Cell culture

The human hepatoma Huh7 cell line (Japanese Collection of Research Bioresources (JCRB) Cell Bank) was cultured in Dulbecco’s Modified Eagle Medium (DMEM) with high glucose (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Corning), 100 U/mL of penicillin-streptomycin (Gibco) and 1% GlutaMAX™ (Gibco) (hereafter referred to as cell culture media) at 37 °C and 5% CO₂. Primary hepatocytes (BioIVT) were cultured in CP Medium (BioIVT, cat no. Z990003) supplemented with 1% PSN (Gibco, cat no. 15640055) (hereafter referred to as hepatocyte media) at 37 °C and 5% CO₂, unless otherwise stated.

Cell line generation

The following DNA plasmids and lentiviruses were provided by VectorBuilder: P_CMV -mCherry-GFP-HT (VB230409-1160fdt), P_EF1a-HT-GFP-FIS1 (VB230406-1238ymp), P_EF1a-HT-mCherry-FIS1 (VB240122-1674bkz), P_UBC-SNAP tag-GFP (VB230607-1547nrb), P_CMV-mCherry-VHL (VB230406-1232ghk), P_CMV-mCherry-Ub (VB230406-1241ura), and P_CMV-UIS4_Pb-C_term-HT-HiBiT (VB240122-1687zve). Briefly, Huh7 cells were seeded at a density of 3 × 10⁵ cells per well in a 6-well plate (Greiner) and transduced the next day with 3 × 10⁵ transduction units (TU) per well. Lentiviruses were added to cells in culture media containing 8 µg/mL polybrene, and cells were spun at 800 × g for 1 h at 37 °C, with low acceleration and break. One day post-transduction, lentiviruses were removed, and fresh cell culture media was added to cells. Transduced cells were then selected for 2–3 weeks in cell culture media containing 2 µg/mL puromycin (Gibco) or 7.5 µg/mL blasticidin (Gibco). The vector IDs listed above can be used to retrieve detailed information about the lentivirus vectors on vectorbuilder.com.

Compounds

HaloPROTAC3 (Promega, cat no. GA3110), the fluorescent HaloTag ligand (Promega, cat no. G2801) and the bifunctional HaXS8 (Tocris, cat no. 4991) used in this study were commercially sourced. The VHL binder compound was produced internally at Novartis as previously described²⁰. HaloPROTAC3, VHL binder and HaXS8 solutions were prepared in DMSO. DMSO-treated control conditions were included in all relevant experiments.

Fluorescence activated cell sorting (FACS)

Huh7 cells that express fluorescent HT proteins were seeded at a density of 1.5 × 10⁴ cells per well in a 96-well flat bottom plate (Greiner) and treated with compounds for 24 h. Cells were detached and resuspended in 5 mM EDTA (Invitrogen) in the 96-well plate prior to measuring intracellular GFP and mCherry fluorescence using the BL1 and YL2 channels on the Attune flow cytometer (Thermo Fisher Scientific). The mean fluorescence of each sample was determined using the software FlowJo (v10.10) and expressed as percentages of DMSO-treated samples. For the mCherry-GFP-HT cells, similar fluorescence profiles were observed for mCherry and GFP, but only data for GFP fluorescence are presented.

Nano-Glo HiBiT lytic assay

WT Huh7 and Huh7 UIS4_Pb-C_term-HT-HiBiT cells were seeded at a density of 8 × 10³ cells per well in a white 384-well plate (Greiner). One day post-seeding, cells were treated with compounds for 24 h before being assayed using the Nano-Glo HiBiT lytic kit (Promega). The luminescence of samples was measured on the CLARIOstar luminescent plate reader (BMG Labtech) 10 min after adding the reagents.

Generation and characterization of P. berghei UIS4-HT

Transgenic P. berghei parasites expressing UIS4-HT were generated using a gene replacement strategy, as previously described with modifications²². Briefly, an insert DNA including the codon-optimized UIS4-HT fusion protein followed by the P. berghei DHFR/TS 3’ downstream region and the UIS4 flanking sequences was generated by DNA synthesis (Invitrogen) (see insert sequence, Supplementary Data S1). This insert DNA was then subcloned into the plasmid pL0005 (MRA-774, BEI Resources, contributed by Andrew P. Waters) using PCR (see primer sequences, Supplementary Table S1) and In-Fusion Cloning (Takara Bio). Positive colonies were identified using PCR analysis and the entire insert sequence of one clone was confirmed using DNA sequencing. The linearized UIS4-HT plasmid was introduced and integrated in P. berghei ANKA Cl15cy1 (MRA-871, BEI Resources, contributed by Chris J. Janse and Andrew P. Waters) parasites using previously described standard methods⁵⁷ and passaged thrice in Swiss Webster mice with selection using pyrimethamine (PYR). Mice were then injected with 200 parasites (from the 3rd passage) and further treated with PYR. Infected blood was collected once parasitemia was > 1% and used to generate blood stocks and extract P. berghei genomic DNA. The genomic integration of UIS4-HT and the absence of contamination with WT parasites was confirmed using PCR analyses (Supplementary Fig. S1 and Supplementary Table S1).

P. berghei liver stage infection assay

Huh7 cells were seeded at a density of 8 × 10³ cells per well in 384-well black plates (Greiner) one day prior to infection. Anopheles stephensi mosquitoes infected with P. berghei WT or P. berghei UIS4-HT were produced by the SporoCore (University of Georgia, UGA) and sporozoites were isolated by microdissection of salivary glands, as previously described with modifications^58,59. Sporozoites were diluted in Roswell Park Memorial Institute 1640 medium (RPMI; Gibco) supplemented with 20% heat-inactivated FBS and Huh7 cells were then spun with 5 × 10³ sporozoites per well at 330 × g for 3 min, with low acceleration and break. Unless otherwise stated, infected cells were treated with compounds 2 h post-infection with media change. Infected cells were fixed using 4% paraformaldehyde (PFA; Electron Microscopy Sciences) 2 days post-infection and stored in DPBS at 4 °C until further processing.

Staining and imaging in Huh7 cells

When required, the MitoTracker deep red probe (Invitrogen, cat no. M22426) was added to live cells and incubated for 45 min at 37 °C before fixation. Uninfected and infected fixed samples were permeabilized for 30 min at room temperature (RT) in blocking and permeabilization (BP) buffer containing 2% Bovine Serum Albumin (BSA) (Sigma-Aldrich) and 0.2% Triton X-100 (Sigma-Aldrich) in DPBS (Gibco). The samples were then incubated with the following for 3 h at RT in BP buffer: goat anti-Pb UIS4 (1:250, Origene, cat no. AB0042-200), rabbit anti-Pc HSP70⁵⁹(1:5,000), mouse anti-HaloTag (1:1,000, Promega, cat no. G9211), rat anti-mCherry (1:100, Invitrogen, cat no. M11217) and/or mouse anti-GFP (1:200, Roche, cat no. 11814460001) antibodies, and/or a fluorescent HaloTag ligand (1:500, Promega G2801). The samples were washed thrice with DPBS and then stained with Hoechst (2 µg/mL, Thermo Fisher Scientific) and the following secondary antibodies (1:1,000) for 1 h at RT in BP buffer: AlexaFluor-488 donkey anti-Rabbit (Invitrogen, cat no. A32790), AlexaFluor-488 donkey anti-mouse (Invitrogen, cat no. A21202), AlexaFluor-568 donkey anti-goat (Invitrogen, cat no. A11057), AlexaFluor-568 donkey anti-rat (Invitrogen, cat no. A78946), AlexaFluor-568 donkey anti-rabbit (Invitrogen, cat no. A10042), AlexaFluor-647P donkey anti-rabbit (Invitrogen, cat no. A32795), AlexaFluor-647P donkey anti-mouse (Invitrogen, cat no. A32787) and/or AlexaFluor-647P donkey anti-goat (Invitrogen, cat no. A32849). The samples were then washed thrice with DPBS and stored in DPBS at 4 °C until imaging. Images were taken with a 20× objective and 25 fields of view were captured from each well of the 384-well plate on the ImageXpress Micro Confocal (IXMC; Molecular Devices) and processed using a Cell Profiler (v4.2.4) analysis pipeline for quantification. Representative pictures of cells and parasites were also taken using either the 20× or 40× objectives on the Laser Scanning Microscope 980 (LSM980) with Airyscan 2 (Zeiss) and processed with FIJI (v1.53t).

P. berghei mouse infection model

Female C57BL/6 mice were injected retro-orbitally with 2.5 × 10⁴ P. berghei WT or UIS4-HT salivary glands sporozoites under isoflurane anesthesia. Blood samples were collected from the tail veins from 72 h post-infection and analyzed, unblinded, on slides using thin blood smears. The slides were then fixed in 100% methanol for 30s, followed by drying and staining with a 10% Giemsa solution for 15 min. After gentle washing with distilled water, the slides were dried and examined using a 100× oil immersion objective on the Nikon Eclipse E200 microscope. Parasitemia was calculated by determining the percentage of infected erythrocytes in 15–20 fields of view. Each field of view typically contained 100–200 total erythrocytes. Parasite DNA was extracted from the blood of sporozoite-infected mice and used to confirm the genotype of transgenic parasites.

P. cynomolgi culture

P. cynomolgi Berok culture was performed as previously described^24,25. Briefly, asexual parasite stages were propagated in rhesus erythrocytes (BioIVT) at a hematocrit of 3% in RPMI-1640 medium supplemented with GlutaMAX (Gibco), 30 mM HEPES (Sigma-Aldrich), 0.2% (w/v) of a 50% D-glucose solution, 200 µM hypoxanthine (Thermo Fisher Scientific), and 20% heat-inactivated Macaca mulatta serum (BioIVT). Parasite cultures were incubated at 37 °C in trimix gas (5% CO₂, 5% O₂, 90% N₂).

Preparation of the plasmid for the generation of P. cynomolgi UIS4-HT

The generation of transgenic P. cynomolgi Berok parasites was achieved as previously described with modifications²⁵. Briefly, the gene editing of parasites was performed using a plasmid referred to as pCJ-110. pCJ-110 is based on the plasmid pYC-L2, designed for CRISPR/Cas9 editing of P. yoelii^60,61. pYC-L2 was extensively remodeled with P. cynomolgi Berok-derived promoter and terminator sequences, amplified from Berok blood stage genomic DNA. Specifically, the resulting pCJ-110 encodes a cassette controlled by the bidirectional Berok EF1α promoter that drives the expression of mutant human dihydrofolate reductase (hDHFR) for positive selection and humanized SpCas9 for gene editing, separated by a 2A self-cleaving peptide. The bidirectional promoter also drives the expression of the bifunctional yeast fusion cytosine deaminase/uracil phosphoribosyltransferase (yFCU) for negative selection. The Berok HSP70 terminator is placed after the SpCas9 gene and the P. falciparum BiP terminator is placed after the yFCU gene. The plasmid also allows the expression of a single guide RNA (gRNA) from the Berok U6 promoter and includes a multiple cloning site facilitating donor template insertion required for the homology-directed double strand break repair. Thus, pCJ-110 allows for CRISPR/Cas9 transgenesis under both positive and negative selection. An illustration of the plasmid is shown in Supplementary Fig. S8. We amplified and sequenced the uis4 locus (corresponding to PcM_0602100) from Berok blood stage gDNA with primers listed in Supplementary Table S1. The construction of the pCJ-110-UIS4-HT plasmid, which was used to tag the 3’ end of P. cynomolgi Berok uis4 with the HaloTag (Supplementary Fig. S2) and to introduce silent shield mutations at the gRNA binding site, was performed in two steps. First, annealed oligonucleotides encoding the gRNA (Supplementary Table S1) were ligated into Esp3I (New England Biolabs, NEB)-digested pCJ-110 using T4 DNA ligase (NEB). Then, the plasmid was digested with SpeI and NotI (NEB) and reassembled with three PCR fragments encoding the 5’ and 3’ uis4 homology regions (amplified from P. cynomolgi Berok genomic DNA) and the HaloTag sequence (amplified from the P. berghei UIS4-HT plasmid) (see primer sequences, Supplementary Table S1) using Gibson cloning (NEB). The sequence of the cloned donor template was confirmed using DNA sequencing.

Generation of transgenic P. cynomolgi UIS4-HT

Transfection of P. cynomolgi was carried out using the pCJ-110-UIS4-HT plasmid as previously described with modifications^25,62. Briefly, schizonts were purified with a Nycodenz gradient²³ and a minimum of 1 × 10⁷ cells were resuspended in 100 µL of supplemented P3 primary cell solution (Lonza). Electroporation was performed using 20 µg of plasmid DNA and an Amaxa 4D electroporator (Lonza) as previously described⁶². To increase reinvasion efficiency, transfected parasites were transferred to 500 µL of pre-warmed culture medium at 20% hematocrit and incubated at 37 °C under shaking conditions for 1 h, prior to being further cultured using standard conditions. Transgenic parasites were selected with 50 nM PYR from 24 h post transfection, for 7 days. A transgenic population was obtained within 4 weeks after transfection and validated through PCR genotyping (Supplementary Fig. S2 and Supplementary Table S1). Blood stage UIS4-HT parasites were shipped to Oregon Health and Science University (OHSU) / Oregon National Primate Research Center (ONPRC) and stored as glycerolyte stocks in liquid nitrogen until use.

P. cynomolgi infection of NHPs (OHSU / ONPRC)

Transgenic UIS4-HT parasites cycled in a Japanese macaque (JMac1, ~ 4.5-year-old male) and through the mosquito vector to a new Japanese macaque (JMac2, ~ 4.5-year-old female) were used to generate blood-stage glycerolyte stocks and to infect a naive Japanese macaque (JMac3, ~ 5-year-old female). Mosquitoes were then fed on the blood from this infected macaque and dissected to isolate salivary gland sporozoites. Sporozoites were counted, washed once in PBS and used to infect a rhesus macaque (~ 5-year-old male) intravenously (with 3.6 × 10⁴ sporozoites). Beginning 6 days post-infection, blood samples were collected for parasitemia quantification (Supplementary Fig. S2) and to monitor hematocrit and hematological factors associated with animal well-being. During peak infection, larger drawings of blood were collected to feed mosquitoes, generate glycerolyte stocks, and genotype parasites. At a maximum parasitemia of 3 × 10⁵ parasites/µL, NHPs were treated with Coartem tablets (20 mg artemether/120 mg lumefantrine, Novartis) twice daily for three days, which clear blood stage parasites but leave hypnozoites intact. Semi-weekly parasitemia monitoring was then conducted to detect relapse infections, arising from hypnozoite activation. Once a relapse was detected, monitoring of the macaque and mosquitoes feeding followed as described above. At the end of the study, the sporozoite-infected NHP was treated with a single dose of tafenoquine (Krintafel, 150 mg, GlaxoSmithKline), to clear any remaining hypnozoites, in addition to a course of Coartem. Blood stage transgenic parasite stocks obtained from a sporozoite-initiated infection were shipped to UGA and further used to produce infected mosquitoes.

Infection of mosquitoes with P. cynomolgi (OHSU / ONPRC)

Anopheles stephensi mosquitoes were reared in-house as previously described⁶³ or purchased from the SporoCore (UGA). Mosquitoes were maintained in a 26 °C incubator with 80% humidity and provided with sugar cubes and water pads until the day of infection. Briefly, blood was drawn from infected macaque into a Lithium-Heparin vacutainer (Greiner Bio-One). To prevent exflagellation of the male gametes prior to uptake by mosquitoes, the blood was maintained at 39 °C during transport from the non-human primate facility, through subsequent washing steps using pre-warmed media and during final transport to the insectary. RBCs were washed twice with RPMI and the serum volume was replaced with human AB⁺ sera. This was used to feed female mosquitoes aged 2–10 days using a glass membrane feeder covered with a single layer of parafilm. Infected mosquitoes were then provided a sugar cube and water pad daily and a supplemental blood meal via an anesthetized mouse at 4–6 days post-infection. Midgut oocysts were quantified in a sample of mosquitoes at 8–10 days post-infection and salivary gland sporozoites were isolated for downstream use at 16–19 days post-infection.

P. cynomolgi infection of NHPs (UGA)

Cryopreserved P. cynomolgi UIS4-HT blood stocks were used to infect male Japanese macaques. After infection, parasitemia was monitored daily as previously described^64,65,66. Parasitemia was monitored after inoculation through the last day of curative treatment using artemether-lumefantrine combination therapy.

Infection of mosquitoes with P. cynomolgi (UGA)

Anopheles dirus mosquitoes were reared at UGA under controlled conditions of 27 °C and 80% relative humidity. Adult mosquitoes were maintained on 20% sucrose for the first three days after pupation and 10% sucrose for the remainder of development. Three- to five-day old female mosquitoes were used in membrane feeding experiments. Once parasitemia reached approximately 1 × 10⁴ parasites/µL, parasitized blood was collected from an infected animal into lithium heparin tubes and maintained at 37 °C throughout processing for feeding mosquitoes. Before membrane feeding, the blood was pelleted via centrifugation at 800 × g for 5 min, and the plasma was removed and discarded. The remaining pellet was resuspended in pre-warmed incomplete RPMI and centrifuged at 800 × g for 5 min. The supernatant was then removed and discarded, and the remaining pellet was resuspended at 50% hematocrit in pre-warmed, malaria-naïve Japanese macaque serum. The blood was placed in a glass-bell covered with parafilm membrane, and mosquitoes were allowed to feed for up to 30 min in the dark at 37 ^oC. Afterwards, unfed mosquitoes were removed and discarded. Mosquitoes that engorged were maintained with 10% sucrose throughout the rest of the study. Six days after feeding, midguts were dissected from 5 to 10 mosquitoes and stained with 0.5% mercurochrome diluted with PBS (w/v). The prevalence of infected mosquitoes and the number of oocysts per midgut were quantified via light microscopy. Only feedings resulting in highly infected mosquitoes were used for liver stage studies.

P. cynomolgi liver stage infection

Primary simian hepatocytes (BioIVT, lots HTV and HMP) were seeded at a density of 2.2 × 10⁴ cells per well in a collagen-coated 384-well black plate (Corning) two days prior to infection, as previously described⁵⁹. Sporozoites were obtained from the salivary glands of mosquitoes infected with P. cynomolgi UIS4-HT (UGA) and collected in RPMI 1640 (KD Medical). Primary hepatocytes were infected with 1 × 10⁴ sporozoites per well in hepatocyte media, spun at 200 × g for 5 min and further incubated in hepatocyte media. Sporozoites were removed 24 h post-infection and hepatocyte media containing 5% PSN was added to infected cells. The 5% PSN hepatocyte media was changed on day four and six post-infection. The samples were fixed with 4% paraformaldehyde eight days post-infection and stored at 4°C in DPBS until processed.

Immunofluorescence assay with primary hepatocytes

Fixed samples were incubated for 1 h at RT in BP buffer. The samples were then incubated at 4 ^oC overnight in BP buffer containing a combination of either rabbit anti-Pc HSP70⁵⁹ (1:5,000) and mouse anti-HaloTag (1:1,000, Promega, cat no. G9211), or rat anti-mCherry (1:200, Invitrogen, cat no. M11217) and mouse anti-GFP (1:100, Roche, cat no. 11814460001) primary antibodies. The next day, the samples were washed thrice with DPBS and stained with Hoechst (2 µg/mL, Invitrogen), AlexaFluor-568 goat anti-rabbit (Invitrogen, cat. A11036), AlexaFluor-488 goat anti-mouse (Invitrogen, cat. A11001) and/or AlexaFluor-568 goat anti-rat (Invitrogen, cat. A11077) (1:1,000) for 2 h at RT in BP buffer. The samples were then washed thrice with DPBS and stored in DPBS at 4 °C until imaging on the LSM 980 equipped with Airysan 2 (Zeiss) or the IXMC (Molecular Devices).

HaloPROTAC3-mediated degradation assay in primary human hepatocytes

Primary human hepatocytes (BioIVT, lots QWK and BGW) were seeded at a density of 2 × 10⁴ cells per well in a collagen-coated 384-well black plate (Corning) and transduced with 1 × 10⁵ TU of the P_CMV-mCherry-T2A-EGFP-HT-HiBiT lentivirus (Vector Builder, VB230405-1468zfc) 48 h post-cell seeding. The construct delivered by this lentivirus allows the host cell to express both a transduction marker (i.e., mCherry) and a degradation reporter (i.e., EGFP-HT-HiBiT), separated by the T2A self-splicing peptide. Briefly, lentiviruses were added to cells in hepatocyte media supplemented with 8 µg/mL of polybrene, and cells were spun at 800 × g for 1 h at 37 ^oC. The next day, lentiviruses were removed, and fresh hepatocyte media was added to the cells. Two days later, transduced cells were treated with compounds for 48 h and then fixed with 4% paraformaldehyde for 30 min at RT, immunostained, and imaged. To evaluate the degradation of EGFP-HT-HiBiT following HaloPROTAC3 treatment, 9 fields of view were imaged per technical duplicate using the 20× objective of the IXMC and analyzed using a Cell Profiler pipeline. Briefly, AlexaFluor488 (immunostained GFP) and AlexaFluor568 (immunostained mCherry) fluorescence intensities were measured in the area surrounding nuclei. Transduced cells were defined as cells with mCherry intensities greater than three standard deviations from the background intensity measured in non-transduced cells. Then, ratios of mean intensities associated with the GFP and mCherry markers were evaluated. We defined 0% GFP expression as the ratio of GFP background in non-transduced cells to the mCherry signal in transduced, DMSO-treated cells. The GFP/mCherry ratio in transduced, DMSO-treated cells was defined as 100% GFP expression.

Preparation of graphs, statistical analysis, and artworks

Using an image-processing Cell Profiler pipeline, number of liver stages, number of host nuclei, fluorescence intensities, and other parameters were quantified before analysis on TIBCO Spotfire (1.0_L_EN_02) and GraphPad Prism (version 9.5). The data generated was then used to create bar charts and response curves. Dose response data were normalized (100%, DMSO-treated samples) and plotted with the curve fit setting “log(inhibitor) vs. normalized response — Variable slope” on Prism. Specific statistical tests, number of technical replicates and number of independent experiments (N) are indicated in figure legends. Prism, FIJI (v1.53t), BioRender (biorender.com), Adobe Illustrator (v28.3), SnapGene® software (from Dotmatics; available at snapgene.com) and ChemDraw (revvity, Signals ChemDraw, v23.1.1.3) were used to create figures. Some micrographs were pseudocolored for representation.