Synthetic RNA-protein modules integrated with native translation mechanisms to control gene expression in malaria parasites

Abstract

Synthetic posttranscriptional regulation of gene expression is important for understanding fundamental biology and programming new cellular processes in synthetic biology. Previous strategies for regulating translation in eukaryotes have focused on disrupting individual steps in translation, including initiation and mRNA cleavage. In emphasizing modularity and cross-organism functionality, these systems are designed to operate orthogonally to native control mechanisms. Here we introduce a broadly applicable strategy for robustly controlling protein translation by integrating synthetic translational control via a small-molecule-regulated RNA-protein module with native mechanisms that simultaneously regulate multiple facets of cellular RNA fate. We demonstrate that this strategy reduces 'leakiness' to improve overall expression dynamic range, and can be implemented without sacrificing modularity and cross-organism functionality. We illustrate this in Saccharomyces cerevisae and the non-model human malarial parasite, Plasmodium falciparum. Given the limited functional genetics toolkit available for P. falciparum, we establish the utility of this strategy for defining essential genes.

ARTICLE

Received 3 Mar 2015 | Accepted 15 Jan 2016 | Published 1 Mar 2016

Synthetic posttranscriptional regulation of gene expression is important for understanding fundamental biology and programming new cellular processes in synthetic biology. Previous strategies for regulating translation in eukaryotes have focused on disrupting individual steps in translation, including initiation and mRNA cleavage. In emphasizing modularity and cross-organism functionality, these systems are designed to operate orthogonally to native control mechanisms. Here we introduce a broadly applicable strategy for robustly controlling protein translation by integrating synthetic translational control via a small-molecule-regulated RNAprotein module with native mechanisms that simultaneously regulate multiple facets of cellular RNA fate. We demonstrate that this strategy reduces leakiness to improve overall expression dynamic range, and can be implemented without sacricing modularity and cross-organism functionality. We illustrate this in Saccharomyces cerevisae and the non-model human malarial parasite, Plasmodium falciparum. Given the limited functional genetics toolkit available for P. falciparum, we establish the utility of this strategy for dening essential genes.

DOI: 10.1038/ncomms10727 OPEN

Synthetic RNAprotein modules integrated with native translation mechanisms to control gene expression in malaria parasites

Suresh M. Ganesan1, Alejandra Falla1, Stephen J. Goldess1,w, Armiyaw S. Nasamu1 & Jacquin C. Niles1

1 Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, 02139 Massachusetts, USA. w Present address: AbVitro Inc., 27 Drydock Avenue, Boston, Massachusetts 02210, USA. Correspondence and requests for materials should be addressed to J.C.N. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727

The ability to precisely, reversibly and temporally regulate gene expression is essential for gaining insights into how biological systems function as well as designing biological

circuits that direct novel cellular behaviours in synthetic biology applications. In model eukaryotic systems, this can be exibly achieved at the transcriptional, translational and posttranslational levels. When available, transcription-based regulatory systems are favoured, as they provide access to large dynamic ranges that allow gene expression to be titrated to a level appropriate for the specic application13. Increasingly, translational and posttranslational control systems are being used to diversify the gene regulation toolkit and ne-tune expression. However, compared with transcription-based systems, the dynamic ranges attainable with these systems are relatively modest, and higher basal expression levels are observed in the repressed state48. This limits their utility as robust and general solutions for controlling gene expression. Thus, straightforward strategies that extend the dynamic range of posttranscriptional control systems would signicantly increase the exibility with which these systems can be broadly used as effective tools.

Towards this goal, we have focused on enhancing the regulatory dynamic range attainable during synthetic control of translation. Current designs emphasize achieving function independent of host cell regulatory mechanisms. Thus, small-molecule-regulated aptamers and riboswitches that directly block translation initiation, and natural or engineered ribozymes that induce cis-cleavage of target transcripts have dominated system design4,68. These approaches enable modular systems that function independently of host cell mechanisms, and, in principle, are broadly applicable in different organisms and cell types. While conceptually simple, this framework is restrictive, as it limits opportunities to achieve improved functionality through direct integration of synthetic parts with the diverse mechanisms cells have evolved to robustly regulate translation.

Natively, translation is controlled at multiple levels. Initiation via cap-dependent recruitment of factors critical for ribosome assembly, messenger RNA (mRNA) subcellular sequestration and turnover are key and highly regulated steps. Many proteins that play key roles in regulating these processes have been previously characterized9,10. For example, eukaryotic translation requires recognition of the mRNA cap structure by eIF4E, which recruits the eIF4F complex and leads ultimately to ribosome assembly and translation. However, the efciency of this process is modulated by eIF4E-binding proteins, which block eIF4F recruitment and inhibit translation. Similarly, decapping and deadenylation enzymes, such as Dcp1p and Pop2p, respectively, in S. cerevisiae, that control mRNA turnover have been characterized11. In many instances, homologues of these characterized factors are found widely across eukaryotes, suggesting they play highly conserved roles. Thus, synthetic gene regulation schemes can potentially be signicantly enhanced if effective strategies for integration with native translation control components can be devised. Furthermore, conservation of the core translation machinery and regulatory components across diverse organisms suggest that this should be attainable without sacricing design modularity and portability between different cell types and organisms.

Herein, we describe a straightforward strategy for directly integrating synthetic and native translation regulatory components to achieve improved inducible control of gene expression. By fusing the Tet repressor protein (TetR) to various translation regulatory effectors, we achieve robust anhydrotetracycline (aTc)-dependent regulation of target transcripts via TetR aptamers genetically encoded in either their 50- or 30-untranslated regions (UTRs). The regulatory dynamic ranges achieved come close to that of some transcription-based approaches when these

aptamer elements ank the target open reading frame. We also illustrate the generality of our approach across two unrelated organisms, namely, the model yeast S. cerevisiae and the human malarial parasite, P. falciparum. In the latter organism, where few tools are available for studying gene function, we show that this technology can be used to stringently classify genes essential for parasite survival.

ResultsMolecular design principles. Existing strategies for achieving synthetic control of eukaryotic translation rely mostly on aptamersmall molecule interactions6, inducible ribozymes7,8 and RNAprotein interactions4,5,12. These control systems are uni-dimensional in design in that the required RNA aptamer element is strategically positioned within a target transcript to predominantly direct a single outcome, such as disruption of translation initiation or induced RNA degradation (Fig. 1a). Such designs emphasize modularity and portability, but likely at the expense of achieving maximal robustness, by limiting intimate coupling between the synthetic module and native regulatory mechanisms.

Examination of natural translation control mechanisms, however, reveals that a diverse set of activities simultaneously converge on a regulated transcript to achieve stringent control over its translation9,10. We hypothesized that synthetic translation control systems that more closely recapitulate natures multi-dimensional strategy might inherently be more robust, and achieve signicantly improved regulatory outcomes than existing uni-dimensional designs. To establish proof of concept, we reasoned that a broadly useful framework should consist of a foundational RNAprotein interaction module in which: (1) a small molecule directly controls the interaction, and allows for transcription-independent translational control; and(2) the protein component is readily engineered to create fusions that bridge the foundational RNAprotein module with native cellular translation control components (Fig. 1b). Successfully validating this approach would provide immediate access to diverse mechanisms used natively to stringently control the cellular fate of RNA, and establish a new and easily adaptable paradigm for designing multi-dimensional translation control systems with improved functional characteristics.

TetR fusion proteins conditionally regulate translation. We selected a previously dened aptamer-TetR protein module that is regulated by tetracycline analogues13 as the foundational framework for this study. This module has been used in a uni-dimensional context (Fig. 1a, right panel) to control translation in the model organism, S. cerevisiae5 and the unrelated human malarial pathogen, P. falciparum14. As such, direct quantitative comparisons between uni- and multidimensional implementation can be made to rigorously quantify improvements in functionality achieved via the latter design. Critical to addressing generalizability of this concept, this selection also facilitates assessment of this multi-dimensional regulatory framework to deliver enhanced translational control in an unrelated organism.

In choosing TetR fusion partners, we emphasized two criteria that the selected proteins must satisfy. Namely, these must(1) have been previously implicated in regulating translation initiation, mRNA stability and/or turnover or mRNA sequestration; and (2) have clearly identiable homologues in other eukaryotes, and, in particular, P. falciparum. The latter criterion provides a stringent test of our central principle in a non-model organism where very little basic mechanistic evidence supporting

2 NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727 ARTICLE

a

Aptamer-small-molecule modules

Ribozyme-small-molecule/protein modules

Aptamer-protein-small-molecule module

TetR

Uni-dimensional

(host cell independent)

designs

Target

AN

Target

AN

Target

AN

AN

Translation ON

Translation OFF

Small-molecule regulator

Translation ON

Small-molecule regulator

aTc

TetR

Target

AN

Target

Target

Translation OFF

Translation OFF

Translation ON

b

P bodies Stress granules

Multi-dimensional

(host cell integrated)

designs

Sequestration

Effector

aTc

Decapping Deadenylation

TetR

Target

A

Effector

Cap

Target

AN

PolyA tail

TetR

Cap

Translation OFF Translation ON

Figure 1 | Integrating a synthetic RNAprotein interaction module with native translation control mechanisms. (a) Several previously implemented approaches for posttranscriptional regulation of gene expression in eukaryotes are summarized. These are classied as uni-dimensional since primarily one activity determines the availability or accessibility of a target transcript for translation. (b) Schematic of the proposed strategy for integrating a synthetic RNAprotein module with host cell translation control mechanisms to achieve multi-dimensional, posttranscriptional regulation of gene expression. Fusing the regulatory protein in the synthetic module to host cell factors involved in controlling RNA turnover, degradation and sequestration is proposed as an approach for extending the dynamic range and reducing leaky expression of the synthetic system. Direct control of gene expression is achieved by toggling the synthetic RNAprotein interaction via a small-molecule inducer.

specic posttranscriptional and translational regulatory mechanisms has been directly elucidated.

The success of this approach is predicated on identifying fusion proteins that preserve both the inducible interaction between TetR and its aptamer, and the partners native role in regulating translation. We performed a focused screen in S. cerevisiae to identify potential candidates by creating fusions of TetR to Dcp1p, Dhh1p, Hrp1p, Pop2p, Upf3p, Caf20p, Eap1p and Cdc33p, which have all previously been implicated in some aspect of native translational regulation in yeast11,1518. We made a reporter construct in which either ve or ten tandem TetR aptamers (referred to hereafter as 5 and 10 aptamer arrays)

were positioned within the 30-UTR of a venus yellow uorescent protein (vYFP)-coding sequence19 (Fig. 2a) to identify TetR fusions that mediated doxycycline (Dox)-inducible changes in vYFP expression relative to TetR alone. Notably, TetR by itself displays no inherent inducible behaviour when its cognate aptamers are installed in a 30-UTR context (Fig. 2a). This is consistent with studies of native translational control mediated by protein binding within the 30-UTR, where factors that modulate cap-dependent translation initiation, decapping, polyadenylation/ deadenylation or mRNA sequestration, as examples, must be recruited to achieve a regulatory outcome9,20. We did not examine the effects of placing tandem aptamers within the 50-UTR, since our previous data indicated that increasingly stable secondary structure within this region signicantly decreases maximal expression without substantially improving the regulatory dynamic range5.

In the 5 aptamer array reporter context, only TetR-Pop2p

resulted in signicant Dox-dependent regulation (B3-fold relative to TetR). In the 10 aptamer array reporter context,

however, TetR-Dcp1p, TetR-Dhh1p and TetR-Pop2p all exhib

ited Dox-dependent regulation of vYFP expression (B24-fold

relative to TetR). All fusions did not yield improved regulation, indicating that simply fusing any protein to TetR is insufcient to achieve an improvement in translational regulation. Overall, these screening data indicate that fusing specic proteins natively involved in regulating translation to TetR can signicantly improve the regulatory dynamic range beyond that attainable with TetR alone, while preserving tetracycline analogue-dependent control. These ndings provided initial validation that integrating our synthetic RNAprotein module with host regulatory mechanisms could yield improved regulatory outcomes beyond that of the RNAprotein module alone when tested in the identical reporter context.

Enhanced regulation of gene expression in P. falciparum. We focused next on systematically understanding how a selected TetR fusion protein that improved translational regulation outcomes in a model system would function in the non-model eukaryotic pathogen, P. falciparum. Our rationale for this approach was twofold. First, native translational regulation in this organism is poorly understood. We reasoned that this choice would provide important insights into the broader applicability of our concept without a need for detailed prior knowledge of fundamental translation control mechanisms in the host organism. Second, the toolkit for robust conditional gene expression inP. falciparum is small, with limitations in both dynamic range and basal leakiness14,2123 We reasoned, therefore, that a host-integrated synthetic control module with substantially enhanced regulatory dynamic range would also have immediate practical value as a functional genetics tool.

Of the three functional fusions identied during our yeast screens, we selected the TetR-Dhh1p context for detailed evaluation in P. falciparum. Several factors inuenced this

NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727

a

vYFP

[ ]

AN

Cap 510

5

4

5 aptamer 10 aptamer

*

Fold induction

*

*

3

* P<0.05*** P<0.0001

*

2

1

None

0 DCP1

DHH1

HRP1

POP2

UPF3

CAF20

EAP1

CDC33

b

hrp2 3

UTR

bsd RL tetR

hsp86 5

UTR

hrp2 3

UTR

bsd RL tetR-dozi

cam

5 hsp86 5UTR

Reporter UTR

UTR

(FLuc)

hsp86 3

Reporter mRNA context

TetR

TetR-DOZI

Fold induction

FLuc

FLuc

FLuc

A B C

70

***

60

20

***

15

10

NS

5

3

1

4

2

0

Context A B C

Figure 2 | TetR fusion proteins enhance translational regulation in bothS. cerevisiae and P. falciparum. (a) In S. cerevisiae, various proteins fused to TetR were evaluated for their ability to enhance doxycycline-dependent regulation of a vYFP reporter via either ve or ten tandem TetR aptamers positioned within the 30-UTR. (b) In P. falciparum, regulated expression of a

FLuc reporter by TetR and the TetR-DOZI fusion was tested. Three reporter contexts were examined, namely: (i) single TetR aptamer in the 50-UTR only; (ii) single mutated TetR aptamer in the 50-UTR and 10 TetR

aptamers in the 30-UTR; and (iii) single TetR aptamer in the 50-UTR and 10 TetR aptamers in the 30-UTR. Functional and mutated (no binding to

TetR) aptamers are illustrated as green and red lollipops, respectively. In both panels, fold induction is calculated as the ratio of reporter expression in the induced state ( aTc) relative to that in the repressed state ( aTc).

Data shown are the means.d. from biological triplicates, and are representative of two (S. cerevisae) and three (P. falciparum) independent experiments. *Po0.05; ***Po0.0001 by t-test. NS Not signicant.

decision. First, we wished to use the P. falciparum homologue rather than assuming the yeast protein functions equivalently inP. falciparum. BLAST sequence analysis comparing the yeast andP. falciparum homologues of Dcp1p, Dhh1p and Pop2p revealed that Dhh1p and PF3D7_0320800 (DOZI) had very high sequence identity and similarity (68% and 80%, respectively), and were similar in size (506 versus 433 amino acids, respectively) (Supplementary Fig. 1a). In contrast, Dcp1p and Pop2p and their P. falciparum homologues are much less conserved (Supplementary Fig. 1b,c). The P. falciparum homologues are also large proteins with poorly conserved regions that potentially encode functionally important regulatory information. Therefore,

creating functional fusions based on these proteins might entail initial screens to identify suitable fragments that preserve both critical biochemical information for efcient integration with the native machinery and TetR function upon fusion. We reasoned that a TetR-DOZI fusion closely reected the context validated in yeast, and, together with its compact size, would be an ideal primary candidate to test in P. falciparum.

Second, the Dhh1p homologue from the related rodent malarial parasite, P. berghei, has been implicated in translational regulation. Deleting this protein (referred to as development of zygote inhibited or DOZI based on the observed phenotype) dysregulates the stability of many transcripts that are normally repressed during asexual-stage development in blood and derepressed during sexual-stage development in the mosquito vector24. A detailed mechanism of how DOZI functions in this process is lacking, though in vitro translation studies using theP. falciparum DOZI suggests a role in regulating eIF4E-dependent translation25. This is likely only a partial explanation of this proteins function in situ, however, since it seems to operate within a larger regulatory protein complex that is still to be functionally dened26. Furthermore, the DDX6 helicase protein family, to which Dhh1p and DOZI belong, is implicated in regulating diverse aspects of RNA biology including decapping, degradation and sequestration11,27,28. Last, Dhh1p homologues in several model organisms including Caenorhabditis elegans, Drosophila melanogaster, Xenopus laevis and humans are all very highly conserved (Supplementary Fig. 1a). Therefore, proof of concept with this TetR fusion would support broad applicability beyond our two primary test organisms.

We rst assessed whether TetR-DOZI functions to regulate translation in P. falciparum, and created several dual expression cassette reporter plasmids (Fig. 2b). One cassette uses a calmodulin promoter (PfCAM) to transcribe a mRNA-encoding rey luciferase (FLuc) with regulatory TetR aptamers in the 50-UTR only, 30-UTR only or both 50- and 30-UTRs to test

TetR-dependent translational regulation. The second cassette drives expression of TetR or TetR-DOZI, Renilla luciferase (RLuc) and the Blasticidin S deaminase (bsd) selection marker from a multi-cistronic transcript using the viral T2A skip sequence29. The RLuc is used as an internal reference signal in our quantitative luciferase assays. We site-specically integrated30 these reporter plasmids at the cg6 chromosomal locus inP. falciparum NF54attB (ref. 31), and obtained clonal parasites by limiting dilution. We grew parasites either in the absence or presence of aTc, and measured inducible FLuc expression in the various regulatory contexts described.

These data showed that both TetR and TetR-DOZI regulated FLuc expression by B5-fold in an aTc-dependent manner when a single aptamer is present within the 50-UTR (Fig. 2b, context A). Thus, TetR-DOZI functions indistinguishably from TetR in this context. Next, we examined the degree of regulation achieved with TetR or TetR-DOZI when a 10 aptamer array is placed

within the 30-UTR, and a non-functional mutated aptamer is placed in the 50-UTR of the FLuc reporter (Fig. 2b, context B). With TetR, we observed no detectable aTc-dependent regulation of FLuc expression (fold regulation B1), which is consistent with our observation in the yeast experiments (Fig. 2a). With TetR-DOZI, however, aTc regulated FLuc expression by B19-fold (Fig. 2b). Thus, TetR-Dhh1p in yeast and TetR-DOZI in P. falciparum both substantially enhance aTc-inducible regulation of gene expression beyond what is attainable with TetR by itself.

Next, we tested whether combining regulation by a single aptamer in the 50-UTR and a 10 aptamer array in the 30-UTR

(Fig. 2b, context C) would synergize to increase the overall

4 NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727 ARTICLE

dynamic regulatory range attainable. With TetR alone, we observed B814-fold aTc-dependent regulation of FLuc expression. This represented a B23-fold increase in dynamic range over that observed during TetR regulation mediated via a single aptamer in the 50UTR. Thus, while TetR did not directly mediate regulation via aptamers in the 30-UTR alone (Fig. 2b), an increase in fold regulation occurred when 50- and 30-aptamers ank the reporter gene. Bridging interactions are believed to bring the 50- and 30-UTRs of eukaryotic transcripts into close proximity with each other during translation32. This nding could reect increased TetR recruitment by the 10 aptamer array to create

locally high TetR concentrations in proximity to the 50-UTR, such that the fractional occupancy of the regulating 50-aptamer by TetR is higher than when no 30-UTR array is present and TetR can only be recruited from a relatively dilute cytoplasmic pool. Intriguingly, with TetR-DOZI, aTc-dependent regulation increased substantially to B4570-fold in the dual-aptamer conguration (Fig. 2b). The improved regulatory outcome seen with TetR-DOZI relative to TetR is consistent with the gain of translation regulation via the 30-UTR aptamer array enabled by the DOZI component of the fusion protein.

Enhanced regulation is independent of 50-UTR context. Many genes exhibit highly regulated and presumably functionally important temporal expression proles during the parasites 48 h intraerythrocytic developmental cycle (IDC)33,34. Therefore, we examined whether enhanced regulation is achieved at different time points within the IDC, while using promoters having distinct temporal expression proles. We selected theP. falciparum chloroquine resistance transporter (PfCRT; PF3D7_0709000) and PfCAM (PF3D7_1434200) promoters as

representative candidates. These exhibit peak transcriptional activity in early (ring)- to mid-(trophozoite), and late (schizont) stages, respectively (Fig. 3a).

With both promoters, substantially enhanced regulation was observed when TetR aptamers in the 50- and 30-UTRs are used together. With the PfCAM promoter, fold expression upon aTc induction ranged from 20- to 70-fold across the IDC. Intriguingly, in the PfCRT promoter context, the observed fold expression was much higher and ranged from B50 to 300 fold across the IDC (Fig. 3b). In general, the fold regulation was lowest in early-stage parasites, but regulation by the TetR-DOZI fusion exceeded that by TetR alone14. Interestingly, regulation increased more substantially in later-stage parasites. This pattern of regulation is likely not simply due to delayed expression of the regulatory TetR-DOZI fusions, as the PfHSP86 promoter driving expression of this construct is constitutively high across the IDC (http://plasmodb.org/plasmo/showRecord.do?name=GeneRecord Classes.GeneRecordClass&source_id=PF3D7_0708400&project_id =PlasmoDB). We hypothesize that this phenomenon might instead be linked to stage-dependent activation of the native parasite translational regulatory pathway(s) in which the DOZI component of our fusion operates33,34. Further studies will be required to precisely delineate the DOZI-dependent regulatory pathways in P. falciparum. However, it is worth emphasizing that even without a detailed fundamental understanding of these native mechanisms, we consistently achieve substantially enhanced regulation by integrating our synthetic system with this native host translational regulatory mechanism.

We also replaced FLuc with an enhanced YFP (EYFP) reporter gene in both the PfCRT and PfCAM promoter contexts to conrm that other genes could stringently be regulated in this framework. In the repressed state (aTc), no EYFP signal above

a b c

278

2,000

4 aTc + aTc

0 Rings

48

*

Background aTc+ aTc

*

Normalized FLuc

expression (102 )

Mean YFP expression

(RFU)

PfCRT (PF3D7_0709000)

3

*

1,500

*

2

300

1,000

RPKM

200

600

400

*

*

1

500

NS NS NS

*

0 0 8 16 24 32 40 48

0 Rings

0 Rings

Hours post invasion

Schizonts

Trophozoites

Schizonts

17 71

4,000

PfCam (PF3D7_1434200)

8 aTc

+ aTc

* *

*

*

Background aTc+ aTc

3,000

Normalized FLuc

Mean YFP expression

(RFU)

600

6

61

RPKM

200

expression (103 )

400

4

2,000

2

1,000

*

NS NS NS

0 0 8 16 24 32 40 48

Hours post invasion

Ring Trophozoite Schizont

Trophzoites

Trophzoites

Schizonts

0 Rings

Trophozoites

Schizonts

Figure 3 | The TetR-DOZI aptamer module provides robust control of gene expression. (a) RNA-Seq data from PlasmoDB showing the stage-dependent transcription prole for two native parasite promoters, PfCRT (PF3D7_0709000) and PfCAM (PF3D7_1434200). (b) Stage-dependent FLuc expression normalized to RLuc expression (non-regulated internal reference) in the absence and presence of aTc for both PfCRT (top) and PfCAM (bottom) promoter contexts. The regulatory dynamic range (ratio of aTc induced to basal expression) is shown in red for ring-, trophozoite- and schizont-stage parasites. (c) Stage-dependent mean YFP expression in the absence and presence of aTc for both PfCRT (top) and PfCAM (bottom) promoter contexts. Both the FLuc and YFP reporters are controlled by a single TetR aptamer in the 50-UTR and 10 tandem TetR aptamers in the 30-UTR. Fold-induction values were not calculated from the uorescence reporter expression data as the signal to noise is relatively low, and in the repressed state the uorescence signal is effectively at the autouorescence background level. Data shown are the means.d. from biological triplicates, and are representative of two to four independent experiments. *Po0.001 by t-test. NS Not signicant.

NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727

a

background could be detected by ow cytometry for both PfCRT and PfCAM promoter contexts at any point during the IDC. However, aTc induction produced a substantial increase in EYFP signal at all IDC stages (Fig. 3c; Supplementary Fig. 2). In contrast, a single 50-aptamer within the PfCAM promoter context regulated by TetR alone results in noticeable uorescence signal above background in the repressed state14. Thus, the dual-aptamer conguration with a host-integrated TetR-DOZI regulatory module stringently represses an orthogonal reporter beyond what is attainable with the standalone TetR module. These data indicate that: (1) DOZI-enhanced translational regulation is not restricted to a specic IDC stage; (2) enhanced regulation is achieved independent of the promoter context and expression prole; and (3) adequate concentrations of TetR-DOZI are produced throughout the IDC to achieve robust regulation.

Leaky expression is reduced with the TetR-DOZI system. A frequent limitation of posttranscriptional regulatory schemes is the challenge of reducing leaky expression, as this can adversely impact the practical utility of these tools for both functional genetics and synthetic biology applications. Given the substantial increase in regulatory dynamic range associated especially with the dual-aptamer conguration, we closely analysed the basis for this outcome.

We observed that inclusion of a 10 aptamer array within the

30-UTR of the target transcript reduced maximal reporter expression by B23-fold (Fig. 4a). This was independent of whether TetR or TetR-DOZI functioned as the regulator, and likely reects the change introduced into the target transcript. We then compared fold reduction in FLuc expression in the repressed state ( aTc) using TetR with a single 50-UTR aptamer

as a reference (Fig. 4b), and found that substituting TetR for TetR-DOZI in this context had no impact on leaky expression. In the 50-mutant aptamer and 10 aptamer in the 30-UTR context,

no reduction in leaky expression is observed with TetR as the regulator. In contrast, leaky expression is reduced 612-fold with TetR-DOZI. In the dual-aptamer context, we observed a 710-fold reduction in leaky expression with TetR. In contrast, when TetR-DOZI is used as the regulator in this context, we observed a 1730-fold reduction in leaky expression. Taken together, these data show that integrating the TetR aptamer module with host translational mechanisms mediated through DOZI results in substantially improved regulation of gene expression, and this is predominantly achieved through a reduction in leaky expression. This feature is especially valuable in functional genetics applications where residual protein expression can mask phenotypes mediated by truly essential genes, or in synthetic biology applications where levels of critical components must be stringently regulated to achieve desired circuit performance. In these applications, the reduction in maximal protein expression observed by installing aptamer arrays within the 30-UTR can be further contextualized. In synthetic designs, the desired protein expression levels can be achieved by choosing appropriate promoters to pair with the aptamer-regulated 30-UTR. In functional genetics applications, it is conceivable that installing a 30-aptamer array in the context of a xed, native promoter may prevent adequate maximal expression of essential genes and preclude recovery of viable organisms for further study. The extent to which this scenario is encountered will only be evident upon broader application of our technology. If encountered, this can be addressed by engineering 50- and 30-UTR aptamer-regulated contexts wherein a quantitatively appropriate promoter strength is selected to simultaneously afford functionally adequate maximal gene expression and

stringently tunable regulation (Figs 2 and 3). However, our PfATP4 vide infra data indicate that this system will be relevant for targeting and studying essential parasite genes via engineering aptamer arrays into the 30-UTR alone.

Our studies, along with previous data, also suggest a possible model for how regulation might be achieved in this system. We previously showed that a functional aptamer capable of interacting with TetR is absolutely required for aTc-dependent regulation of gene expression, as when a mutated aptamer that does not interact with TetR is used, no aTc-dependent regulation is observed5. Alternatively, if expressing the TetR fusion protein alone were sufcient to repress reporter gene expression, this would not be aTc dependent. Furthermore, for the same TetR aptamer conguration, absolute FLuc expression levels in the presence of aTc were indistinguishable in the presence of TetR or TetR-DOZI (Fig. 4a). This indicates that any repression by direct interaction of the DOZI fusion component with the reporter is either minimal or does not have a signicant effect on expression. This observation is analogous to our previous study in which TetR fusions to She2p and She3p in yeast were used to conditionally localize target transcripts to daughter cells in yeast19. Taken together, these data support a model in which

25

aTc + aTc

*

20

Normalized FLuc

expression (103 )

*

15

10

*

NS

5

* *

0

5-aptamer WT MUT MUT WT WT

WT

3-aptamers + + + +

Regulatory protein TetR-DOZI TetR TetR-DOZI TetR TetR-DOZI

TetR

b

5-aptamer 3-aptamers

TetR-DOZI

WT

MUT +

TetR-DOZI

MUT +

WT +

WT +

Regulatory protein

TetR

TetR

TetR-DOZI

0

* *

*

NS NS

Fold change in basal

and maximal expression

10

* *

20

Basal expression Maximal expression

30

*

40

Figure 4 | Multi-dimensional synthetic control of translation achieves signicant reduction in leaky expression. (a) FLuc expression levels in the absence and presence of aTc normalized to RLuc expression (non-regulated internal reference) when TetR and TetR-DOZI are used to regulate FLuc expression via aptamers positioned within the 50-UTR, 30-UTR or both as shown. (b) Fold change in maximal FLuc expression upon induction( 0.5 mM aTc) and in basal FLuc expression ( aTc) is shown relative to

the context in which TetR regulates FLuc expression via a single aptamer in the 50-UTR. Data shown are the means.d. from biological triplicates, and are representative of two to four independent experiments. *Po0.001by t-test. NS Not signicant.

6 NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727 ARTICLE

TetR aptamers in a target transcript recruit the TetR-DOZI (or other fusion partner), and this facilitates downstream regulatory outcomes mediated through the fusion partner to occur in a transcript-specic manner. Future studies will be needed to quantitatively dissect which combination(s) of the various translational regulatory roles mediated by DOZI/DDX6 family proteinssuch as degradation, deadenylation and sequestrationcontribute most to the TetR-DOZI-dependent regulation of targeted transcripts. This information could be useful for designing even more efcient and broadly applicable regulators, in addition to expanding our currently poor fundamental understanding of translational regulation in asexual-stage parasites.

Another native protein mediates regulation of translation. We next examined whether the extended regulatory dynamic range observed with the TetR-DOZI fusion in P. falciparum was limited to DOZI. We selected the P. falciparum CAR-1/trailer hitch homologue (CITH) as another fusion partner to test for three key reasons. First, CAR-1 and trailer hitch in C. elegans and D. melanogaster, respectively, have been implicated as key translational regulators, and this protein family is widely conserved across eukaryotes35. Second, like DOZI, CITH has been implicated in regulating translation in Plasmodium berghei. Both proteins seem to function within the same translation regulatory complex, though additional orthogonal pathways cannot be excluded26. Third, from a practical standpoint, CITH homologues are relatively small proteins, which we reasoned should produce fairly compact and biochemically well-behaved fusions to TetR.

We generated a reporter plasmid in which we replaced TetR-DOZI with a TetR-CITH fusion, and FLuc expression was regulated by aptamers in both 50- and 30-UTRs. We then compared aTc-dependent regulation achieved by TetR/50-aptamer only (benchmark), and TetR, TetR-DOZI and TetR-CITH in the dual-aptamer context. Our data show that TetR-DOZI and TetR-CITH both produced substantially improved translational regulation over TetR alone in either the single or dual-aptamer context (Supplementary Fig. 3). These data show that translational regulators from distinct protein families can be incorporated into our conceptual framework for successfully integrating synthetic and native translational regulation to achieve substantially improved conditional control over target gene expression.

Establishing gene essentiality in P. falciparum. Having established quantitative proof of concept for our proposed strategy, we sought to establish that this approach could have useful applications. To demonstrate this, we used our framework to directly establish for the rst time the essential function of a difcult-to-study P. falciparum membrane protein, PfATP4 (PF3D7_1211900), which is of increasing interest as a new antimalarial drug target/resistance mechanism. This P-type ATPase protein has been proposed as the putative target of a new, potent and clinically promising antimalarial drug class, the spiroindolones36,37. However, the available evidence does not denitively demonstrate that loss of PfATP4 function should, in fact, impair parasite survival. Thus, a critical piece of information required for rigorously assigning mechanism of action for a new drug target could not be easily satised due to technological limitations. The availability of this information assumes even greater signicance given recent data indicating that many diverse chemical scaffolds converge on PfATP4 as a common direct or indirect mechanism of parasite sensitivity and/or resistance to these compounds38,39.

To begin addressing this knowledge gap, we implemented our TetR-DOZI system to conditionally regulate PfATP4 expression. Given the large size (B3.8 kb) of PfATP4, we genetically encoded a 10 TetR aptamer array in a 30-UTR context at the native

locus as illustrated (Fig. 5a). Clonal parasites were genotyped by PCR (Fig. 5b; Supplementary Fig. 4) and sequencing of the diagnostic products, and Southern analysis (Fig. 5c; Supplementary Fig. 5). We also conrmed stringent aTc-dependent regulation of PfATP4 expression by western blot analysis (Fig. 5d; Supplementary Fig. 6). Altogether, these data conrmed that we had successfully modied the PfATP4 genomic locus and could stringently control the expression of this integral membrane protein target36.

We next determined whether downregulating PfATP4 expression adversely impacted parasite growth. We selected two isogenic clones that had been grown continuously in the presence of aTc to ensure maximal PfATP4 expression, and split these into

aTc and aTc conditions to monitor parasite growth over

multiple IDCs. Over the rst two IDCs, we observed no signicant aTc-dependent difference in growth between the G2 and 8c clones. However, beginning in the third IDC and continuing thereafter, there was a signicant decrease in growth of the G2 and 8c clones in the no aTc condition compared with when aTc was present (Fig. 5e). As the parental NF54 line grew similarly both in the absence and presence of aTc, this eliminates an inherent proliferative effect of aTc on parasites as the explanation for the growth phenotype. Altogether, these ndings establish, for the rst time, that loss of PfATP4 function is indeed deleterious to blood-stage parasite growth.

Here we demonstrate that synthetic RNAprotein modules can be effectively integrated with native cellular mechanisms for controlling translation, and in so doing, more robust regulation of target gene expression can be achieved. While preservation of modularity and cross-organism compatibility criteria has typically motivated designs intended to minimize interactions with native control mechanisms, we show that these properties can be preserved even while intentionally pursuing host-cell-integrated designs. A key outcome of this conceptual framework is a substantial reduction in leaky/basal expression, which has been a major challenge in leveraging robust translation control tools for functional genetics and synthetic biology applications in eukaryotes.

We envision this strategy will be broadly useful as the principles for implementation are straightforward and apply in unrelated organisms, as illustrated here. Many posttranscriptional regulatory proteins operating either through the same or distinct mechanisms have been described11. These can serve as potential library components that can be systematically evaluated to establish a core set that can serve as gateways for efciently accessing specic modes of native translational regulation. By focusing on components with high protein sequence conservation and presumably regulatory mechanism(s) across different genera, a high rate of successful portability among organisms should be attainable. Novel and orthogonally regulated synthetic RNAprotein interaction pairs can be efciently discovered using existing methods4,13. We envision that implementing such orthogonal units using the framework validated here will facilitate multiplexed, stringent and directly tunable control of translation to motivate increasingly sophisticated and robustly controlled functional genetics and synthetic biology applications.

Methods

Molecular cloning. Plasmids used in this study were cloned by restriction/ligation, yeast homologous recombination or Gibson assembly methods29. All restriction enzymes were purchased from New England Biolabs and reagents were from Sigma-Aldrich, Research Products International, GoldBio Technology or Alfa

NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727

a

b

1.2 kb

p1 p3

PfATP4

Native locus

Legend

P E

2.5 kb

Clone G2 Clone 8c NF54attB

1.2 kb PfATP4 locus

p1+p2

p1+p3

p1+p2

p1+p3

p1+p2

p1+p3

X

5UTR

3UTR

cMyc epitope tag

Aptamers

bsd2ARL2Atd

TP4

1.2 kb MQO

pMG75

c

NF54attB (parental)

Clone G2

Clone 8c

Single crossover integration

3.0 kb 5.1 kb

P P E E

PfATP4

bsd2ARL2Atd

5.1 kb

p1 p2 p31.3 kb Modified locus

3.0 kb

2.5 kb

d

e

10,000

* *

1,000

Cumulative fold

expansion

Clone G2 Clone 8c NF54attB aTc + +

Anti-cMyc (tagged PfATP4)

Anti-GAPDH

NF54 (+ aTc) NF54 ( aTc) Clone G2 (+ aTc) Clone G2 ( aTc) Clone 8c (+ aTc) Clone 8c ( aTc)

100

10

150 kDa 100 kDa

37 kDa

1

0.1 0 1 2 3 4 5 6

Generation

Figure 5 | Establishing essentiality of a native parasite gene using TetR-DOZI with a genetically encoded 10 aptamer array in the 30-UTR.

(a) Schematic of the single crossover event to modify the native PfATP4 chromosomal locus to install a C-terminal epitope tag, a 10 aptamer array and the

regulatory TetR-DOZI expression cassette in a single step. Diagnostic PCR primers to assess the status of the locus are designated p1p3, and the expected sizes of PCR products derived from the native and modied locus are indicated. The blue bar corresponds to the probe used in Southern blot experiments in which genomic DNA was digested simultaneously with EcoR1 (E) and Pst1 (P). The expected diagnostic band sizes for the native and modied loci are indicated. (b) Diagnostic PCR analysis of two isolated clones (G2 and 8c) with modied PfATP4 loci. The malate:quinone oxidoreductase (MQO) gene is used as a positive control for the presence of genomic DNA in all samples tested. (c) Southern blot analysis for the parental NF54attB strain and clones G2 and 8c. (d) Western blot using anti-cMyc antibody to detect aTc-dependent expression of PfATP4 in clones G2 and 8c. The parental NF54attB strain (no cMyc-tagged PfATP4) is included as a negative control. (e) ATc-dependent growth of the parental NF54attB strain and Clones G2 and 8c over multiple generations. Data shown are the means.d. from biological triplicates, and are representative of three independent experiments. *Po0.001 by t-test.

Aesar, unless otherwise specically mentioned. PCR amplication of parasite DNA was performed on either crude lysates from the parasite fractions or puried DNA templates using Hemo KlenTaq mixed 15:1 (v:v) with PfuTurbo (Agilent). Supplementary Tables 1 and 2, respectively, summarize the plasmids and primers used in this work. Plasmid DNAs for transfections were prepared using the Xtra Midi Kit (Clontech). Plasmids were propagated in DH5a bacteria, and grown on solid or in liquid media at 30 C to minimize the potential for spontaneous truncations of the 30-UTR TetR aptamer arrays.

Plasmid DNA constructs sequences and sources. GenBank les for the constructs used in this study are provided as supplementary les. FLuc is wild-type FLuc DNA carrying the K549E mutation for cytosolic targeting in yeast.

Yeast inducible expression assays. The various TetR-containing regulatory proteins were cloned as C-terminal fusions to TetR in the YCpSUP yeast expression vector5. We created a vYFP reporter plasmid with either 5 and 10 tandem repeats of the TetR aptamer 51.2 in the 30-UTR by replacing the non-uorescent vYFPD in vYFPD51.2(5 ) and vYFPD51.2(10 )19,

respectively, with vYFP. S. cerevisiae W303-1B cells harbouring both a repressor and an appropriate reporter plasmid were grown to saturation at 30 C in Synthetic Dened Media #1 (SD1) (6.7 g l 1 yeast nitrogen base without amino acids (RPI), 20 mg l 1 adenine, 30 mg 1l lysine, 100 mg 1l leucine, 20 mg l 1 histidine) 20 g l 1 glucose. Cells were diluted 1:80 into SD1 20 g l 1 rafnose

and grown for 4 h. Glucose (to repress TetR or TetR fusion protein expression) or galactose (to induce TetR or TetR fusion expression) was added to 20 g l 1, and cells were grown 16 h at 30 C with shaking before measurement. To measure vYFP expression by ow cytometry, cells were grown in triplicate either in the presence of glucose (22 mM doxycycline) or galactose (22 mM doxycycline), and analysed on a C6 Flow Cytometer (Accuri). For all samples, B50,000 events were captured and vYFP uorescence measured in the FL1 channel.

P. falciparum culture and transfection. The P. falciparum NF54attB parasites were a gift from David Fidock (Columbia University). These were grown in human erythrocytes (Research Blood Components, Boston, MA) at 5% haematocrit under 5% O2 and 5% CO2 in RPMI 1640 media supplemented with 5 g l 1 Albumax II (Life Technologies), 2 g l 1 NaHCO3, 25 mM HEPES-K pH 7.4, 1 mM hypoxanthine and 50 mg l 1 gentamicin. For transfections, we used 5060 mg of each plasmid per 200 ml packed red blood cells (RBCs), adjusted to 50% haematocrit. We used a Bio-Rad Gene Pulser II to either directly electroporate DNA into ring-stage parasites40 or preload uninfected RBCs41. For direct electroporation of ring-stage parasites, 0.2-cm electroporation cuvettes were loaded with 0.3 ml of parasitized RBCs at 50% haematocrit and 50 mg of plasmid DNA in incomplete cytomix solution that contains 120 mM KCl, 0.15 mM CaCl2, 10 mM K2HPO4/KH2PO4, 25 mM HEPES, 2 mM EGTA and 5 mM MgCl2. Electroporation conditions were0.31 kV and 960 mF. For spontaneous DNA uptake, eight square-wave pulses of 365 V for 1 ms, separated by 100 ms were used. Transfected parasites were selected with 2.5 mg l 1 Blasticidin S beginning 4 days after transfection. In the case of

Bxb1-mediated integration, no selection for the integrase-expressing plasmid was applied. In allelic replacement experiments, drug on/off cycles (28 days each) were completed until the desired integration event was detectable by PCR. Parasites were cloned by limiting dilution thereafter and genotyped by PCR, sequencing of the diagnostic product and Southern analysis.

Induction experiments and luciferase assays. Ring-stage parasites were tightly synchronized using 0.3 M alanine for three consecutive IDCs before induction. Synchronized parasites were induced using 0.5 mM aTc at ring-stage and allowed to go through a complete IDC before collecting ring-, trophozoite- and schizont-stage samples for luciferase assays. Firey and RLuc measurements were made with the Dual Luciferase Assay kit (Promega) on a GloMax 20/20 luminometer (Turner Biosystems).

Western blot. Approximately, 106 late trophozoite-stage parasites were lysed by saponin treatment (0.5 g l 1 saponin in PBS). These lysates were prepared for

8 NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727 ARTICLE

western blot by heating in Laemmli sample buffer at 95 C for 10 min. After separation by SDSpolyacrylamide gel electrophoresis, proteins were transferred to a polyvinylidene diuoride membrane and probed with a rabbit monoclonal antibody (71D10) against the cMyc tag (Cell Signaling Technology, catalogue no 2278 S; 1:2,000 dilution) or rabbit polyclonal antibody against GAPDH (Abcam, ab9485; 1:1,000 dilution). Blots were imaged using a horseradish peroxidase-coupled secondary antibody and SuperSignal West Femto substrate (Thermo Scientic, catalogue no 34095).

Flow cytometry. Ring-stage parasites were synchronized and analysed by ow cytometry at the trophozoite stage. Cells were stained for nucleic acid content with 1 mM SYTO 61 (Life Technologies) and analysed using an Accuri C6 ow cytometer (BD Biosciences). EYFP and SYTO 61 signals were monitored in the FL1 and FL4 channels, respectively. Parasites were selectively analysed by gating on events with high FL4 signal intensity. Adding 0.5 mM aTc to either uninfected RBCs or parasites not expressing EYFP had no effect on FL1-H signal intensity.

Southern blot. Southern blots were carried out using genomic DNA isolated from schizont-stage parasites at B25% parasitemia in a 30 ml, at 2% haematocrit culture using the QIAamp DNA blood mini kit (Qiagen). RBCs were lysed using a 0.1% saponin solution. Between 2 and 3 mg of genomic DNA was restriction enzyme digested overnight with PstI and EcoRI (New England Biolabs), and probed with a PfATP4 PCR product obtained using primers SMG291 and SMG292. The probe was labelled with biotin-11-dUTPs using the Pierce Biotin Random Prime kit (Thermo Scientic). Blots were processed using the TurboBlotter kit (Whatman) for transfer and the North2South kit (Thermo Scientic) for development.

Growth assays. Parasites were synchronized to rings using 0.3 M alanine in10 mM HEPES (pH 7.4) at day 0, adjusted to 1% parasitemia and seeded in 96-well microtiter plates in quadruplicate at 2% haematocrit in 200 ml of RPMI complete media in the presence or the absence of aTc (0.5 mM). The parental NF54attB strain and engineered PfATP4 lines were all included in the assay. Expansion was measured over six IDCs, and samples were analysed every 48 h to determine parasitemias at the end of each IDC. After each measurement, all cultures were split by the same dilution factor required to keep the pre-invasion parasitemia of the control lines below 2% to avoid culture over-expansion. Parasitemias were measured by incubating the cells with 1:1,000 dilution of SYBR Green I for 15 min at 37 C, before ow cytometry on an Accuri C6 instrument (BD Biosciences) and analysis of data collected in the FL1 channel.

References

1. Forster, K. et al. Tetracycline-inducible expression systems with reduced basal activity in mammalian cells. Nucleic Acids Res. 27, 708710 (1999).

2. No, D., Yao, T. P. & Evans, R. M. Ecdysone-inducible gene expression in mammalian cells and transgenic mice. Proc. Natl Acad. Sci. USA 93, 33463351 (1996).

3. Rnicke, V., Graulich, W., Mumberg, D., Mller, R. & Funk, M. Use of conditional promoters for expression of heterologous proteins in Saccharomyces cerevisiae. Methods Enzymol. 283, 313322 (1997).

4. Auslander, S. et al. A general design strategy for protein-responsive riboswitches in mammalian cells. Nat. Methods 11, 11541160 (2014).

5. Goldess, S. J. et al. Direct and specic chemical control of eukaryotic translation with a synthetic RNA-protein interaction. Nucleic Acids Res. 40, e64 (2012).

6. Suess, B. et al. Conditional gene expression by controlling translation with tetracycline-binding aptamers. Nucleic Acids Res. 31, 18531858 (2003).

7. Win, M. N. & Smolke, C. D. A modular and extensible RNA-based gene-regulatory platform for engineering cellular function. Proc. Natl Acad. Sci. USA 104, 1428314288 (2007).

8. Wittmann, A. & Suess, B. Selection of tetracycline inducible self-cleaving ribozymes as synthetic devices for gene regulation in yeast. Mol. Biosyst. 7, 24192427 (2011).

9. Besse, F. & Ephrussi, A. Translational control of localized mRNAs: restricting protein synthesis in space and time. Nat. Rev. Mol. Cell Biol. 9, 971980 (2008).

10. Rocak, S. & Linder, P. DEAD-box proteins: the driving forces behind RNA metabolism. Nat. Rev. Mol. Cell Biol. 5, 232241 (2004).

11. Decker, C. J. & Parker, R. P-bodies and stress granules: possible roles in the control of translation and mRNA degradation. Cold Spring Harb. Perspect. Biol. 4, a012286 (2012).

12. Saito, H. et al. Synthetic translational regulation by an L7Ae-kink-turn RNP switch. Nat. Chem. Biol. 6, 7178 (2010).

13. Belmont, B. J. & Niles, J. C. Engineering a direct and inducible protein-RNA interaction to regulate RNA biology. ACS Chem. Biol. 5, 851861 (2010).14. Goldess, S. J., Wagner, J. C. & Niles, J. C. Versatile control of Plasmodium falciparum gene expression with an inducible protein-RNA interaction. Nat. Commun. 5, 5329 (2014).

15. Cosentino, G. P. et al. Eap1p, a novel eukaryotic translation initiation factor 4E-associated protein in Saccharomyces cerevisiae. Mol. Cell. Biol. 20, 46044613 (2000).

16. de la Cruz, J., Iost, I., Kressler, D. & Linder, P. The p20 and Ded1 proteins have antagonistic roles in eIF4E-dependent translation in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA 94, 52015206 (1997).

17. He, F. & Jacobson, A. Upf1p, Nmd2p, and Upf3p regulate the decapping and exonucleolytic degradation of both nonsense-containing mRNAs and wild-type mRNAs. Mol. Cell. Biol. 21, 15151530 (2001).

18. Kessler, M. M. et al. Hrp1, a sequence-specic RNA-binding protein that shuttles between the nucleus and the cytoplasm, is required for mRNA 3&apos;-end formation in yeast. Genes Dev. 11, 25452556 (1997).

19. Belmont, B. J. & Niles, J. C. Inducible control of subcellular RNA localization using a synthetic protein-RNA aptamer interaction. PLoS ONE 7, e46868 (2012).

20. Kuersten, S. & Goodwin, E. B. The power of the 3&apos; UTR: translational control and development. Nat. Rev. Genet. 4, 626637 (2003).

21. Armstrong, C. M., Armstrong, C. M., Goldberg, D. E. & Goldberg, D. E. An FKBP destabilization domain modulates protein levels in Plasmodium falciparum. Nat. Methods 4, 10071009 (2007).

22. Meissner, M. et al. Tetracycline analogue-regulated transgene expression in Plasmodium falciparum blood stages using Toxoplasma gondii transactivators. Proc. Natl Acad. Sci. USA 102, 29802985 (2005).

23. Prommana, P. et al. Inducible knockdown of Plasmodium gene expression using the glmS ribozyme. PLoS ONE 8, e73783 (2013).

24. Mair, G. R. et al. Regulation of sexual development of Plasmodium by translational repression. Science 313, 667669 (2006).

25. Tarique, M., Ahmad, M., Ansari, A. & Tuteja, R. Plasmodium falciparum DOZI, an RNA helicase interacts with eIF4E. Gene 522, 4659 (2013).

26. Mair, G. R. et al. Universal features of post-transcriptional gene regulation are critical for Plasmodium zygote development. PLoS Pathog. 6, e1000767 (2010).

27. Ernoult-Lange, M. et al. Multiple binding of repressed mRNAs by the P-body protein Rck/p54. RNA 18, 17021715 (2012).

28. Rouya, C. et al. Human DDX6 effects miRNA-mediated gene silencing via direct binding to CNOT1. RNA 20, 13981409 (2014).

29. Wagner, J. C. et al. An integrated strategy for efcient vector construction and multi-gene expression in Plasmodium falciparum. Malar. J. 12, 373 (2013).

30. Nkrumah, L. J. et al. Efcient site-specic integration in Plasmodium falciparum chromosomes mediated by mycobacteriophage Bxb1 integrase. Nat. Methods 3, 615621 (2006).

31. Adjalley, S. H. et al. Quantitative assessment of Plasmodium falciparum sexual development reveals potent transmission-blocking activity by methylene blue. Proc. Natl Acad. Sci. USA 108, E1214E1223 (2011).

32. Wells, S. E., Hillner, P. E., Vale, R. D. & Sachs, A. B. Circularization of mRNA by eukaryotic translation initiation factors. Mol. Cell 2, 135140 (1998).33. Bozdech, Z. et al. The transcriptome of the intraerythrocytic developmental cycle of Plasmodium falciparum. PLoS Biol. 1, 85100 (2003).

34. Le Roch, K. G. et al. Discovery of gene function by expression proling of the malaria parasite life cycle. Science 301, 15031508 (2003).

35. Decker, C. J. & Parker, R. CAR-1 and trailer hitch: driving mRNP granule function at the ER? J. Cell Biol. 173, 159163 (2006).

36. Rottmann, M. et al. Spiroindolones, a potent compound class for the treatment of malaria. Science 329, 11751180 (2010).

37. Spillman, N. J. et al. Na( ) regulation in the malaria parasite Plasmodium

falciparum involves the cation ATPase PfATP4 and is a target of the spiroindolone antimalarials. Cell Host Microbe 13, 227237 (2013).38. Flannery, E. L. et al. Mutations in the P-Type cation-transporter ATPase 4, PfATP4, mediate resistance to both aminopyrazole and spiroindolone antimalarials. ACS Chem. Biol. 10, 413420 (2014).

39. Lehane, A. M., Ridgway, M. C., Baker, E. & Kirk, K. Diverse chemotypes disrupt ion homeostasis in the malaria parasite. Mol. Microbiol. 94, 327339 (2014).

40. Wu, Y., Sifri, C. D., Lei, H. H., Su, X. Z. & Wellems, T. E. Transfection of Plasmodium falciparum within human red blood cells. Proc. Natl Acad. Sci. USA 92, 973977 (1995).

41. Deitsch, K., Driskill, C. & Wellems, T. Transformation of malaria parasites by the spontaneous uptake and expression of DNA from human erythrocytes. Nucleic Acids Res. 29, 850853 (2001).

Acknowledgements

This work was funded by the National Institutes of Health Directors New Innovator Award (1DP2OD007124, J.C.N.); a grant from the Bill and Melinda Gates Foundation through the Grand Challenges Explorations initiative (OPP1069759, J.C.N.); the National Institute of Environmental Health Sciences Predoctoral Training Grant (5-T32-ES007020, S.J.G.); National Institute of General Medical Sciences Center for Integrative Synthetic Biology Grant (P50 GM098792); Siebel Scholars Foundation Award (S.J.G.); and MIT startup funds.

Author contributions

S.M.G. designed and performed most experiments. A.F. cloned and transfected the PfCRT-EYFP plasmid construct, performed YFP induction studies, analysed PfATP4

NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 9

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10727

clones by Southern blot and did the aTc-dependent growth assays on the regulated PfATP4 lines. S.J.G. designed and did the initial TetR fusion protein screening inS. cerevisiae. S.M.G. and A.S.N. performed quantitative luciferase assays. S.M.G., A.F. and J.C.N. designed experiments and analysed data. S.M.G. and J.C.N. wrote the manuscript. J.C.N. supervised the project.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: S.J.G. and J.C.N. are inventors of the genetically encoded protein-binding RNA aptamer technology described and have led patent applications with other co-inventors. The remaining authors declare no conict of interest.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Ganesan, S. M. et al. Synthetic RNAprotein modules integrated with native translation mechanisms to control gene expression in malaria parasites. Nat. Commun. 7:10727 doi: 10.1038/ncomms10727 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

10 NATURE COMMUNICATIONS | 7:10727 | DOI: 10.1038/ncomms10727 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications