Abstract
Automated Analysis in Life Sciences
Automated 3D-Printed Unibody Immunoarray for Chemiluminescence Detection of Cancer Biomarker Proteins
Advances in 3D printing allow simple, functional microfluidic devices to be fabricated (relatively) inexpensively and quickly. Stereolithography (SLA) is one such method to print transparent acrylate resins by using focused light to photocure polymer precursor. The authors use SLA to fabricate a clear plastic flow sensor to measure DNA by electrochemiluminescence (ECL) and for the multiplexed detection of cancer biomarkers. While advances have been made in high-throughput protein detection, further automation is necessary to minimize operator error and simplify assays in a cost-effective manner.
By reacting luminol with hydrogen peroxide and peroxidase enzyme, chemiluminescence (CL) signals are simply and sensitively detected by cameras without a light or power source. Herein, the authors use an SLA 3D-printed sandwich immunoassay to detect prostate-specific antigen (PSA) and platelet factor (PF-4) by simultaneous CL detection. The immunoreagents are first forced into a 3D network mixer (6.5 min) prior to downstream capture. Flow resumes after a 15 min incubation to remove excess reagents. Thereafter, CL reagents are added and the protein array image is captured (60 s). This entire protocol lasts no more than 30 min, and detection limits of 0.5 pg/mL are achieved for both PSA and PF-4 biomarkers. This assay is then validated with traditional enzyme-linked immunosorbent assays performed on human patient serum samples (three prostate cancer patients and one healthy patient). It achieves correlation values of 1.1 and 0.95 for PSA and PF-4, respectively.
These SLA 3D-printed devices result in a cost of ~$2 for two biomarker proteins for a 30 min assay time. They provide fast, quality results at low cost, making this technique suitable as a point-of-care cancer diagnostic for global health applications. (Tang, C. K.; et al. Lab Chip
Automated Protein Biomarker Analysis: On-Line Extraction of Clinical Samples by Molecularly Imprinted Polymers
Automated analysis of low sample volumes with minimal sample handling is of particular interest to clinical proteomics. Integrating automated sample preparation with mass spectrometry (MS) bioanalysis is seen as an alternative to conventional immunoassays. Different workflows have been formulated, such as replacing antibodies with synthetic receptors and molecularly imprinted polymers (MIPs) using “plastic antibodies,” which are a rapid and economical alternative to immunocapture assays.
Clinically, the detection of the biomarker progastrin-releasing peptide (ProGRP) in serum samples is of high diagnostic and prognostic value for small-cell lung cancer (SCLC). Within this context, immunoextraction is particularly critical to discriminate between healthy and patient donors via ProGRP, which is normally present at low reference levels (7.6 pM). The authors thus utilize ProGRP peptides as templates to produce thin MIP films on silica bead surfaces. This is highly desirable in separation science since these molecular-imprinted silica beads are robust, uniform, reproducible, and easily packed into extraction cartridges and chromatography columns. Crucially, the authors explore a novel MIP-based workflow to extract ProGRP from patient serum and quantify the peptide concentration using liquid chromatography–mass spectrometry (LC-MS).
While optimizing protocol workflows, pH conditions, and biomaterials, the target peptide, NLLGLIEAK, is ascertained to be retained on the MIP particles. The entire automated ProGRP extraction process is a cost-effective workflow that lasts 50 min (both extraction and chromatography processes), requiring merely a 50 µL sample. This method is linear over three orders of ProGRP concentration magnitude, with a correlation of R2 > 0.97. The estimated limit of detection (LOD) is ~17.2 pM, which is substantially lower than that of similar methods (625 pM) and was validated on patient serum samples.
Coupling a MIP trap with an analytical column and tandem MS detection provides the means for the automated determination of ProGRP in patient samples. Its low sample volume, short analysis time, and low detection and quantification limits are achieved without using antibodies and bring automated protein detection with MS bioanalysis closer to conventional immunoassays. (Rossetti, C.; et al. Sci. Rep.
Fe-Nitrilotriacetic Acid Coordination Polymer Nanowires: An Effective Sensing Platform for Fluorescence-Enhanced Nucleic Acid Detection
The need for highly sensitive and selective DNA sensors is becoming critically important for a number of industries (e.g., food, environment, and medicine). Carbon nanostructures (i.e., carbon nanotubes and graphene) have received significant interest due to their effective DNA sensing. However, they are limited by their composition and structural properties, leading to overall inadequate performance. Thus, there is a need to develop nanoquenching technologies from novel precursor materials via bottom-up synthesis.
In previous studies, the authors explored various structures, such as nano-C60, nanobelts, nanofibers, nanorods, and nanoparticles, to identify and explore novel nanostructures suitable for nucleic acid detection. Nitrilotriacetic acid (NTA), a chelating agent, is used to generate coordination polymer nanowires. To detect nucleic acids, Fe-NTA coordination polymer nanowires are used as a rapid (15 min), fluorescence-based detection technology. As a proof of concept, an oligonucleotide sequence associated with human immunodeficiency virus (HIV) is chosen.
In the presence of Fe-NTA, a FAM-labeled oligonucleotide sequence (PHIV) leads to an 89% decrease of fluorescence intensity, which shows its effective fluorescence quenching. The authors attribute this to the high-quenching ability of Fe(III). Increasing quantities of target oligonucleotides (0–300 nM) lead to increasing fluorescence intensity. Fluorescence quenching occurs in 5 min, whereas fluorescence is recovered within 2 min. This suggests that a complete cycle of signal recovery and quenching can be completed within 15 min—shorter than the time required for other nanostructures (e.g., nanobelts, nanofibers, nanoparticles, nanospheres, and nanosheets), which ranges between 20 and 90 min.
The newly developed probe distinguishes perfectly matching targets from single-base-pair mismatched and from unmatching oligonucleotide targets (525 nm fluorescence emission). Further efforts facilitate the experimentation of Fe-NTA content to optimize quenching and recovery kinetics. In conclusion, Fe-NTA polymer nanowires are shown to be an effective fluorescence quenching nanostructure. Its recovery and requenching can be completed within 15 min, with a low detection limit (0.2 nM), even capable of identifying single-base-pair mismatch. Its low cost and scalability make it suitable for the sensitive and selective detection of nucleic acids. (Zhou, Y.; et al. Nanotechnology
An Automated and Portable Microfluidic Chemiluminescence Immunoassay for Quantitative Detection of Biomarkers
The development of lab-on-a-chip (LoC) devices has seen a variety of biomarkers, including proteins, nucleic acids, and cells in blood or urine, rapidly, reliably, and sensitively detected within integrated microdevices with sample-in-answer-out capability. A bottleneck in their usage as clinical diagnostics concerns microchannel fluid transport. Sophisticated on-chip valves (e.g., pneumatic, solenoid, and screw valves) facilitate fluidic actuation but require bulky external equipment that impedes their integration, automation, and portability. A second challenge lies in having an automated, reusable, and portable readout system to sensitively detect biochemical signals and provide quantitative readings. Smartphone diagnostics are one such area, although they are unsuitable for complicated biochemical assays that require multiple washing and reaction steps. Disc platform devices are another option, although their costs can be prohibitive. The authors propose an integrated and self-contained microfluidic-chemiluminescence immunosensor that is advantageous for its quantitation, sensitivity, versatility, cost-effectiveness, portability, and ease of use for diagnostic procedures.
Overall, the diagnostic device consists of injection-molded microfluidic chips assembled with a microchannel layer, patterned antibody/antigen layer, substrate layer, and on-chip valve layer. The microfluidic chip bearing the sample is inserted into a portable, automated instrument that allows chemiluminescence detection. To evaluate the assembled platform (containing five microfluidic chips), a negative pressure (–3.5 kPa) is applied onto the outlet port. The flow rates within the chips are consistent (~4.4 µL/s) with good stability and reliability. To evaluate its multiplex detection ability, C-reactive protein (CRP) and testosterone biomarkers are simultaneously evaluated by direct sandwich and competitive immunoassays, respectively. Both biomarkers are detected following a series of ~6 steps consisting of adding reagents, sample, and substrates interspersed with washing. The entire automated process lasts <70 min, and the chemiluminescence is detected via instrumentation.
Both CRP and testosterone exhibit linear dependencies with signal intensity.
The intra-assay standard deviation is <15%, while the inter-assay deviation is between 9.4% and 13.1%. The limit of detection (LoD) is 4.27 and 0.45 ng/mL for CRP and testosterone, respectively. The automated platform is sufficiently versatile to simultaneously incorporate interleukin-6 detection and achieve multiplexed detection with good reproducibility and high sensitivity. This platform is further tested on patient serum samples (16 subjects) for CRP (bacterial infection) and testosterone (healthy donors). This technique is then validated against the conventional enzyme-linked immunosorbent assay (R2 > 0.9).
Thus, the authors are able to demonstrate an automated and portable chemiluminescence immunoassay platform to quantitate CRP and testosterone in serum samples. They foresee improvements to facilitate long-term storage of prepatterned chips (containing antibodies and antigens) with preloaded reagents to facilitate point-of-care diagnostics. (Hu, B.; et al. Lab Chip
3D Bioprinting
A Generalizable Strategy for the 3D Bioprinting of Hydrogels from Nonviscous Photo-Crosslinkable Inks
3D bioprinting has attracted great interest in the deposition of materials (e.g., bioinks) and cells to build tissue structures at high resolution with the required structural complexity. Despite advances in the field, viscous bioinks induce shear forces on cells and compromise cell viability, making it challenging to balance between printability and biological functionality. One approach combines photo-crosslinkable hydrogels with polymers that increase precursor viscosity, which involves pre- and post-crosslinking. Pre-crosslinking methods cause heterogeneous printed structures and low cell viability, whereas post-crosslinking using light improves cell viability, but the structure is found to be unstable.
To circumvent such issues, a bioprinting method involving photo-crosslinking just prior to deposition (in situ crosslinking) is developed to reduce extrusion forces. This approach is advantageous because it does not require additional materials for copolymerization, retains high cell viability, and facilitates the printing of structures. Tuning and controlling photo-crosslinking parameters (i.e., light intensity, exposure time, and average ink velocity) is critical to maintaining structural integrity and preserving cell viability. As proof of concept, fibroblasts are encapsulated in a 5 wt% methacrylated hyaluronic acid (MeHA) and are found to maintain high cell viability (~95%) after printing into filaments of different sizes (200–700 µm). The authors attribute the high cytocompatibility to simultaneous extrusion and crosslinking that protects cell exposure to high shear stresses during the printing process.
This 3D printing technique is further explored using a range of different materials, such as gelatin methacryloyl (GelMA), poly(ethylene glycol) diacrylate (PEGDA), and various norbornene-functionalized hyaluronic acid (NorHA) formulations, including substituting photoinitiators. To further illustrate the versatility of this technique, NorHA hydrogels are modified with nondegradable dithiothreitol (DTT) or matrix metalloproteinase–degradable (MMP-deg) crosslinkers either with or without arginine–glycine–aspartic acid (RGD) functionalization. Importantly, the addition of RGD sequences leads to cell spreading throughout each printed filament, indicating successful functional incorporation.
This method is also amenable to printing more sophisticated structures, such as core–shell structures, using different bioinks with or without cells. By controlling the on/off status of core and shell channels, heterogeneous filaments can be printed with a programmable distribution of various inks or cell types. Finally, using various bioink formulations, the authors demonstrate the printing of hollow filaments that can be perfused.
Crucially, this study demonstrates a generalizable technology for 3D bioprinting that is not hindered by the inherent properties of the bioink (i.e., ink viscosity). By designing a printing process with a photopermeable capillary (in situ crosslinking), the authors overcome the limitations of cell viability and structural integrity commonly found in 3D bioprinting. This achievement facilitates the printing of complex 3D material–cell structures for biomedical applications. (Ouyang, L.; et al. Adv. Mater.
Fabrication of Fillable Microparticles and Other Complex 3D Microstructures
3D microstructures for fabrication are typically chosen based on the size, shape, and composition of microdevices since each technique has limitations in spatial resolution, device geometry, materials compatibility, and throughput scale. Often, there is a trade-off between printing resolution and material selection. Some techniques (e.g., particle replication in nonwetting template [PRINT]) have attractive nanoscale resolution and throughput but cannot readily fabricate nontrivial structures with internal architecture or top-narrowing 3D shapes. To readily achieve printable microstructures, a bottom-up, high-resolution fabrication technique is used to create complex microdevices using commercially and clinically relevant biomaterials, like lactide-glycolide copolymers.
The approach, termed stamped assembly of polymer layers (SEAL), combines computer chip manufacturing with soft lithography and sintering to produce structures of <400 µm. To fabricate high-fidelity microdevices, the technique aligns layers with high precision during sintering with a photomask aligner. It is retrofitted with a heating element, temperature controller, relay, and voltage source to enable simultaneous alignment and thermal bonding. This process is monitored by observing the disappearance of light diffraction patterns upon cooling below the glass transition temperature. This facilitates the creation of large arrays of complex microstructures, including a 3D star; two-layered letters; angular words, such as MiT; and other shapes with high aspect ratios.
The motivation for developing this process is based on a desire to create poly(lactic-co-glycolic acid) (PLGA) microdevices to deliver timed pulses of antigens for any vaccine from a single injection. Fillable bases are molded using thermoplastic polymers before filling with a model drug solution using a BioJet Ultra picoliter dispenser before sealing. These core–shell particles are useful for biomedical applications because they have small, injectable sizes and can be created with biodegradable materials that induce minimal foreign body reaction. The fabrication process also allows greater control over wall thickness to yield larger cores for controlled drug release upon stimulation.
This process facilitates the generation of microdevices with pulsatile release for single-injection immunizations, which the authors confirmed while studying fluorescent dextran (model drug) release from PLGA microparticles. This proof of concept potentially allows the release of antigen in discrete pulses at time points designed to match vaccination schedules. When tested in vitro and within live mice, fluorescence signal is detected at distinct time points, as indicated by a ~50-fold increase in signal upon release. This is significant because carriers with different release kinetics could be introduced within a single administration and engineered to release their cargo as required. Finally, as a proof of concept, the SEAL technique is used to deliver inactivated polio vaccine and ovalbumin.
Prior to their delivery, the released proteins are found to retain their prefabrication properties, a marked improvement compared with emulsion-based fabrication, which typically compromised protein payload quality. A payload containing two different SEAL ovalbumin (OVA) formulations is injected in a single dose that accumulates a higher peak titer than two bolus injections comparable to a double OVA dosage. Additionally, low-temperature storage (1 month) and desiccation do not compromise the OVA payload.
The flexibility of the SEAL approach also allows the generation of pH-responsive biomaterials that bypass the stomach (low pH) and deliver their cargo to the intestines (neutral pH).
In summary, the SEAL technique overcomes many limiting factors for 3D printing. It allows complex structure fabrication without the limitation of biopolymer selection and is highly versatile with high throughput. These techniques allow the fabrication of particles with discrete and engineered drug release profiles. (McHugh, K. J.; et al. Science
Optical µ-Printing of Cellular-Scale Microscaffold Arrays for 3D Cell Culture
Even though micro- or nanoscale surface topography has been shown to have a significant influence on cell behavior and fate, adherent human mesenchymal stem cells (hMSCs) have not been observed in defined 3D microenvironments. The ability to directly study hMSCs can lead to a greater understanding of their multipotency, allowing researchers to better understand how physical cues, such as pattern size, roughness, and topography, influence stem cell and fate determination (e.g., osteoblasts, chondrocytes, and adipocytes).
Despite its initial promise, 3D printing technology has a variety of limitations in generating defined topographical features, including feature size, low fabrication yield, and biocompatibility.
The authors utilize a rapid bioprinting technique (micrometer resolution) using optical maskless exposure. Instead of using hydrogels for fabrication, they chose SU-8 photoresist for its strong mechanical properties, chemical resistance, and nonirritant features. Thereafter, gelatin methacrylate (GelMA) is selectively deposited to guide cell adhesion and spreading. The combination of both materials allows the tuning of biochemical and structural properties for controlled cell and migration studies.
This technique achieves 3D microprinting through a high-speed spatial light modulator (digital micromirror device [DMD]). Building on a previous scheme for optical maskless exposure, a digital camera and a motorized stage are integrated for machine-vision metrology and high-precision alignment for in situ printing. By segmenting the scaffold into 100 layers of slices, the optical patterns sequentially irradiate the photoresist to generate sophisticated 3D microstructures with 600 nm resolution. To demonstrate the direct visualization of cell interaction with microtopography, the scaffolds are coated with polydopamine before seeding hMSCs. Within 3 h of seeding, hMSCs attach to the surfaces and achieve a spread morphology.
The fine resolution achieved from the DMD printing process facilitates the study of hMSC behavior in single-cell microcubicles. Whereas larger surface areas lead to lower aspect ratios (elongated shape) and cell spreading, smaller areas lead to more squarish and smaller shapes.
With this well-defined system to control cell shape/spreading, the yes-associated protein (YAP) is found to be strongly expressed in the larger microcubicles, where cells have the most spread morphology. It also significantly directs hMSCs toward an early osteogenic lineage (e.g., alkaline phosphatase and 7-day ascorbate and dexamethasone treatment). With the increased precision in 3D fabrication, the authors deposit gelatin on selected areas of the microcubicle. By depositing GelMA on different parts of the microscaffold, the seeded cells preferentially attach to the GelMA deposits.
This DMD microprinting method allows the printing of fine, high-resolution features. It allows the precise tailoring of adhesion features to control cell shape and size with high precision. This presents a novel investigational route for cell behavior, such as stem cell lineage determination. (Ouyang, X.; et al. Sci. Rep.
Nanobiomedicine
Synthetically Lethal Nanoparticles for the Treatment of Endometrial Cancer
Endometrial cancer (EC), the most prevalent gynecological malignancy in the United States, is on the rise due to an increasingly obese population. In fact, its survival outcomes have even diminished since the 1970s.
Among the phenotypes, type II EC has a poor prognosis and disproportionately contributes to mortality rates. The standard of care consists of chemotherapy and/or radiotherapy, while numerous monotherapy and combination therapeutics have been explored. Previously, the authors established that combining paclitaxel (PTX) with tyrosine kinase inhibitors (TKIs) induces synergistic cell death through loss-of-function (LOF) mutations in TP53, the tumor suppressor gene.
Their effectiveness is derived from stopping cell cycle progression, causing cells to arrest in mitosis and die. This phenomenon is known as synthetic lethality, because it involves blocking an alternative gene alongside a single-gene mutation that results in cell death. Each mutation by itself is insufficient to cause fatality.
In this case, blocking TK activity, which is critical in vascular endothelial growth factor receptors, platelet-derived growth factor receptors, and fibroblast growth factor receptors, prevents the compensatory survival pathways from keeping P53-mutated cells viable. This combination is inherently highly specific to P53-mutated EC cells, minimizing detrimental effects on normal cells. A triple angiokinase molecular inhibitor, BIBF 1120 (BIBF; also known as nintedanib), is employed for TKI.
To facilitate delivery, a polymeric nanoparticle is employed to enhance dissolution (i.e., improve solubility of PTX and BIBF), improve pharmacokinetics, and minimize side effects (i.e., eliminate off-target effects) by passively targeting tumors through the enhanced permeability and retention (EPR). The PTX and BIBF drug combination is first assayed on various EC cell cultures. Only the P53 mutant cell line, Hec50co (loss of function), has significantly higher loss of cell viability than Ishikawa (wild-type) and KLE (gain-of-function) cells. Simultaneous addition of both drugs also generates significant killing synergy compared with sequential drug treatment at concentrations of >100 nM.
Biocompatible poly(lactic-co-glycolic acid) (PLGA; 75:25) particles (<175 nm) are generated by nanoprecipitation, found to be uniform and readily taken up by Hec50co cells. Compared with PTX formulations, particularly
Through a series of molecular experiments, it was determined that the combination of PTX and BIBF delivered through nanoparticles sends the majority of the LOF P53 cells into a G2-M DNA damage checkpoint status. Mutation-free parental Hec50co cells, instead, are not susceptible to the combined therapy.
This therapy is then extended to an in vivo xenograft comprising Hec50co cells. Following 32 days of treatment, the PTX and BIBF combination is delivered using PLGA (75:25) particles intravenously. Not only are the tumors significantly smaller, but also the median survival increases to 51 days compared with 41 and 39 days for PTX and saline treatments, respectively.
Promisingly, a stable body weight is observed throughout, alongside the absence of necrosis in the heart, lung, liver, spleen, or kidneys, suggesting that treatment is sufficiently tolerable. Notably, this system leads to the increased accumulation of intratumoral drug.
In summary, the authors demonstrate that combinatorial therapeutics can improve the outcomes of lethal cancers compared with single-inhibitor agents. This study shows that synthetic lethality is a principle with great promise to treat LOF P53 tumors. It potentially prolongs survival with minimal toxicity and may extend its usage beyond EC to TP53 mutant cancers, such as non-small-cell lung cancer and ovarian cancer. (Ebeid, K.; et al. Nat. Nanotechnol.
Amphiphilic Nanocarrier-Induced Modulation of PLK1 and miR-34a Leads to Improved Therapeutic Response in Pancreatic Cancer
RNA interference is an increasingly effective means for treating cancer. Various cancers, such as pancreatic ductal adenocarcinoma (PDAC), are well suited due to the extent of their miRNA dysregulation. Unlike siRNA, which targets specific genes, miRNA potentially regulates hundreds of mRNA targets at once. Using The Cancer Genome Atlas (TCGA), miR-34a and polo-like kinases (PLK1) are chosen following their validation from formalin sections obtained from short-term and long-term survivor PDAC patients. It is hypothesized that combining their targeting will lead to a synergistic anticancer effect, including MYC inhibition.
Since high miR-34a and low PLK1 expression levels correlate with favorable outcomes, the authors supply a miR-34a mimic, together with silencing PLK1, as a strategy to improve therapeutic response and prolong survival. Furthermore, a poly-(α)glutamic acid (PGA) nanocarrier that is water soluble, nonimmunogenic, and biodegradable by cathepsin B (a common enzyme in many cancers) is chosen for oligonucleotide delivery. The cathepsin B selectivity design principle is further validated by evidence that most pancreatic cell lines and corresponding xenograft models express it strongly, while healthy tissue does not.
To efficiently deliver oligonucleotides, an amphiphilic polyglutamate amine (APA) polymeric nanocarrier composed of repeating units of PGA is employed. Cell internalization studies then find that the APA oligonucleotide vehicles are taken up in 4 h, before peaking at 48 h. The proportion of APA-siRNA transfected cells is 42.5%, which is higher than conventional lipofectamine transfection (32.7%). The oligonucleotide-loaded APA carriers further transfect another five cell lines. When the oligonucleotide-APA carriers are introduced into cells (MiaPaCa2), miR-34a levels are significantly higher, while PLK1 expression is reduced (assessed by gene/protein expression). Cell viability, cell migration, and colony-forming potential are significantly inhibited by the double-oligonucleotide nanocarrier strategy. To test biocompatibility, the oligonucleotide-filled APA nanocarriers are tested on freshly isolated peripheral bone marrow cells (pBMCs) without triggering significant cytokine release (interleukin-6 or tumor necrosis factor-A). Similarly, these nanocarriers do not release oligonucleotides in animal serum. The nanocarrier material (PGAamine polyplex) similarly does not cause hemolysis (red blood cell lysis). In vivo, the APA nanocarrier does not significantly change blood glucose levels or pancreas morphology following intravenous infusion.
The oligonucleotide-APA nanocarriers are then tested on an orthotopic pancreatic cancer mouse model that contains mCherry-labeled MiaPaCa2 cells implanted into SCID mice pancreas. These nanocarriers are then administered intravenously. Following passive delivery, there is nanocarrier accumulation in the tumor sites, whereas other organs, such as kidneys, spleen, heart, lungs, and liver, have minimal accumulation. Resected tumor tissue further confirms increased miR-34a levels and their respective target genes (e.g., Bcl2, CDK6, MET, and Notch1).
Tumor size is tracked through its fluorescence signal, which shows that the nanocarrier combination therapy (miR-34a and siRNA PLK1) inhibits tumor growth (3.85% of the untreated group). Other combinations, such as single-oligonucleotide or noncoding controls, do not inhibit tumor growth as effectively. Sampled tumor tissues express lower proliferation (ki67) and angiogenesis (CD31), whereas apoptosis (cleaved caspase 3) concomitantly increases. Overall, this therapeutic leads to increased mouse survival compared with no treatment or single-agent treatment. Further investigations also reveal that dual-agent therapy inhibits MYC expression in combination. Promisingly, long-term pancreatic survivor tumors have reduced MYC expression compared with that of short-term survivor tumors.
In summary, the authors demonstrate the efficacy of oligonucleotide combinations to suppress pancreatic tumors. They identify this to be driven by suppressing MYC—a cancer biomarker that correlates with aggressive phenotypes. Crucially, these agents are successfully delivered via an APA nanocarrier to enable treatment in an orthotopic xenograft model. (Gibori, H.; et al. Nat. Commun.