Apoptotic dysregulation mediates stem cell competition and tissue regeneration – Nature Communications

Mice

All animal studies were approved by the Committee on the Ethics of Animal Experiments of the Technion, Israel institute of Technology. Mice were purchased from the Jackson Laboratory. These mutant mice have targeted mutations of Bax and Bak1. For targeted mutation of Bax, Dr. Stanley J. Korsmeyer’s lab (Harvard Medical School) designed a vector to flank exons 2–4 with loxP sites. This construct was electroporated into 129 × 1/SvJ-derived RW4 embryonic stem (ES) cells.

Mice resulting from germline transmission of this Bax^fl allele (on a mixed B6;129 genetic background) were generated. For targeted mutation of Bak1, Dr. Craig B. Thompson’s lab (University of Pennsylvania) designed a vector to replace exons 3–6 (encoding the Bcl2 homology domains) with a neo cassette. This construct was electroporated into (129 × 1/SvJ x 129S1/Sv)F1-derived R1 ES cells. Mice resulting from germline transmission of this allele (on a mixed B6;129 genetic background) were generated. These two mutant strains were bred together and maintained as homozygotes by Dr. Stanley J. Korsmeyer’s lab (Harvard Medical School) prior to arrival at The Jackson Laboratory. Bak-Baxfl/fl (B6;129–Bax^tm2Sjk Bak1^tm1Thsn/J; JAX stock #006329), K15-CrePGR (B6;SJL-Tg(Krt1-15-cre/PGR)22Cot/J), and ROSA-26-Confetti (B6.129P2-Gt (ROSA) 26Sortm1 (CAG-Brainbow2.1)Cle/J) mice were purchased from Jackson Laboratories. Foxn1^nu/nu mice and C57BL/6 black WT mice were purchased from Invigo.

Baxfl/fl mice were generated using standard homologous recombination floxing exon 6 surrounding the catalytic region QACGG of Bax. FRT flanked Neo cassette was inserted between the LoxP sites and FLP recombinase was used to remove the Neo cassette. Embryonic stem (ES) cells were targeted with the vector via electroporation. The screening of the ES cells was done by Genentech. K15-CrePGR/Baxfl/fl mice were prepared by crossing K15-CrePGR with Baxfl/fl mice. The promoter-specific expression of CrePGR (Cre fusion to progesterone receptor) enables conditional deletion of the Bax gene in the Krt15⁺ HFSC population, as well as controllable activation of CrePGR upon treatment with the synthetic PGR ligand, RU486. Only mice that were Cre positive were used for K15-CrePGR and K15-CrePGR/Baxfl/fl strains and only mice homozygous for the floxed and R26 allele were used for Baxfl/fl. Harvested tissue samples were obtained from both female and male mice.

Cre-recombinase induction

Prior to induction with RU486, mice were shaved with electric clippers and treated topically with hair removal cream (Nair). To activate Cre-recombinase, RU486 was dissolved in DMSO (30 mg/mL) and diluted in PBS (1.5 mg/mL). A final concentration of 13.5 mg/kg was injected at the tip of the tail base, in conjunction with topical application (30 mg/ml RU486 in 80% ethanol). Injections were performed on 8-week-old telogenic mice for 5 consecutive days and skin was harvested 5 days post injection or 14 days post injection. Primers used for genotyping are outlined in Table S1.

Wound infliction

For wounding experiments, induction was performed as described above. At 10.5 weeks of age, full-thickness excision wounds (1.0 cm²) were inflicted on dorsal skin or 3 mm² punch biopsy (Medex Supply) wounds on the tail skin. Wounds were monitored daily and wound size was measured using a transparent sheet. At harvesting timepoints, mice were sedated with isoflurane during injections and wounding and were administered Buprenorphine (0.1 mg/kg) prior to wounding for three days PWI. At the harvesting point, mice were euthanized with CO₂ and the wounded skins were harvested and either embedded in OCT, paraffin or prepared for wholemounts.

Depilation

Three days prior to depilation (day −3), 8-week-old mice were subcutaneously injected with 200uL of BAI-1 (1 mg/kg)⁹¹ (MedChemExpress) along dorsal skin. Injections were continued daily for a total of five consecutive days (until day 2 post-depilation). On the day of depilation (day 0), dorsal hair of mice was trimmed with electrical clippers and wax strips were used to remove remaining hair from the dorsal skin. Mice were monitored daily for skin coloration changes. At day 9 post depilation, skin was harvested and embedded in OCT for further sectioning and histological staining.

Hair grafting

Grafting was performed based on protocols described^68,69,70 with modifications. Briefly, silicon chambers were implanted onto the dorsal skin of female 8-week-old (P56) nude mice prior to grafting. For preparation of fresh neonatal dermal fibroblasts (DFs), WT (C57BL/6 black) pups were collected at P0–P2 and whole skin was removed and processed with dispase solution overnight at 4 °C. Dermis and epidermis were then separated, and the dermal layer was chopped and treated with collagenase I solution for 1 h at 37 °C for downstream processing as previously described⁶⁹. In place of whole epidermal cell fractions, WT-mRFP1 and BaxKD-GFP HFSCs (up to passage 20) were expanded under standard HFSC culture conditions. Immediately prior to grafting, HFSCs were harvested through trypsinization for 10 min at 37 °C, followed by centrifugation for 5 min at 500xg. HFSCs and dermal cells were then counted and combined at a 1:3 ratio (~3 million HFSCs and ~9 million dermal cells per graft) in 1X sterile PBS at a volume of 150 μL per graft. Cell mixtures were then gently transplanted in a drop-wise manner into silicon chambers. Harvested cells were kept on ice for the entire duration of time. Silicon chambers were removed after 8–10 days, and grafts were collected 28 days after post-transplantation for downstream processing.

In vivo neutralization of TNFα

For grafting experiments, experimental groups were treated with either non-specific Rat LEAF purified IgGI (10 μg/g, Biolegend) or LEAF purified anti-mouse TNFα (10 μg/g). Antibodies were administered daily through direct injection onto grafting sites in 200 μL of sterile 1X PBS from days 0-4 post-grafting.

Cloning of lentiviral vectors for RNA silencing and overexpression

To generate lentiviral vectors, we used an H2B-GFP empty PLKO.1 lentiviral vector suitable for silencing (Addgene #25999). We next digested the PLKO.1 lentiviral vector with RsrII and EcoRI enzymes. Synthesized shRNA oligonucleotide target sequences (IDT) were annealed and ligated into the digested PLKO.1 vector. Knockdown targeting sequences were designed by Sigma as outlined in Table S1 (Supplementary Information). The pHIV-EGFP plasmid, a gift from Bryan Welm & Zena Werb (Addgene plasmid# 21373)⁹², was used as a backbone for gene overexpression. The backbone lentiviral vector was digested using SmaI HF (NEB) enzyme. The Bcl2 gene was amplified from mouse HFSC cDNA using primers with adaptors for Gibson Assembly (Table S1), and amplicons were run on a 2% agarose gel and extracted using the QIAquick Gel Extraction kit (Qiagen). Amplicons were then ligated into the backbone using the Gibson Assembly Master Mix (NEB).

Ligated vectors were transformed into chemically competent Stable3 cells and cultures were grown on agar plates with Carbenicillin (50 μg/mL) overnight at 37 °C. Individual clones were picked and grown overnight in standard Luria Broth (LB) at 37 °C, and plasmids were extracted using the Qiagen miniprep kit. Samples were sent for standard sequencing through Macrogen Europe. Positive clones were expanded overnight at 37 °C and plasmids were harvested using the PureYield Plasmid Maxi Prep kit (Promega).

Viral production

For viral production, HEK293FT cells (passage 10 or under) were grown in high glucose DMEM media supplemented with 10% fetal bovine serum (FBS), 1% Pen/Strep, 1% L-glutamine, 1% Sodium Pyruvate, 1% Sodium Bicarbonate, and 1% non-essential amino acids (NEAA). Prior to transfection, cells were grown to 85% confluence and passaged 1 day prior onto poly-L-lysine (PLL)-coated 15 cm dishes.

Cells were transfected with the lentiviral target vector together with 2^nd generation viral packaging vectors using 1 mg/mL polyethylenimine (PEI) transfection reagent at 1:1.5 DNA:PEI ratio and incubated overnight in starvation media. The following morning, media was changed to standard growth media (as described above) and incubated overnight. Day 1 of viral-containing media was collected the following morning and media was replaced. This was repeated for Day 2 of collection. Viral media was concentrated using Amicon Ultra Centrifugal filtration units (Millipore), aliquoted, and stored at −80 °C until further use.

Generation of stable cell lines

HFSCs at 60% confluence were transduced with virus using polybrene reagent at 1:1000 (10 mg/mL stock) and incubated overnight in standard media. 72 h post-transduction, transduced cells were isolated based on fluorescence expression using FACS.

RNA extraction, reverse transcription, and real-time quantitative PCR (RT qPCR)

RNA was isolated using TRIzol (Sigma) and cDNA was synthesized using up to 1ug of RNA (qScript cDNA Synthesis Kit). Real-time quantitative PCR (RT qPCR) was carried out on the CFX Connect Real Time PCR Detection System (BioRad) with the PerfeCTa SYBR Green FastMix (Quanta), with gene-specific primers outlined in Table S1. Amplicon levels were analyzed in triplicates and quantified relative to a standard curve comprising cDNA. Values were normalized to housekeeping gene levels (RPLP0 or GAPDH). All experiments were performed in biological and technical triplicates. Reactions were: 3 min at 95 °C, 40 cycles of 10 s at 95 °C and 30 s at 60 °C with addition of a melt curve step: 10 s at 95 °C, and increments of 0.5 °C every 5 s between 65 °C and 95 °C. Fold change in gene expression was calculated after normalization of target to housekeeping genes, and comparison between conditions using the delta delta Ct method.

Histology

Isolated dorsal wounds at 18 or 30 days PWI were fixed for 2 h in 4% paraformaldehyde (PFA). Wounds were embedded in paraffin in the following conditions: 80% EtOH (45 min), 95% EtOH (45 min x2), 100% EtOH (45 min x2), 1:1 EtOH:xylene (45 min), xylene (45 min), xylene (30 min), paraffin (45 min, 54 °C, X3) and samples were then embedded in paraffin. Deparaffinization was achieved in the following conditions: xylene (5 min, x2), 100% EtOH (2 min, x3), 95% EtOH (2 min), H₂O wash. For histological analysis of depilation and grafting experiments, skin samples were harvested and fixed as described above, followed by embedding into OCT for preparation of cryosections. 

For Hematoxylin and Eosin staining, deparaffinized or cryopreserved 20μm sections were treated with: Hematoxylin (1 min; Sigma HHs32), H₂O rinse, differentiator (0.3% alcoholic HCl, 2 dips), H₂O rinse, 95% EtOH (1 min), Eosin (Sigma, HT110116, 2 min), 95% EtOH (10 s), 100 EtOH (1 min, x2), xylene (1 min, x3).

Immunofluorescence

HFSCs were grown on PLL-coated glass coverslips for downstream analysis. Samples were fixed in 4% PFA at room temperature (RT) for 30 min. For wounded skin sections, harvested skin was embedded in either OCT or paraffin. For OCT sections, 15 μm sections were mounted onto microscopy slides and fixed in 4% PFA at RT for 30 min. For paraffin sections, 7 μm sections were mounted onto microscopy slides and deparaffinization was achieved in the following conditions: xylene (5 min, x2), 100% EtOH (2 min, x3), 95% EtOH (2 min), H₂O wash.

For preparation of wholemount samples, isolated tail and dorsal tissue were treated with 20 mM EDTA for 4 h (tail) and 6 h (dorsal) at 37 °C for efficient separation of the epidermis from the dermis. Following separation, samples were fixed in 4% PFA for 2 h at RT.

After fixation or deparaffinization, all samples were washed in 1X PBS (5 min, x2) and blocked for 2 h in blocking buffer consisting of 10% goat serum, 2% BSA, 0.2% Triton-X. Primary antibodies were diluted in blocking buffer and cells or tissues were incubated overnight at 4 °C. Samples were washed at least three times with 1X PBS. Secondary antibodies were incubated for 1 h at RT followed by 4 washes with 1X PBS. The following primary antibodies were used: Bax (Mouse, 1:100, Thermo Scientific, cat. #MA5-14003, lot #QK2110159; 1:100, Cell Signaling, cat. #5023), cleaved Caspase-3 (Rabbit, 1:100, Cell Signaling, cat. #9661 S, lot #47), cleaved Caspase-8 (Rabbit, 1:100, Cell Signaling, cat. #8592), Ki67 (rat, 1:100, eBioscience: cat. #14-5698-82, lot # 2196796), TNFα (Mouse, 1:100, abcam, cat. #1793, lot #GR2370127-1), cIAP1 (Rabbit, 1:100, Santa Cruz, cat. #7943), NFkB p65 (Mouse, 1:100, Santa Cruz #8008), Keratin-15 (Mouse, 1:100, Abcam, cat. #ab80522), and CD34 (Rat, 1:100, Pharmingen, cat. #553731).

Antibody staining was visualized using secondary antibodies conjugated to Alexa Fluor dyes: 488, 546, and 633 antibodies (1:250, Life Technologies, cat. #s: A11001, A11003, A11006, A11008, A11010, A11039, A11040, A11081, A21103, A21050, A21070, A21094). Additional stains were performed using Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL; ApopTag TdT kit, Millipore) and mitochondrial labeling MitoTracker Red CMXRos (ThermoFisher). Analyses were performed on a Zeiss LSM-880 confocal microscope.

Protein extraction and immunoblotting

For whole cell lysate extraction, cells were washed with ice-cold 1XPBS and collected by scraping on ice, followed by centrifugation (1800x g, 5 min at 4 °C). Cell pellets were lysed using RIPA lysis buffer containing protease inhibitor cocktail (1:200, Abcam) and phosphatase inhibitor cocktail (1:100, Sigma), followed by incubation on ice for 30 min and centrifugation (22,000 x g, 15 min at 4 °C). The protein-containing supernatants were collected for downstream processing. For nuclear fractionation, cells were washed with ice-hold 1XPBS and collected by scraping on ice and centrifuged at 500 x g for 3 min. Downstream steps were performed using the NE-PER Nuclear and Cytoplasmic Extraction kit (ThermoFisher) according to manufacturer protocols.

Protein concentration was measured (Bradford reagent, BioRad). Protein samples were denatured through addition of 5X LSB containing β-mercaptoethanol and 5-minute incubation at 95 °C. Protein samples were resolved by running on 12.5 % SDS-PAGE and electrotransferred to a nitrocellulose membrane (Whatman). Transfer of proteins was visualized using Ponceau Red dye (Sigma). Membranes were blocked in 5% dry skimmed milk or 3% BSA in TBS-T for 1 h at RT and incubated with primary antibodies (1:1000) overnight. Membranes were then incubated with secondary antibody (ms-HRP or rb-HRP, 1:10,000) for 1 h at RT, washed 3X with TBS-T for a minimum of 10 min each, and chemiluminescence was visualized using the iBright Imaging Systems (Invitrogen). Densitometric values were obtained through normalization of proteins of interest using either the iBright Imaging Systems or ImageJ V.1.53. The following antibodies were used: Bax (1:1000, Cell Signaling, cat. #5023), cleaved Parp1 (1:1000, Cell Signaling, cat. #9544), cleaved Caspase-3 (1:1000, Cell Signaling, cat. #9661 S, lot #47), GAPDH (1:10000, Sigma, cat. #G9545, lot #128M4817V), α-tubulin (; 1:10000, Santa Cruz, cat. #23948, lot #C0415), b-actin (1:10000, Santa Cruz, cat. #81178, lot #J1116), b-tubulin (1:10000, Cell Signaling, cat. #2128), Histone-3 (1:10000, Abcam, cat. #18521; 1:10000, Cell Signaling, cat. #4499), cIAP1 (Santa Cruz, Cat. #7943), cIAP2 (Santa Cruz, cat. #sc-7944), IkBa (1:1000, Santa Cruz, #1643), NFkB p65 (1:1000, Santa Cruz cat. #8008, lot #H1819; 1:1000, Cell Signaling #8242), and TNFR2 (1:1000, Abcam, cat. #15563).

Flow cytometry

Cell competition assays

Co-cultured HFSCs were isolated using either fluorescent activated cell sorting (FACS) performed on a BD FACSAria IIIu for downstream RNA processing, or analyzed on a BD LSR I or Attune NxT flow cytometer. Cells were collected at distinct timepoints to quantify changes in cell population ratios between 0 and 10 DPS.

Quantification of CD34⁺ HFSC populations in vivo

Epidermal cell populations were isolated from telogenic dorsal skin of mice, and HFSC populations were subjected to immunostaining as previously described^28,29. Briefly, single cell suspensions of epidermal cells from control or RU486-induced BakcBax^−/− mice were were obtained after digestion with trypsin and filtration, followed by immunostaining against the following antibodies: Integrin-a6 PE (1:200, eBioscience, Cat. #12-0495-82), ScaI FITC (1:200, eBioscience, Cat. #553335), and CD34 eFluor 660 (1:100, eBioscience, Cat. #50-0341-82), alongside DAPI (1:10,000). Cells were incubated with antibodies diluted in staining buffer (3% chelated FBS) on ice for 30 min, followed by washing and centrifugation steps at 500x g for 5 min at 4 °C before FACS analysis (pertaining to Fig. 6k). n = 3 mice were pooled for each condition. The CD34⁺ population is represented as a percentage from the parental Integrin-α6⁺/Sca1^– epidermal pool.

Downstream analysis of FACS data was performed using FCS Express v.7.0, FlowJo v.10.9, BDFACS Diva v.9.0. For gating strategy, please see Supplementary Fig. 11.

Cell culture

Cell maintenance

Feeder free isolated a6⁺ScaI^–CD34⁺ HFSCs were isolated and grown as previously described²⁸. HFSC medium was prepared using Dulbecco’s modified Eagle medium (DMEM)/F12 3:1 (Biological Industries) containing L-glutamine (Biological Industries; 1:100), penicillin/streptomycin (Biological Industries; 1:100), 10% chelated fetal bovine serum, 5 μg/ml insulin (Sigma I-5500), 5 μg/ml transferrin (Sigma T-2252), 2 × 10⁻¹²M T3 (3,30-triiodo-L-thyronine; Sigma T-2752), 400 ng/ml hydrocortisone (Sigma H0888), cholera toxin (10⁻¹⁰M), and CaCl₂ (50 μM).

Media was replaced every 2 days and cells were grown to 80–85% confluence prior to passaging. Cells were passaged by incubating with Trypsin for 10 min followed by centrifugation at 500 x g for 5 min.

Cellular treatments were performed as described below. Cell lines were passage matched across all experiments. Media from cells underwent routine testing to ensure culturing under mycoplasma-free conditions.

Apoptotic induction assays

Cells were treated with ABT199 (Bcl-2 inhibitor) (50 μM) or TNFα (20 ng/mL) + cycloheximide (CHX) (protein synthesis inhibitor) (10 μg/mL) immediately prior to acquisition of initial brightfield (BF) images. BF images and samples were taken every 30 min for up to 4 h for ABT199 treatment and up to 6 h for TNFα + CHX treatment for downstream processing. The percentage of live cells were analyzed at each timepoint using Trypan Blue to determine cell survival. MitoTracker Red CMXRos (200 nM, ThermoFisher) was used to visualize mitochondria during ABT199 treatment. Untreated cells were visualized under the same conditions.

Colony formation assay (CFA)

Cells were seeded at low density (1,000 cells per well of a 6-well plate) and media was changed every two days. Cells were monitored daily and colony size and growth were measured at specified time points between 1 and 8 days post seeding (DPS). For histology, colonies were fixed with 4% PFA for 15 min at RT, followed by washing with 1X PBS. Toluidine Blue O dye (Acros Organics) (0.1% in ddH₂O) was then directly added to wells and incubated for 20 min at RT, followed by several washes with ddH₂O prior to image acquisition.

Cell competition co-culture assays

Cell populations specified in each experiment were counted, mixed at 1:1 cell ratio, and seeded in co-culture (eg., WT and BaxKD co-cultures alongside WT only and BaxKD only control cultures; WT and BaxKD-TNFαKD co-cultures alongside WT and BaxKD control co-cultures, etc.). Co-cultures were monitored daily and maintained under standard growth conditions as described above (unless otherwise noted, such as in the case of treatments) for the duration of the experiment (up to 10DPS). For live imaging experiments, co-cultures were grown in a standard incubator (37 °C, 5% CO₂) until substantial cell-cell contact was observed, after which culture plates were transferred into an incubation chamber for live imaging under the same conditions for the specified duration of time. For co-culture experiments requiring additional treatments or other downstream analysis (eg., FACS, RNA extraction, in situ image analysis, TNFα neutralization, TNFα stimulation, etc.), cells were processed as described for each method.

BAI-1 chemical inhibition

Cells were treated with BAI-1 (100 μM)⁹¹ (MedChemExpress) and samples were harvested at timepoints including 15, 30, 45, 60, and 90 min alongside vehicle (NT) controls for downstream protein, RNA, and IF assays. All experiments were performed under the same experimental conditions.

TNFα stimulation

Both WT and BaxKD SCs were incubated in starvation media for 12 h prior to stimulation, in order to synchronize cell cycles⁹³. TNFα (20 ng/mL) was added into media, and samples were visualized or harvested for downstream analysis at specified timepoints. All experiments were performed under the same experimental conditions.

Conditioned media assays

In the first assay (pertaining to Supplementary Fig. 2e), cells were seeded at low confluency (20%) and treated daily with 50% conditioned media (CM) from WT or BaxKD cells for 5 days. CM was harvested from high confluence (80%) WT or BaxKD cells. Cells were monitored daily for morphological changes characteristic of apoptosis. In the second assay (pertaining to Supplementary Fig. 2f, g), CM was harvested from high confluence (80%) WT cells, BaxKD cells, or competitive co-culture cells. Next, individual WT or BaxKD cells were treated for up to 24 h with each type of CM and harvested at various timepoints as previously described³⁴. Downstream immunofluorescent analysis of cCp8 expression in treated cells was used as an indicator of extrinsic death induction.

TNFα neutralization in vitro

Competitive WT and BaxKD co-cultures were seeded at equal ratios in media containing the TNFα inhibitor Etanercept (20 μg/mL)⁹⁴ (Sigma) or vehicle (NT). Cultures were harvested at 0DPS, 1DPS, and 3DPS for FACS quantification of population ratios.

Live imaging

Imaging of cell competition assays was performed at 5DPS, and proliferation at 3DPS using a ZEISS inverted microscope system containing an incubator for optimal cell growth conditions (37 °C, 5% CO₂). Cells were visualized using endogenous fluorescence together with BF imaging for a duration of 12–24 h.

In situ analysis of cell competition

In situ analysis of cell competition was performed by harvesting competitive co-cultures at various timepoints after seeding (in parallel to FACS analysis), followed by image analysis using a fluorescent microscope and quantification of the number of cells pertaining to each cell population via ImageJ v.1.53.

RNA sequencing data analysis

SC RNA-seq data from Haensel et al. (GEO #GSE142471)¹⁹ was analyzed in RStudio v.1.3.1093 using the Seurat package v.3.2.2. After identification of a HFSC bulge cell cluster based on canonical marker expression, CD34⁺Krt15⁺ expressing cells were subset and assessed for expression of various genes within this population.

Quantification and data analysis

Data are mean ± SEM. The n values represent biological repeats measured independently as specified in each figure legend. Each experiment included at least three biological replicates and three technical replicates where appropriate, and was repeated a minimum of two times. Significance was determined by performing parametric unpaired two-tailed Student’s t-test, where *p < 0.05, **p < 0.01 and ***p < 0.001. Analysis software includes: ImageJ v.1.53, ZEN v.3.0, and iBright Imaging Systems programs.

Image-based analyses

Identical parameters were utilized for analysis of photos within the same experiment using ZEN v.3.0 (Carl Zeiss Microscopy) and/or ImageJ v.1.53.

RNA sequencing analyses

Analyses derived from single cell RNA-sequencing dataset(s) specified were performed using Seurat v.3.2.2. in RStudio v.1.3.1093.

Statistics and reproducibility

All experiments were repeated at least twice with similar results obtained. Each independent experiment included at least three biological repeats unless otherwise indicated in the figure legends. No statistical methods were used to predetermine sample sizes. Data distribution was assumed to be normal, but this was not formally tested. Data collection and analysis were not performed blind to the conditions of the experiments. No animals or data points were excluded from the analyses. Data are presented as mean ± s.e.m. Statistical significance was determined by unpaired two-tailed Student’s t-test, where *P < 0.05, **P < 0.01 and ***P < 0.001. The following software were used: Excel v.16.19 and GraphPad Prism v.9.5.1.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

• SEO Powered Content & PR Distribution. Get Amplified Today.
• PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
• PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
• PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
• PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
• Source: https://www.nature.com/articles/s41467-023-41684-x