an increase in P. malariae prevalence (Cao et al., 2016; Yman et al., 2019; Agonhossou et al., 2022; Nainggolan et al., 2022). Due to improvements and increased usage of molecular diagnostics, the number of P. malariae infections has been shown to be substantially higher than previously described (Woldearegai et al., 2019; Mbama Ntabi et al., 2022; Nguiffo-Nguete et al., 2023). However, treatment as well as epidemiological studies and vaccine development are mainly directed towards P. falciparum and more recently also P. vivax, while other species are almost entirely neglected (Groger et al., 2022). A deeper understanding of the biology and epidemiology of the non-falciparum species would contribute to our understanding of transmission dynamics which is a prerequisite for achieving malaria elimination.

P. malariae causes a more benign form of malaria characterized by comparably mild symptoms and low parasitemia. However, serious complications have also been reported, mainly including severe anemia (Langford et al., 2015) and acute kidney injury (Ehrich and Eke, 2007). Infections are described as more chronic, persisting in the human host for a long time, indicating that P. malariae is successfully adapted to its human host.

A powerful tool to further characterize a parasite population and its transmission dynamics is the usage of genotyping markers. With this approach, diverse genetic regions are used to distinguish parasite genotypes. For P. malariae two genotyping marker panels have been identified previously (Bruce et al., 2007; Mathema et al., 2020). Both approaches focus on microsatellite markers, that represent genomic repetitive regions displaying a high genetic diversity. Few other P. malariae genes have been investigated additionally regarding their variability (Guimarães et al., 2015; Lalremruata et al., 2017; Srisutham et al., 2018; Putaporntip et al., 2022). In the present study, we aimed to establish a panel with a minimum number of genotyping markers for P. malariae yet with a high discriminative power. We assessed four markers from previous studies - two microsatellites, Pm02 and Pm09, as well as P. malariae thrombospondin-related anonymous protein (pmtrap), and P. malariae merozoite surface protein 1 fragment 2 (pmmsp1 F2) - the most promising candidate markers based on published diversity and successful amplification (Bruce et al., 2007; Srisutham et al., 2018; Mathema et al., 2020). We optimized the workflow, confirmed their diversity in a large set of P. malariae samples from Gabon, a malaria-endemic country in sub-Saharan Africa, and analysed dynamics and complexity of infections, by following-up asymptomatically infected individuals from Gabon in a tight schedule over one week.

The number of P. malariae infections in endemic countries has been shown to be substantially higher than previously acknowledged. Nevertheless, there are many open questions regarding the genetic epidemiology of this species. In the present study, we assessed four genotyping markers and selected two of them for establishment of a genotyping panel for P. malariae with a minimum number of markers and a high discriminative power.

First, we analysed the individual performance of the four markers using a sample set of 95 P. malariae infected individuals from Gabon. Our analysis showed that pmtrap, pmmsp1 F2 and Pm02 are the most diverse among the chosen loci in the population studied. They showed the highest number of distinct alleles - 11, 15 and 15, respectively -, whereas only 5 distinct alleles were found with Pm09 in the 95 samples. The mean MOI was highest for pmtrap (1.61), showing that this marker can very well reflect the multiplicity of P. malariae infections. The expected heterozygosity was highest for pmmsp1 F2 (0.81), demonstrating that this marker can detect a wide variety of different P. malariae genotypes. In terms of theses parameters, Pm02 performs comparatively well. However, this marker seems unsuitable for the fragment analysis using automated capillary gel electrophoresis with a resolution of 3-5 bp. The repeat size of only 4 bp is too small to reliably distinguish single alleles in infections with multiple genotypes, a limitation of this method. Similar results were found in another study on P. vivax genotyping, where automated capillary gel electrophoresis also reached its limit in distinguishing alleles that are close to each other in size (Manrique et al., 2015). The observed differences in the respective PCR efficiency of each marker can be attributed to various influencing factors. Low DNA concentrations in the sample are most likely the major reason why some samples were not amplified at all. We found a mismatch in the primer region in one of the pmmsp1 F2 sequences. This further underlines the high rate of polymorphisms in this region and might have led to the low amplification success for this marker. If primer binding is not efficient, less DNA is amplified. PCR conditions were thoroughly optimized to increase PCR efficiency to a maximum.

Overall, the range of heterozygosity (0.56 – 0.80) is comparable to other published marker panels by Bruce et al. (0.331 – 0.839) and Mathema et al. (0.530 – 0.922); the mean MOI for the four markers (1.11 – 1.61) is even slightly higher (Bruce et al., 2007; Mathema et al., 2020). However, both parameters depend on the characteristics of the investigated population, as geographical origin, clinical status, participants age, or seasonality. The samples used here for the validation of genetic diversity are from one geographical region (Lambaréné and surrounding villages, Gabon) and were collected within two months (study 1). The same is true for the P. malariae sample set from Myanmar of Mathema et al. In contrast Bruce et al. tested samples from diverse geographical origin (Thailand, Malawi, and Gambia), different clinical status and transmission seasons. Therefore, a direct comparison of the different marker panels using different settings can be misleading.

A notable observation for the pmmsp1 F2 marker was that comparatively more amplicons with a small size were detected and that the MOI was lower than for the other three markers. This effect of preferential amplification of shorter over longer amplicons has been seen in P. falciparum genotyping before, targeting the msp1 locus as well (Messerli et al., 2017). It can lead to an underestimation of the MOI. However, we saw no evidence of preferential amplification of smaller fragments for the pmmsp1 F2 locus ( Supplementary File S2 ). Also, it is unlikely that a low parasite density of the minor genotype leads to these findings in infections with multiple genotypes, as the other three markers have a higher MOI. Another explanation for the seen effect could be mutations in the primer region of the bigger alleles.

We optimized our genotyping approach based on automated capillary gel electrophoresis. To validate the method, we sequenced amplicons from samples that were infected with one genotype only and compared the amplicon lengths from both methods. The high concordance observed for all four markers shows that our chosen analysis method is very reliable ( Supplementary File S3 ). This approach has been proven efficient in previous Plasmodium genotyping approaches as well (Manrique et al., 2015; Nguyen et al., 2019; Tadele et al., 2022). General advantages of the method include the fast and high throughput, as well as the straightforward analysis. The high deviation of up to a maximum of 21 bp seen for pmmsp1 F2 can be explained by the lower resolution of 20 bp given by the used method. As the allele size range was bigger for this marker (92 – 694 bp), a method with a lower resolution (20 bp) had to be used.

Previous studies on P. malariae genotyping have been carried out in samples from South America and/or Southeast Asia (Guimarães et al., 2015; Srisutham et al., 2018; Putaporntip et al., 2022). The pmtrap and pmmsp1 sequences generated here originate from Gabon and therefore represent a valuable set of genotypes from Africa. A comparison of spatial dynamics of P. malariae genotypes reveals that some genotypes appear in all three continents mentioned: for pmtrap 3/10 identified genotypes have been found in Southeast Asia as well as Africa; for pmmsp1 F2 4/25 genotypes have been identified either in Southeast Asia and South America (2 genotypes), in South America and Africa (1 genotype) or in all three continents (1 genotype). However, most genotypes are shared among samples from the same region, forming spatial genotype clusters. As most of the available sequences are from Southeast Asia, the biggest cluster was identified from this region for both markers (4/10 using pmtrap and 14/25 using pmmsp1 F2). Another genotype cluster exclusively seen in samples from South America, consists of three genotypes and was identified using pmmsp1 F2. We found three new genotypes using both markers in our samples from Gabon, Africa, that were not seen in the sequencing data from any other continent before.

In the end, a combination of the two markers pmmsp1 F2 and pmtrap resulted in the panel with the best discriminative power. While pmmsp1 F2 can detect a high variety of diverse genotypes, pmtrap is able to best cover the multiplicity of P. malariae infections. In combination the two markers provide a powerful tool for studies on genetic diversity of P. malariae.

Based on the results from the sample set of the two-weekly follow-up study, both markers demonstrated a very consistent performance over time. Genotypes were detected consistently throughout the different sampling timepoints, which shows the method`s effectiveness. We detected up to 8 genotypes in one individual at the same timepoint. This surprisingly high number of circulating P. malariae genotypes in an asymptomatic population is particularly interesting. In a previous study from Gabon including symptomatic individuals, P. malariae was the species with the highest number of genotypes (Lalremruata et al., 2017). To which extend these findings are related to the clinical status needs further investigation.

Throughout the follow-up period, fluctuations of genotypes were evident, indicating that not all genotypes are present at every sampling occasion. This observation is likely because the amount of DNA from these minor genotypes approaches the limit of detection, so that they are detectable at one time point but not at the other. However, it is also possible that some of the appearing genotypes are new infections. The 72-hour cycle of P. malariae could not be seen in the pattern of detected genotypes from our data.

We followed the P. malariae infections for a limited time, providing valuable insights into their dynamics. As P. malariae infections are known for their longevity and to persist in the human host for even decades (Ciusa et al., 2022), studying infections over longer periods would provide even more information on the biology of this species. However, the high complexity of asymptomatic P. malariae infections with multiple genotypes and variable detection, shown by our findings, makes longitudinal investigations challenging to interpret since a genotype may not be detected on one sampling timepoint but be present in another. This could be eventually addressed with next generation sequencing (NGS) approaches that can increase sensitivity and the probability to detect minor clones, potentially adding further insights into P. malariae genotypes, infection dynamics and biology (Gruenberg et al., 2019; Wamae et al., 2022).

Some individuals in our study were still PCR-positive on day 14, seven days after treatment with artemether-lumefantrine. We do not think that these are treatment failures, but rather residual DNA that can still be detected with our highly sensitive PCR. Similar results were also found for P. falciparum (Inoue et al., 2024). In addition, no parasites were detected microscopically at those timepoints. However, in order to be certain that these are no treatment failures and that the individuals turn PCR negative later, follow-up samples from later timepoints would have been required.

We were able to successfully implement a new P. malariae genotyping panel consisting of only two markers, pmtrap and pmmsp1 F2. This set of markers will add a valuable tool to characterize the genetic diversity of P. malariae infections from different regions and to better elucidate the transmission dynamics of this neglected malarial parasite species. It can be implemented in many settings as its usage is straightforward, having a short turnaround time as only two markers are required.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/genbank/, OR829554-OR829580.

The studies involving humans were approved by Institutional Ethics Committee of CERMEL. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.

MRo: Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. KK: Investigation, Methodology, Validation, Writing – review & editing. LS: Investigation, Writing – review & editing. LB: Investigation, Writing – review & editing. MS: Investigation, Methodology, Writing – review & editing. AL: Methodology, Writing – review & editing. TW: Resources, Writing – review & editing. PB: Resources, Writing – review & editing. MG: Resources, Writing – review & editing. RZ: Resources, Writing – review & editing. DE: Resources, Writing – review & editing. GM-N: Resources, Writing – review & editing. SA: Resources, Writing – review & editing. MRa: Resources, Writing – review & editing. BM: Resources, Writing – review & editing. JI: Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing. JH: Conceptualization, Project Administration, Resources, Supervision, Writing – original draft, Writing – review & editing.