A workflow is demonstrated for the absolute quantification of drug carrier-cell interactions using flow cytometry to allow better rational evaluation of novel drug delivery systems. This workflow is applicable to drug carriers of any type.
A major component of designing drug delivery systems concerns how to amplify or attenuate interactions with specific cell types. For instance, a chemotherapeutic might be functionalized with an antibody to enhance binding to cancer cells ("targeting") or functionalized with polyethylene glycol to help evade immune cell recognition ("stealth"). Even at a cellular level, optimizing the binding and uptake of a drug carrier is a complex biological design problem. Thus, it is valuable to separate how strongly a new carrier interacts with a cell from the functional efficacy of a carrier's cargo once delivered to that cell.
To continue the chemotherapeutic example, "how well it binds to a cancer cell" is a separate problem from "how well it kills a cancer cell". Quantitative in vitro assays for the latter are well established and usually rely on measuring viability. However, most published research on cell-carrier interactions is qualitative or semiquantitative. Generally, these measurements rely on fluorescent labeling of the carrier and, consequently, report interactions with cells in relative or arbitrary units. However, this work can be standardized and be made absolutely quantitative with a small number of characterization experiments. Such absolute quantification is valuable, as it facilitates rational, inter- and intra-class comparisons of various drug delivery systems-nanoparticles, microparticles, viruses, antibody-drug conjugates, engineered therapeutic cells, or extracellular vesicles.
Furthermore, quantification is a prerequisite for subsequent meta-analyses or in silico modeling approaches. In this article, video guides, as well as a decision tree for how to achieve in vitro quantification for carrier drug delivery systems, are presented, which take into account differences in carrier size and labeling modality. Additionally, further considerations for the quantitative assessment of advanced drug delivery systems are discussed. This is intended to serve as a valuable resource to improve rational evaluation and design for the next generation of medicine.
The design of drug delivery constructs that exhibit specific, designed behavior depending on what cell type they encounter has attracted substantial research interest. Potential drug delivery constructs or "carriers" include lipid formulations, nano-grown inorganics, polymeric assemblies, extracellular vesicles, functionalized bacterial cells, or modified viruses. All of these can exhibit organ, tissue, or cell specificity due to physical properties, surface properties, or engineered chemical functionalizations such as antibody attachment1,2.
A nearly ubiquitous step in in vitro carrier evaluation is to incubate cells with a suspension containing said drug-loaded carrier. Post incubation, carrier performance is measured via a functional readout of the drug cargo's performance, for example, transfection efficiency or toxicity. Functional readouts are useful, as they are a downstream measure of carrier effectiveness. However, for more complex drug delivery constructs, it is increasingly important to move beyond functional readouts and separately quantify the degree of carrier interaction with the cell of interest. There are a few reasons for this.
First, there is increasing interest in discovering (and iteratively improving) "platform" carrier technologies, which can carry a variety of cargo. For example, lipid nanoparticles (LNPs) designed to encapsulate RNA can exchange one RNA sequence for another with few caveats3. Thus, to iteratively improve the carrier technology, it is critical to quantify its performance independent of the cargo functionality. Second, functional readouts may not be straightforward for the cargo of interest, compromising the ability to rapidly iterate and evaluate carrier formulations. While one could perform in vitro optimization using a model cargo with a straightforward functional readout (for instance, fluorescence), changing the cargo can change the biological response to a carrier4 and may, thus, not yield representative results. Third, many carriers are designed to interact with and be taken up by a specific cell type. Such targeting capability of a carrier can and should be differentiated from the performance of its therapeutic cargo post targeting. To continue the LNP example, an RNA cargo might be extremely potent, but if the LNP is unable to bind to the cell, be internalized, and release the RNA, no downstream functional effect will be observed. This can be an issue particularly for carriers intended to target hard-to-transfect cell types, such as T cells5. Conversely, an LNP could target extremely effectively, but the RNA cargo might not function. A downstream assay that just measures cargo functionality will be unable to differentiate between these two situations, thus complicating the development and optimization of carrier drug delivery systems.
In this work, how to absolutely quantify carrier association is discussed. Association is a term that refers to the experimentally measured degree of interaction between a carrier and a cell. Association does not differentiate between membrane binding and internalization-a carrier may be associated because it is bound to the cell surface or because the cell has internalized it. Association is commonly measured as part of cell-carrier incubation experiments. Historically, association has been reported either in arbitrary fluorescent units (typically "median fluorescence intensity" or MFI) or as "percent association," metrics whose limitations have been previously discussed6. In short, these measurements are not comparable between experiments, laboratories, and drug carriers due to differences in experimental protocols, flow cytometer settings, and the labeling intensities of different carriers. Efforts have been made to overcome the former by calibrating the cytometer, thereby converting the relative measure of MFI into an absolutely quantitative measure of fluorescence7. However, this method does not account for the variability in the labeling intensity of various carriers and, thus, does not allow the rational comparison of various carrier performances in a target cell of choice8.
Here, how to practically convert from relative, arbitrary fluorescent units to the absolute quantitative metric of the "number of carriers per cell" is demonstrated by performing a small number of additional characterization experiments. If another metric of carrier concentration is desired (e.g., carrier mass per cell or carrier volume per cell), it is straightforward to convert from carriers per cell, provided carrier characterization has been done. For brevity and to avoid jargon, the word "carrier" is used within this work to refer to the vast assortment of drug delivery constructs. These quantification techniques are equally applicable, whether applied to a nano-engineered gold particle or a bio-engineered bacteria.
A few facts enable the conversion from arbitrary fluorescent units to carriers per cell. First, the measured fluorescence intensity is proportional to the concentration of a fluorophore9 (or a fluorescently labeled carrier), assuming the fluorescence is within the detection limits of the instrument and the instrumentation settings are the same. Thus, if the fluorescence of a carrier and the fluorescence of a sample are known, one can determine how many carriers are present in that sample if all the measurements were performed under the same settings and conditions. However, especially for smaller carriers, it may not be possible to measure carrier fluorescence, cell autofluorescence, and cell-associated-with-carriers fluorescence on the same instrument with the same settings. In this case, there is a second requirement to make it possible to convert between measured fluorescence on one instrument and measured fluorescence on another. To do so, a standard curve of fluorophore concentration can be established to measure the fluorescence intensity on both instruments, taking advantage of the Molecules of Equivalent Soluble Fluorochrome (MESF) standard9. This then allows measurement of the carrier fluorescence in bulk on a non-cytometer, a measurement that can be done on carriers of any size or characteristic. When such bulk quantification is done on a carrier suspension of known concentration, the number of carriers per cell of a sample can, once again, be calculated.
While this work demonstrates the process for measuring carrier association (as determined by measured fluorescence intensity), an analogous protocol could be performed for other measures of cell-carrier interaction (e.g., an experimental protocol that differentiates internalized and membrane-bound carriers). Additionally, this protocol would be largely the same if association was measured through a non-fluorescent assay (for instance, through mass cytometry).
1. Choosing the appropriate stream
Figure 1: Workstream decision tree. The decision as to which Stream to use depends primarily on the carrier type of interest. Larger carriers and carriers with high scattering properties can more easily be detected individually on cytometers, thus making them suitable for quantification using the Cytometer Stream. The Bulk Stream is suitable for all other carrier types. Please click here to view a larger version of this figure.
Figure 2: Overview of workstreams. This protocol is split into two different Streams. The Cytometer Stream uses a sensitive cytometer to count the carriers in suspension, measure their individual fluorescence, and then determine the fluorescence of cells incubated with carriers. The Bulk Stream uses non-cytometry-based techniques, such as Nanoparticle Tracking Analysis, to count the carriers in suspension. The individual carrier fluorescence is then quantified using a microplate reader or spectrofluorometer. The use of the flow cytometer is, therefore, restricted to measuring the final fluorescence of cells incubated with carriers, a measurement that can be done on a wider range of cytometers and that is independent of the carrier type used. Abbreviations: MESF = Molecules of Equivalent Soluble Fluorochrome; MFI = median fluorescence intensity. Please click here to view a larger version of this figure.
2. The Cytometer Stream
3. The Bulk Stream
As discussed previously, different drug carrier types require the use of different techniques for the absolute quantification of cell-carrier association. For example, 633 nm disulfide-stabilized poly(methacrylic acid) (PMASH) core-shell particles are large and dense enough for detection using a sensitive flow cytometer. As such, these particles were labeled fluorescently, then gated and counted using side-angle light scattering (SALS, analogous to SSC), as well as the appropriate fluorescent channel (Figure 3). The difference in event count in both channels was 1.98%, well within the acceptable range.
In contrast, 100 nm superparamagnetic iron oxide nanoparticles are too small to detect individually and were, thus, analyzed using the Bulk Stream. These nanoparticles were counted and characterized using Nanoparticle Tracking Analysis (Figure 4). The mean nanoparticle size of 136 nm reflects the hydrodynamic diameter of the nanoparticle in water. The nanoparticle concentration measured-still uncorrected for the dilution performed-falls within the dynamic range of the instrument, suggesting successful and accurate determination of the nanoparticle concentration.
Continuing in the Bulk Stream, the absolute fluorescence intensity of a carrier suspension is to be converted into an MFI value on the flow cytometer used for the final cell-carrier incubation. In this experiment, quantitation beads labeled with the same fluorescent dye as the carrier were used on multiple days to generate standard curves on a flow cytometer (Figure 5). The correlation between the measured MFI and the MESF value of the quantitation beads is linear and largely similar between the dates measured. However, slight differences between dates can be observed, and, as such, it is recommended to regenerate a standard curve as part of the readout of the carrier-cell experiments.
Once the MFI of individual carriers has been determined, the cell-carrier association data can be absolutely quantified and more accurately interpreted. Performing time course experiments, e.g., incubating HeLa cells fluorescently labeled with 235 nm PMASH capsules, for various timepoints between 0 h and 24 h (Figure 6) is recommended. As expected, median cell fluorescence increases over time, indicating the capsules are associating with HeLa cells. While such experiments can be used to compare the relative carrier performance at various timepoints, these results are not absolutely quantitative.
The importance of absolute quantitation, whether done via the Cytometer Stream or Bulk Stream, becomes clear when comparing the association of two carrier types. Figure 7 depicts the same two experiments, analyzed by either relative quantitation (Figure 7A) or absolute quantitation. The difference in the apparent cellular response to carriers is stark, depending on the analysis performed; carriers should not be directly compared when using relative quantification (Figure 7A), whereas absolute quantification is independent of labeling intensity and, thus, more comparable (Figure 7B).
Figure 3: Counting particles using a flow cytometer (Cytometer Stream). An Apogee flow cytometer was used. (A) Particle counting using an optical channel (SALS), resulting in an event count of 200,659. (B) Carrier counting using the carrier's fluorescent channel, resulting in an event count of 204,636. In both cases, an arcsinh transform followed by gating was applied (at 6 and 4, respectively). Figures created from raw data of 633 nm core-shell particles first published in Faria et al.10. Abbreviation: SALS = small-angle light scatter. Please click here to view a larger version of this figure.
Figure 4: Counting carriers using Nanoparticle Tracking Analysis (Bulk Stream). Superparamagnetic iron oxide nanoparticles of 100 nm in size were characterized and counted using Nanoparticle Tracking Analysis. Left, concentration/size histogram of individual results of five replicate measurements. Right, average concentration/size distribution ± SEM of five replicates. In this particular sample, a 1:2,000 dilution from stock was performed to obtain a concentration within the dynamic range of the instrument. Please click here to view a larger version of this figure.
Figure 5: Converting MESF to MFI using quantitation beads on a flow cytometer (Bulk Stream). Fluorescent beads with four different MESF values were analyzed on a flow cytometer on various days, and standard curves were drawn. Abbreviations: MESF = Molecules of Equivalent Soluble Fluorochrome; MFI = median fluorescence intensity. Please click here to view a larger version of this figure.
Figure 6: Final cell experiment on flow cytometer-time course series (Cytometer Stream and Bulk Stream). Representative images of flow cytometry data from a time course experiment. THP-1 cells were incubated with a 235 nm polymeric capsule for 0 h, 1 h, 2 h, 4 h, 8 h, 16 h, and 24 h (left to right). Carrier fluorescence is shown on the x-axis, while an optical channel is shown on the y-axis. In all cases, an arcsinh transform was performed. No gating was performed (other than choosing the limits of the graph). Created from raw data in Faria et al.10. Abbreviation: SALS = small-angle light scatter. Please click here to view a larger version of this figure.
Figure 7: Relative versus absolute quantification of carrier-cell association (Cytometer Stream and Bulk Stream). Both figures visualize data from the same two experiments, a 24 h incubation between RAW264.7 macrophages and 150 nm (blue) or 633 nm core-shell carriers (orange), created from raw data in Faria et al.10. The medians of both technical replicates are plotted. (A) Relative quantification, reported as MFI in a.u. (B) Absolute quantification, reported as the number of carriers per cell. Abbreviations: MFI = median fluorescence intensity; a.u. = arbitrary units. Please click here to view a larger version of this figure.
Characterizing the interactions between drug carriers and cells is becoming increasingly important in the development of novel drug delivery systems. Specifically, to allow the rational evaluation and comparison of various carrier constructs, absolute quantification of the performance of said carrier to interact with target and off-target cells is critical. This protocol describes a two-stream methodology that allows any researcher working with a drug carrier to convert relative, semiquantitative flow cytometry data on cell-carrier association into absolute quantitative results. The outlined process is applicable to any type of carrier — small, large, organic, inorganic — provided they are fluorescently labeled. However, different carriers require different approaches to calculate the carriers per cell. This is due to the limits of various instrumentation used for the required characterization measurements.
As such, this protocol is split into two Streams: the Cytometer Stream and the Bulk Stream. The first, the Cytometer Stream, is the more straightforward approach and requires only a flow cytometer, but this cytometer needs to be sensitive enough to detect the individual carriers used, or, in other words, the drug carrier of interest needs to be large and dense enough to be detected. The Bulk Stream is compatible with any carrier type, as the need to detect individual carriers is circumvented through the use of alternative instrumentation. However, the Bulk Stream requires more characterization experiments to be done.
To assist with choosing the appropriate Stream (and as such, the appropriate technologies), the above considerations have been summarized into a decision tree (Figure 1). To reiterate, the Bulk Stream is suitable for all researchers irrespective of the carrier type used and can, thus, always be reverted to as a backup. The workflows of the two Streams are outlined in Figure 2. Notably, with the ongoing advances in the sensitivity of flow cytometers, it is possible that the Cytometer Stream can be used for more <300 nm carriers.
A few notes should be made on this procedure. First, the described strategy for converting the cell fluorescence into an absolute number of carriers per cell relies on the measurement of the individual carrier fluorescence in suspension. However, fluorescence is influenced by the immediate chemical environment of the fluorochrome (e.g., the carrier diluent, the cell surface, or various intracellular compartments). In particular, the acidic environment of the endosomal compartment is known to affect the fluorescence intensity of certain primarily protein-based dyes11. As such, irrespective of the carrier type studied, it is recommended to use pH-insensitive and highly photostable dye ranges for the labeling of drug carriers (see the Table of Materials for a recommendation). The additional benefit of this dye range is its general brightness, which enhances the sensitivity of drug carrier detection.
Second, the methods discussed above are intended to provide guidance but by no means form an exclusive list of techniques that can be used to perform the various steps in the quantitation workflow. In particular, a variety of techniques exist to count smaller or low-scattering carriers (as those in the Bulk Stream). These include other instruments able to perform the same Nanoparticle Tracking Analysis12, as well as other methods altogether, such as multi-angle dynamic light scattering, mass spectroscopy, electron microscopy13, gravimetry, or optical density measurements (further reviewed by Shang and Gao14). Furthermore, this experimental quantification-moving from arbitrary fluorescence units to an absolute number of carriers per cell-is only part of what might be termed a "quantitative biology workflow". Absolute experimental quantification is necessary but not sufficient. Merely quantifying and reporting the carrier association does not account for differences in particle association due to the experimental setup, such as the incubation time, dose, and concentration of drug carrier added, or the number of cells used per experimental container.
Additionally, carriers with different physicochemical properties may have very different dosages delivered to cells, even when the experimental setup is identical15. Carrier performance-even in vitro-cannot be determined without accounting for all these factors. In other words, merely quantifying the carrier-cell association does not, in general, allow for an unbiased comparison of carriers between researchers and laboratories. In this light, there has been prior discussion of the importance of performing time-course experiments of carrier-cell association and subsequent in silico mathematical modeling to derive kinetic parameters (i.e., rate constants) as an unbiased and quantitative measure of the affinity between a particular carrier-cell pair10. Experimental quantification is also a prerequisite for other in silico techniques to be applied, such as identifying potential sources of biological heterogeneity16 or determining penetration kinetics in a complex biological model17. Combined with the Minimum Information Reporting in Bio-Nano Experimental Literature guidelines for standardized reporting18, the quantitative evaluation of carrier performance can accelerate the development of novel nanomedicine.
The authors have nothing to disclose.
This work was supported by the Australian National Health and Medical Research Council (NHMRC; Program Grant No. GNT1149990), the Australian Centre for HIV and Hepatitis Virology Research (ACH2), as well a gift from the estate of Réjane Louise Langlois. F.C. acknowledges the award of a National Health and Medical Research Council (NHMRC) Senior Principal Research Fellowship (GNT1135806). Figure 1 and Figure 2 were created with BioRender.com.
Alexa Fluor 647 C2 Maleimide | Invitrogen | A20347 | pH-stable dye used to label 150 nm, 235 nm, or 633 nm PMASH carriers; example of good dye to use in cell-carrier association studies |
Apogee A50 Microflow | Apogee | Sensitive flow cytometer capable of detecting small carriers for counting | |
CytoFLEX S Flow Cytometer | Beckman Coulter | Sensitive flow cytometer capable of detecting small carriers for counting and read out for final cell-barrier experiments | |
FCS Express | De Novo Software | Software used to analyze flow cytometry data, i.e., perform gating and derive median fluorescence intensity values of populations of choice. Alternatives include FlowJo, OMIQ, Python | |
Infinite 200 PRO | Tecan Lifesciences | Standard microplate reader instrument used for bulk fluorescence measurements of carriers in solution | |
LSRFortessa Cell Analyzer | BD Biosciences | Less sensitive flow cytometer, but one more generally available to researchers. Can be used to read out final cell-carrier experiment | |
NanoSight NS300 | Malvern Panalytical | Instrument used for Nanoparticle Tracking Analysis | |
Prism 8 | GraphPad | Software used to graph and calculate standard curves. Alternatives include Microsoft Excel, Origin, Minitab, Python amongst many others | |
Quantum MESF kits Alexa Fluor 647 | Bangs Laboratories | 647 | Absolute quantitation beads for flow cytometery. Used to convert fluorescence intensities measured in bulk on a microplate reader to fluorescence intensities measured on a flow cytometer using the MESF standard |