Contents

4 Tools and Workflows to Advance Antibody Discovery

As technologies and methods continue to evolve, more mAb therapeutics will
become available.

7 Functional Characterization of Monoclonal

Antibodies Using Kinetic Live-Cell Analysis
Novel workflow provides real-time data and actionable insights for antibody discovery.

16 Pharmacological Characterization of the Antibody Drug

Conjugate, Kadcyla®, Using Live-Cell Analysis
Advanced method facilitates analysis of pharmacological effects through
binding affinity, internalization, and cytotoxicity studies.

26 Real-Time 96-Well Antibody Internalization Assays

Using Incucyte® Fabfluor-pH Antibody Labeling Dye
Integrated assay solution addresses limitations of traditional methods.

36 Antibody Internalization: Advanced Flow Cytometry

and Live-Cell Analysis Give Rich Insights During
Antibody Profiling
Novel antibody screening method focuses on speed and insight.

46 Resources

2
More Colors. Simplifying Progress

More Insights.
More Possibilities.
See more information in every sample with
the new Incucyte® SX5 featuring patent-
pending optics. Do more with up to 5 colors
specifically designed for live-cell analysis.
www.sartorius.com/sx5
Specifications subject to change without notice. © 2022. All rights reserved. Incucyte and all names of Sartorius products
are registered trademarks and the property of Sartorius AG.
Tools and Workflows to Advance
Antibody Discovery
As technologies and methods continue to evolve,
more mAb therapeutics will become available.

Antibodies are ideal therapeutic candidates for However, despite these early successes, the ad-
many different medical disorders due to their un- verse side effects resulting from the production of
rivaled specificity and affinity for a broad range of anti-drug antibodies (ADAs) prompted researchers
targets. Following the U.S. Food and Drug Admin- to develop alternative mAb formats.
istration (FDA) approval in 1986 of the first ever
Chimeric monoclonal antibodies were among the
monoclonal antibody (mAb) therapeutic, over 100
next therapeutic mAbs to reach the market. Con-
mAbs have been approved for treating conditions
sisting of mouse mAbs in which the constant re-
including cancer, chronic inflammation, and infec-
gion is replaced by a human equivalent, these went
tious diseases. In this eBook, we look at the rise
some way toward addressing ADA production,
of antibody therapeutics and showcase some of
although the frequency of unwanted side effects
the tools and technologies that are empowering
remained problematic due to the largely “foreign”
antibody discovery. We also take a deep dive into
nature of these products. Humanized mAbs (anti-
how kinetic live-cell analysis is transforming mAb bodies in which only the complementarity-deter-
screening workflows. mining regions are non-human) showed greater
promise, and many continue to progress through
The rise of antibody therapeutics clinical trials today.
The very first monoclonal antibody therapeutic to Other types of antibody therapeutics that have
be approved by the FDA was a mouse mAb target- since entered development include antibody-drug
ing CD3, the T cell co-receptor. Termed muromon- conjugates (ADCs) for delivering cytotoxic drugs
ab-CD3, it was originally indicated for preventing directly to tumor cells; bispecific antibodies ca-
acute rejection after kidney transplantation (by pable of simultaneously binding two different
blocking cytotoxic T cell function) and was later ap- antigens to accomplish novel functionalities; and
proved for other types of solid organ transplants. antibody fragments, which offer improved tissue

4
penetration and fewer Fc-related effects compared Next, high-throughput binding and functional as-
to full-length antibodies. As technologies for anti- says are used to identify around 50 leading mAbs
body engineering and characterization continue for scale-up and further evaluation. Finally, a small
to evolve, further mAb therapeutics will become number of mAbs (<10) are humanized and sub-
available to patients. jected to more detailed characterization. Ultimate-
ly, the goal is for one or more of these humanized
Challenges for the mAb discovery mAbs to enter into clinical trials.
workflow In addition to ELISA, fluorescence-activated cell
sorting (FACS) and microscopy-based methodol-
A typical mAb discovery workflow begins with gen- ogies have long been central to mAb discovery
erating thousands of hybridomas against an anti- workflows. Antibody internalization is a key mech-
gen of interest. These are then assessed for target anism for evaluation, usually via a largely manual
recognition, often by enzyme-linked immunosor- process. This begins with labeling each antibody
bent assay (ELISA), which narrows down the pool with a fluorescent tag and removing any free label
to approximately 500–1,000 different candidates. via a column or wash step before incubating the

5
labeled antibodies with the cells. Once the incu-
Enabling real-time live-cell analysis
bation is complete, the cells are transferred to 4°C
(to prevent further antibody uptake) and washed Sartorius developed Incucyte® Live-Cell Analysis
multiple times before quantitation of the fluores- Systems to enable visualization and quantification
cent signal. of cellular behaviors in real-time within the envi-
Other types of assays that are commonly per- ronmentally stable conditions of a standard labo-
formed using these methods include monitoring ratory incubator. Notably, by moving the camera
of antibody-dependent cell-mediated cytotoxicity rather than the culture vessel, Incucyte® systems
(ADCC), antibody-dependent cellular phagocytosis eliminate the risk of perturbing sensitive cell types.
(ADCP), and antibody-driven immune cell killing. Although Incucyte® Live-Cell Analysis Systems
These involve incubating the relevant cell types to- are already widely used for applications such as
gether in the presence of each antibody and track- tracking cell proliferation and health, monitoring
ing the resultant events. An alternative approach advanced cell models including organoids, and
entails engineering one of the cell types to express performing functional assessment of immune cell
a reporter, whereby a luminescent readout (mea- killing, they are increasingly being integrated into
surable with a plate reader) directly correlates to mAb discovery workflows.
phagocytosis or cell killing. In a mAb discovery setting, Incucyte® Live-Cell
As well as having laborious and time-consum- Analysis Systems are used for applications that
ing workflows, a major limitation of the methods include quantifying the expression of cell surface
just described is that they report only end-point epitopes in real-time, characterizing ADCs, and
measurements. This prohibits researchers from performing high-throughput screening of anti-
fully investigating a mAb’s therapeutic potential. body internalization. To further support the latter
For example, being able to track mAb internaliza- application, Sartorius also offers Incucyte® Fabflu-
tion over time is critical for determining whether or-pH Antibody Labeling Reagents for conjugating
a mAb is likely to achieve its intended purpose. fluorescent, pH-sensitive probes to test mAbs via
While rapid internalization is desirable for ADCs, a one-step, no-wash protocol. At neutral pH, each
which must efficiently deliver a cytotoxic pay- Fab-mAb complex has little or no fluorescence.
load, inducing ADCC requires that the antibody But, as the complex is internalized and processed
remains at the cell surface. via acidic lysosomes and endosomes, a fluorescent
signal is observed and can be quantified using In-
For these reasons, platforms that enable real-time
cucyte® real-time live-cell analysis.
live-cell analysis—and are capable of measuring
parameters such as how much mAb is bound to In this eBook, we share real-life examples of how
the target cells, how long the mAb remains at the Incucyte® Live-Cell Analysis Systems and Fabflu-
cell surface, and how quickly the mAb is inter- or-pH Antibody Labeling Reagents are being used
nalized—are fast becoming the go-to option for for empowering antibody discovery. Read on to
mAb discovery. find out more.

6
Functional Characterization
of Monoclonal Antibodies Using
Kinetic Live-Cell Analysis
Novel workflow provides real-time data and actionable insights for
antibody discovery.

by Nicola Bevan, Gillian Lovell, Jasmine Trigg, Kirsty McBain, and Kalpana Barnes

Introduction
Antibodies are the ideal therapeutic biomole-
cules due to their ability to bind to a variety of
targets with high affinity and specificity. Innova-
tions in hybridoma and phage display technol-
ogies have allowed drug developers to exploit Figure 1: The Antibody Discovery Screening Pathway.
the natural diversity in antibodies, leading to Highlighting the use of both binding and functional
the wide use of monoclonal antibodies (mAbs) analysis for the characterization of leads in antibody
in treating cancers, chronic inflammation, and discovery and development.
infectious diseases.
played a major role in antibody screening work-
The antibody development workflow begins by
flows. However, these approaches are labor inten-
identifying thousands of candidates against a tar-
get of interest (Figure 1). Mutagenesis is used to sive, time consuming, and report only data “snap-
create large antibody libraries for the purpose of shots” in the form of end-point measurements.
optimizing specific properties, such as binding or Data-rich screening technologies that help scien-
specificity. Next, candidates are characterized us- tists identify the top leads early are critical to the
ing high-throughput binding and functional assays antibody development process.
to identify the top leads for further development.
We have discussed at length the capabilities
Traditionally, fluorescence-activated single cell of our advanced flow cytometry platform for
sorting (FACS), enzyme-linked immunosorbent as- high-throughput, multiplexed analysis of antibod-
say (ELISA), and microscopy methodologies have ies in a separate publication.1 Here, we demon-

7
strate how an Incucyte® Live-Cell Analysis System complex cell models have become more popular in
provides real-time data and actionable insights pre-clinical research.
in antibody discovery workflows, including an-
tibody internalization, antibody-dependent cel- Incucyte® Live-Cell Analysis Systems are well-es-
lular phagocytosis (ADCP), antibody-dependent tablished platforms for visualizing and quantifying
cell-mediated cytotoxicity (ADCC), and anti- cell behavior over time at an industrial scale. Each
body-driven immune cell killing. instrument includes a microscope that resides in-
side a standard tissue culture incubator for full en-
Kinetic Live-Cell Analysis vironmental control. It can auto- matically capture
and analyze time-lapse images from up to 2,304
Popular end-point techniques such as FACS, ELI- assay wells in parallel (6 x 384-well plates) over
SA, and microscopy are not amenable to detailed pre-determined time intervals, enabling continu-
kinetic analysis of time-dependent processes, such ous monitoring of cells. A wide variety of integrat-
as internalization and phagocytosis, and have ed application solutions (software, reagents, proto-
some limitations depending on the application. cols) are available, including assays for apoptosis,
immune cell killing, phagocytosis, and 3D tumor
FACS, for example, can be used to sort hundreds
growth and viability.
of thousands of cells, but it cannot be used to per-
form high- throughput, time-dependent studies We will describe how the Incucyte® Live-Cell Anal-
on individual cells, as the cells must be sorted into ysis Systems and validated assays can be used to
single colonies first.2 ELISA has historically been provide real-time kinetic insights about antibody
used to screen hybridomas and other libraries for function and cell behavior during the antibody tri-
binding of antibodies to their targets.3 However, age process.
this approach poses challenges for screening anti-
bodies that bind to cell surface antigens because Cell Surface Expression of Epitopes
only purified versions of the antigens can be ad-
sorbed to the ELISA plate, potentially altering the While other methods, like advanced flow cytome-
conformation of its epitope. The assay also does try, are ideal for binding studies in antibody devel-
not report on how the therapeutic antibody affects opment workflows, live-cell analysis can be used to
the health, behavior, and function of the test cells. quantify the expression of cell surface epitopes.

Microscopy is a powerful technique for localiza- Immunocytochemistry (ICC) is a powerful labora-

tion and temporal studies in single cells and offers tory imaging technique for visualizing the location
a more thorough functional analysis of a smaller of specific proteins in cells using targeted fluores-
set of antibody candidates. Minimally invasive as- cently labeled antibodies. However, conventional
says that capture the behavior of cells under more ICC is less informative for tracking the distribution
natural conditions are the best predictors of how a and abundance of proteins in real time and in liv-
therapeutic might perform in a physiological set- ing cells, and the subsequent linking of these ob-
ting. This is partly why non-perturbing assays and servations to cell morphology and function.

8
Here, we used a live-cell ICC assay, developed on expression of macrophage cell surface markers
an Incucyte® Live-Cell Analysis System, for dynamic CD11b (integrin αM, part of the complement C3
monitoring of surface proteins in living cells over receptor), CD14 (an endotoxin receptor), and CD40
several days. (a co- stimulatory protein) using the Incucyte® Fab-
fluor-488 Antibody Labeling Dye complexed to an-
In this protocol, an antibody targeting the protein
of interest is first tagged with an isotype-matched tibodies for those proteins. Time-lapse images of
Fc-targeted fluorescently labeled antigen binding cellular fluorescence were gathered continuously
fragment (Fab) via a simple mix, no wash protocol. over days, and automatically analyzed to provide
Next, this complex is added directly to living cells the levels and pattern of expression over time.
and used to visualize the expression of proteins on
The hTHP-1 cells showed little marker expres-
cells on the Incucyte® Live-Cell Analysis System.
sion for up to 72 hours under control conditions
We used live-cell ICC to probe the differentia- (Figure 2A). Differentiation agents—vitamin D3
tion of human monocytic THP-1 cells into mac- and phorbol myristate acetate (PMA)—caused a
rophage-like cells (Figure 2).4 We quantified the notable upregulation of CD markers to different

Figure 2: Tracking the Differential Expression of Surface Markers. Monocyte to macrophage differentiation was captured using
hTHP-1 cells exposed to media (undifferentiated), vitamin D3 (50 nM), or PMA (100 nM) in the presence of Incucyte® Fabfluor-488
Dye complexed to CD11b, CD14, or CD40 Abs (+ Incucyte® Opti-Green). Incucyte® images (top, 20X, every three hr) were analyzed for
marker expression (green fluorescent area), normalized for confluency. Data shown as mean ± SEM, n = 4 wells.

9
extents over 12–72 hours. Most notably, PMA up- Following a 12-hour incubation with AU565
regulated CD11b and CD40, but not CD14, while vi- (HER2-positive) target cells, we observed orange
tamin D3 increased expression of CD11b and CD14 fluorescence (internalization) with both Kadcyla®-
in a time-dependent manner. Vitamin D3 did not and trastuzumab-treated cells, which is expected
upregulate CD40. Based on the Incucyte® images, as they both recognize the same epitope (Figure
only PMA caused a change in morphology, pro- 3A, data not shown). Both antibodies showed a
ducing large, flattened, adherent macrophage-like concentration-related effect with similar EC50 val-
cells. Interestingly, PMA, but not control or vitamin ues after 24 hours (Figure 3B and C). No internaliza-
D3, also yielded a decrease in cell proliferation tion was observed for the isotype immunoglobulin
(confluence) associated with macrophage pheno- G (IgG) control.
type (data not shown). The DM1 payload attached to Kadcyla® is a tubu-
lin inhibitor that results in cell cycle arrest and cell
Characterization of death. Next, we quantified the pharmacological
Antibody-Drug Conjugates efficacy of Kadcyla® by monitoring cell health. In
this assay, AU565 cells were transduced with Incu-
Antibody-drug conjugates (ADCs) take advantage of cyte® Nuclight Green Lentivirus to express a nucle-
the specificity of mAbs to deliver cytotoxic payloads ar-restricted green fluorescent protein. Using the
directly to tumor cells expressing a specific marker. integrated Incucyte® Basic Analyzer software, we
The mechanism of action of ADCs involves two stag- quantified the number of green nuclei present as
es: (1) the binding of the antibody to the cell-specif- an indicator of cell proliferation and health.
ic surface protein, and (2) the internalization of the As expected, cells treated with Kadcyla® appeared
entire molecule, releasing cytotoxic drugs inside the unhealthy and stopped proliferating (reduced
cell and leading to targeted cell death. green fluorescence), while cells treated with either
trastuzumab or the isotype IgG control resumed
We investigated the internalization and cytotox-
normal growth (Figure 3D, data not shown). We ob-
ic effects of the clinically approved breast cancer
served a concentration-dependent decrease in nu-
drug Kadcyla®, a combination ADC of trastuzum-
clear count over time in cells treated with Kadcyla®,
ab (Herceptin®), an antibody against the human
indicating poor cell health (Figure 3E). We did not
epidermal growth factor receptor 2 (HER2), and measure any effect on cell number in the presence
emtansine (DM1), a cytotoxic agent.5 We tracked of trastuzumab or the isotype control IgG, suggest-
the localization of Kadcyla® and trastuzumab ing then cytotoxic affects are a result of the DM1
alone with the Incucyte® Fabfluor-pH Orange An- payload (Figure 3F).
tibody Labeling Dye, which is a Fc-targeted Fab,
labeled with a pH-sensitive dye. Once bound to Real-Time Analysis of Phagocytosis
the epitope, the complex is internalized where it
encounters the low-pH environment of the lyso- Phagocytosis is a critical component of innate and
some, causing the dye to fluoresce. The fluores- adaptive immune responses by which micro-or-
cence is visualized and quantified using the Incu- ganisms and unwanted cells are internalized, de-
cyte® Live-Cell Analysis System. stroyed, and removed. This process can be trig-

10
 Figure 3: Characterization of an Antibody-Drug Conjugate. Incucyte® Nuclight Green–expressing AU565 cells were treated with either
Kadcyla®, trastuzumab or isotype IgG complexed to Incucyte® Fabfluor-pH Orange. Images (A, 12 hr) depict the orange fluorescent
signal in cells displaying internalization. The time course (B) shows the quantification of this internalization signal and potency
compared to trastuzumab (C). The isotype control is shown in teal (B), dotted line (C). Viability was also assessed, by quantifying the
green fluorescence. Images (D, 72 hr) show the health of cells post treatment. The time-course (E) shows the quantification of cell
numbers after Kadcyla® treatment and compares it to trastuzumab (F). The isotype control is shown in teal and camptothecin (CPT)
cytotoxic control in black. Potency of effect of Kadcyla® (F) is quantified after 72 hr. Data shown as mean ± SEM, n = 3 wells.

gered using mAbs to promote clearance of viable The anti-CD20 mAb rituximab is known to induce
tumor cells via a process called antibody-depen- destruction of certain cancer cells via ADCP. We
dent cellular phagocytosis (ADCP).6 used the Incucyte® Phagocytosis Assay to per-
form a high- throughput study of engulfment us-
Standard approaches to studying phagocytosis
commonly rely on end-point assays like flow cy- ing native and engineered rituximab Fc mutants.
tometry, which offer limited kinetic information. We treated pHrodo®-labeled Ramos cells with in-
Additionally, the protocols are destructive to cells, creasing concentrations of native and engineered
preventing visual assessment of morphological human rituximab isotypes: anti-CD20 IgG1, an-
changes. The non-invasive Incucyte® Phagocyto- ti-CD20 IgG1 Fut (non-fucosylated Fc Mut), and an-
sis Assays use pH-sensitive pHrodo® dyes and in- ti-CD20 IgG1 NQ (non-glycosylated Fc Mut). Next,
tegrated image-based analysis tools for real-time we co-cultured them with bone marrow derived
quantification of phagocytosis. macrophages (BMDMs).

11
The microplate view provides easy visualization of clonal expansion and activation of T cells as well
engulfment across all wells in real time, as indicated as direct contact between CD3-positive T cells and
by an increase in fluorescence (Figure 4A). We ob- CD19-positive tumor cells, resulting in tumor-spe-
served a concentration-dependent ADCP response cific cell lysis.
for anti-CD20 IgG1 mAb and Fc mutated IgG1 Fut,
but not IgG1 NQ or IgG1 control wells (Figure 4B). We investigated the cytotoxicity induced by a CD3
This aligns with previous observations that anti- x CD19 BiTE antibody using an Incucyte® Live-Cell
body glycosylation is essential for Fc-receptor-me- Analysis System. We incubated Ramos target cells
diated effector function. stably expressing a nuclear targeted GFP (Incucyte®
Nuclight Green Lentivirus) with anti-hCD3 x CD19
Immune-Cell Killing of Tumor Cells (Invivogen, Cat. No. bimab-hCD19CD3) or control
antibody targeting CD3 alone (Ctrl Ab 1: anti- hCD3
Bispecific T cell engager (BiTE) antibodies are a x βGAL). Next, we co-cultured with PBMCs.
type of immunotherapy that recruit T cells to tu-
mor cells.8 The CD3 x CD19 BiTE antibody (Blina- We analyzed the green fluorescent intensity on the
tumomab), which engages the CD3 on T cells and Incucyte® as a measure of target cell death. Cell
the tumor-associated antigen CD19 on B cells, is images at the 72-hour time point showed healthy
approved for treatment of B-lymphocytic leuke- proliferation of cells treated with the control Ab 1
mia and other late-stage cancers. It causes the (anti-hCD3 x βGAL), but a reduction in cell growth

Figure 4: High-Throughput Quantification of ADCP. pHrodo®-labeled Ramos cells were treated with native and engineered human
rituximab isotypes anti-CD20 IgG1, anti-CD20 IgG1 Fut (non-fucosylated Fc Mut), and anti-CD20 IgG1 NQ (non-glycosylated
Fc Mut), and co-cultured with BMDM cells (1:2.5K effector-to-target ratio). Microplate graph (A) shows kinetic response for
phagocytosis, total red object area (μm2/ image) over 4 hr. Data (B) demonstrates a concentration-dependent ADCP response
for anti-CD20 IgG1 mAb and Fc mutated IgG1 Fut, but not IgG1 NQ or IgG1 control. This aligns with previously observed effector
functions (table).7 Data shown as mean ± SEM, n = 3 wells.

12
for the BiTE antibody-treated cells (Figure 5A and vanced Flow Cytometry Platform. The CD3 x CD19
C). The BiTE antibody elicited a concentration-de- BiTE antibody induced expression of IFNγ and TNFα
pendent cytotoxicity of the Ramos cells (IC50 val- in a concentration-dependent manner, reaching
ue of 0.09 ng/mL), with the highest concentration maximal concentrations of 1.3 ± 0.3 ng/mL and 1.5
of BiTE antibody causing a 93% decrease in target ± 0.3 ng/mL, respectively (Figure 5E and F). Inter-
cell viability compared to control Ab 1 (Figure 5C estingly, CD3/CD28 Dynabead™ stimulation led to
and D). vastly increased production of both cytokines. De-
spite the high levels of target cell death, the BiTE
We also removed supernatant samples daily and antibody induced relatively low levels of cytokines
measured cytokine production on the iQue® Ad- compared to CD3/CD28 stimulation (Figure 5G).

Figure 5: Quantification of CD3 x CD19 Bispecific Antibody Driven Immune Cell Killing. Ramos cells labeled with Incucyte® Nuclight
Green Lentivirus were seeded (15K/well) with PBMCs (1:5 target to effector) and activated with BiTE antibody (anti-hCD3 x CD19,
0.6 pg/mL to 10 ng/mL) or control antibody (anti-hCD3 x βGAL, 10 ng/mL). Representative images at 72 hr in the presence of (A)
control or (B) BiTE antibody. Incucyte® time-course data tracked the target cell quantity (C). The fold change in green integrated
intensity metric, which accounts for both brightness and size (GCU x μm2), was used to quantify target cells. Concentration-response
curves at 72 hr for target cell killing (D) are shown in the presence of the BiTE antibody compared with the control antibody (teal).
Temporal cytokine production was also quantified, activated either with increasing concentrations of BiTE antibody or CD3/CD28
Dynabeads™ (75K/well). Daily supernatant samples were (10 μL) taken for analysis of cytokine (IFNγ and TNFα) concentrations by
the iQue® (E and F). Temporal cytokine concentrations were compared to the level of immune cell killing as measured by fold change
in green integrated intensity over time (G). Data shown as mean ± SEM, n = 3 wells.

13
 assess ADCC within tumor multi-spheroids. The en-
Quantification of ADCC in Multi-
hanced depth of focus brightfield (DF brightfield)
Spheroid (MS) Model image acquisition enables long-term imaging of
multiple tumor spheroids grown on a bed of extra-
Scientists are increasingly choosing multicellular
cellular matrix (Matrigel®).
tumor spheroids for oncology and immuno-oncol-
ogy research as they provide more predictive data We evaluated the impact of Herceptin®-induced
compared to 2D monolayer cell models.9 Current ADCC in HER2-positive (BT474) and HER2-neg-
methods for assessing the morphology of tumor ative (MCF7) multi-spheroids. We co-cultured
spheroids are often time-consuming and expen- tumor multi-spheroids, either stably expressing
sive, or they involve end-point readouts that can
nuclear restricted RFP (Incucyte® Nuclight Red
miss time-dependent insights.
Lentivirus) or cytoplasmic GFP (Incucyte® Cytoli-
We used an Incucyte® Live-Cell Analysis System ght Green Lentivirus), with freshly isolated periph-
and Incucyte® 3D Multi-Tumor Spheroid Assays to eral blood mononuclear cells (PBMCs) (E:T, 5:1)

Figure 6: Quantification of Trastuzumab Driven ADCC Using an

Advanced Cell Multi-Spheroid Model. Impact of trastuzumab-
induced PBMCs on MS proliferation. (A) Incucyte® brightfield
and fluorescence images taken at seven days (MCF-7) or ten
days (BT-474) show the effect on spheroid proliferation in the
absence (top panel) and presence (bottom panel) of PBMCs.
The brightfield outline mask is shown in yellow.

(B) Time-courses show multi-spheroid death quantified as a

loss of fluorescence intensity within the spheroid brightfield
object. Increased cytotoxicity was seen with the addition of
treatments activating T cell populations (anti-CD3 and IL-2).
Data was collected over ten days at six hr intervals. Data shown
as mean ± SEM, n = 4 wells.

14
and treated with Herceptin® (trastuzumab). We
References
quantified multi-spheroid proliferation and ADCC
kinetically using fluorescence intensity within 1. High throughput, multi-parametric analysis accelerates
brightfield boundary, which does not require antibody discovery workflows. White paper. Published June
masking of fluorescent tumor multi-spheroids. 24, 2022. Accessed July 15, 2022. www.sartorius. com/en/
products/flow-cytometry/flow-cytometry- resources/white-
Based on change in fluorescence intensity, Her- paper-high-throughput-multi- parametric-analysis-accelerates-
ceptin® alone had little or no effect on the size or vi- antibody-discovery- workflows
ability of multi-spheroids (Figure 6A). However, Her-
2. Doerner A, Rhiel L, Zielonka S, Kolmar H. Therapeutic antibody
ceptin®-activated PBMCs, showed a marked loss of engineering by high efficiency cell screening. FEBS Lett.
viability in HER2-positive spheroids. Herceptin® in- 2014;588(2):278-287. doi:10.1016/j.febslet.2013.11.025
duced a concentration-dependent, cytotoxic effect
3. Saeed A, Wang R, Ling S, Wang S. Antibody engineering
in HER2-positive multi-spheroids, but not HER2-neg-
for pursuing a healthy future. Front Microbiol. 2017;8(495).
ative multi-spheroids (Figure 6B). The highest test doi:10.3389/fmicb.2017.00495
concentration of Herceptin® (1 μg/mL) produced ap-
proximately 80% growth inhibition (Figure 6B, con- 4. Aldo PB, Craveiro V, Guller S, Mor G. Effect of culture conditions
centration-response curve). Activation of the PBMC on the phenotype of THP-1 monocyte cell line. Am J Reprod
Immunol. 2013;70(1):80-86. doi: 10.1111/aji.12129
T cell population with an anti-CD3 antibody and IL-2
cocktail (10 ng/mL of each) resulted in maximal kill- 5. von Minckwitz G, Huang CS, Mano MS, et al. Trastuzumab
ing across both spheroid models. emtansine for residual invasive HER2- positive breast cancer. N
Engl J Med. 2019;380(7):617- 628. doi:10.1056/NEJMoa1814017

Conclusion 6. Kim D, Wang J, Willingham SB, et al. Anti-CD47 antibodies

promote phagocytosis and inhibit the growth of human
Cell analysis is a powerful technique for evaluating myeloma cells. Leukemia. 2012;26(12):2538-2545. doi: 10.1038/
the pharmacological efficacy of therapeutic anti- leu.2012.141
bodies and monitoring for effects on cell health, 7. InvivoGen. Reagents and tools for cell biology research.
behavior, and morphology. Importantly, systems Accessed July 15, 2022. www.invivogen.com/anti-hcd20-
that allow for unperturbed analysis in real time higg1nq
provide physiologically relevant insights that bet- 8. Ross SL, Sherman M, McElroy PL, et al. Bispecific T cell engager
ter translate to behavior in the clinic. (BiTE®) antibody constructs can mediate bystander tumour cell
killing. PLoS ONE. 2017;12(8:e0183390. doi: 10.1371/journal.
We have demonstrated the utility of an Incucyte® pone.0183390
Live- Cell Analysis System for functional assess-
ment of leads in high-throughput antibody devel- 9. Verjans ET, Doijen J, Luyten W, Landuyt B, Schoofs L. Three-
dimensional cell culture models for anticancer drug screening:
opment workflows. The system combines non-per-
worth the effort? J Cell Physiol. 2018; 233(4);2993-3003. doi:
turbing cell visualization and data acquisition, 10.002/jcp.26052
with integrated analysis to provide kinetic insights
about mechanism of action and other effects.
About the authors
Importantly, validated Incucyte® reagents and as-
says streamline workflows for routine studies in Nicola Bevan, Gillian Lovell, Jasmine Trigg, Kirsty
both 2D and complex 3D cell models. McBain, and Kalpana Barnes are with Sartorius U.K.

15
Pharmacological Characterization
of the Antibody Drug Conjugate,
Kadcyla®, Using Live-Cell Analysis
Advanced method facilitates analysis of pharmacological effects
through binding affinity, internalization, and cytotoxicity studies.

by Nicola Bevan, Kirsty McBain, Daryl Cole and Tim Dale

Introduction
In recent years, increasing development in antibody
therapeutics has led to the discovery of new treat-
ments for multiple diseases. Of particular interest
in the cancer therapy field is the development of
antibody-drug conjugate (ADC) complexes. ADCs
exploit the specificity of monoclonal antibodies
(mAbs) to achieve targeted cell death by delivering
a cytotoxic drug directly to tumor cells expressing
a characteristic marker or antigen.

A few examples of clinically relevant ADCs are

Kadcyla®, Mylotarg®, Blenrep® and Zynlonta™, Figure 1: Kadcyla® Molecule
which have been used in the treatment of breast,
leukemia, multiple myeloma and lymphoma can- There are multiple mechanisms of action (MoA)
cers.1-4 Kadcyla®, clinically used against some forms for Kadcyla®: (1) the selective delivery of DM1 to
of breast cancer, is a combination ADC of trastu- HER2 positive cancer cells, (2) the trastuzumab an-
zumab (Herceptin®), an antibody targeting the hu- tibody-mediated inhibition of HER2 signalling, (3)
man epidermal growth factor receptor 2 (HER2), the inhibition of HER2 extracellular domain shed-
and emtansine (DM1), a cytotoxic agent (Figure 1). ding and (4) the induction of antibody-dependent

16
cell-mediated cytotoxicity (ADCC).5 The first MoA Live-Cell Analysis System. Test antibodies (Kadcyla®,
can be described in two key stages: the effective trastuzumab, or an isotype control IgG) and Incu-
binding of the antibody to HER2 resident on the cyte® Fabfluor-pH Orange Dye were incubated to-
target cell surface, and the internalization of the gether (15 min, 37 °C, 12 µg/mL) at a 1:3 molar ratio,
entire molecule releasing cytotoxic DM1 within which resulted in the formation of an antibody-Fab-
the cell, leading to targeted cell death. The cyto- fluor-pH complex. AU565 (HER2+) or MDA- MB-231
cells (HER2 low) were seeded overnight in 96-well
toxic component of the ADC is completely reliant
plates before the serially diluted antibody complex
on the binding affinity and specificity of the anti-
(1 in 2) was added to wells (final assay concentration
body component. For this reason, it is important
6 µg/mL to 3 ng/mL) and imaged using the Incu-
to analyze the pharmacological effects of Kadcy-
cyte® SX5 Live-Cell Analysis System.
la®, through binding affinity, internalization, and
cytotoxicity studies to characterize its efficacy as a Cell health was assessed using AU565 cells stably
clinical treatment agent. This document exempli- expressing a nuclear restricted green fluorescent
fies how the various biological activities of Kadcy- protein (Incucyte® Nuclight Green Lentivirus). Cells
la® can be characterized using the Incucyte® Live- were seeded overnight in 96-well plates before
Cell Analysis System in combination with multiple Kadcyla® or trastuzumab (6 µg/mL to 3 ng/mL)
functional assessments. were added to the well. Isotype control IgG (3 µg/
mL) and the cytotoxic drug camptothecin (1 µM)
Methods were included as negative and positive controls for
the assay. Images were collected by the Incucyte®
The internalization of Kadcyla®, along with its ef- and Green Nuclear Count used as a readout for cell
fects on cell health and cell cycle, were fully charac- numbers in the well.
terized using Incucyte® live-cell imaging and anal-
Effects on cell cycle were assessed using AU565
ysis; trastuzumab was included for comparison. For
cells stably expressing Incucyte® Cell Cycle Lenti-
all studies Kadcyla® was obtained from Midwinter
virus. This indicator has been developed to distin-
Solutions and trastuzumab was purchased as a re-
guish between cells in the G1 and S | G2 | M cell
search grade biosimilar (Absolute Antibody).
cycle phase without altering cell function. Cells
The internalization of both Kadcyla® and trastuzum- fluoresce green during S | G2 | M and red or or-
ab was assessed using the Incucyte® Human Fabflu- ange (depending on construct used) during G1;
or-pH Antibody Labeling Dye. This reagent compris- cells are colorless during the transition from M to
es a pre-conjugated antigen-binding fragment (Fab) G1 and yellow (expressing green and red, or green
against the Fc region of IgG, labeled with a pH-sen- and orange, simultaneously) in transition from G1
sitive dye and enables easy conjugation to an anti- to S phase. Cells were seeded overnight in 96-well
body of interest. Upon binding to the specific epi- plates before Kadcyla® (6 µg/mL to 3 ng/mL) was
tope, the complex is internalized in the cell where it added to the wells. An IgG isotype control (3 µg/
enters the low pH environment of the lysosome. mL) was added as a negative control.

The change in pH causes the dye to fluoresce, which ADCC was assessed as a co-culture of AU565 Nu-
can be visualized and quantified on the Incucyte® clight Green cells and pre-isolated natural killer

17
 was incubated with HER2 positive breast carci-
noma AU565 cells. The images in Figure 3A show
that after incubation (12 h) of cells with the labeled
antibody, an orange fluorescence can be seen for
both Kadcyla® and trastuzumab treatments but
not in IgG isotype control-treated cells. Quantifi-
cation of the response using orange fluorescence
area normalized for cell phase area demonstrates
that Kadcyla® is rapidly internalized in AU565 cells,
reaching a plateau in < 6 h for the maximal con-
centration tested (6 µg/mL, Figure 3B). A similar in-
ternalization profile was measured for trastuzumab
in AU565 cells. Results show a concentration-relat-
Figure 2: iQue® Human NK Cell Killing Kit
ed effect producing similar EC50 values after 24 h
(Kadcyla® 0.38 µg/mL = 2.48 nM, trastuzumab 0.22
cells (StemCell Tech) using a 1:5 target to effector
µg/mL = 1.42 nM, Figure 3C). Demonstrating the
cell ratio in the presence of Kadcyla® (6 µg/mL to specificity of the response, minimal internal- iza-
3 ng/mL). Incucyte® Annexin V NIR Dye was also tion was detected for either antibody complex
added for quantification of target cell apoptosis. when incubated with MDA-MB-231 cells (low HER2
Images were collected by Incucyte® and ADCC expressing by flow, Figure 3D).
was quantified as the loss of target cell green nu-
clei over time. Replicate plates using AU565 wild Quantification of Cell Health
type cells were set up for further analysis using the The DM1 payload attached to Kadcyla® is a tubulin
iQue® Advanced Flow Cytometry Platform utilizing inhibitor that inhibits the assembly of microtubles
the iQue® Human NK Cell Killing Kit to assess acti- causing disruption of the cell cycle resulting in mi-
vation marker expression on the NK cells and se- totic arrest and then death. Upon internalization of
creted cytokine levels (Figure 2). Kadcyla® into lysosomes, the DM1 is cleaved and
released into the cells where it can take effect. To
Characterization of Kadcyla® in measure this effect, AU565 cells transduced with
HER2 Positive Monocultures Incucyte® Nuclight Green Lentivirus to express a
nuclear restricted green fluorescent protein were
Pharmacological, Kinetic Quantification treated with either Kadcyla® or trastuzumab and
of Antibody Internalization imaged using Incucyte® Live-Cell Analysis System.
The images below (Figure 4A) clearly show that
As Kadcyla® and trastuzumab share the same epi- after 72 h, cells treated with Kadcyla® (1.5 µg/mL)
tope recognition, specificity, and base antibody appear unhealthy and are no longer proliferating
construction, it is expected they should have sim- while cells treated with either trastuzumab (1.5
ilar internalization profiles when mixed with HER2 µg/mL) or an isotype control IgG (3 µg/mL) look
positive target cells. This was investigated using an healthy and continue to proliferate, as expect-
antibody–Fabfluor-pH sensitive complex, which ed. Analysis of the images using the integrated

18
 Figure 3: Internalization Profiles

Note. AU565 cells were treated with either Kadcyla®, trastuzumab or isotype IgG complexed to Incucyte® Fabfluor-pH Orange.
Images (A) show a fluorescent signal in cells treated with either Kadcyla® or trastuzumab but not IgG. The time course (B) shows a
rapid concentration related internalization signal for Kadcyla®, which has a similar potency to trastuzumab (C). No internalization
is seen with the isotype control (teal [B], dotted line [C]). The bar graph (D) demonstrates the specificity of the signal with no signal
measured in HER2 low MDA-MB-231 cells. All data shown as the mean ± SEM for 3 wells.

Incucyte® Basic Analyzer software enables quan- The potency of the effect of Kadcyla® on cell
tification of the number of green nuclei present health was quantified to be 0.24 µg/mL, similar to
in the well as an indicator of cell proliferation and the potency of internalization and shows a com-
health. Those cells treated with Kadcyla® demon- parable maximal effect to camptothecin. Similar
strate a rapid concentration-dependent decrease quantification of cell apoptosis with Kadcyla® was
in nuclei numbers over time (Figure 4B) indicat- measured using the cell heath indicator Incucyte®
ing poor cell health. No effect on cell number was Annexin V Dye (0.30 µg/mL, data not shown).
measured in the presence of trastuzumab (Figure
Effect of Kadcyla® on Cell Cycle
4C) or the isotype control IgG. Camptothecin (1
µM), a cytotoxic control compound, was included The effect of DM1 on cell cycle can be further in-
to act as positive control for maximum cell death. vestigated using cells expressing the Incucyte® Cell

19
 Figure 4: Effects on proliferation

Note. Nuclight Green-expressing AU565 cells were treated with either Kadcyla®, trastuzumab, isotype IgG or camptothecin (CPT).
Images (A) show the poor health of cells treated with Kadcyla® and lack of effect of trastuzumab or IgG. The time course (B) shows
a rapid concentration related decrease in cell numbers after Kadcyla® treatment while trastuzumab (C) has no effect. The isotype
control (teal line) has no effect on cell proliferation while CPT (black line) causes maximal cell death. Potency of effect of Kadcyla® (D)
is quantified after 72 h. All data shown as the mean ± SEM for 3 wells.

Cycle indicator, in combination with the Incucyte® yellow (G1 | S, 15 ± 2%) or colorless (M | G1, 20 ±
Cell-by-Cell Analysis Software Module. This soft- 4%) cells. Images (Figure 5A) and quantification
ware enables segmentation of individual cells and (Figure 5C) of AU565 cells treated with Kadcyla® (3
quantification of their fluorescence so sub-popula- µg/mL) show a shift over the first 24 h, with a loss
tions can be identified and tracked. of orange cells (down to 13%) and an increase in
colorless cells (up to 27%), indicating a stalling of
Under normal conditions, AU565 cells expressing the cycle in the M | G1 phase. At 36 h, the image
the Incucyte® Green | Orange Cell Cycle Lentivirus shows mainly green (24%) and non-fluorescent
display a mix of cell cycle phases shown as a mix of cells (52%) which, coupled with their increasing-
green (S | G2 | M, 40 ± 3%), orange (G1, 24 ± 2%), ly rounded morphology, indicates that the stall is

20
around the mitosis stage of the cycle. By 48 h, the combination with direct killing) was compared
cells look unhealthy and are dying. There is little to its direct cytotoxic effects in monoculture. To
effect on the populations with the isotype control quantify this, AU565 cells expressing a nuclear re-
IgG (Figure 5B and 5D) or with trastuzumab (data stricted green fluorescent protein (Incucyte® Nu-
not shown). clight Green Lentivirus), were either treated with
Kadcyla® in monoculture or as a co-culture with
Effect of Kadcyla® in an
pre-isolated natural killer cells (1:5, target:effec-
ADCC Co-Culture Model
tor ratio). Incucyte® Annexin V NIR was added to
Kadcyla® has been designed to combine the tar- define target cell apoptosis. The images (Figure
geting ability of trastuzumab with the delivery of 6A) display a concentration-related decrease in
a cytotoxic payload to specific cells. To demon- green fluorescence in both the monoculture and
strate the combined power of Kadcyla®, its ability co-culture wells, indicating loss of target cells with
to kill target cells through conventional ADCC (in Kadcyla® treatment. It is clear from the images that

Figure 5: Effect of Kadcyla® on Cell Cycle

Note. AU565 cells expressing the Incucyte® Cell Cycle indicator were treated with Kadcyla® or isotype IgG. Series of images (A)
depicting the changes in cell cycle phases after treatment with Kadcyla® or IgG (3 µg/mL, B). The time course graphs (C and D)
display the change in populations post-treatment. All data shown as the mean ± SEM for at least 3 wells.

21
the additional ADCC effect in the presence of NKs is directly compared to the effect in monoculture a
more potent than the direct cytotoxic effect alone. clear difference can be seen (Figure 6C). The kill-
The pseudo-colored blue fluorescence in the imag- ing of target cells occurs faster through ADCC and
es indicates the Annexin V response and shows a is some 25-fold more potent (EC50 direct cytotoxic
concentration-related increase in the presence of
0.27 µg/mL compared to 0.011 µg/mL for ADCC).
Kadcyla®. There is no effect seen with the isotype
control IgG in either assay setup. Quantification of Similar data was reported following quantification
the green cell fluorescence in the images confirms of the Annexin V signal (EC50 direct cytotoxic 0.38
the rapid, concentration-related killing of target µg/mL compared to 0.012 µg/mL for ADCC, graphs
cells (Figure 6B). When the effect in co-culture is not shown).

Figure 6: Enhanced effect of Kadcyla® with ADCC

Note. Nuclight Green-expressing AU565 cells were tested as a monoculture or as a co-culture with NK cells in a 1:5 target:effector
cell ratio. Cells were treated with either Kadcyla® or isotype IgG in the presence of Incucyte® Annexin V NIR Dye. Images (A) depict the
concentration-related decrease in green target cells and increase in Annexin V signal (blue, 72 h) in both assay formats. The effect
is enhanced in the presence of NK cells (ADCC). The time course (B) demonstrates the quantification of this ADCC response. When
directly compared, there is an enhanced target cell killing effect due to ADCC in the co-culture compared to monoculture (C). The
isotype control has no effect on cell death (A, B and C). All data shown as the mean ± SEM for at least 3 wells.

22
The data at 96 h suggests an improved clearance of µg/mL) indicating activation of the NK cells (Figure
target cells in the ADCC co-culture model, which is 7A). Kadcyla® also induced a reduction in the levels
in line with Kadcyla’s® high clinical efficacy for tu- of CD16 on the NK cells (Figure 7B). This shedding
mor cell clearance. event was enhanced after 48 h compared to 24 h
with a slight shift in mid-point determination (EC50
To gain further insight on the biological action of
24 h 0.17 µg/mL, 48 h 0.07 µg/mL).
Kadcyla® in the ADCC assay, replicate plates were
set up with wild type AU565 cells and NK cells. CD16 shedding has been linked to activation of NK
These plates were analyzed using the iQue® Hu- cells where it is involved in the detachment of NK
man NK Cell Killing Kit in combination with the and target cells during ADCC and can lead to an in-
iQue® Advanced Flow Cytometry Platform to en- crease in serial target engagement.6 There was also
able analysis of the activation state of the NK cells. a concentration-dependent increase in IFNγ secre-
The co-culture plates were treated with Kadcy- tion in response to Kadcyla® treatment (EC50 0.13
la® for either 24 h or 48 h before the contents of µg/mL, Figure 7C) only measured in the co-culture
the well were lifted and stained with the antibody wells indicating a link to NK activation. Taken to-
cocktail provided in the kit. The data shows a clear gether the data supports that Kadcyla® treatment
concentration-dependent increase in CD25 expres- of HER2 expressing target cells in the presence of
sion on the CD3-CD56+ NK cells induced by Kadc- NK cells results in the activation of NK cells to in-
yla® at both 24 h and 48 h (EC50 0.08 µg/mL, 0.05 duce the ADCC event.

Figure 7: Quantification of NK Activation During ADCC

Note. Co-cultures of AU565 and NK cells in a 1:5 target:effector cell ratio were set up and treated with Kadcyla® for 24 h or 48 h. The
iQue® Human NK Cell Killing Kit was used to assess activation of the NK cells. The graphs show the concentration dependent effect
of Kadcyla® on CD25 expression on NK cells (A), shedding of CD16 (B) and secreted IFNγ levels (C). All data shown as the mean ± SEM
for at least 3 wells.

23
Conclusion and Summary References
The datasets displayed in this document clearly ex- 1. von Minckwitz G, Huang CS, Mano MS, et al. Trastuzumab
emtansine for residual invasive HER2 positive breast
emplify the detailed characterization of Kadcyla®
cancer. N Engl J Med. 2019;380(7):617-628. doi:10.1056/
as an example of an ADC complex using live-cell NEJMoa1814017
analysis on the Incucyte® Live-Cell Analysis System.
The data demonstrates how the various analysis 2. Baron J, Wang ES. Gemtuzumab ozogamicin for the
treatment of acute myeloid leukemia. Expert Rev Clin
readouts can be applied in the characterization of
Pharmacol. 2018;11(6):549-559. doi:10.1080/17512433.
this class of biotherapeutics. 2018.1478725

The assays enable simple, clear visualization and 3. Joseph NS, Tai YT, Anderson KC, Lonial S. Novel approaches
quantification of various effects of Kadcyla®. to treating relapsed and refractory multiple myeloma with
a focus on recent approvals of belantamab mafodotin and
• Kadcyla® binds specifically to HER2 selinexor. Clin Pharmacol. 2021;13:169-180. doi:10.2147/
expressing cells and is internalized CPAA.S288840
with similar efficiency and potency as 4. Ahmed N, Hamadani M. Evaluating efficacy and safety
trastuzumab. of loncastuximab tesirine injection for the treatment of
adult patients with relapsed or refractory large B-cell
• Unlike trastuzumab, once internalized lymphoma. Expert Rev Anticancer Ther. 2021;1-8. doi:10.1
Kadcyla® results in directed cell death 080/14737140.2021.1988853
through the release of the cytotoxic payload
5. Hunter FW, Barker HR, Lipert B, et al. Mechanisms of
DM1, a potent tubulin inhibitor.
resistance to trastuzumab emtansine (T-DM1) in HER2-
positive breast cancer. Br J Cancer. 2020;122(5):603-612.
• Kadcyla’s® combined effect of ADCC activity
doi:10.1038/s41416-019-0635-y
and direct killing results in improved efficacy
for tumor cell clearance. 6. Srpan K, Ambrose A, Karampatzakis A, et al. Shedding
of CD16 disassembles the NK cell immune synapse and
The biological analysis of these assays is further boosts serial engagement of target cells. J Cell Biol.
enhanced by the combination with flow cytomet- 2018;217(9):3267-3283. doi:10.1083/jcb.201712085
ric analysis to quantify immune effector activation
states and cytokine secretion profiles. Utilizing About the authors
both technologies together they offer great poten-
tial for use in the development and research for fu- Nicola Bevan, Kirsty McBain, Daryl Cole and Tim Dale
ture ADC therapeutics. are with Sartorius U.K.

24
Empower Your Research With Simplifying Progress

Incucyte® Live-Cell Analysis

The Incucyte® Live-Cell Analysis System speeds scientific discovery by
combining lab-tested protocols and reagents with powerful, automated image
acquisition and analysis. Gain dynamic insights into the health, morphology,
movement and function of cell models, all from the stable environment of a
tissue culture incubator. Select an Incucyte® for your research at
www.sartorius.com/incucyte

Specifications subject to change without notice. © 2022. All rights reserved. Incucyte and all names of Sartorius products
are registered trademarks and the property of Sartorius AG.
Real-Time 96-Well Antibody
Internalization Assays Using
Incucyte® Fabfluor-pH
Antibody Labeling Dye
Integrated assay solution addresses limitations of
traditional methods.
by Nicola Bevan, Tim Dale, and Del Trezise

mechanisms is a key determinant of the half-life in

Introduction
the body, and small modifications to monoclonal
Monoclonal antibodies are now widely used as antibody structure can have profound effects on
anti-cancer, anti-inflammatory, and anti-viral duration of activity. For these reasons, understand-
therapeutic agents. An important property of an- ing the internalization of antibodies into cells, and
tibodies is the extent, location, and rate of inter- being able to compare the uptake of different anti-
nalization into cells. For example, antibody-drug bodies, is a key requirement in antibody selection
conjugates (ADCs) deliver cytotoxic payloads to and optimization for biologics drug discovery.
cells and the specificity and degree of cellular
Current methods to directly quantify antibody in-
uptake is critical to the efficacy and safety of the
ternalization are laborious, time-consuming, and
molecule. Monoclonal antibodies are also used in
not amenable to testing multiple antibodies. These
immunotherapy to label tumor cells for immune
assays generally require labeling each antibody
cell clearance (e.g., ADCC, ADCP1-3) and to drive
with a fluorescent tag—in most cases the labeled
internalization of receptors associated with tumor
antibody must be separated from the free label
survival (e.g., EGFR4,5).
via a column or wash step. The analysis requires
In each case, the rate of removal from the cell sur- that the signal from the internalized antibody can
face will strongly influence the therapeutic profile be robustly isolated from that outside the cells.
of the antibody. More generally, internalization of Wash steps, blocking dyes and/or a reduction in
antibodies by either specific (target mediated) or temperature to slow cellular activity are common.
non-specific (e.g., pinocytosis in endothelial cells) Almost all (e.g., flow cytometry, high content im-

26
aging) provide only end point assays and multiple Jurkat and Raji cells (both RPMI 1640+10% FCS)
experiments are required to follow internalization were grown in 75 cm2 tissue culture treated flasks.
over time. All cell lines were provided by ATCC and culture
reagents were obtained from Life Technologies.
In this application note, we describe an integrated
Commercially available test antibodies were CD71
assay solution, based on Incucyte® live-cell imag-
clone 1a, 2, CD20 (Sigma), clone 3-5 (Abcam), CD3,
ing and analysis and a novel pH-sensitive dye cou-
CD45 (Biolegend), human IgG isotype control
pled antibody fragment (Incucyte® Fabfluor-pH An-
(Absolute Antibody), mouse IgG1 isotype control
tibody Labeling Dyes), that is designed to address
(R&D Systems).
these limitations. This solution provides a simple,
turnkey method for comparing the internalization Antibody Labeling
of large numbers of antibodies and is amenable for
use throughout the industrial antibody drug dis- Test antibody at known concentration was mixed
covery process and basic scientific research. with isotype matched, Incucyte® Fabfluor-pH
Dye at a molar ratio of 1 to 3 in complete growth
Assay Principle media. Reactions were performed at twice the final
required assay concentration in either a round
Incucyte® Fabfluor-pH Antibody Labeling Dyes are bottom 96-well plate (Costar Cat. No. 3799) or an
Fc-region targeting Fab fragments conjugated to amber microtube, depending on the volumes
a pH-sensitive fluorescent probe. These reagents required, and incubated for 15 min at 37° C.
enable a generic one-step, no-wash, labeling Any required dilutions (e.g., for construction of
protocol for all isotype matched, Fc-containing concentration response-curves) were performed
test antibodies. At pH 7.0, the Fab-Ab complex post conjugation to maintain the labeling ratio.
has little or no fluorescence. When labeled anti- Fabfluor-pH labeled antibody (50 μL) was then
bodies are added to cells, a fluorogenic signal is added directly to pre-plated cells (in 50 μL).
observed as the Fab-Ab complex is internalized
and processed via acidic (pH 4.5–5.5) lysosomes Incucyte® Antibody
and endosomes (Figure 1). The full time course Internalization Experiments
of internalization is visualized and automatically Cells were plated (BT-474 10K per well, 16 h, HT-
quantified using Incucyte® real-time live-cell anal- 1080 8K per well 4 h) on flat-bottom 96-well plates
ysis. The labeling method and assay workflow is (Costar Cat. No. 3595) prior to assay. For assays using
commensurate with comparing internalization
non-adherent cell types, plates were coated with
rates of large numbers of antibodies (10–100’s) in
poly-L-ornithine (PLO, Sigma Cat. No. P4957) for 1 h
miniaturized (96/384-well) plate formats.
and allowed to dry for 30 min prior to cell seeding.
Cells were seeded at 30 K/well, 1 h prior to assay.
General Methods and Materials Following addition of Ab/Fab complex assay plates
were immediately placed in an Incucyte® Live-Cell
Cell Culture and Reagents
Analysis System and scanned for HD phase and red
BT-474 cells (DMEM+10% FCS), HT-1080 cells fluorescence images at 10X or 20X magnification
(F12K+10% FCS, 1% glutamax, 1u/mL pen/strep), every 15–30 minutes for up to 48 h. Images were

27
Figure 1: Principle of antibody Internalization assay using Incucyte® Fabfluor-pH labeling dye (A). Fluorogenic signal as internalized
antibody is processed into the acidic endosome and lysosome. Data to show the pH sensitivity of the labeling probe (B and C). Note
the relatively low fluorescence of Fabfluor-pH at pH 7.0.

28
automatically analyzed using the integrated Incu- A rapid time-dependent increase in red fluores-
cyte® software for the following metrics: (1) Phase cence was observed with anti-CD71, but not iso-
Confluence (measure of cell area), (2) Red Fluores- type or media control, from the first time point (15
cence Object area (index Fabfluor-pH labeled inter- min) of the assay (Figure 2). At 12 h, the signal:back-
nalized antibody). Top Hat subtraction was used to ground of the assay was > 15 fold. The red signal
minimize any background fluorescence signal. was observed in the cytosolic compartment of the
cells but not in the nucleus, consistent with the
Validation Data expected localization of the internalized antibody
to lysosomes and endosomes. Fluorescence area
Proof of Concept (µm2/well) was used to quantify the specific signal.
Similar proof of concept assays were conducted
As a proof of concept, internalization of an anti- using the Incucyte® Fabfluor-pH Orange Antibody
body to CD71, the transferrin receptor, was mea- Labeling Dye as well.
sured in HT-1080 fibrosarcoma cells. The anti-CD71
antibody (clone MEM189) and a mouse IgG1 iso- Normalization for Cell Number
type control were labeled with isotype matched In theory, the antibody internalization signal
Incucyte® Fabfluor-pH Red Mouse IgG1 Antibody should increase as a function of cell number. To ver-
Labeling Dye as described. Fabfluor-pH labeled an- ify this, and to understand the contribution of cell
tibodies (4 μg/mL) were added to cells and moni- proliferation to the signal over time, experiments
tored (phase contrast, red fluorescence) for 12 h in were conducted on cells plated at different den-
an Incucyte® Live-Cell Analysis System. sities (1–20 K per well). Anti-CD71 internalization

Figure 2: Internalization of Incucyte® Fabfluor-pH labeled α-CD71 in HT-1080 cells. HT-1080 cells were treated with either Incucyte® Fabfluor-pH
labeled α-CD71 or IgG1 isotype control (4 μg/mL), HD phase and red fluorescence images (10X) were captured every 15 min over 12 h. Images
of cells treated with Fabfluor-pH- α-CD71 display red, cytosolic fluorescence (A). Cells treated with labeled isotype control display no cellular
fluorescence (B). Time course data shows a rapid increase in red object area over time in cells treated with labeled α-CD71 but not with isotype
control (C). Images shown taken at 6 h post treatment, data shown as mean of 3 wells ± SEM.

29
(red fluorescence area) was detectable at 1 K cells the red fluorescence area was normalized to the
per well and markedly increased at higher plating total cell area (Phase Confluence) to account for
densities (Figure 3 A and B). Over time (0–12 h) proliferation, the time course signals were highly
the internalization signal continued to rise. When similar. Importantly, after 4 h there was no further

Figure 3: Antibody internalization response is cell number dependent. An increasing density of HT-1080 cells were seeded (1–20
K/well) and treated with Incucyte® Fabfluor-pH labeled α-CD71 (4 μg/mL). HD phase and red fluorescence images (10X) were
captured every 30 min over 12 h. The time- course of red object area data demonstrates an increasing internalization signal with
increasing cell number (A and B). When the red object signal is normalized for phase area, it is clear the internalization response
size is dependent on cell number (C and D). All data shown as a mean of 3 wells ± SEM, bar graphs show area under the curve (AUC)
calculated from time course data.

30
increase in the normalized signal, indicating that calized with green, supporting the premise that
antibody internalization had reached equilibrium. the Fabfluor labeled antibody was internalized to
This normalization method helps minimize the the lysosome.
impact of well-to-well cell seeding variation and
isolate the true internalization rate signal from
Applicability Across Different Cell Types
cell proliferation.
and Antibodies
To demonstrate the broad application and specific-
Signal Co-Localization With
ity of the method, internalization was assessed for
Lysosomal Marker
a range of test antibodies targeted against specific
To confirm the presence of the internalized anti- CD markers expressed in different cell lines (Figure
body in the lysosome, we conducted dual labeling 5). Anti-CD20 was internalized in the B cell line Raji,
experiments with LysoSensor® Green (Thermo Sci- but not Jurkat, a T cell line. Conversely, anti-CD3
entific), an end-point lysosomal marker. HT-1080 was internalized in Jurkat but not Raji. Antibodies
cells were treated with Fabfluor-pH-labeled-CD71 to CD71 and CD45, general lymphocyte markers,
for 3 h and monitored for antibody internaliza- were internalized in both cell types. Importantly,
tion. LysoSensor Green was added, and the plate IgG was not internalized in either. These data are in
then returned to the Incucyte® to measure red alignment with the known CD surface marker ex-
(CD71) and green (LysoSensor) fluorescence. A pression of these cell lines, and provide strong con-
strong co-incident signal was observed in the two fidence in the signal specificity and generic utility
signal channels, 74% of the red signal was co-lo- of the method.

Figure 4: Co-localization of Incucyte® Fabfluor-pH labeled α-CD71 and LysoSensor® Green in HT-1080 cells. Internalization of Incucyte®
Fabfluor-pH labeled α-CD71 (4 μg/mL) was established for 3 h in HT- 1080 cells before addition of LysoSensor® Green DND-189 (Thermo,
0.25 μM). Images show individual LysoSensor® Green and Fabfluor-pH labeled α-CD71 red signal (A and B), co-localization of red and
green signals (C), and the co-localized analysis mask shown in yellow (D). (E) Incucyte® analysis of the coincidence of the red and green
fluorescence confirms co-localization of 74% of the red signal with the green signal. Images captured at 20X magnification, 30 min post
LysoSensor® addition, data shown as mean of 4 wells ± SEM.

31
 Figure 5: Internalization of CD surface marker targeted antibodies in lymphocytic cell lines. Jurkat (T cell-like) and Raji (B cell-like) cells
(30 K/well) were treated with different Incucyte® Fabfluor-pH labeled antibodies (4 μg/mL). HD phase and red images were captured
every 30 min using a 20X objective over 12 h. Time course data (A and B) and area under the curve (AUC, C) analysis demonstrates the
response profile in both cell lines. All data shown as a mean of at least 4 wells ± SEM, time course data shown as normalized red area.

= 2.1 nM, Figure 6). In Raji cells, the EC50 value for
Quantitative Pharmacological Analysis
Rituxan was 426 ng/mL = 2.6 nM, Figure 7). These
Pharmacological, Kinetic Quantification EC50 values are similar to the known KD values for
of Antibody Internalization Herceptin and Rituxan for their target receptors
(both approximately 5 nM).
To illustrate the quantitative nature of the method
and suitability for analyzing therapeutic antibod- Comparison of Multiple Test Antibodies
ies, we sought to determine EC50 values for the in- for High-Throughput Screening
ternalization of Herceptin (Trastuzumab) and Rit- The features of the Incucyte® Antibody Internal-
uxan (Rituximab), two clinically used monoclonal ization Assay are such that it should be facile to
antibodies. After labeling of each antibody with parallel label many antibodies and compare their
the Fabfluor-pH reagent, the antibody was serially internalization. To validate this, we took 6 different
diluted (1:2) prior to addition to the cells to enable commercially available anti-CD71 antibodies and
construction of a concentration-response curve. compared their internalization properties head-
Handling the labeled antibody this way is good to-head. The antibodies were plated in 96-well
practice to eliminate any variation in Fab labeling plates and labeled in full media with the Incucyte®
efficiency which could occur across the concentra- Fabfluor-pH Dye. Serial dilutions were performed
tion range. In BT-474 Her2-positive breast carcino- in full media (8 point, 1:2). Labeled IgG and Fab-
ma cells, clear time and concentration dependent fluor-pH alone were added to control wells. La-
internalization of Herceptin was observed over 48 beled antibodies were then added to pre-plated
h. From an area under the time-course (AUC) analy- HT-1080 cells and monitored for internalization
sis, the EC50 value for internalization was 323 ng/mL for 12 h.

32
Figure 6: Quantitative pharmacological analysis of Incucyte® Figure 7: Quantitative pharmacological analysis of Incucyte®
Fabfluor-pH labeled Herceptin. BT-474 Her2-positive cells Fabfluor-pH labeled Rituxan. Raji cells were treated with
were treated with increasing concentrations of Fabfluor-pH increasing concentrations of Fabfluor-pH labeled Rituxan. The
labeled Herceptin. The time course graph displays an increase time course graph displays an increase normalized red area over
normalized red area over time with increasing Herceptin time with increasing Rituxan concentrations (A). Area under the
concentrations (A). Area under the curve analysis of this curve analysis of this response displays a clear concentration
response displays a clear concentration dependent response dependent response with an EC50 of 426 ng/mL (B). All data
with an EC50 of 323 ng/mL (B). All data shown as a mean of 3 shown as a mean of 3 wells SEM, time course data shown as
wells ± SEM, time course data shown as normalized red area. normalized red area.

Of the 6 antibodies, 3 (Ab 1a, Ab2 and Ab 1b) and gave similar internalization responses. Abs
produced large internalization signals and 3, 4, and 5 were internalized more weakly and
were detected at low concentrations (< 0.05 µg only at higher con- centrations (Figure 8). From
mL-1). Reassuringly, Ab 1a and Ab 1b were the the control responses, a mean Z’ value of 0.82
same antibody clone from different suppliers was determined (2 plates 0.75, 0.87) indicating

33
a microplate assay with high robustness. These antibody per se. Indeed, the assay precision and
data confirm the suitability of the method workflows are such that hundreds of different
for comparing the internalization of multiple antibodies could be compared at once and
antibodies at a single target, and illustrate that further throughput could be achieved through
the internalization profile is a property of the miniaturization to 384-well format.

Figure 8: Screening test Abs for internalization. Six different CD71 antibodies including one clone from 2 different suppliers (clone 1a
& 1b) were tested head-to-head in HT-1080 cells. The antibodies were labeled with Incucyte® Fabfluor-pH Dye prior to addition to cells
and the internalization signal captured every 30 min over 12 h using a 10X magnification. Plate views taken from Incucyte® show clear
positive and negative control responses in column 11 and 12 with concentration dependent responses for each antibody across two
plates (A). Head-to-head analysis of antibody data shows a range of responses across these clones (B) control responses at 12 h display
a clear positive response. All data shown as mean of 3 wells ± SEM, controls shown as mean of 8 wells.

34
 covery process, and will prove valuable in efficacy,
Conclusions
safety, and pharmacokinetic optimization of novel
The key features of the approach described in this therapeutic antibodies. In addition, the method
application note are: is suited to understanding basic mechanisms of
endocytosis, pinocytosis, and receptor turnover
• A single step labeling protocol for easily where antibodies can be employed.
tagging antibodies of interest with an
Fc-targeted Fab coupled pH-sensitive dye References
(Incucyte® Fabfluor-pH). The labeling
method is conducted in full media and is 1. Ochoa, M.C., Minute, L., Rodriguez, I., Garasa, S.,
suitable for purified antibodies and Perez- Ruiz, E., Inogés, S., Melero, I., & Berraondo,
antibody supernatants. P. (2017). Antibody-dependent cell cytotoxicity:
immunotherapy strategies enhancing effector NK
• An automated, image-based and real time cells. Immunology and Cell Biology, 95 (4), 347-355.
analysis method (Incucyte® live-cell imaging
and analysis) for monitoring internalization 2. Weiskopf, K., & Wiessman, I. (2017). Macrophages are
in multiple 96-well microplates at once. The critical effectors of antibody therapies for cancer.
format is amenable to both adherent and mABs, 7 (2), 303-310.
non-adherent cells. 3. Beck, A., Goetsch, L., Dumontet, C., & Corvaïa, N.
• An assay system that follows the full (2017). Strategies and challenges for the next
time course of the biology and reports generation of antibody-drug conjugates. Nature
internalization with high specificity, sensitivity, Reviews Drug Discovery, 16 (5), 315-337.

and morphological information. The use of a 4. Tomas, A., Futter, C.E., & Eden, E.R. (2014). EGF receptor
pH-sensitive dye provides for low background trafficking: consequences for signaling and cancer.
signal and obviates the need to separate out Trends in Cell Biology, 24 (1), 26-34.
fluorescence arising from antibody on the cell
surface or in bulk solution. 5. Harding, J., & Burtness, B. (2005). Cetuximab: An
epidermal growth factor receptor chimeric human-
Taken together, these attributes provide a simple, murine monoclonal antibody. Drugs Today, 41 (2),
integrated and quantitative solution for directly 107-127.
studying internalization of antibodies into cells
that can easily be scaled to compare many anti- About the authors
bodies (10–100’s) in parallel. This method enables
antibody internalization measurements to be im- Nicola Bevan, Tim Dale, and Del Trezise are with
plemented at earlier stages in the biologics dis- Sartorius U.K.

35
Antibody Internalization:
Advanced Flow Cytometry and
Live-Cell Analysis Give Rich
Insights During Antibody Profiling
Novel antibody screening method focuses on speed and insight.

by Clare Szybut, Caroline Weldon, Nicola Bevan, and Lori King

are desirable properties for antibody-drug conju-

Introduction
gates (ADCs), as cells must be selectively targeted
The natural characteristics of antibodies, such as and killed via delivery of a cytotoxic payload. In
high binding affinity, specificity to a wide variety of contrast, if the goal is to induce antibody-depen-
targets, and good stability, make them ideal ther- dent cell-mediated cytotoxicity (ADCC), the anti-
apeutic candidates for many diseases. Monoclonal body must remain bound to the cell surface, rath-
antibodies (mAbs), in particular, deliver promising er than being internalized, in order to activate an
therapeutic results in several different disease ar- immune response.
eas, such as autoimmunity, oncology, and chron-
When designing ADCs, one key attribute to pre-
ic inflammation. Researchers’ abilities to improve
dicting efficacy is the antibody internalization (ABI)
the breadth of antibodies have been aided by in-
rate and the associated kinetics. These internal-
novative technologies for antibody discovery, for
ization kinetics are influenced by factors such as
instance, through humanization of mouse anti-
the epitope on the target antigen, affinity of the
bodies and phage display. However, advanced an-
ADC-antigen interaction, and intracellular traffick-
tibody design techniques create the need for new
ing. Evaluating for these factors is critical through-
screening methods so that lead candidates can be
out the antibody screening pathway (Figure 1);
quickly and effectively identified as early in the de-
however, here, we concentrate on evaluating an-
velopment process as possible.
tibodies’ ABI during functional profiling via rate
Part of an effective screening strategy is to identify comparisons during the development process, as
the desired therapeutic goal for your candidates. well as multiplexed mechanistic studies later in the
For instance, binding and rapid internalization screening process.

36
Traditionally, antibody screening workflows have post-translational modifications for their activity
required labor intensive, time consuming, end- (Chames et al., 2009). Therefore, researchers face
point-only methods, such as FACS, ELISA, or mi- several challenges when engineering and produc-
croscopy. One solution to address these screening ing mAbs as therapeutics. Engineering antibod-
challenges is to use advanced screening methods ies to optimize their biological potencies during
that offer maximum insight through high-content the discovery phase can address many of these
data. Herein, we demonstrate an advanced anti- challenges. However, these attempts to optimize
body screening method that focuses on speed and one attribute can have profound and unintended
insight by using a novel, pH-sensitive reagent for consequences on other antibody attributes. For
characterizing ABI via both advanced flow cytome- instance, optimizing an antibody’s specificity may
try and live-cell imaging. negatively affect activity (Tiller and Tessier, 2015).

One way of simultaneously optimizing multiple

Challenges with Traditional antibody properties is by using mutagenesis to
Antibody Screening Workflows produce large screening libraries. However, large
screening libraries necessitate a thorough in vitro,
Monoclonal antibodies (mAbs) are large (about high-throughput screening method to quickly
150 kDa), complex biologic molecules that require identify the most suitable drug candidates for fur-

Figure 1: Inclusion of ABI in the screening pathway for antibodies. Selecting for ABI throughout the antibody screening pathway,
whether during functional profiling and rate comparisons or mechanistic studies.

37
ther development early in their discovery process. tigens to confirm their specificity and cross- re-
activity. Second, ELISA is not the best method for
Workflows for antibody screening commonly in- screening antibodies that bind to cell surface an-
clude fluorescence-activated cell sorting (FACS), tigens because these antigens are extracted from
enzyme-linked immunosorbent assays (ELISA), the cell membrane and purified before adsorb-
and microscopy (such as confocal) techniques. Yet, ing to the plastic ELISA plate. Extracting antigens
these in vitro antibody screening methods have from the membrane often leads to disruption of
several drawbacks: (1) they can be labor intensive conformational epitopes that can be important
with limited throughput, (2) they do not allow di- targets for therapeutic antibodies. Finally, to min-
rect, head-to-head comparisons of antibodies, and imize background signal, ELISA requires multiple
(3) they need large amounts of reagents. wash steps to remove unbound antibodies and
For instance, even though FACS can be used to sort detection reagents, resulting in long, labor-inten-
hundreds of thousands of cells based on their flu- sive screening workflows.
orescence properties, it cannot be used to perform Also, although ELISA can provide data on the im-
high-throughput, time-dependent studies on indi- munoglobulin G (IgG) titer, it offers no reflection
vidual cells because the cells must be sorted into on how the therapeutic antibody candidates affect
single colonies before any further analysis can be the health of the test cells. i.e. how quickly or ef-
performed (Doerner et al., 2014). ficiently they induce death. Thus, candidates that
ELISA, on the other hand, is adaptable to appear to be productive based on ELISA screening
high-throughput screening (Saeed et al., 2017), may be carried forward into the next step of the
production process even though they are unideal
and thus has historically been used to screen hy-
candidates in terms of function or vigour.
bridomas and other libraries for antibody binding
to each of its targets. ELISAs are performed by Microscopy techniques, in contrast to FACS, are
coating a single target antigen onto the wells of useful for single-cell analysis, as well as for such
assay plates, followed by the addition of individu- localization and temporal studies as antibody in-
al samples from an antibody library (for instance, ternalization and live-cell imaging for monitoring
from hybridoma or phage display). Antibodies individual cell behavior. Microscopy techniques,
that bind to the immobilized antigen are detect- however, have limited throughput due to data ac-
ed by a color change due to an indirect enzyme/ quisition time, and they have limited multiplexing
substrate reaction. capabilities. Thus, for antibody discovery, some
other method, preferably high-throughput, is gen-
In addition to its adaptability to high-throughput
erally used for the initial selection and screening of
screening, ELISA is rapid, consistent, and relatively
large libraries, followed by a more thorough func-
easy to analyze (Saeed et al., 2017). However, ELI-
tional analysis of a smaller set of antibody candi-
SA has several disadvantages that can limit its suc-
dates using microscopy (Doerner et al., 2014).
cessful use in a modern antibody screening lab.
First, primary screens that test binding to a single Like FACS, microscopy techniques also require la-
antigen often require subsequent secondary and beling each antibody with a fluorescent tag, which
sometimes even tertiary screens with control an- must be separated from the free label via a column

38
or wash step because analysis requires robust iso-
A Simple Way of Labeling Antibodies
lation of internalized antibodies from those out-
side the cells. To aid isolation of the positive signal,
for Addressing Challenges with FACS,
researchers often resort to perturbing techniques, ELISA, and Microscopy
such as washing cells, using blocking dyes, and re-
One of the challenges presented above with FACS,
ducing the temperature to slow cellular activity.
ELISA, and microscopy techniques is the require-
However, cells can be lost during washing steps, ment for wash steps to minimize background sig-
and the associated reductions in temperature per- nal. Among other undesirable effects, such as in-
turb the cellular environment. creased time to results, this need for wash steps
A further drawback to almost all the techniques also increases reagent requirements (and cost) and
makes cell loss inevitable. However, an advanced
used for antibody screening is that they only en-
method for labeling cells could reduce reagent re-
able end-point analysis, which means that multiple
quirements, as well as simplify protocols when an-
experiments are required to follow an antibody at-
alyzing antibody internalization.
tribute, such as internalization, over time.

The best approach to address all the limitations ABI Assays Made Simple, With
of traditional antibody screening is to combine No-Wash Labeling and Low
data from advanced screening methods; using ad- Reagent Requirements via a Novel,
vanced methods, researchers can choose their can- Ph-Sensitive Reagent
didates based on a thorough evaluation of all rel-
evant characteristics, such as IgG titer, cell health, The key to this simple, no-wash protocol is a novel
internalization, and their associated kinetics early reagent composed of Fc-region targeting Fab frag-
in their discovery process. ments conjugated to a pH-sensitive fluorescent

Figure 2: The pH-sensitive fluorescent probe principle. A novel pH-sensitive fluorescent probe enables one-step, no-wash labeling of
isotype- matched antibodies. A fluorescent signal is generated as internalized antibody is processed into the acidic endosome and
lysosome pathway.

39
 Figure 3: The assay consists of 3 components, each 10 μL: labeled test antibody, cells (encoded or not), and Cell Membrane Integrity Dye
(B/Green). Each component is prepared at 3X before addition for a final concentration of 1X and an assay volume of 30 μL.

probe (Nath et al., 2016). This type of novel reagent

Functional Profiling for
enables a generic, one-step, no-wash labeling pro-
tocol for all isotype-matched, Fc-containing test Comprehensive Cell and Antibody
antibodies when optimized for use on specific in- Characterization Early in the
struments. Figure 2 shows how this reagent works: Process with Multiplex Analysis
labeled antibodies are added to cells, and a fluoro-
genic signal is produced as the Fab-Ab complex is Combining the above-described pH-sensitive dye
internalized and processed via acidic (pH 4.5- 5.5) with other reagents in one assay on an advanced
lysosomes and endosomes. flow cytometry platform enables simultaneous
Antibodies of interest are quickly and effectively la- analysis of a variety of cell and antibody charac-
beled, with low reagent requirements, by incubating teristics within the same workflow. For instance,
in growth media with this novel, pH-sensitive dye you can measure cell viability using a membrane
(Figure 3). Cells are then added to 384-well plates, integrity dye to assess general cell health, as well as
along with the dye-conjugated antibodies, and in- cell death due to cargo delivery, when optimizing
cubated again. This reagent, when used with an ADCs. Or you can characterize cell specificity using
advanced flow cytometry platform, provides a com- encoding dyes (for cell lines) or directly conjugat-
prehensive, integrated solution for rapid profiling of
ed fluorescent antibodies (for complex cell models
antibody internalization and other critical antibody
with a variety of cell types). Other reagents, such as
attributes using small sample volumes in 96- or even
those for assessing cytokine release, are also avail-
384-well plate formats (Riedl et al., 2016). Much of
the data acquisition and analysis, including gener- able for more detailed antibody assessments on
ation of serial dilution curves and EC50 calculations, the sample. Therefore, analysis with an advanced
is automated with the help of advanced software flow cytometry platform can be optimized to deliv-
packages, for example, as included in the iQue® ad- er rich content very quickly while only using a low
vanced flow cytometry platform. sample volume.

40
 Figure 4: Serial dilution curves for internalization-labeled antibodies with a top concentration of 1 mg/mL in different cell types after
a three-hour incubation. Multiplexed positive and negative cell lines may be used in an ABI assay to generate high-content data in
one assay. Median fluorescent intensity (MFI) for Internalization Reagent-labeled antibodies after three hours. A serial dilution of each
antibody with a top concentration of 1 mg/mL was prepared and incubated with encoded Jurkat, Raji, and Ramos cells in the same
well. Jurkat cells (a T lymphocyte cell line) show a concentration- dependent increase in internalization of anti-CD3, but not the two
B cell markers. Conversely, Raji cells show a concentration-dependent increase in internalization of anti-CD19 and anti-CD22, but
not anti-CD3. Ramos cells show a concentration-dependent increase in internalization of anti-CD79b, an ADC drug target for non-
Hodgkin’s lymphoma.

Sartorius produces such a pH-sensitive reagent for robotic interface, and integrated Forecyt® software
use on our iQue advanced flow cytometry plat- for multiparametric data analysis and visualization
form. The combination of non-perturbing and val- expedite the process of screening drug candidates
idated reagents for multiplexing, no-wash proto- for potential efficacy and toxicity to accelerate an-
cols, high-throughput capabilities, flexibility for a tibody discovery and development.

41
To demonstrate the multiplexing capabilities of
Full Concentration Profiling with
this novel pH-sensitive dye on the iQue platform,
we used Ramos and Raji cells stained with two in- Live-Cell Imaging and Analysis
tensities of violet encoding dye, combined with
A pH-sensitive dye-coupled antibody fragment
unstained Jurkat cells. We incubated this for three
designed according to the same principle as that
hours with a serial dilution of dye-conjugated spec-
used above (Figure 2) can also be optimized and
ificity antibodies (isotype-matched anti-CD3 as a T
used in other instruments. For instance, using this
cell marker, anti-CD19, anti-CD20, anti-CD22, or an-
pH-sensitive reagent in a real-time, live-cell imag-
ti-CD79b as B cell markers, anti-CD71 as a positive
ing system such as the Incucyte® Live-Cell Analysis
control, and IgG as a negative control), then add-
System, allows visualization and automatic quan-
ed a cell membrane integrity dye before acquiring
data from the plate. Using this strategy, we were tification of the full time-course of ABI. This com-
able to identify viable cells, then spectrally sepa- bination of reagent and platform thus provides a
rate Ramos, Raji, and Jurkat cells. simple method for directly profiling and compar-
ing ABI for a large number of antibodies (10–100s
We then assessed antibody internalization for each at a time in a miniaturized format).
cell line. We generated series dilution curves for
the specificity markers of each cell type (Figure 4A). To demonstrate the power of the pH-sensitive dye
As expected, Jurkat cells showed internalization of approach for high-throughput antibody internaliza-
anti-CD3, but not anti-CD19 or anti-CD22, whereas tion assays in a real-time, live-cell analysis system,
the Raji cells internalized anti-CD19 and anti-CD22, we performed a head-to-head comparison of the
but not anti-CD3. Only Ramos cells showed a con- internalization properties of six different commer-
centration-dependent increase in internalization of cially available anti-CD71 antibodies into HT1080
anti-CD79b, an ADC drug target for non-Hodgkin’s fibrosarcoma cells. We labeled the anti-CD71 with
lymphoma. In the three-hour assay time frame, our pH-sensitive reagent before adding to cells in
we did not observe anti-CD20 internalization, but 96-well plates. We then captured the internalization
we did see an increase in the two B cell lines by 24 signal in a live-cell analysis system every 30 minutes
hours (data not shown). over 12 hours using a 10X magnification.

Importantly, we saw little difference when we as- The plate view in Figure 5A shows clear positive and
sessed the cells for internalization alone or when negative control responses in column 11 and 12,
we mixed them. This result shows that multiplex- with concentration-dependent responses for each
ing positive and negative cell lines does not inter- antibody across the two plates (antibodies 1a and
fere with the antibody internalization assay. 1b are the same clone from two different sources).

Compared to performing a series of singleplex as- We found that three antibodies (Ab1a, Ab2, and
says, a multiplexed assay approach enables you to Ab1b) gave internalization signals that we detect-
analyze multiple readouts (internalization, viability, ed at low concentrations (< 0.05 μg/ml) (Figure 5).
cell type) from a single well, decreasing the num- Reassuringly, Ab1a and Ab1b gave similar internal-
ber of tests needed to perform a comprehensive ization responses. Antibodies 3, 4, and 5 were in-
functional characterization of the Ab candidate. ternalized more weakly and only at higher concen-

42
trations. From the control responses, we calculated
Showing a Simple Pharmacological
a mean Z’ value of 0.82 (two plates: 0.75, 0.87), in-
dicating high robustness for this microplate assay.
and Kinetic Quantification of Antibody
Internalization Using Herceptin
These data show our method is suitable for com-
paring the internalization of multiple antibodies In addition, we experimented to determine EC50
at a single target, or one antibody in various cell values for the internalization of a clinically used
types. The assay precision and workflow is such monoclonal antibody, Herceptin (Trastuzumab).
that it would be possible to compare 100s of dif- We constructed a concentration-response curve
ferent antibodies at once, and further throughput by labeling Herceptin with the pH-sensitive re-
could be achieved through miniaturization to a agent, then serially diluting (1:2) before adding to
384-well format. BT-474 cells.

Figure 5: Screening test of Abs for internalization. The pH-sensitive reagent is suitable for high-throughput antibody internalization
assays in a real-time, live-cell analysis system.

43
In BT-474 Her2-positive breast carcinoma cells, we tion criteria for ADC candidates. It can be used for
saw definite time and concentration-dependent functional profiling and rate comparisons, as well
internalization of Herceptin over 48 hours. From an as mechanistic studies when coupled with addi-
area under the curve (AUC) time-course analysis, tional multiplexed readouts, thus reducing the
we calculated the EC50 value for internalization at time required for lead generation. To illustrate the
323 ng/mL = 2.1 nM (Figure 6). Our calculated EC50 capabilities of advanced antibody screening, we
value is similar to the known KD value for Herceptin have described our solution for efficiently interro-
for its target receptor (approximately 5 nM). gating libraries of candidates early in the screen-
ing process.
Speed and Insight through Advanced For comprehensive cell and antibody characteriza-
Antibody Screening tion, we used our novel, pH-sensitive reagent and
the iQue advanced flow cytometry platform. This
Quick and accurate identification of suitable drug combination is best for screening and early full
candidates is key to the development of therapeu- concentration profiling because it enables simul-
tic antibodies. ABI is an essential part of the selec- taneous analysis of a variety of cell and antibody
characteristics within the same workflow. For quan-
titative, pharmacological analysis and direct, head-
to- head comparisons of ABI, we used a version of
our novel, pH-sensitive reagent and the Incucyte
live-cell analysis system. This combination is useful
for further functional profiling that requires spatial
and temporal resolution. This advanced antibody
screening process is a simple, fast, and insightful
way to identify candidates that meet your thera-
peutic goals early in the drug discovery process.

Combining Advanced Flow

Figure 6: Quantitative pharmacological analysis of Cytometry and Live-Cell Imaging
pH-sensitive dye-labeled Herceptin shows time and and Analysis for Complete Antibody
concentration- dependent internalization and an EC50
value of 2.1 nM. Quantitative pharmacological analysis Profiling: A Real-World Screening
of pH-sensitive dye-labeled Herceptin. BT-474 Her2- Strategy Employing Our pH-
positive cells were treated with increasing concentrations
of pH-sensitive dye-labeled Herceptin. The time course
Sensitive Dye and Instrument
graph displays an increasing normalized red area over Platforms
time with increasing Herceptin concentrations (A). Area
under the curve analysis of this response displays a clear Researchers at LifeArc used the iQue platform and
concentration dependent response with an EC50 of 323 ng/ Incucyte Live-Cell Analysis System as part of their
mL (B). All data shown as a mean of 3 wells ± SEM, time
strategy to develop a new ADC targeting neuro-
course data shown as normalized red area.
blastoma. Neuroblastoma is a rare cancer that nev-

44
ertheless is the most common extra-cranial solid try and live-cell imaging, they gained maximum in-
tumor in children, with a 5-year survival rate of 50% sight through high-content data and were able to
for patients with high-risk disease. The researchers make critical decisions early in their process, thus
found they were able to use these systems in com- saving time, material, and money.
bination, not only for screening, but also for assay
development, lead candidate profiling, and char- References
acterization. They found that the iQue platform
offered fast, high-content analysis, while the Incu- 1. Tiller KE and Tessier PM. Advances in antibody design.
cyte live-cell analysis system offered kinetic, im- Annu Rev Biomed Eng, Feb 5; 17: 191-216 (2015)
age-based analysis. Combining data from the two PMID: 26274600
systems gave them a complete antibody profile. 2. Doerner A, et al. Therapeutic antibody engineering
In brief, the researchers identified anaplastic lym- by high efficiency cell screening. FEBS Lett, Jan 21;
phoma kinase (ALK) as a target for their therapeu- 588(2); 278-87 (2014) PMID: 24291259
tic approach because level of ALK expression cor- 3. Nath N, et al. Homogeneous plate based antibody
relates with disease stage and ALK antibodies show internalization assay using pH sensor fluorescent
surface expression in patient samples. In collabora- dye. J Immun Meth, Feb 3; 431: 11-21 (2016) PMID:
tion with colleagues at Mt. Sinai, researchers at Li- 26851520
feArc produced 1,152 candidate hybridoma clones
that they were able to narrow to 53 candidates that 4. Reidl T, et al. High-Throughput Screening for
ELISA and flow cytometry showed bound to ALK. Internalizing Antibodies by Homogeneous
Using the advanced flow cytometry capabilities of Fluorescence Imaging of a pH-Activated Probe.
the iQue platform, the researchers were able to fur- J Biomol Screen, Jan; 21(1): 12-23 (2016) PMID:
ther narrow this to 20 candidates that bound ALK 26518032

at the surface of cells, since this is a critical attribute 5. Chames P, et al. Therapeutic antibodies: successes,
of the mechanism of action of ADCs. limitations and hopes for the future. Br J Pharmaco,
May; 157(2): 220-33 (2009) PMID: 19459844
Of particular interest, these researchers used inter-
nalization assays on both the iQue and Incucyte 6. Saeed A, et al. Antibody Engineering for Pursuing a
cell analysis platforms to further narrow their lead Healthy Future. Front Microbiol, Mar; 8(495) (2017)
candidates to two. Critically, they gained kinetic PMID: 28400756
imaging data and high-content analysis using mul-
tiplexed cell viability assays early in their ADC dis- About the authors
covery process. Using advanced screening meth-
ods that incorporate a novel, innovative reagent for Clare Szybut, Caroline Weldon, Nicola Bevan, and Lori
characterizing ABI via both advanced flow cytome- King are with Sartorius

45
 Resources

Incucyte® Fabfluor-pH High-Throughput

Antibody Labeling Quantification of
Reagents Antibody-Dependent
Phagocytosis Using
Live-Cell Analysis

Antibody Internalization Incucyte® Live-Cell

Assays for Live-Cell Analysis Systems
Analysis

Incucyte® Reagents Incucyte® Software

and Consumables

46