Human microRNAs translocate from host erythrocytes to Plasmodium falciparum parasites. Here, the techniques used to transfect synthetic microRNAs into host erythrocytes and isolate all RNAs from P. falciparum are described. In addition, this paper will detail a method of polysome isolation in P. falciparum to determine the ribosomal occupancy and translational potential of parasite transcripts.
The genetic variation responsible for the sickle cell allele (HbS) enables erythrocytes to resist infection by the malaria parasite, P. falciparum. The molecular basis of this resistance, which is known to be multifactorial, remains incompletely understood. Recent studies found that the differential expression of erythrocyte microRNAs, once translocated into malaria parasites, affect both gene regulation and parasite growth. These miRNAs were later shown to inhibit mRNA translation by forming a chimeric RNA transcript via 5' RNA fusion with discreet subsets of parasite mRNAs. Here, the techniques that were used to study the functional role and putative mechanism underlying erythrocyte microRNAs on the gene regulation and translational potential of P. falciparum, including the transfection of modified synthetic microRNAs into host erythrocytes, will be detailed. Finally, a polysome gradient method is used to determine the extent of translation of these transcripts. Together, these techniques allowed us to demonstrate that the dysregulated levels of erythrocyte microRNAs contribute to cell-intrinsic malaria resistance of sickle erythrocytes.
Malaria, caused by apicomplexan parasites of the genus Plasmodium, is the most common human parasitic disease, globally infecting approximately 200 million people each year and causing around 600,000 deaths1. Of the five Plasmodium species that infect humans, the most relevant to human disease are P. falciparum and P. vivax, due to their widespread distributions and potential for severe malaria complications. The life cycle of the malaria parasite requires infection of both mosquitoes and humans. When an infected mosquito bites a human, the parasites travel through the bloodstream to the liver, where an initial round of replication occurs. After merozoites rupture from the host hepatocyte, they infect nearby red blood cells, initiating either asexual or sexual replication. The asexual stage of replication, which lasts 48 hr in P. falciparum, is the focus of this study since it is both the source of most malaria symptoms and is easily recapitulated in vitro.
While a number of public health initiatives, including improved anti-malarial therapies, have somewhat reduced the burden of malaria globally, the continuing emergence of drug resistant parasites presents a problem for malaria control efforts. One area which may suggest new therapeutic approaches is the study on how various genetic variants confer resistance to malaria. In malaria-endemic regions, a variety of erythrocytic polymorphisms are quite common2,3. These mutations, with sickle cell being perhaps the most prominent, are often associated with substantial resistance to the onset of symptomatic malaria infection4. The underlying mechanisms by which they cause erythrocytes to resist malaria infection are incompletely understood. Parasitized erythrocytes with hemoglobin mutations are subject to enhanced phagocytosis through enhanced cellular rigidity and dehydration, which is associated with decreased invasion by P. falciparum5. The HbC allele also affects protein expression at the erythrocyte surface and with the remodeling of the cytoskeleton, further inhibiting parasite development6,7. Finally, P. falciparum grows poorly within homozygous sickle (HbSS) erythrocytes8,9 in vitro, suggesting intrinsic erythrocytic factors of malaria resistance. However, while all of these mechanisms appear to play a role, they do not fully explain the mechanisms behind sickle cell resistance to malaria.
One potential set of erythrocytic factors which remain poorly understood is the large pool of miRNA present within mature erythrocytes. MicroRNAs are small non-coding RNAs, 19-25 nt in size, which mediate translation and/or stability of target mRNAs by base pairing within the 3' UTR. They have been implicated in the control of mammalian immune responses, including the suppression of virus replication10, and were shown to confer resistance to viruses in plants They have also been shown to regulate several erythrocytic processes, including erythropoiesis11,12 and iron metabolism13. Previous studies identified an abundant and diverse population of erythrocytic miRNAs, whose expression was dramatically altered in HbSS erythrocytes14,15. Since mature erythrocytes lack active transcription and translation, the functional role of these erythrocyte miRNAs remains unclear. As significant material exchange occurs between the host cell and P. falciparum during the intraerythrocytic developmental cycle (IDC)16, it was speculated that the altered miRNA profile within HbS erythrocytes may directly contribute to cell-intrinsic malaria resistance.
These studies ultimately led to the development of a pipeline to isolate, identify and functionally study the role of human miRNA within the malaria parasite, P. falciparum, which indicated that those host/human miRNAs ultimately covalently fuse and then translationally repress parasite mRNA transcripts17. This provided an example of the first cross-species chimeric transcripts formed by trans-splicing and implicates that this miRNA-mRNA fusion could be occurring in other species, including other parasites. All trypanosome mRNAs are trans-spliced with a splice-leader (SL) to regulate the separation of polycistronic transcripts18. Since P. falciparum lacks orthologs for Dicer/Ago19,20, it is possible that erythrocyte miRNAs hijack similar SL machinery in P. falciparum to integrate into target genes. Recent studies in P. falciparum have in fact indicated the presence of 5' splice leader sequences21. This study details the methods that led to the discovery of human-parasite miRNA-mRNA fusion transcripts, including both transcriptomic and translational regulation techniques. The overall goals of these methods are to investigate the effects of small RNAs in the gene regulation, phenotypes and translation potential of P. falciparum transcripts.
The initial identification of human-parasite chimeric transcripts relied upon usage of RNA analysis techniques, such as real-time PCR, transcriptome sequencing and EST library capture, which included both total and small RNAs, rather than using techniques which only isolated small RNAs. Isolating all RNA together in one large pool, rather than separately, allowed the identification of both translocated human small RNAs in the parasite as well as the presence of these small RNA sequences as part of a larger sequence. This then required an analysis of the translation state of these fusion mRNAs to determine the functional consequences of these fusions.
While extensive efforts on characterization of the parasite's genome and transcriptome have added to the understanding of the parasite's biology22-25, far less is known about the translational regulation of the mRNA transcriptome across the life cycle of P. falciparum26. This limited understanding of the parasite's proteome has hindered both understanding of the parasite's biology and the ability to identify new targets for the next generation of anti-malarial therapeutics. This gap in the understanding of the parasite's cellular biology has persisted largely due to the lack of adequate techniques to investigate translational regulation in P. falciparum. One recent paper described the use of ribosomal footprinting of P. falciparum to determine the global translation status21. One well established measurement of translational potential of transcripts is the number of associated ribosomes determined by polysome profiling. However, when this technique is applied to P. falciparum, it is unable to recover most polysomes and captures predominantly monosomes. Recently, several groups27,28 have optimized P. falciparum polysome techniques by lysing the erythrocyte and parasite simultaneously to preserve the polysomes and characterize the ribosomal occupancy and translational potential of these malaria parasites during their asexual development in host red cells28.
Collectively, these methods demonstrate that the observed fusion of human miRNA and parasite mRNAs modulates parasite protein translation of those fusion mRNAs, which was demonstrated using previously reported methods27, and is a major determinant of malaria resistance in HbAS and HbSS erythrocytes17. These methods would be useful in any system looking to identify and functionally explore RNA splicing events, whether those fusion RNAs are within P. falciparum or other eukaryotic systems.
1: Isolation of Small-sized RNAs from P. falciparum During the IDC
2: Quantitation of Small-sized RNAs from P. falciparum During the IDC
3: Introduction of Small RNAs into Malaria Parasites
4: Determination of Erythrocyte microRNAs Effect Upon Parasite Infection Rate
5: Biotin-tagging and Elution of miRNA-mRNA Fusion Products
6: Polysome Separation to Determine the Ribosomal Occupancy of P. falciparum
Global profiling of human microRNA in P. falciparum
The techniques presented here were used to extract microRNAs from parasites in a variety of conditions. One thing to note is that RNA extractions for uninfected RBCs were performed as indicated in32, which often served as a reference point for the microRNA data presented in both this article and the original study29. Those previous studies also demonstrated that HbSS erythrocytes possessed far greater levels of certain miRNAs, miR-451 in particular. The techniques indicated above were sufficient to demonstrate the changes in miRNA abundance in HbS (HbAS or HbSS) infecting parasites. RNA samples were extracted as indicated above and normalized against 18S rRNA. Parasite samples from HbSS RBCs, for example, do show a significant increase in levels of several microRNAs, including miR-451 (Figure 1). The uptake of miRNA from the host erythrocyte is also consistent with previous studies34,35.
Transfection of HbAA erythrocytes with miR-451 increased HbAA miR-451 to levels similar to HbSS erythrocytes (Figure 1). Flow cytometry was then used to measure parasite infection, in order to assess the effect of miRNA translocation upon parasite growth. Based upon that, HbAA erythrocytes transfected with miR-451, miR-223 and miR-181a, along with an ssDNA oligo, were examined for the effects of microRNAs upon parasite growth. Transfection of ssDNA or miR-181a had no effect upon parasitemia (Figure 2A–D) while transfection with miR-451 markedly reduced parasitemia, which could be mitigated only by co-transfecting an antisense miR-451 oligo.
Blocking of microRNAs by transfection of 2'-O-Me antisense microRNAs demonstrated a reciprocal effect. HbSS erythrocytes, with naturally higher miR-451 levels, were transfected with antisense miR-451 or 2'-O-methyl oligonucleotides (Figure 3). Inhibition of miR-451 increased parasite growth in both HbSS erythrocytes (Figure 3C–F). Parasite percentages were calculated from at least 3 independent transfections, then averaged and presented as fold change relative to HbSS growth (Figure 3F).
Capture of miRNA-mRNA fusion RNAs by biotin capture assay
After transfection of 5'-desthiobiotin-miR-451 (5'Db-miR-451) and 5'Db-miR-181, as indicated in methods section 5, it was determined that the observed miRNA sequence originated with human microRNA. qRT-PCR analysis indicated that transfection of 5'Db-miR-451 resulted in enrichment of PKA-R fusion transcripts (Figure 4).
Polysome separation to determine the ribosomal occupancy of P. falciparum
A typical A254 trace is shown in Figure 5A (with northern confirmation of rRNA content in Figure 5B), with the fraction density increasing from left to right on the graph (i.e. lighter fractions to the left). The initial A254 is quite high due to absorbance by hemoglobin in the early fractions, but quickly drops to the baseline. The first small peak corresponds to the small ribosomal subunit (40S), followed quickly by the second peak, which is slightly larger and corresponds to the large ribosomal subunit (60S). The third peak, which is generally the largest for P. falciparum, corresponds to the monosome peak (80S). Due to its height, the 80S should be used as a guide to adjust the sensitivity of the UV detector for subsequent runs (Step 6.17). Also, due to the size of the 80S peak, it is common for the 60S signal to blur like a "shoulder" into the start of the monosome peak, if both the 60S and 80S signals are large.
Following the monosome peak are the peaks corresponding to polysomes. Each successive peak represents a polysome fraction representing progressively larger numbers of ribosomes (2, 3, 4, etc.). Each polysome peak is also smaller than the previous peak, decreasing in height by 2-3 fold with each successive polysome number. Typical runs will allow the resolution of polysomes composed of 5-7 ribosomes, though resolution of up to a 9-mer polysome has been observed in some runs with a large amount of starting material. Higher order polysomes compose the remainder of the trace, and cannot be resolved under these gradient conditions.
Figure 1: MiR-451 are elevated in HbSS erythrocytes. Intraparasitic miR-451 levels, as determined by real-time PCR and normalized against 18S rRNA, from parasites isolated from the indicated erythrocyte types. Values are mean ± SE.
Figure 2: Overexpression of miR-451 in normal erythrocytes inhibits parasite growth. (A–C) Representative FACS plots indicating infection rates, as a percentage of total RBC, are indicated by m1, for the shown erythrocyte types. (D) Composite infection rates derived from FACS and normalized against untransfected HbAA RBCs. Values in panel A-C are representative FACS plots while D is mean ± SE.
Figure 3: Inhibition of miR-451 in sickle cell erythrocytes elevates parasite growth. (A–E) Representative FACS plots indicating infection rates, as a percentage of total RBC, are indicated by m1, for the shown erythrocyte types. (F) Composite infection rates derived from FACS and plotted as a normalized value against untransfected HbAA RBCs. Values in panel A-E are representative FACS plots, F is mean ± SE.
Figure 4: Transfection of miR-451 enriches for PKA-R fusion transcripts and shows that the fused miRNA is erythrocytic in origin. Enrichment of the indicated desthiobiotin-conjugated miRNA-mRNA fusion transcripts that have been captured by streptavidin. Values were measured using real-time PCR and plotted as a relative value versus non-transfected parasites. Values are mean ± SE.
Figure 5: Parasite ribosomes can be isolated via sucrose gradient. (A–B) Example traces of a polysome profile extracted from Plasmodium falciparum. (A) Trace of a polysome profile extracted from Plasmodium falciparum, with the 40S, 60S, 80S and Polysome peaks (with #s indicating the number of ribosomes associated) displayed. (B) Northern blot of 18S and 28S rRNA across the ribosomal fractions. Values are mean ± SE.
HbS is one of the most common hemoglobin variants in malaria endemic areas, largely because it provides protection against severe malaria caused by P. falciparum. The techniques needed to characterize the role of human microRNA in the gene regulation of P. falciparum are detailed throughout this manuscript. By extracting total RNA in such a way as to include all small RNAs, and by performing a relatively straightforward parasite lysis procedure, these fusion RNAs were able to be identified through a variety of independent techniques.
These steps are relatively straightforward but are sensitive to problems and contamination if not followed properly. During the parasite extraction, it is critical to perform saponin lysis to remove contaminating host erythrocytes. This is particularly true during the saponin lysis step of the RNA isolation. The miRNA component of the host red blood cells is several orders of magnitude greater than that of the parasite, in part due to most red blood cells being uninfected, so it is vital to remove as much erythrocytic contamination as possible. It is also important to maintain the malaria culture in a healthy and proper stage (if a specific point in the infection cycle is required) to obtain an accurate miRNA profile. Similarly, during the phenol-chloroform extraction, it is vitally important to make sure the aqueous phases are clear and, if not, repeat the phenol-chloroform extraction to avoid any remaining contamination.
Also, in the desthiobiotin capture experiment, it is critical to elute with an excess of biotin and to use the minimum of streptavidin beads, so as to ensure proper specific elution. The elution of bound RNAs through competition with biotin, rather than complete denaturation of the beads themselves, served to dramatically reduce the background RNA enrichment. Finally, the desthiobiotin capture experiment was highlighted as one important way to demonstrate how these miRNAs were captured, but several other techniques were also used in the previous studies to demonstrate these fusion RNAs, such as northern blots and ribonuclease protection assays29. Such methods of using the transfected microRNAs to retrieve the modified transcripts may have broader application, such as the purification of associated protein complexes that may mediate the transport and processing of the transfected miRNAs.
While various genome-wide approaches have been used to study the malaria proteome and transcriptome22-25 , there are few methods to globally examine translational regulation in P. falciparum21,26. This methodological limitation precludes the global analysis of the translational regulation in P. falciparum. One of the most common experimental approaches to study translational regulation is polysome profiling, which assesses overall translational activity based on the ribosomes loaded onto specific mRNAs of interest. Detailed in this manuscript is an optimized polysome profiling method for use in malaria parasites that lyses both the infected erythrocyte and the parasite within. Previous methods did not recover the majority of polysomes and made it seem as though malaria parasites did not have substantial populations of polysomes. The use of a lysis buffer with a high concentration of potassium acetate and magnesium made it possible to lyse both red blood cells and parasites while solubilizing membrane-bound ribosomes to allow the capture of intact polysomes.
This cost-effective method of ribosomal purification and profiling will help to bridge the gap in existing knowledge between the malaria transcriptome and proteome and enable the genome-wide analysis of the translational state of P. falciparum. This approach has been used to compare the translation state of the early and late stage blood stage parasites28, which have tightly coordinated gene expression. Data on the ribosomal loading of transcripts can be readily obtained and analyzed to determine the density of ribosomes on specific mRNAs. Furthermore, when combined with sorbitol synchronization, changes in ribosomal density can be used to establish how translation is regulated throughout the parasite life cycle. Isolation of ribosomes using this approach requires a large amount of culture, which unfortunately limits the stages of the malaria life cycle in which it can be performed. The large amount of input material will also limit its use both to in vitro lab strains of Plasmodium falciparum, and may restrict its applicability to other infectious organisms. Anyone using this procedure will also be able to screen for genes that show unusual translation profiles, in particular transcripts which are not associated with ribosomes, reflecting alternative modes of gene regulation. Finally, this approach will enable us to evaluate how novel antimalarial agents affect protein translation and to identify translation-based mechanisms of drug resistance. This pipeline has great utility in identifying and determining the post-transcriptional regulation in many experimental systems.
The authors have nothing to disclose.
This research was funded by Doris Duke Charitable Foundation, Burroughs Wellcome Fund, NIH R21DK080994, Duke Chancellor’s pilot project fund, the Roche Foundation for Anemia Research and The Bill and Melinda Gates Foundation. G.L. is supported by Duke’s UPGG and the NIH (Grant # 5R01AI090141-03) and K.A.W. by the NSF Graduate Research Fellowship Program.
REAGENTS-All reagents must be RNAse-free. |
Diethyl pyrocarbonate (DEPC; Sigma, cat. no. D5758) |
DEPC-treated water (see REAGENT SETUP) |
Yoyo-1 DNA fluorescent dye (Thermo Fisher) |
Gene Pulser II (or comparible) electroporator (Bio-Rad) |
0.2 cm Electroporation cuvettes |
4 M potassium acetate (Mallinckrodt, cat. no. 6700) |
2 M potassium HEPES (pH 7.2; Sigma, cat. no. H0527) |
1 M magnesium acetate (Sigma, cat. no. M-2545) |
1 M dithiothreitol (DTT; Research Products International, cat. no. D11000) |
100 mM phenylmethylsulfonate fluoride (PMSF; dissolved in isopropanol; Sigma, cat. no. P7626) |
10% (v/v) Igepal CA-630 (Sigma-Aldrich, cat. no. I3021) |
10% (w/v) sodium deoxycholate (DOC; Sigma, cat. no. D-6750) |
Cycloheximide (CHX; Sigma, cat. no. C-7698) |
Sucrose (Mallinckrodt, cat. no. 8360-04) |
RNAseOUT (Invitrogen, cat. no. 100000840) |
RPMI-1640 w/ L-glutamine (Cellgro, cat. no. 10-040-CV) |
10% AlbuMAX I (w/v in sterile water) (Invitrogen, cat. no. 11020-039) |
Gentamicin (Gibco, cat. no. 15750-060)) |
HT supplement (100x) (Invitrogen, cat. no. 11067030) |
45% (w/v) sucrose (Sigma, cat no. G8769) |
1 M HEPES buffer (Gibco, cat. no. 15630-080) |
1x phosphate buffered saline (PBS; Cellgro, cat. no. 2109310CV) |
Corning 500 ml vacuum filter flask (VWR, cat. no. 430769) |
Glass slides (VWR, cat. no. 48311-702) |
Giemsa stain (50X) (VWR, cat no. m708-01) |
mirVana miRNA isolation kit (Ambion, ThermoFisher Scientific). |
5' Biotin conjugated mature miRNA (sequence varies by miRNA of interest) (Dharmacon) |
Biotin powder (Sigma-Aldrich) |
1M Potassium Chloride |
Streptavidin Sepharose High Performance Beads (GE Healthcare) |
Ribosome resuspension buffer (see REAGENT SETUP) |
Lysis buffer (see REAGENT SETUP) |
15% (w/v) sucrose gradient solution (see REAGENT SETUP) |
50% (w/v) sucrose gradient solution (see REAGENT SETUP) |
0.5 M sucrose cushion (see REAGENT SETUP) |
60% (w/v) sucrose chase solution (see REAGENT SETUP) |
Plasmodium cell culture media (see REAGENT SETUP) |
RNA capture Wash Buffer (see REAGENT SETUP) |
RNA Elution Buffer (see REAGENT SETUP) |
REAGENT SETUP |
Solutions |
Plasmodium cell culture media |
500 ml RPMI-1640 w/ L-glutamine |
0.5 ml gentamicin |
5 ml HT supplement |
2.5 ml 45% glucose |
25 ml 10% AlbuMAX I |
12.5 ml 1 M HEPES |
Sterile filter with 0.2 mm vacuum filter flask. |
Plasmodium cell culture media containing 10x CHX |
Same recipe as Plasmodium cell culture media above, with the addition of: |
2 mM cycloheximide |
Media is prepared as normal, then add CHX to a final concentration of 2 mM and filter. |
1x PBS containing cycloheximide |
Cycloheximide is added fresh to 1x PBS to a final concentration of 200 mM, and kept at 4 OC until use. |
Ribosome resuspension buffer |
400 mM potassium acetate |
25 mM potassium HEPES, pH 7.2 |
15 mM magnesium acetate |
200 mM cycloheximide (add fresh) |
1 mM DTT (add fresh) |
1 mM PMSF (add fresh) |
40 U/ml RNaseOUT (add fresh) |
Lysis buffer |
Same recipe as ribosome resuspension buffer above, with the addition of: |
1% (v/v) Igepal CA-630 |
0.5% (w/v) DOC |
Sucrose solutions |
Same recipe as ribosome resuspension buffer above, with the addition of varying amounts of sucrose: |
15% (w/v) sucrose to make 15% sucrose gradient solution |
50% (w/v) sucrose to make 50% sucrose gradient solution |
0.5 M sucrose to make 0.5 M sucrose cushion |
As above, cycloheximide, DTT, PMSF, and RNaseOUT must be added fresh. |
The 60% sucrose chase solution requires only sucrose and water, no other components |
60% (w/v) sucrose in DEPC-treated water to make 60% sucrose chase solution |
RNA Capture Wash Buffer |
20mM KCl |
5unit/ml Rnase Out (Invitrogen) |
RNA Capture Elution Buffer |
Same recipe as RNA Capture Wash Buffer, with the addition of: |
2mM Biotin |
EQUIPMENT |
SW55 Ti ultracentrifuge rotor (Beckman, cat. no. 342194) |
SW41 Ti ultracentrifuge rotor (Beckman, cat. no. 331362) |
Ultra-ClearTM ½ x 2 in (13 x 51 mm) ultracentrifuge tubes for the SW55 (Beckman, cat. no. 344057) |
Polyallomer 9/16 x 3 ½ in (14 x 89 mm) ultracentrifuge tubes for the SW41 (Beckman, cat. no. 331372) |
L8-80M ultracentrifuge (Beckman) |
Steel blunt syringe needle, 4 inch, 16G (Aldrich, cat. no. Z261378) |
1 ml syringe with 27G½ needle (Becton Dickinson, cat no. 309623) |
5 ml syringe (Becton Dickinson, cat. no. 309646) |
Parafilm |
Tube rotator, end-over-end |
Microcentrifuge, refrigerated |
Density gradient fractionation system (Teledyne Isco, cat. no. 69-3873-179) |
TracerDAQ (Measurement Computing) |
Eppendorf 5810R centrifuge |
Eppendorf A-4-81 rotor (Eppendorf, cat. no. 022638807) |