MRC FAPESP: Defining the role of the hematopoietic parasite reservoir in Plasmodium vivax infection and pathology
Lead Research Organisation:
University of Glasgow
Department Name: College of Medical, Veterinary, Life Sci
Abstract
Plasmodium vivax is the most widely distributed malaria parasite and a major public health burden. Recent studies suggest that the majority of parasites is present outside of circulation, making it difficult to track and target them. We have demonstrated that bone marrow in particular represents an under-appreciated reservoir which supports P. vivax growth and differentiation to transmission stages. Parallel studies have also reported major parasite accumulation in the spleen. Based on these findings we hypothesize that the haematopoietic niche of bone marrow and spleen represents the main parasite reservoir during infection and drives disease severity. In this ambitious research program, we will analyze infected bone marrow and spleen tissue from a series of cohorts of naturally exposed patients in endemic areas in Brazil. We will perform histological, molecular and phenotypic characterization of sequestered and circulating parasite and host cell populations to systematically investigate and quantify the role of bone marrow and spleen for parasite infection, transmission, diagnosis and pathology. This work will thus contribute much needed insights and critical tools for the ongoing global malaria elimination campaign.
Technical Summary
Plasmodium vivax (P. vivax) places a heavy disease burden across endemic regions worldwide and presents a major challenge for malaria eradication. Recent studies from our laboratories and others have demonstrated that the extravascular bone marrow niche and potentially the spleen represent a major tissue reservoir for parasite proliferation and transmission stage maturation across Plasmodium spp. lineages, with significant implications for malaria diagnosis and treatment success. P. vivax is restricted to invasion of the youngest red blood cells, known as immature reticulocytes, which originate in the bone marrow. This restriction results in a distinct and poorly understood biology that we believe has precluded previous efforts to establish in vitro parasite culture and systematic study of P. vivax. Importantly, P. vivax elicits a potent host response and causes severe and fatal manifestations at much lower parasitaemia than P. falciparum, suggesting that P. vivax restriction to host cells enriched in the hematopoietic niche of BM and spleen could induce alterations in host homeostasis such as anaemia, lymphopenia, thrombocytopenia and splenomegaly. This collaborative proposal will employ for the first time a systematic investigation of the biology of the P. vivax reservoir in BM and spleen and its role in disease pathogenesis, infection and transmission in a series of cohorts of patients in Brazil. We will characterize parasite and host biology in infected tissues and investigate their behaviour ex vivo using a series of cutting edge technologies and in collaboration with world leading experts. The proposed work will thus close a major knowledge gap in our understanding of P. vivax biology and pathogenesis and provide critical tools for the development of new P. vivax control measures to support the ongoing malaria elimination agenda.
| Title | Protocol 1: SOP for Lab Processing of human peripheral blood samples. |
| Description | 1- Key observations: 1- Consistency and cell viability are absolutely key so important to process samples immediately after they have been collected. 2- Sample aliquots for Flow cytometry, Cytof and scRNASeq: should be cell suspensions only exposed to ice cold CryoStor10® (cell cryopreservation media - Sigma Aldrich and STEM Cell are the manufacturers) - never room temperature. 2.1- With blood samples: the samples constitute already dissociated cells and it is really important that cells are immediately put in Mr. Frosty and into -80C after being suspended in ice cold CryoStor10®. At the same time, cells are very delicate and essential to be gentle as possible when they are handled. 2.2- With tissue samples: work to minimise time in HypoThermosol® to less than 30mins. Leave 15 - 30mins in CryoStor10® before transferring to -80C: samples need a small amount of time (ideally 15mins) in CryoStor10® to penetrate samples but not more than 30mins. 3- Ideally 2 people working to enable this. If not, determine when best to collect samples considering "gaps" in the collection and processing workflow. 4- Storage: Thermoblocks store in -20 and Mr. Frosty's should be stored in cold room. 2- Preparation, Reagents and equipment: When blood and BM biopsy collection consent obtained and confirmation that biopsy will proceed, go to lab; Put ice machine on and waterbath; Prepare 3 boxes: 1 box with cool blocks and thermoblock with 1st 4 tissue specimens in, 1 small box with just a rack - final box prepared when biopsy started with ice and thermoblock; Put cold block and thermoblocks in the cool box. Put 4ml of HypoThermosol® into 50ml labelled falcon tubes and put in the thermoblock. Put into cold room until ready to be collected; Put liquid nitrogen in the Dewar; If don't have FBS thawed, then thaw a fresh 10% FBS aliquot (in waterbath); Set up Centrifuges; Large: set spin speed to 400g and temp to 21C; Small set spin speed to 1500g (1.5 and ensure g is pressed); Check Mr. Frosty's - ensure that they have sufficient isopropanol and that replaced if needed (every 5 freezes - change at case 6). Should be in cold room.; Ensure you have all the cryovials set up and labelled; Ensure you have all reagents ready: ficol, 10% FBS (in waterbath - take out when defrosted), Saponin on ice, PBS, Methanol on ice, Aliquot of PBS on ice, Glycerolyte aliquot, Cryostor10 on ice. Get ready and in the hood all the equipment you will need including: Cryovials - label, 15ml and 50ml Falcon's, Tips (p1000; p200;), serological pippetes, eppendorfs, pipettes (p200, p1000), pipette boy, transfer pipettes, discard beaker with virkon, petri dishes, scalpel or razor, forceps. 3- Processing Peripheral Blood samples (more details in Protocol - "PBMC and plasma protocol"): Blood sample (usually 2-4mL); Balance and spin samples for 400g for 10 mins at room temp; Prepare Ficol tubes while spinning; Remove plasma, aliquot into eppendorfs to respin - follow protocol for platelet poor plasma preparation. Aliquot each patient plasma samples in cryovials (500uL/cryovial) and label. This samples will be used for Luminex (200uL), ELISA pvLDH (20uL), Protein Array (20uL), Metabolomics and Lipidomics (200uL); Cell pellet tube: Add 10% FBS/ PBS and layer ficol; Put in centrifuge for PBMC separation; Isolate PBMCs and aliquot and collect red blood cells (RBCs); Aliquots of the PBMCs pellet and labelling: 3.1- From the pellet of PBMCs aliquot and label accordingly: A)- Resuspend in Trizol (at least 100uL of pellet in 500uL trizol) and snap-freeze in liquid nitrogen; for host RNASeq; B) - Resuspend in Cryostor10: freeze in Mr. Frosty 24h and transfer to liquid nitrogen (at least 400uL of pellet - following the Protocol 2). 3.2- Aliquots of the red blood cells pellet and labelling: A)- Prepare Smears (1-2uL per each slide): for each patient prepare a good number of slides for Giemsa staining and 100% methanol-fixed smears for 30sec at room temperature; B)- Resuspend in Trizol (at least 200uL of pellet in 800uL trizol) and snap-freeze in liquid nitrogen; C)- Resuspend in Cryostor10: freeze in Mr. Frosty overnight and transfer to liquid nitrogen (at least 600uL of pellet - following the Protocol 3). |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to streamline processing of human peripheral blood samples from Plasmodium vivax patients and healthy donors to prepare aliquots of different sample formats. It also allowed us to cryopreserve cells in the endemic field to create a local biobank of samples and also to ship samples to other study centres to be used in different experimental approaches, such as, bulk and single-cell RNA sequencing, suspension mass cytometry, flow cytometry, multiplex bead-based profiling, metabolomics, lipidomics and microscopy. |
| Title | Protocol 2: SOP for Lab Processing of Bone Marrow (BM) samples. |
| Description | 1- Key observations: Consistency and cell viability are absolutely key so important to process samples immediately after they have been collected. Sample aliquots for Flow cytometry, Cytof and scRNASeq: should be cell suspensions only exposed to ice cold CryoStor10® (cell cryopreservation media - Sigma Aldrich and STEM Cell are the manufacturers) - never room temperature. With BM samples: the samples constitute already dissociated cells and it is really important that cells are immediately put in Mr. Frosty and into -80C after being suspended in ice cold CryoStor10®. At the same time, cells are very delicate and essential to be gentle as possible when they are handled. With tissue samples: work to minimise time in HypoThermosol® to less than 30mins. Leave 15 - 30mins in CryoStor10® before transferring to -80C: samples need a small amount of time (ideally 15mins) in CryoStor10® to penetrate samples but not more than 30mins. Ideally 2 people working to enable this. If not, determine when best to collect samples considering "gaps" in the collection and processing workflow. Storage: Thermoblocks store in -20 and Mr. Frosty's should be stored in cold room. 2- Preparation, Reagents and equipment: When blood and BM biopsy collection consent obtained and confirmation that biopsy will proceed, go to lab; Put ice machine on and waterbath; Prepare 3 boxes: 1 box with cool blocks and thermoblock with 1st 4 tissue specimens in, 1 small box with just a rack - final box prepared when biopsy started with ice and thermoblock; Put cold block and thermoblocks in the cool box. Put 4ml of HypoThermosol® into 50ml labelled falcon tubes and put in the thermoblock. Put into cold room until ready to be collected; Put liquid nitrogen in the Dewar; If don't have FBS thawed, then thaw a fresh 10% FBS aliquot (in waterbath); Set up Centrifuges: Large: set spin speed to 400g and temp to 21C; Small set spin speed to 1500g (1.5 and ensure g is pressed); Check Mr. Frosty's - ensure that they have sufficient isopropanol and that replaced if needed (every 5 freezes - change at case 6). Should be in cold room. Ensure you have all the cryovials set up and labelled; Ensure you have all reagents ready: ficol, 10% FBS (in waterbath - take out when defrosted), Saponin on ice, PBS, Methanol on ice, Aliquot of PBS on ice, Glycerolyte aliquot, Cryostor10 on ice. Get ready and in the hood all the equipment you will need including: Cryovials - label 15ml and 50ml Falcon's, Tips (p1000; p200;), serological pippetes, eppendorfs, pipettes (p200, p1000), pipette boy, transfer pipettes, discard beaker with virkon, petri dishes, scalpel or razor, forceps. 3- Processing Bone Marrow aspirate samples: BM aspirate sample (usually 2-4mL); Balance and spin samples for 400g for 10 mins at room temp; Remove supernatant and aliquot into eppendorfs to respin - follow protocol for platelet poor plasma preparation: Centrifuge the plasma at 100 g for 10 minutes for removal of residual leukocytes. Add Prostaglandin E1 300nM to minimize platelet aggregation and subsequently centrifuge at 800g for 20 minutes to obtain the platelet pellet. The supernatant of this centrifugation was centrifuged at 1000 g for 10 minutes to obtain platelet poor plasma (PPP). Aliquot each patient plasma samples in cryovials (500uL/cryovial) and label. This samples will be used for Luminex (200uL), ELISA pvLDH (20uL), Protein Array (20uL), Metabolomics and Lipidomics (200uL). 4- Aliquots of the BM cell pellet : Aliquot 1 - Prepare Smears: 1-2uL per each slide. Aims: for Giemsa staining Immunofluorescence assay (fix with 100% methanol for 30sec at room temperature); Aliquot 2 - Resuspend in Trizol: At least 200uL. Snap-freeze in liquid nitrogen. Aims: Split sample and perform leukodepletion in one sample? Aliquot 2.1- Non-leukodepleted sample - host bulk transcriptomics Aliquot 2.2- Leukodepleted sample - parasite bulk transcriptomics Aliquot 3 - Resuspend in Cryostor10: 600-800uL in Cryostor10 (following protocol below); Aims: Aliquot 3.1 - for 10X Library preparation; Aliquot 3.2 - Cytof in cell suspension (Helios); Aliquot 3.3 - to prepare FFPE cell blocks (RNAScope and imaging cytof - Hyperion). Cryostor10 1. Keep on ice until you are putting the cells in Cryostor. Then resuspend RBC pellet in 1ml of cryostor - put in 2 aliquots and put on ice (do RBC pellet first but do together with PBMCs to avoid cells sitting in Cryotstor). Up to 30min on ice 2. Freeze in Mr. Frosty: once a Mr. Frosty is in Freezer leave in for >4hours or overnight until samples are ready for liquid nitrogen; 3. AFTER 4hs OR THE NEXT DAY FIRST THING: transfer samples for 10X Library preparation from Mr. Frosty to liquid nitrogen - this needs to be done very quickly to avoid any thaw. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to streamline processing of human peripheral blood samples from Plasmodium vivax patients and healthy donors to prepare aliquots of different sample formats. It also allowed us to cryopreserve cells in the endemic field to create a local biobank of samples and also to ship samples to other study centres to be used in different experimental approaches, such as, bulk and single-cell RNA sequencing, suspension mass cytometry, flow cytometry, multiplex bead-based profiling, metabolomics, lipidomics and microscopy. |
| Title | Protocol 3: PBMC and Plasma Isolation and Cryopreservation. |
| Description | 1. Introduction: PBMCs are sensitive to cryopreservation. Length and type of storage conditions are some of the reasons for variability in cell recovery and cell viability of frozen PBMCs. It is critical to limit variability to ensure optimal yield and thus function in the immunoassays. The success of a protocol depends upon the adequate collection processing, cryopreservation and retrieval of specimens. Guidelines for sample collection and storage need to anticipate the requirements of future studies that are yet to be designed or technological advances, which are in the early stages of development. While this is not always possible, certain basic tenets exist. For example, the cryopreservation process should be very carefully controlled and all specimens should be collected and processed using aseptic techniques and Universal Safety Precautions. These include the use of STERILE tubes, pipette tips and reagents, and a work environment that is designed to prevent contamination of samples and provide adequate safety measures for everyone in the lab. 2. Purpose: The Purpose of this SOP is to describe the procedure for the isolation and freezing of plasma, PBMCs and iRBC. 3. Materials and Equipment: 3.1- Anticoagulated blood. Blood specimens will be kept at 15-20°C and must be processed within 2 hours of specimen draw. Blood samples should be processed as quickly as possible to ensure maximal PBMC viability. Take the temperature of the 15-20°C sample holder on arrival. Warm cooled blood samples in a 21°C water bath. 3.2- Reagents: Sterile 1x PBS (Ca-free and Mg- free). Observe manufacturer's expiration date. Label bottle with open date; use opened bottle within three months. Ficoll-Paque Premium; FBS - heat inactivated; 10% FBS; Cryostor10 freezing medium (StemCell 7930); Virkon. 3.3 Equipment: 0.22uM filters for sterilization; 2mL Sterile Cryopreservation Vials suitable for use in liquid nitrogen; Class 2 Bio-safety cabinet; Mr. Frosty freezing unit. Store in the cold room when not in use. Replace Isopropanol every 5 times. If the container is accidentally left out of the fridge, cool it down at -20°C for 10 minutes before use; Centrifuge Sorvall ST40R; Refrigerator 4°C; -80°C freezer; Liquid nitrogen storage tank and boxes; Ice - turn on ice machine 30 minutes prior to intended use; 5mL, 10mL and 25mL sterile pipettes; 15mL & 50 mL sterile conical tubes; 200µL and 1000µL micropipettes; Disposable micropipette tips; Disposable transfer pipettes (Fisherbrand Cat# 13-711-20); Timer; Interval timers; Disposable gloves; Water bath; Safety glasses; Lab coat; Plastic 1 L beaker; Ice Bucket; Microscope - counting cells; Hemocytometer and cover slip. 4- PBMC separation: Turn on tissue culture hood, wipe down with ethanol. Turn on ice machine. Turn on centrifuge and set to 21°C. Put frozen 10% FBS in waterbath to defrost, put 0.15% saponin aliquot on ice to thaw, put methanol aliquot on ice to cool; Record in CRF the blood volume in the blue (sodium citrate) tube. Balance and spin for 10 mins at 400g; Acceleration 9; decel 9; With 1000ul pipette aspirate 1.5ml of plasma (for 5ml blood volume); 1ml if less than 5ml and 0.5ml for 2ml blood volume being absolutely sure not to disturb the cells. 4.1- Follow protocol for platelet poor plasma preparation: Centrifuge the plasma at 400 g for 10 minutes for removal of residual leukocytes. Add Prostaglandin E1 300nM to minimize platelet aggregation and subsequently centrifuge at 800g for 20 minutes to obtain the platelet pellet. The supernatant of this centrifugation was centrifuged at 1000 g for 10 minutes to obtain platelet poor plasma (PPP). Aliquot each patient plasma samples in cryovials (500uL/cryovial) and label. This samples will be used for Luminex (200uL), ELISA pvLDH (20uL), Protein Array (20uL), Metabolomics and Lipidomics (200uL). 4.2- Replace plasma with equal volume of 10% FBS with PBS (PBS should be calcium and magnesium free); Transfer the blood + FBS to a 15ml conical tube and note volume; .6 Dilute with equal volume of PBS; gently pipette 2 times to mix (no bubbles); Prepare 15mL conical tube with 4mL of ficoll. Layer diluted blood on ficoll as follows: Tip the tube over at a 30 degree angle. With a 10mL serological pipette, layer the blood very carefully onto the ficoll so as not to mix the ficoll layer with the blood. Start off dripping until there is a layer of blood on the ficoll and then you can slowly increase the rate, gradually bringing the tube to an upright position; Spin at 1500 RPM (450g) on Sorvall ST40R for 27 minutes at 21°C with acceleration=3 and deceleration=2 (program #1 on Sorvall ST40R). After centrifugation, you will see the following layers (from top to bottom): Plasma, then a cloudy interface layer containing PBMCs, then the ficoll, then an RBC layer with the granulocyte layer on top. While waiting for the spin to end, prepare 15 mL conical tube containing 6 mL of PBS; Using a sterile transfer pipette or a p1000 micropipette very carefully remove the cloudy interface layer (PBMC layer). Use a circular motion just over the top of the cloudy interface layer. About 1 mL or so will be removed. It doesn't matter if serum is mixed in, but minimize aspiration of Ficoll. Avoid aspirating the ficoll solution by maintaining the pipette tip ABOVE the cell layer and SLOWLY drawing the cells up into the pipette. Ideally give the remaining tube containing the RBC pellet to a second person. If this isn't possible put the tube containing the RBC and the Ficol carefully on ice while you prepare the PBMCs. Place the cloudy interface layer into the 15mL tube with PBS. If the PBMC layer gets mixed with RBCs while collecting, discard the tip, do NOT combine the mixed sample with the rest of the PBMCs already collected. Add PBS to top off the tubes at 15 mL. Cap the tube and mix by gently inverting tube 2 times. Do NOT create air bubbles. Air bubbles reduce cell viability. Spin at 1000 RPM (200g) at 21? C for 15 minutes with acceleration=6 and deceleration=3 (program #2 on Sorvall ST40R). Discard supernatant into waste container by slowly tilting tube in one smooth motion. Leave 2ml supernatant not to disturb the pellet. Do NOT shake the tube while aspirating the supernatant. If the PBMC pellet is large remove with a serological pipette instead. Because the spin is at such a low speed the pellet is loose and very easy to resuspend. Gently loosen the cell pellet by flicking the bottom of the tube on the palm of your hand until all clumps are gone. Top off with PBS and gently invert the tube twice. Spin 15 mL conical tube with the cell suspension for 11 minutes at 1300 RPM (350g), with acceleration=9 and deceleration=8 (program #3 on Sorvall ST40R). Discard supernatant into waste container by tipping in one smooth motion until it is upside down. While upside down briefly blot with clean paper towel. Gently loosen cell pellet by flicking on your palm until all clumps are gone. Put on ice. 5- CRYOPRESERVATION - work quickly from this point. Minimal time between putting cells in Cryostor10 and putting in -80C is essential to maximise cell viability. Keep the cells for 5 mins on ice while you collect the Mr. Frosty and ice cold Cryostor10. Aliquot 1 - Resuspend in Trizol: at least 100-200uL of PBMC pellet in 500uL Trizol in pre-labelled tube. Snap-freeze and storage in liquid nitrogen. Aim: RNAseq Aliquot 2 - Resuspend in Cryostor-10: Aim: for 10X Library preparation; Cytof in cell suspension (Helios); Cryostor10: Put 2ml of ice cold Cryostor10 into the 15ml conical. Resuspend 600uL of the pellet using a wide bore tip and very gently pipette back and forth 5 times. Using a new p1000 tip aliquot 500ul into each of the pre-labelled cryovials; carefully screw on lids. Put cryovials in Mr. Frosty, ensure that the lid is screwed tight and put in - 80. Quick transfer is essential to ensure maximal cell viability. Cells should remain in -80°C for >4 hours. In afternoon (for early morning sample) OR the morning after (for afternoon sample), transfer cells RAPIDLY from the -80°C freezer to the liquid nitrogen storage. This step must be done quickly because even a brief increase in temperature can reduce cell viability. Record time of transfer to -80°C freezer, time of transfer to liquid nitrogen and location of specimens in blood quality control log. 6- Quality Control/Assurance: PBMC are frozen in Cryostor10 which contains DMSO that is a toxic chemical that allows the cells to freeze without lysing the cell membrane. Thus, during the cryopreservation procedure, cells must be processed quickly to minimize the length of exposure to DMSO to preserve the function of the cells and insure a good recovery of live cells. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to streamline processing of human peripheral blood samples from Plasmodium vivax patients and healthy donors and isolation of PBMCs. It also allowed us to cryopreserve mononuclear cells from peripheral blood in the endemic field to create a local biobank of samples and also to ship samples to other study centres to be used in different experimental approaches, such as, bulk and single-cell RNA sequencing, suspension mass cytometry, flow cytometry, multiplex bead-based profiling, metabolomics, lipidomics and microscopy. |
| Title | Protocol 4: Infected red blood cell (RBC) separation, enrichment and fixation. |
| Description | This is only done for malaria positive patients 1. After the PBMCs have been carefully transferred to a separate 15ml falcon with a transfer pipette remove and discard the supernatant containing Ficol and FBS from the red blood cell (RBC) pellet, careful not to disturb the RBC pellet. 2. Top up with PBS; 3. Spin at 500g for 3mins; 4. Remove and discard the supernatant; 5. You should have about 0.8 - 1.5ml of pellet. Divide the pellet into aliquots: Aliquot 1 - Prepare Smears: 1-2uL per each slide; Aims: for each patient prepare a good number of slides for Giemsa staining; Immunofluorescence assay (fix with 100% methanol for 30sec at room temperature); Aliquot 2 - Resuspend in Trizol: 150-200uL of pellet in 800uL trizol and snap-freeze in liquid nitrogen; Aims: for parasite bulk transcriptomics; Aliquot 4 - for storage in glycerolyte: 200uL; for ex vivo assays. Glycerolyte: 1. Measure the volume of packed RBC using a pipette. 2. The total volume glycerolyte to be added is twice the packed RBC volume. e.g. 2 ml for 1ml pellet 3. Add 1/3 volume of glycerolyte drop wise while gently shaking the tube, and allow to stand for 3-5 min. e.g. 0.67ml for 1 ml pellet 4. Add remaining glycerolyte drop-wise while gently shaking the tube. e.g. 1.37 ml for 1 ml pellet 5. Aliquot into 2 cryovials 6. Freeze immediately at -80oC. After > 4hrs transfer into liquid nitrogen. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to streamline processing of human peripheral blood samples from Plasmodium vivax patients and healthy donors to purify red blood cells (infected of not) for assays using fresh parasites or cryopreservation. It also allowed us to create a local biobank of samples in the endemic area and also to ship samples to other study centres to be used in different experimental approaches, such as, bulk and single-cell RNA sequencing. |
| Title | Protocol 5: Thawing and Resuspension of Bone Marrow Mononuclear cells (BMMCs) for single-cell RNAseq. |
| Description | 1- List of reagents/equipment required: GIBCO IMDM; UltraPure Bovine Serum Albumin (BSA, 50mg/mL); BSA in DPBS (10%); PBS with 10% BSA; Seradigm Premium Grade Fetal Bovine Serum (FBS); MACS SmartStrainers 30um; Flowmi Cell Strainer 40um; RPMI 1640; PBS without Calcium and Magnesium; Trypan Blue Stain (0.4%); Countess II FL automated cell counter; Countess II FL automated counting chamber slides; DNA LoBind Tubes, 2mL; Wide-bore pipette tip. 2- Thawing & Resuspension: Set up a water bath to 37°C before staring cell thawing. All cell washes are performed at room temperature; Remove cryovials from storage and immediately thaw in the water bath at 37°C for 2-3 min. DO NOT submerge the entire vial in the water bath. Remove from the water bath when a tiny ice crystal remains; In a biosafety hood, slowly transfer thawed cells to a 50-ml conical tube using a wide-bore pipette tip. Rinse the cryovial with 1 ml warm complete growth medium (10% FBS in IMDM or RPMI) and add the rinse dropwise (1 drop per 5 sec) to the 50-ml conical tube while gently shaking the tube; Sequentially dilute cells in the 50-ml conical tube by incremental 1:1 volume additions of media for a total of 5 times (including dilution at step b) with ~1 min wait between additions (see Appendix). Add media at a speed of 1 ml/3-5 sec to the tube and swirl; Centrifuge at 300 rcf for 5 min; Remove most of the supernatant, leaving ~1 ml and resuspend cell pellet in this volume using a regular-bore pipette tip. Cell pellet may be present on the side or on the bottom of the tube; Add an additional 9 ml complete growth medium (at a speed of 1 ml/ 3-5 sec) to achieve a total volume of ~10 ml; Determine the cell concentration and viability using a Countess II Automated Cell Counter or hemocytometer; Transfer ~2 x 106 cells into a new 50-ml tube; Centrifuge at 300 rcf for 5 min; Remove the supernatant without disrupting the cell pellet; Using a wide-bore pipette tip, add 1 ml 1X PBS + 0.04% BSA and gently pipette mix 5x. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us recover cryopreserved mononuclear cells from bone marrow aspirates with high viability. The cells are in good condition to be used in different downstream experimental approaches where high viability is a requirement and limiting factor, such as single-cell RNA sequencing and suspension mass cytometry. |
| Title | Protocol 6: Cell surface protein labelling for single-cell RNA sequencing and antibody-detected tag experiments. |
| Description | 1- List of reagents/equipment required: Human TruStain FcX (Fc receptor blocking solution); TotalSeq Antibody-Oligonucleotide Conjugates; Cell Staining Buffer; UltraPure Bovine Serum Albumin (BSA, 50mg/mL); Fetal Bovine Serum, heat inactivated; Bovine Serum Albumin in DPBS (10%); Trypan Blue Stain (0.4%); PBS or DPBS without Calcium and Magnesium; Countess II FL automated cell counter; Countess II FL automated counting chamber slides; DNA LoBind Tubes, 2mL. 2- Cell surface protein labeling protocol: • This protocol was optimized using TotalSeq-B/C antibody-oligonucleotide conjugates from BioLegend. • This protocol was demonstrated using 0.2-2 x 106 cells. Wash cells according to the appropriate 10x Genomics Demonstrated Protocol for the cell type being prepared. • All steps can be performed in 5-ml centrifuge tubes, 15-ml centrifuge tubes or round-bottom FACS tubes. 2.3.1- Buffers - Preparation For samples containing >70% viable cells • Chilled (4°C): PBS + 1% BSA • Chilled (4°C): PBS + 0.04% BSA For samples containing <70% viable cells • Chilled (4°C): PBS + 10% FBS 2.3.2- Prepare Antibody Mix Supernatant • Add appropriate/manufacturer's recommended amount of antibody-oligonucleotide conjugates to a 1.5-ml microcentrifuge tube. • If using a custom lyophilized antibody: Resuspend the antibody-oligonucleotide conjugates in an appropriate volume of labeling buffer. • Centrifuge the mix at 14,000 rcf for 10 min at 4°C. • Transfer the supernatant (containing Antibody Mix) to a new tube and maintain at 4°C. 2.3.3- Cell labeling • Transfer cells to a new 5-ml tube and add chilled PBS + 0.04 % BSA for a total 1 ml volume (cells are already in this solution in the end of Protocol 1). For samples containing <70% viable cells, use chilled PBS + 10% FBS. • Centrifuge cells at 4°C. Use of swinging-bucket rotor is recommended for higher cell recovery. Centrifugation speed and time depends upon the sample type. Larger or fragile cell types may require slower centrifugation speeds. • Remove the supernatant without disturbing the pellet. • Resuspend cell pellet in 50 µl chilled PBS + 1% BSA or Cell Staining buffer. For samples containing <70% viable cells, resuspend in chilled PBS + 10% FBS. If staining in small tubes use LoBind DNA tubes. • Add 5 µl Human TruStain FcX. Gently pipette mix. • Incubate for 10 min at 4°C. • Add the prepared Antibody Mix supernatant. If also performing FACS enrichment, add FACS antibody pool. • Add chilled PBS + 1% BSA or Cell Staining buffer to the cells to bring the total volume to 100 µl. Gently pipette mix 10x (pipette set to 90 µl). For samples containing <70% viable cells, add chilled PBS + 10% FBS. • Incubate for 30 min at 4°C. If using FACS antibodies, incubate without light exposure. • Proceed to cell washing. Cell washing protocol and wash & resuspension buffers depends upon the sample. Choose appropriate protocol based on starting sample viability. Wash protocol below for samples with >85% viable cells. 3- Wash: • Thorough washing of cells post labeling is critical to obtain high-quality data. Optimization of centrifugation speed/time may be needed based on cell type. Buffers - Preparation: • Chilled (4°C): PBS + 0.04% BSA; • Chilled (4°C): PBS + 1% BSA; • Cell Staining Buffer. A- Wash 1: • Wash by adding 3.5 ml chilled PBS + 0.04% BSA to the labeled cells. Gently pipette mix. • Centrifuge at 4°C. Centrifugation speed and time depends upon the sample type. Use Table 2 for guidance. • Remove the supernatant without touching the bottom of the tube to avoid dislodging the pellet. Leaving behind excess supernatant may cause non-specific binding, which may result in increased background reads during sequencing. • Resuspend the pellet in 100 µl room temperature PBS (Ca/Mg-free) and transfer to a new 5-ml tube. Incubate for 5 min at room temperature. B- Wash 2: • Using a pipette tip, resuspend the pellet or cells in 3.5 ml chilled PBS + 1% BSA. • Centrifuge at 4°C. Centrifugation speed and time depends upon the sample type. Use Table 2 for guidance. • Remove the supernatant without touching the bottom of the tube to avoid dislodging the pellet. C-Wash 3 and 4: • Repeat wash 2 2x for a total of four washes. D- Final step: • Slowly filter cells through a 40 µm Cell Strainer. • Verify cell concentration and viability after filtration. • Based on starting concentration and assuming ~50% cell loss, add an appropriate volume chilled PBS + 1% BSA to obtain a concentration of 700-1,200 cells/µl. If necessary leave suspension in DNA LoBind tubes. • Determine cell concentration and viability using a Countess II Automated Cell Counter or a hemocytometer. • Proceed immediately to relevant single cell RNA Sequencing protocols with Feature Barcode technology. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to prepare highly viable bone marrow mononuclear cells for single-cell RNA sequencing and suspension mass cytometry. The output data confirmed the success of the method in producing cells with high quality for such experimental approaches. |
| Title | Protocol for ex vivo Plasmodium vivax culturing. |
| Description | 1- Parasite enrichment: 10X KCl stock buffer was made as 100 mM HEPES, 1,150 mM KCl, and 120 mM NaCl in water, and 100% KCl Percoll was made by diluting the 10 KCl stock buffer 1:9 with stock Percoll (GE Healthcare Life Sciences, Pittsburgh, PA). The dilution buffer was made as 20 mM HEPES, 1 mM MgCl2 , 1 mM NaH2 PO4 , 10 mM glucose, 0.5 mM EGTA, 115 mM KCl, and 12 mM NaCl in water, with the pH adjusted to 7.4. Next, 1.080 g/ml KCl Percoll was made by mixing 100% KCl Percoll with the dilution buffer at a ratio of 64.03% to 35.97%, and the density was veri?ed at ambient temperature using a DMA 35 portable density meter (Anton Paar, Graz, Austria). Samples (0.100 ul to 1500 l) of packed parasitized blood were resuspended to 3 ml using incomplete Iscove's modi?ed Dulbecco's medium (IMDM) and layered on 3 ml of 1.080 g/ml KCl Percoll at ambient temperature in a 15-ml conical tube. This was centrifuged at ambient temperature for 15 min at 1,200 g with slow acceleration and no brake. The interface was removed, washed in incomplete IMDM, and applied to assays. 2- Thawing P. vivax isolate and ex vivo culture: First, prepare complete IMDM media with 10-20%v/v AB+ heat-inactivated human serum with 50ug/mL gentamicin. Then, retrieve cryovial from liquid nitrogen and thaw at RT° for 1-2min. Transfer to 15mL tube and measure volume (should be around 200-500uL). Drop by drop, add 0.1x volume of 12% NaCl (50uL for 500uL parasites) while gently shaking the tube. Incubate at RT 2min. Drop by drop, add 10x volume of 1.6% NaCl (5mL for 500uL parasites) while gently shaking the tube. Centrifuge at 1500rpm 5min. Discard supernatant. Add 10x volume of complete IMDM media. Centrifuge at 1500rpm 5min. Discard supernatant. Resuspend pellet with 5mL of complete IMDM. Make a smear to determine the parasitemia. Add sufficient media to make 1% hematocrit. Distribute iRBCs in a 24 or 48-well plate accordingly to the number of time points to be collected. Prepare cultures with 1% hematocrit at 37°C in 5% CO2, 1% O2, and 94% N2. Change media every 12h and smear culture every time point to monitor growth and to prepare methanol-fixed smears. |
| Type Of Material | Biological samples |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | This method allowed us to culture and grow ex vivo cryopreserved isolates of Plasmodium vivax parasites to be used in transcriptomic experiments, such as bulkRNAseq, to establish a reference dataset of transcriptional changes during the intraerythrocytic cycle. It allows also to test and optimise protocols for single-cell RNAseq of cultured fresh parasites for experiments described in the Aim 2 of our research grant. |
| Title | In silico pipeline for imaging mass cytometry analysis. |
| Description | 1- Preprocessing: Quantitative analysis of the molecules captured by multiplexed imaging requires processing of multi-channel images. This involves image extraction and pre-processing, image segmentation, and the quanti?cation of biological objects such as cells. We have developed a modular computational work?ow to process and analyze multiplexed imaging data. The presented work?ow is customizable and integrates with a variety of downstream analysis strategies by employing standardized data formats. First, after image acquisition, a single MCD ?le can hold raw acquisition data for multiple regions of interest, optical images providing a slide level overview of the sample ("panoramas"), and detailed metadata about the experiment. Raw IMC data pre-processing can be performed using different python packages such as the readimc and imctools, both developed by Bernd Bodenmiller's group at ETH Zurich. These packages allow extracting the multi-modal (IMC acquisitions, panoramas), multi-region, multi-channel information contained in raw IMC images. While imctools contains functionality specific to the generate input files for the segmentation pipelines, the readimc package contains reader functions for IMC raw data and should be used for this purpose. Starting from IMC raw data and a "panel" file, individual acquisitions are extracted as TIFF and OME-TIFF files. The panel contains information of antibodies used in the experiment and the user can specify which channels to keep for downstream analysis. In this step, random tiles of the single-channel TIFF files can be scaled 2X and cropped to facilitate pixel labelling. 2- Denoise: The next step in the analysis pipeline is the denoise of the image files. The signal-to-noise ratio for particular markers in the IMC dataset can be low despite optimization of staining conditions, and the presence of pixel intensity artifacts can deteriorate image quality and the subsequent performance of downstream analysis. To improve the signal-to-noise ratio and remove artifacts, we use the IMC-Denoise package, developed by Peng Lu et al at (Daniel L.J. Thorek's group) from the Washington University School of Medicine. This is an automated content-aware pipeline to restore IMC images. In general this pipeline deploys a differential intensity map-based restoration (DIMR) algorithm to detect and remove hot pixels. Hot pixels are concentrated areas of high counts which are uncorrelated with any biological structures. These can result from deposition of metal-stained antibody aggregates. In IMC images, single hot pixels are the most common outliers, and small hot clusters with several consecutive pixels may also exist. Next, the pipeline runs a self-supervised deep learning algorithm for filtering shot or background noise called DeepSNF. Shot noise exists because of imaging process relies on ion counting, this is pixel-independent, but signal-dependent. Additionally, noise levels are related to other factors, such as variations in conjugated metal isotopes, antibody concentration and arrangement. DeepSNF is capable of unmixing the signal and background and filter the background out. IMC-Denoise is very efficient in hot pixel and background removal and delivers significant image quality improvement that leads to improved downstream analysis, with limited manual user manipulation. The automated DIMR approach results in improved cell segmentation, through robust removal of artifacts caused by hot pixels, correction of false positive data and all cell profiles. It enhances the sensitivity and specificity of cell phenotyping. 3- Cell segmentation: This step is the principal challenge in the analysis of tissue imaging data and it serves as the foundation for subsequent single-cell analysis. Unlike flow cytometry or single-cell RNA sequencing, where cells are in suspension and already at "single-cell" partition, tissue imaging is performed with intact specimens. Thus, to extract single-cell data, each pixel must be assigned to a cell in a process known as cell segmentation. And the features extracted through this process are the basis for downstream analyses, so inaccuracies at this stage can have far-reaching consequences for interpreting image data. We have been using different methods and packages to perform cell segmentation. Some methods are semi-automated pixel classification-based image segmentation while other are fully automated deep learning-based segmentation approaches. Random forest-based image segmentation is performed by training a pixel classifier. Pixels are classified as nuclear, cytoplasmic, or background. Then employing a customizable pipelines, the probabilities are then thresholded for segmenting nuclei, and nuclei are expanded into membrane regions to obtain cell masks. DeepCell is a deep learning-based image segmentation as presented by (Greenwald et al. 2021). Briefly, the user first aggregates image channels to generate two-channel images representing nuclear and membrane signals. Next, the DeepCell Python package is used to run Mesmer, a deep learning-enabled segmentation algorithm pre-trained on TissueNet, to automatically obtain cell masks without any further user input. TissueNet is a dataset built with trained segmentation models that contains more than 1 million manually labeled cells. (Mesmer is more accurate than previous methods, generalizes to the full diversity of tissue types and imaging platforms in TissueNet, and achieves human-level performance. Mesmer enabled the automated extraction of key cellular features, such as subcellular localization of protein signal). Segmentation masks are single-channel images that match the input images in size, with non-zero grayscale values indicating the IDs of segmented objects (e.g. cells). These masks are written out as TIFF files after segmentation. 4- Single-cell feature extraction: Following image segmentation, features of the detected cells are quanti?ed. Using the segmentation masks together with their corresponding multi-channel images, different pipelines such as imctools and ark-analysis can be used to extract cell-specific features. These include the mean pixel intensity of each protein per cell, morphological features (e.g. object area) and the objects' locations. Some packages can measure region properties, such as area, eccentricity, aggregated marker intensities (e.g., mean, median), and spatial neighbors. The measurement of spatial neighbors yields spatial object graphs, in which nodes correspond to cells, and cells in spatial proximity are connected by an edge. Distances between cells are computed based on their centroids or borders. These distances are used to construct spatial object graphs by thresholding the distance between cells or by detection of the k-nearest neighbors (k-NN). These graphs serve as a proxy for interactions between neighbouring cells. 5- Single-cell and spatial analysis: We have been using different python and R packages for the single-cell analysis, such as, Giotto, Scanpy, Seurat, ASTIR (Automated supervised assignment of cell types for highly multiplexed imaging). ASTIR employs a statistical machine learning model base don deep recognition neural networks) to assign cell type probabilities to each cell. We basically input an expression matrix containing the measured expression data (anndata object) and a priori speci?ed set of marker proteins for all expected cell types to be found (yaml file). For spatial analysis, we have been using Giotto, Squidpy, ATHENA, lisacLust, imcRtools. Cytomapper is used for imaging visualisation and quality control. 6- Spatial graph construction: The spatial graph is a graph of spatial neighbors with cells as nodes and neighborhood relations between spots as edges. We use spatial coordinates of the cells or spots to identify neighbors among them. Different approaches of defining a neighborhood relation among cells and spots are used for different types of spatial datasets but in imaging-based molecular data, a node can be defined as a cell (or pixel). The neighborhood set can be chosen based on a fixed radius, expressed in spatial units, rom the centroid of each cell of interest. Alternatively, other approaches, such as Euclidean distance or Delaunay triangulation, can be utilized to build the neighbor graph. For a fixed number of the closest spots for each spot, it leverages the k-nearest neighbors search from Scikit-learn. 7- Spatially variable features with Moran's I: There are several methods that aimed at to identify proteins that show spatial patterns, based on point processes or Gaussian process regression framework (SPARK, Spatial DE, trendsceek, HMRF). We used a simple approach based on the well-known Moran's I statistics which is in fact used also as a baseline method in the spatially variable gene analysis. 8- Neighborhood enrichment: Computing a neighborhood enrichment help us identify cell clusters that share a common neighborhood structure across the tissue. We can compute such score using an enrichment score on spatial proximity of clusters: if spots belonging to two different clusters are often close to each other, then they will have a high score and can be defined as being enriched. On the other hand, if they are far apart, and therefore are seldom a neighborhood, the score will be low and they can be defined as depleted. This score is based on a 1000x permutation-based test. 9- Co-occurrence across spatial dimension: In addition to the neighbor enrichment score, we compute cluster co-occurrence in spatial dimensions. This is a similar analysis of the one presented above, yet it does not operate on the connectivity matrix, but on the original spatial coordinates. The score is computed across increasing radii size around each observation (i.e. spots here) in the tissue. 10- Ligand-receptor interaction analysis: To identify the molecular instances that could potentially drive cellular interactions, we apply ligand-receptor interaction analysis based on CellPhoneDB [Efremova et al., 2020] and Omnipath [Türei et al., 2016]. 11- Heterogeneity quanti?cation: (i) spatial statistics scores that quantify the degree of clustering or dispersion of each phenotype individually, (ii) graph-theoretic scores that examine the topology of the tissue graph, such as, modularity; (iii) information-theoretic scores that quantify how diverse the tissue is with respect to different phenotypes present and their relative proportions, such as, Shannon index, Simpson index and Quadratic Entropy, and (iv) cell interaction scores that assess the pairwise connections between different phenotypes in the tissue ecosystem, such as cell infiltration and interaction scores. 12- Spatial community analysis: The detection of spatial communities was proposed by (Jackson et al. 2020). Here, cells are clustered solely based on their interactions as defined by the spatial object graph. In the following example, we perform spatial community detection separately for tumor and stromal cells. The general procedure is as follows: create a colData in the SpatialExperiment object that specifies if a cell is part of a tissue compartment. Use the detectCommunity function of the imcRtools package to cluster cells within the those compartments solely based on their spatial interaction graph. The spatial communities are stored in the colData of the SpatialExperiment object under the spatial_community identifier. We set the seed argument within the SerialParam function for reproducibility purposes. 13- Cellular neighborhood (CN) analysis: We use the imcRtools package to detect cellular neighborhoods. This approach has been proposed by (Goltsev et al. 2018) and (Schürch et al. 2020) to group cells based on information contained in their direct neighborhood. We perfom Delaunay triangulation-based graph construction or K-nearest neighbor graph, followed by neighborhood aggregation and then clustered cells into cellular neighborhoods using k-means clustering. When clustering cells based on the mean expression within the direct neighborhood, tissue patches are split across CNs. 14- Spatial domain analysis: We use the lisaClust package to compute local indicators of spatial associations (LISA) functions and clusters cells based on those. The package summarizes L-functions from a Poisson point process model to derive numeric vectors for each cell which can then again be clustered using kmeans. After creating the SegmentedCells object, the lisa function computes LISA curves across a given set of distances. We calculate the LISA curves within a 10µm, 20µm and 50µm neighborhood around each cell. Increasing these radii will lead to broader and smoother spatial clusters. However, a number of parameter settings should be tested to estimate the robustness of the results. Then, we can determine the cell type composition per spatial domain. 15- Spatial Context (SC) Analysis: Downstream of CN assignments, we analyze the spatial context (SC) of each cell using three functions from imcRtools. While CNs can represent sites of unique local processes, the term SC was coined by Bhate and colleagues (Bhate et al. 2022) and describes tissue regions in which distinct CNs may be interacting. Hence, SCs may be interesting regions of specialized biological events. First detect SCs using the detectSpatialContext function. This function relies on CN fractions for each cell in a spatial interaction graph (originally a KNN graph), which is calculated using buildSpatialGraph and aggregateNeighbors. Of note, the window size (k for KNN) for buildSpatialGraph should reflect a length scale on which biological signals can be exchanged and depends, among others, on cell density and tissue area. In view of their divergent functionality, we recommend to use a larger window size for SC (interaction between local processes) than for CN (local processes) detection. Different parameters should be tested. Subsequently, the CN fractions are sorted from high-to-low and the SC of each cell is assigned as the minimal combination of SCs that additively surpass a user-defined threshold. The default threshold of 0.9 aims to represent the dominant CNs, hence the most prevalent signals, in a given window. Next, SCs can be filtered based on user-defined thresholds for number of group entries (here at least 3 samples) and/or total number of cells (here minimum of 100 cells) per SC with filterSpatialContext. Lastly, we can use the plotSpatialContext function to generate SC graphs, analogous to CN combination maps in (Bhate et al. 2022). 16- Patch Detection Analysis: The patchDetection function allows the detection of connected sets of similar cells as proposed by (Hoch et al. 2022). 17- Interaction analysis - similar to Neighbourhood Enrichment Analysis: It allows statistically testing the pairwise interaction between all cell types of the dataset. For this, we use the testInteractions function in the imcRtools R package. Per grouping level (e.g., image), the testInteractions function computes the averaged cell type/cell type interaction count and computes this count against an empirical null distribution which is generated by permuting all cell labels (while maintaining the tissue structure). The returned DataFrame contains the test results per grouping level, "from" cell type (from_label) and "to" cell type (to_label). The sigval entry indicates if a pair of cell types is significantly interacting (sigval = 1), if a pair of cell types is significantly avoiding (sigval = -1) or if no significant interaction or avoidance is detected.These results can be visualized by computing the sum or average of the sigval entries across all images. The imcRtools package further implements an interaction testing strategy where the hypothesis is tested if at least n cells of a certain type are located around a target cell type (from_cell). This type of testing can be performed by selecting method = "patch" and specifying the number of patch cells via the patch_size parameter. These results are comparable to the interaction testing presented above. The main difference comes from the lack of symmetry. |
| Type Of Material | Data analysis technique |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | With this model we could build a scalable and modular computational framework for analysis of spatial proteomics and transcriptomics datasets to generate spatially-resolved single cell data. |
| Title | In silico pipeline for single-cell RNAseq/CITEseq analysis. |
| Description | 1- Alignment and quantification: scRNAseq gene expression (GEX) and antibody-detected tag (ADT) libraries were processed using Cell Ranger v7.1. Reads were aligned to the GRCh38 human genome and unique molecular identifiers (UMIs) deduplicated. CITE-seq UMIs were counted for GEX and ADT libraries simultaneously to generate feature-X droplet UMI count matrices. 2- Sample matrices were loaded in Seurat R package (v4.0) and Seurat objects for each sample were created. Background antibody-specific and nonspecific staining was subtracted from ADT counts in each sample using the dsb package using the default parameters. Empty/background droplets were de?ned by either clear breaks in the protein library size distribution or droplets de?ned as negative/background. The denoise.counts argument was set to TRUE which carries out the denoising cell-cell technical variations by estimating and regressing out the technical component for each cell and the use.isotype.control argument set to TRUE, de?ning each cell's technical component by combining isotype control antibody values and the mean of background counts. 3- Merged data from the samples were filtered to remove cells that expressed fewer than 200 genes and more than 8,000 genes and >10% mitochondrial reads. Non-empty droplets were called within each sample using the emptyDrops function implemented in the Bioconductor package DropletUtils (v1.10.3), using a UMI threshold of 200 and FDR of 1%. The probability of being a doublet was estimated for each cell per sample (that is, one 10x lane) using the DoubletFinder (v2.0.3) function to empirically determine an optimal 'pK' value for doublet detection. DoubletFinder analysis was performed on each sample separately using ten principal components (PCs), a 'pN' value of 0.25, and the 'nExp' parameter estimated from the fraction of ground-truth doublets and the number of samples. For all analyses, we employed standard pre-processing for all single-cell RNA-seq datasets. We ?rst performed log-normalization of all datasets, using a size factor of 10,000 molecules for each cell. We next standardized expression values for each gene across all cells (z-score transformation), as is standard prior to running dimensional reduction tools such as principal component analysis (PCA). Highly variable genes (HVGs) were identified using the Seurat vst algorithm. Harmony was used to integrate PCs by sample ID and experiment date and used to generate the neighborhood graph and embedded using UMAP. Clustering was performed using the Louvain algorithm with an initial resolution of 1. For initial clustering, differentially expressed genes were calculated using the Wilcoxon rank-sum test. 4- The query integrated BM dataset was mapped using Seurat to 2 reference datasets: Reference 1, from NIH Human Biomolecular Atlas Project (HuBMAP). It is a scRNAseq dataset from 297,627 BM cells from 39 donors and 3 different studies: Oetjen et al., 2018, Granja, Klemm, McGinnis, et al. 2019, and the Human Cell Atlas Immune Cell Census. All samples, which were generated using 10x Genomics v2, were integrated together and used to define a reference UMAP visualization and list of cell type annotations; Reference 2: Multi-modal dataset (RNA+protein) from 33,454 BM cells from 1 donor: Stuart et al., 2018. The following human Totalseq BioLegend antibodies were included in the pool: CD34, CD11a, CD11c, CD123, CD14 , CD16 ,CD161, CD56, CD57, CD3, CD4, CD8a, CD127-IL7Ra, CD197-CCR7, CD25, CD27 , CD278-ICOS, CD28, CD69, CD19, CD79b, HLA-DR, CD45RA, CD45RO, CD38. The expression levels of these protein markers can be "imputed"/"predicted" in the query datasets. The anchors between the query and reference datasets were found based on a CCA with 50 dimensions and the labels were transferred using a PCA for the weight reduction of the query dataset. The anchors were used for cell type annotation of clusters and to impute the ADT markers for the query cells which were then merged with the CITE-seq data and centered in order to run a UMAP analysis to visualize all the cells together. Numbers of cells of each cell type were quantified in each sample to compute a matrix of cell type × sample counts. 5- The weighted nearest neighbor (WNN) procedure implemented in Seurat v4 was applied to integrate the GEX and ADT data from each sample to de?ne a single uni?ed representation of single-cell multimodal data. For each cell, the procedure learns a set of modality weights, which re?ect the relative information content for each data type in that cell. This enables the generation of a WNN graph: for each cell, this graph denotes the most similar cells in the dataset based on a weighted combination of protein and RNA similarities. The WNN graph was used as input for common downstream analytical tasks including UMAP visualization, graph-based clustering, and the identi?cation of developmental trajectories. 6- We used expert immunological knowledge to guide a curated integration of the data from the different modalities (GEX and ADT ) to identify and label the cell sub-populations present in the CITE-seq dataset. First, it was performed separate clustering of gene expression and clustering of surface protein expression. Next, led by expert understanding of the two feature spaces, we prioritized use of ADT surface phenotype for de?nition of major cell lineages and subsets where de?nitive marker expression was available. Cell types and subsets were further re?ned using information from the GEX and ADT layers, or in the absence of de?nitive ADT information, they were identi?ed by GEX cluster phenotype. |
| Type Of Material | Data analysis technique |
| Year Produced | 2023 |
| Provided To Others? | Yes |
| Impact | It allowed us to confirm that our protocols for sample processing, cryopreservation of cells and antibodies staining work. In addition, we could also generate a reference of healthy BM single-cell dataset by combining our data with published data to be used when we obtain single-cell RNAseq/CITEseq data from Plasmodium vivax-infected patients. |
| Description | Poster presentation at the American Society of Tropical Medicine and Hygiene (ASTMH) |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Presenting data at this event was good for networking, sharing ideas and protocols with other researchers. We could strength collaborations with groups in Australia/Indonesia. Researchers showed interested regarding our approaches to establish the prospective longitudinal study with BM aspirate samples. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Talk at the 8th International Conference on Plasmodium vivax Research, |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Participation in this event as an invited speaker gave us the opportunity to network, share ideas and protocols with other researchers. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Talk at the Annual Society Conference of the British Blood Transfusion Society |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Participation in this event as an invited speaker gave us the opportunity to network, share ideas and protocols with other researchers. Researchers showed interest in establishing collaborations regarding access to blood samples enriched in reticulocytes from the NHS Blood Bank. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Talk at the EMBL Conference: BioMalPar XVIII: biology and pathology of the malaria parasite |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Participation in this event as an invited speaker gave us the opportunity to network, share ideas and protocols with other researchers. Researchers showed interest in establishing collaborations regarding the investigation of neutrophils and trained immunity in the bone marrow of Plasmodium vivax patients. Researchers also were interested regarding our approaches to establish the prospective longitudinal study with BM aspirate samples. |
| Year(s) Of Engagement Activity | 2022 |