Fluorescent genetic barcoding in mammalian cells for enhanced multiplexing capabilities in flow cytometry

Authors


  • CAS and BJH contributed equally to this work.

Abstract

The discovery of the green fluorescent protein from Aequorea victoria has revolutionized the field of cell and molecular biology. Since its discovery a growing panel of fluorescent proteins, fluorophores and fluorescent-coupled staining methodologies, have expanded the analytical capabilities of flow cytometry. Here, we exploit the power of genetic engineering to barcode individual cells with genes encoding fluorescent proteins. For genetic engineering, we utilize retroviral technology, which allows for the expression of ectopic genetic information in a stable manner in mammalian cells. We have genetically barcoded both adherent and nonadherent cells with different fluorescent proteins. Multiplexing power was increased by combining both the number of distinct fluorescent proteins, and the fluorescence intensity in each channel. Moreover, retroviral expression has proven to be stable for at least a 6-month period, which is critical for applications such as biological screens. We have shown the applicability of fluorescent barcoded multiplexing to cell-based assays that rely themselves on genetic barcoding, or on classical staining protocols. Fluorescent genetic barcoding gives the cell an inherited characteristic that distinguishes it from its counterpart. Once cell lines are developed, no further manipulation or staining is required, decreasing time, nonspecific background associated with staining protocols, and cost. The increasing number of discovered and/or engineered fluorescent proteins with unique absorbance/emission spectra, combined with the growing number of detection devices and lasers, increases multiplexing versatility, making fluorescent genetic barcoding a powerful tool for flow cytometry-based analysis. © 2013 International Society for Advancement of Cytometry

Since the isolation and cloning of the green fluorescent protein (GFP) from Aequorea victoria [1], fluorescent proteins have revolutionized all aspects of biomedical research particularly the field of flow cytometry. The expression of these proteins in mammalian cells and others has enabled tracking of individual cells within a large population, enabling the study of cell fate. Moreover, they have been crucial for the study of gene regulation, and their use as tags within fluorescent fusions have dramatically facilitated the investigation of their biological functions and consequences [2]. The introduction of retroviral technology that enables protein expression in mammalian cells in a stable manner has extended the advantages of fluorescent proteins [3-7]. The ability to stably express genes in mammalian cells together with the discovery of an increasing number of fluorescent proteins and genetic manipulations of GFP [8], has further enhanced the utility of flow cytometry for cell analysis [9, 10]. Novel fluorescent proteins with broader absorbance/emission spectra and larger Stokes shifts [11-14] have been introduced in conjunction with additional probes, dyes, and lasers of varying wavelengths [14-18]. This has allowed for the analysis of an ever-increasing number of parameters that can theoretically be analyzed concomitantly in the same sample at the same time. However, the multiparameter aspects of the current instrumentation do not always match the experimental design nor reflect the appropriate technological capabilities.

Multiparameter, multifaceted applications are thus made available by flow cytometry and should allow for more complex analysis or utilities than classical detection of gene expression, cell cycle, apoptosis, phosphorylation events, or any general biological question at the single cell level [10, 19-24]. The introduction of robotics in plate reader systems, together with new imaging and flow cytometry coupled applications, has significantly increased high-throughput capabilities in biomedical research [21, 25], as demonstrated by novel applications involving, but not restrained to, the use of cell-based assays as a platform for drug screening [22, 26-29], as well as multiparameter analysis of signaling cascades [20, 30, 31]. Typically, the preparation of these samples includes time-consuming protocols and significant amounts of costly reagents including antibodies, beads and/or stains [22, 32, 33]. While many of these procedures can be calibrated in advance to reduce cost and time, such optimization is not always feasible and requires a higher degree of expertise.

A growing number of biological applications in clinical and/or research settings, in parallel to the growing technological capabilities of the available instrumentation such as acoustic cytometry, multispectral imaging flow cytometry, or mass spectrometry [25, 34-43], demand new methodologies that can efficiently couple both. Multiplexing, as defined by the simultaneous evaluation of several experimental elements, can accomplish this goal [33, 44]. Multiplexing allows for a significant increase in the number of samples analyzed per unit of time. When high-throughput screening (HTS) is paired with multiplexing, time efficiency is enhanced while cost can be considerably reduced [22, 45, 46]. Krutzik and Nolan [22, 33] describe an elegant way of multiplexing cell analysis aimed at distinguishing different cell populations based on increasing amount of antibody/stain. While this approach does decrease time, it relies on previous careful laborious calibration of the staining technique, whether it is antibody or dye based. Moreover, the approach may be compromised with rapidly dividing cells [47] or with cells overexpressing transporter systems that interact with dyes (Cannon and Sklar, unpublished). Retaining multiplexing capabilities without the need for dyes, stains, antibodies, quantum dots or biolabels in general can be accomplished by utilizing retroviral technology instead.

Retroviral technology for gene delivery has transformed the field of protein expression in mammalian cells and since its development has been widely used in a broad range of gene products and cell types [3, 6, 48-51]. Moreover, retroviral-based technology allows for stable expression of ectopic genetic information for long periods of time as the information carried inside the retroviral particle is integrated into active sites of transcription, as part of the virus' natural life cycle [52].

GFP, mCherry, td Tomato, E2 Crimson, Cerulean, and mBanana represent only a few examples of the ever increasing number of fluorescent proteins available for biomedical applications. Retroviral technology allows for the engineering of cells encoding a diverse range of fluorescent gene products, thus generating cell populations distinguishable by their fluorescence characteristics. A distinct fluorescence profile identifies one cell from its counterpart and can thus be exploited for what we refer to as “genetic barcoding.” Tiemann et al. [53], for example, have elegantly utilized different fluorescent protein genes coupled to individual canonical reprogramming factors to track their expression toward pluripotency in somatic cells. Livet et al. [54] have genetically labeled neurons with multiple, distinct colors for the analysis of the neuronal network architecture.

Genetic barcoding further allows the mixing of unique fluorescent cells, dramatically increasing multiplex capabilities. Naturally, the chosen fluorescent proteins should have minimal spectral overlap for multiplexing. To further enhance the power of multiplexing, each cell population can harbor a number of fluorescent proteins which can also be selected on the basis of varying fluorescence intensities. Thus, a matrix with a larger number of distinguishable populations can be obtained by combining different fluorescent proteins and intensities. Established populations of barcoded cells with fluorescent genetic markers can be used in tandem with an array of cell-based assays to address a variety of biological questions. We have genetically barcoded cells with different fluorescent proteins, tested their stability across multiple generations, and obtained distinct clonal populations based on differential fluorescent intensities. Moreover, to demonstrate biological applications, established genetically barcoded cells have been adapted to two of our existing cell-based assays [55, 56]. The coupling of genetically barcoded cells to cell-based assays will enhance high-throughput capabilities by reducing the number of screens needed.

Materials and Methods

Cell Maintenance

Human embryonic kidney (HEK) 293T and Phoenix GP cell lines were maintained at 37°C and 5% CO2 in Dulbecco's modified eagle medium (DMEM; Cellgro) supplemented with 10% fetal calf serum (FCS), penicillin–streptomycin, and 2 mM L-glutamine. SupT1 cells were maintained at 37°C and 5% CO2 in RPMI (Cellgro) supplemented with 10% FCS, penicillin–streptomycin, and 2 mM L-glutamine. Cells were routinely screened for mycoplasma contamination. Phoenix GPs were provided by Gary Nolan from Stanford University.

Plasmid Construction

The construct pBMN-i-td Tomato was created by digesting a previously constructed plasmid pBMN-i-enhanced GFP (eGFP). Td Tomato (kindly provided by Roger Tsien at UCSD, was polymerase chain reaction amplified using the forward primer TATAACATGTCAATTGCCACCATGGTGAGCAAGGGCGAGGAG, which contains a PciI site, and the reverse primer ATGGACCAGCTGTACAAGTAGGTCGACTATA, which contains a SalI site. The amplicon was digested with PciI and SalI and used to ligate into pBMN-i-eGFP digested with NcoI and SalI, which removes eGFP. pBMN-i-E2 Crimson was constructed similarly. The forward primer used to amplify E2 Crimson (obtained from Clontech) was TATACCACCATGGATAGCACTGAGAACGTC, containing an NcoI site and the reverse primer CGCCACCACCTGTTCCAGTAGTCTAGAGTCGACTATA, which contains a SalI site. Both pBMN-i-eGFP and E2 Crimson products were digested with NcoI and SalI for ligation.

Generation of Infectious Viral Particles

For production of murine leukemia virus, the Phoenix GP cell line at 60–70% confluence was transfected with 3 μg of transfer vector [pBMN-i-E2 Crimson, pBMN-i-td Tomato, pBMN-i-td mCherry, pBMN-i-eGFP, and pBMN-i-enhanced cyan fluorescent protein (eCFP)] and 3 μg of Vesicular Stomatitis Virus Envelope glycoprotein vector (pCI-VSVg). DMEM media was replaced 24 h post-transfection and viral supernatant was collected 48 h and at 72 h after transfection. All viral stocks were filtered with 0.45 μm polytetrafluoroethylene filters (Pall Corporation) and frozen at −80°C in 2 mL aliquots.

Transductions

Huh 7.5.1 and HEK 293-T cells grown in DMEM at 250,000 cells/well in a six well plate were prepared for transduction. The 24 h after plating, cells were treated with 5 μg/mL polybrene (hexadimethrene bromide; Sigma) and transduced with viral stocks by hanging bucket rotors centrifuge (Becton Dickinson) at 1,500g, for 120 min at 32°C. The 24 h post-transduction media was changed. 5 × 106 SupT1 cells/well in a six well plate grown in RPMI supplemented were treated with 5 μg/mL Polybrene (hexadimethrene bromide; Sigma) and infected with viral stocks by centrifugation in a hanging bucket rotors centrifuge (Becton Dickinson) at 1,500g, for 120 min at 32°C. Cells were then analyzed for expression 72 h postinfection.

Staining for Analysis

Cells were pelleted, re-suspended in 200 μL and incubated with mouse anti-FLAG (Sigma Aldrich, St. Louis, MO) and rabbit anti-HA (Cell Signaling, Beverly, MA) at 1:400 dilution for 20 min and then washed with phosphate-buffered saline (PBS). Cells were then incubated with anti-mouse Alexa Fluor 488 and anti-rabbit Alexa Fluor 647 (Cell Signaling) antibodies at 1:200 dilutions for 20 min and washed with PBS. All cells were stained in suspension with PBS.

Flow Cytometry and Sorting

Cell samples were washed twice with 1× PBS before loading into the instrument. Adherent cell types were detached using 0.25% trypsin and 1 mM ethylenediaminetetraacetic acid, resuspended and neutralized with DMEM (10% FCS) before washing. The Flow Cytometry Core Facility at San Diego State University performed analysis of cells on BD fluorescent-activated cell sorting (FACS) Aria at 405 nm, 488 nm, and 633 nm lasers, as well as the BD FACS Canto using the 488 nm blue laser, and the 633 nm red laser. Excitation of E2 Crimson was performed with a red laser and emission detection with 660/20 band pass filter. Excitation of td Tomato/mCherry was performed with a blue laser and emission detection with 585/42 band pass filter separated by a 556LP dichroic filter. Excitation of eGFP and eCFP was performed with a blue laser and a violet laser, respectively, with a 530/30 band pass filter separated by a 502LP dichroic filter. Data were collected FACS Diva 6.1.1 software, and analyzed on FlowJo 7.6.5.

Results

Genetically Barcoded Mammalian Cells Can Distinguish Different Populations

Individual genetically barcoded cells must be independently obtained before being able to discriminate single cell populations within a mixture of cells. Expression of fluorescent proteins in mammalian cells has been extensively performed in the past [2, 57-60], but not in the context of “genetic barcoding.” Here, we have genetically engineered cells with different fluorescent proteins to achieve genetic multiplexing capabilities. We initially selected the Huh 7.5.1 hepatocytic cell line and the commonly used HEK 293T cell line as examples of adherent cells to demonstrate the versatility of genetic barcoding. Cells were transduced with retroviral particles carrying an individual fluorescent protein chosen from a variety of fluorescent proteins such as mCherry, td Tomato, and E2 Crimson. Following a process of transduction and amplification, individual cells were collected in single wells of 96-well plates using FACS. A series of Huh 7.5.1 and HEK 293T cell clones expressing a single fluorescent protein were obtained. Mammalian cells genetically barcoded with fluorescent proteins, as shown in Supporting Information Figure 1A, can be identified through flow cytometry. Visualization is only possible if cell populations are uniform and have fluorescent intensities distinguishable from the nonbarcoded naïve cells, in the same channel.

Barcoding Mammalian Cells Allows for Multiplex Analysis

The generation of individual genetically barcoded clones with different fluorescent characteristics allows us to mix them together without losing their distinct characteristics. Moreover, individual clones can be engineered to express more than one fluorescent protein, provided their spectrum does not overlap. When the fluorescence characteristics are separated, one can then obtain any of their combinations together to further increase the number of populations with unique signatures. Here we have genetically barcoded a third cell type to establish the principle for higher throughput applications that avoid resuspension of adherent cells. We chose the nonadherent SupT1 T-cell line. Supporting Information Figure 1B shows that with two fluorescent proteins one can obtain a matrix of up to four unique populations; the naïve (Population #1), the single positive (Populations #2 and 3), and the double positive (Population #4) expressing both CFP and mCherry.

Genetically Barcoded Cells Can be Further Differentiated Based on Fluorescence Intensity

A panel of distinct genetically barcoded cells can be further expanded for multiplexing by exploiting the level of protein fluorescence intensity. This can be achieved if the clonal populations carrying individual proteins are selected based on expression at different intensities. To make them distinguishable, the expression level needs to be such that the value of the mean fluorescence intensity (MFI) of each population is far enough apart from each other, typically one log scale. SupT1 cells were analyzed 72 h following simultaneous retroviral transduction with particles carrying td Tomato and particles carrying E2 Crimson. As expected, upon transduction, cells were obtained that express either td Tomato, E2 Crimson, or both (Supporting Information Fig. 2, left panel). Gates were then set to obtain distinct clonal populations based on fluorescence channel and intensity, as shown in Supporting Information Figure 2, left panel. While E2 Crimson is very stable and populations with different intensities can be easily obtained (Supporting Information Fig. 3), we have chosen as proof of principle, a single intensity for E2 Crimson and two for td Tomato (dim and high). After sorting and amplification, a matrix of six distinctive populations was obtained, including E2 Crimson (Population #4), td Tomato dim and high (Populations #2 and 3), and two populations expressing E2 Crimson in conjunction with mid and high td Tomato (Populations #5 and #6, respectively, Fig. 2, right panel).

Genetically Barcoded Cells Are Stable for Long Periods of Time

The functionality of genetically barcoded cells can be further exploited for biological applications as long as the expression levels and fluorescent characteristics remain stable. To ensure the stability of genetically barcoded cells and the reproducibility of the instrumentation to identify and track populations over long periods of time, we performed a time course experiment. SupT1 clonal populations were analyzed at Day 0, and passaged for 6 months to determine whether protein expression is stable. Figure 1 (right upper panel) shows that the selected populations remain distinguishable over at least 6 months, proving the stability of barcoded cells and ability for performing multiplex analysis for long-term usage. While Population #5 (dim td Tomato/E2 Crimson) and #6 (bright td Tomato/E2 Crimson) start to merge, they are still distinguishable; with phycoerythrin (PE)-A channel MFI values of ∼3,400 and 23,000, respectively. The rest of the cell populations drifted minimally and Populations #2 and #3 remained identical after 6 months. To corroborate that freeze-thaw does not disturb signal stability, the same cells analyzed at Day 0 were frozen for a period of 6 months, thawed and re-analyzed. Results show that populations do not differ from the cells passaged for the same period of time (Fig. 1, left panel versus right lower panel). Genetically barcoded cells retain their unique fluorescent profiles and can be used in assays immediately upon thawing.

Decoding in Different Channels Allows to Further Increase Multiplexing and to Pinpoint Masked Populations

Figure 1.

Analysis of SupT1 clonal populations over a period of 6 months. Cells were analyzed at Day 0 for E2 Crimson and td Tomato (left panel), and reanalyzed after passaging for 6 months (right upper panel) or frozen at −70°C for the same period of time and thawed (right lower panel). [Color figure can be viewed in the online issue which is available at wileyonlinelibrary.com.]

Figure 2.

Decoding in different channels further increases the power of multiplexing and a panel of populations can actually mask a higher number of distinct populations. (A) The two six-population panels (eGFP negative: top panel, and eGFP positive: lower panel) are indistinguishable when analyzed for E2 Crimson and td Tomato. (B) A mixture of nine selected populations [circled numbers in (A)] were analyzed for E2 Crimson and td Tomato. (C) When analyzed in the FITC channel for eGFP expression, the masked populations (Populations #7, 9, and 11) are revealed. Gating the eGFP positive populations allows observation of all nine populations in the same dot plot panel through artificial color juxtaposition (right panel). (Overlaying histograms results in the loss of the scale).

Multiplexing, as assessed in Supporting Information Figure 2 with a panel of six populations, can be further enhanced with an additional fluorescent marker. The six populations obtained with a combination of td Tomato and E2 Crimson were transduced with retroviral particles carrying eGFP. Clonal populations were obtained as described previously and analyzed for E2 Crimson and td Tomato. The original populations (#1–6) were compared to the eGFP-expressing populations (#7–12). When analyzing by td Tomato and E2 Crimson the populations have identical signatures (Supporting Information Fig. 4, dot plots in left panels); however, when analyzed for eGFP, six new populations are revealed (Supporting Information Fig. 4, histograms in right panels).

The addition of eGFP allowed expansion of the matrix from a six-population (E2 Crimson and td Tomato) to a 12-population panel (E2 Crimson, td Tomato and eGFP) in the allophycocyanin (APC), PE, and fluorescein isothiocyanate (FITC) channels, respectively. A matrix of 12 populations could thus be theoretically obtained by combining the original six-population panel (Supporting Information Fig. 4A) with the same panel expressing eGFP (Supporting Information Fig. 4B). To demonstrate the power of revealing masked populations in genetically barcoded cells, we also mixed the set of eGFP negative six populations (#1–6) with three of the eGFP expressing populations (#7, 9, and 11, Fig. 2A). While the new panel of nine is not distinguishable from the original panel of six analyzed for td Tomato and E2 Crimson (Fig. 2B), the masked populations are revealed when analyzed for eGFP (Fig. 2C). Gating of the eGFP positive populations allows tracking them back and reveals them when analyzed with the original channels, as seen by juxtaposition of the initial colored populations with the green populations (right panel in Fig. 2C).

Multiplexing through Genetic Barcoding for Biological Applications—Adaptation to Cell-Based Assays

As described in Supporting Information Figures 2 and 4, a panel of distinct populations based on two fluorescent proteins was further expanded with the use of a third fluorescent protein. While the third fluorescent marker was used purely as an additional genetic marker, it can be exploited as part of biological application such as cell signaling, phenotypic outcomes, or a cell-based assay. For these types of applications eGFP expression is intended to be a function of biological activity. As proof of principle we have utilized a panel of four SupT1 populations including naïve, E2 Crimson, td Tomato, and E2 Crimson/td Tomato barcoded cells. Each of the populations within the panel was used to express one of four different HIV-1 protease variants in an inducible manner. In the assay, as described elsewhere [56], the protease variants can induce eGFP expression if inhibited. Figure 3 shows the panel of genetically barcoded cells, each expressing a distinct HIV-1 protease variant. When untreated or when treated with doxycycline, neither of the populations is fluorescent as they do not express eGFP. When inhibited with a known protease inhibitor such as Darunavir, all populations express eGFP, as observed upon secondary analysis. Such an experiment allows simultaneous monitoring of the activity of four selected HIV-1 protease mutants using eGFP as a biosensor for protease inhibition. Decoding in a different channel can thus reveal which population or protease variant was affected by that specific treatment, even though in this example all populations respond to inhibitor.

Figure 3.

Genetically barcoded SupT1 cells expressing HIV-1 protease variants. A mixed population of SupT1 cells expressing combinations of fluorescent proteins (E2 Crimson and td Tomato) and PR variants were analyzed for E2 Crimson and td Tomato (dot plot). Following activation with Dox and Dar, all clones turn green fluorescent (histograms in lower panels). Dox: Doxycycline, Dar: Darunavir.

When appropriate for evaluating biological function, genetic barcoding can be used in tandem with classical methods of antibody staining. In yet another example, enhanced genetic barcoding was utilized for a cell-based assay that monitors cleavage in the classical secretory pathway and relies on the expression of one or two tags on the cell surface (HA or HA and FLAG). Td Tomato barcoded cells (Supporting Information Fig. 2, Populations #2 and 3) were chosen to independently express viral proteins known to be cleaved by host enzymes. If cleaved, only HA can be detected on the cell surface through APC-coupled anti-HA antibody (Populations #1 and 3). However, both HA and FLAG (FITC-coupled anti-FLAG antibody) will be detected if uncleaved (pop #2). The genetic elements of the assay with three different viral proteins known to be cleaved in the secretory pathway were introduced into naïve, dim, and high td Tomato-expressing cells. The three populations carrying the assay are shown in the PE channel (Fig. 4A). When the mixed sample was analyzed in the FITC and APC channels, a combination of APC only–and APC and FITC-positive populations was observed differentiating between cells that expressed cleaved versus uncleaved products (Fig. 4A). To further demonstrate the power of genetic barcoding in the context of a cell-based assay, these cells were gated based on HA-single positive or HA and FITC-double positive staining, and analyzed in the PE channel (Fig. 4B). Distinct phenotypes could thus be observed in different subsets, which could be pinpointed and tracked through decoding in different channels.

Figure 4.

Genetically barcoded SupT1 cells expressing cell surface markers through a cell-based assay. (A) A mixed population of SupT1 cells expressing td Tomato at different intensities were analyzed in the PE channel (left dot panel). (B) Following staining with anti-FLAG and anti-HA antibodies, cells were analyzed in the FITC and APC channels (right dot plot). The HA-single positive and the HA and FITC-double positive populations were gated and analyzed in the PE channel to determine their barcoded identity. [Color figure can be viewed in the online issue which is available at wileyonlinelibrary.com.]

Discussion

Fluorescent protein-based genetic barcoding provides a cell with an inherited characteristic that can distinguish it from its counterparts. While labor intensive at first, it becomes valuable once cell lines are developed, as from then on, no further manipulation or staining is required, dramatically decreasing time, nonspecific background associated with staining protocols, and, of course cost. Here we have utilized the power of retroviral technology to genetically engineer a number of cell types, both adherent and nonadherent, including HEK 293T and Huh7.5.1 cells, as well as SupT1 cells, respectively. We have utilized CFP in combination with mCherry or E2 Crimson in combination with td Tomato to obtain at least a minimum of four independent populations which include naïve, two single positive and a double positive population. Expression of fluorescent proteins through retroviral technology, which exploits the ability of viral particles to insert their genome into the genome of the host, has proven to be stable for long periods of time, at least for up to 6 months with nonadherent SupT1 cells. Some of the fluorescent proteins might be less stable in adherent cells such as HEK293T and Huh 7.5.1 cells and should be tested. Stable expression further allowed us to choose a larger number of genetically barcoded cells not only based on the type of fluorescent protein but on fluorescence intensity as well.

Variations in fluorescence intensities are expected in genetic barcoding due to the inherited characteristic of retroviral technology. Effects of insertional preferences into the host genome will lead to differential expression of the fluorescent protein marker. The nature of the insertion can be exploited to increase the multiplex power of genetic barcoding through retroviral technology. Combining fluorescent proteins with distinguishable physical parameters (absorbance/emission spectra), each at different intensities, allows to obtain a matrix of a large number of distinct populations. Here we have shown matrices of 4, 6, 9, and 12 populations, which included two six-population panels of E2 Crimson and td Tomato, with or without eGFP. As proof of principle in our example we utilized only one intensity for E2 Crimson and eGFP, and two for td Tomato. Theoretically, one could achieve a panel of up to 27 populations with three fluorescent proteins, each at two intensities. As mentioned, at least with SupT1 cells, it might be possible to expand the matrix to 27 populations considering the uniformity and the range of MFI values of the individual populations. Table 1 shows the multiplexing capabilities of combining up to three fluorescent proteins (proteins A–C) with up to two intensities each (low and high). We have shown panels of 4, 6, and 12 possible combinations (indicated by a star in Table 1), within the many possible combinations. The third and/or fourth color (proteins C and/or D) could be used as a marker of cell response or biological function. A panel of 81 distinct populations is theoretically possible with four colors, each at two intensities (not shown in Table 1).

Table 1. Multiplexing power of up to three proteins with up to two intensities each
 #ProteinsProtein AProtein BProtein CProtein D#Multiplex
  1. The #multiplex column shows the number of possible distinguishable populations in each panel. Protein C and/or D freed for biological function. A panel of 81 is theoretically possible with four colors, each at two intensities (not shown). L, low; H, high.

  2. a

    Examples shown in the article.

Intensity1L2
L/H3
2LL4a
L/HL6a
L/HL/H9
3LLL8
L/HLL12a
L/HL/HL18
L/Hl/HL/H27

While we have obtained two clear distinct intensities for td Tomato (dim and high in Supporting Information Fig. 2), up to three distinct intensities should be easily achievable for each protein. The theoretical multiplexing capabilities using proteins with up to three intensities each (low, medium, and high) are depicted in Table 2. A major advantage of utilizing fewer proteins at more intensities rather than more proteins at fewer intensities while maintaining the same multiplexing power, is the freedom of additional channels that may be needed for biological outcomes.

Table 2. Multiplexing power of up to three proteins with up to three intensities each
 #ProteinsProtein AProtein BProtein CProtein D#Multiplex
  1. The #multiplex column shows the number of possible distinguishable populations in each panel. Protein C and/or D freed for biological function. L, low; M, medium; H, high.

Intensity1L/M/H4
2L/M/HL8
L/M/HL/H12
L/M/HL/M/H16
3L/M/HLL16
L/M/HL/HL24
L/M/HL/HL/H36
L/M/HL/M/HL32
L/M/HL/M/HL/H48
L/M/HL/M/HL/M/H64

The single positive td Tomato populations of low and high intensities (Supporting Information Fig. 2) exhibit a PE-A MFI value of ∼5,250 and 31,500, respectively. However, the E2 Crimson positive cells display a maximum APC-A MFI value of ∼11,850, which leads us to believe that much higher expressing populations could be obtained. While the E2 Crimson intensity of Population #5 (Fig. 1) decreased over time, with APC-A MFI values decreasing from ∼11,800 to ∼4,200, the reduction did not hinder the ability to distinguish the populations when analyzed simultaneously. It is important to mention that different MFI values obtained from one fluorescent protein can be easily compared. In contrast, MFI values obtained from different fluorescent proteins can be compared to each other only with the use of classical methodologies such as western blotting for protein analysis or other nonclassical flow cytometry technologies based on intracellular staining against the protein of interest, which can determine the amount of protein at the single cell level.

When choosing combinations of different fluorescent proteins, we recommend the use of proteins that not only have a distinct absorbance/emission spectrum but also in their monomeric versions. This will avoid the formation of heteroassociations that can occur between different versions of monomeric GFP mutants or fluorescent proteins known to dimerize or tetramerize. Heteroassociations between CFP and GFP, for example, will most probably complicate the efforts of obtaining distinct populations with specific intensities.

We foresee that in many real life experiments a panel of two, three, or four distinct genetically barcoded populations will answer all the requirements and fulfill the advantages of multiplexing. For example, in a screen aimed at finding antivirals against Dengue virus, a panel of four would suffice to perform a screen against all the main Dengue virus serotypes. This is in considerable contrast to performing four independent screens against each of the individual serotypes. When needed, 6, 12, or even a larger number of populations can be obtained.

Genetic barcoding can be easily adapted to a broad range of biological applications. While in many of these applications, such as the study of signaling cascades or detection of specific endogenous markers, it is difficult if not impossible to avoid staining protocols, coupling it to genetic barcoding can still dramatically decrease cost and time. Cell-based assays that rely by themselves on fluorescent genetic markers can be coupled to genetically barcoded cells to increase the power of multiplexing, as shown in our example with HIV-1 protease variants (Fig. 3). This is also true when coupled to assays that rely on staining protocols, as demonstrated by the cell-surface marker assay (Fig. 4). The power of multiplexed genetic barcoding is of great value for HTS applications as the number of screens required proportionally decreases with the number of genetic barcoded populations that compose the mixed sample under analysis. High-throughput assays as the ones described here are just two examples to demonstrate the value that multiplexing through genetic barcoding has to accelerate drug screening and biological assays with flow cytometry-based read outs. Figure 5 depicts the model of genetic barcoding for multiplexing, where analysis in different channels can reveal masked populations.

Figure 5.

Depiction of genetic barcoding for multiplexing, where masked populations can be revealed through decoding in a different channel. FCM: Flow cytometry.

While the high-throughput repeatability of the experiment relies on the instrument and its capacities, reproducibility of the experiment relies on the operator. Genetic barcoding thus decreases the variability introduced by the human factor. The inherited properties of genetically fluorescent barcoded cells make their detection and analysis quick, simplified, and straightforward. An individual genetic background represented by a specific fluorescent protein, distinguishes the barcoded cell from its counterparts. Increasing genetic barcoding-based multiplexing capabilities is only possible when certain criteria are met as described by Shaner et al. [61]. The growing number of existing fluorescent proteins and derivatives with distinct absorbance/emission spectra, combined with the growing number of affordable detection devices and lasers, increases the versatility of multiplexing, making fluorescent genetic barcoding a powerful tool for flow cytometry-based analysis. While we focused here on flow cytometry, multiplexed genetic barcoding can be further exploited for applications involving imaging-coupled flow-cytometry and microscopy.

Acknowledgments

We would like to thank Dr. Garry Nolan from Stanford University for providing the Phoenix GP packaging cell line for the production of retroviral particles. We thank Dr. Roger Tsien at UCSD for providing td Tomato. We would also like to thank the San Diego State University Flow Cytometry Core Facility for their service.

Ancillary