Peter van Endert (ed.), Antigen Processing: Methods and Protocols, Methods in Molecular Biology, vol. 960,
DOI 10.1007/978-1-62703-218-6_23, © Springer Science+Business Media, LLC 2013
Tracking Antigen-Speci fi c CD8 + T Cells Using MHC
Class I Multimers
Cécile Alanio , Isabelle Bouvier , Hélène Jusforgues-Saklani ,
and Matthew L. Albert
The tracking of epitope-speci fi c T cells is a useful approach for the study of adaptive immune responses.
This protocol describes how Major Histocompatibility Complex Class I (MHC-I) multimers can be used to
stain, enrich, and enumerate (rare) populations of CD8 + T cells speci fi c for a given antigen. It provides
the detailed steps for multimer labeling, magnetic enrichment, and cytometric analysis. Additionally, it
provides informations for multiplexing experiments in order to achieve simultaneous detection of multiple
antigenic speci fi cities, and strategies for coupling the protocol with functional assays (e.g., intracellular
cytokine staining). Future developments in cytometric systems (e.g., mass spectroscopy-based cytometry) and
gene expression studies (e.g., single cell PCR) will extend these approaches and provide an unprecedented
assessment of the immune repertoire.
Key words: CD8 + T cells , MHC Class I multimers , Antigen-speci fi c T cells , T-cell receptor
Mature CD8 + T lymphocytes bear α β T cell receptors (TCR) that
are speci fi c for a major histocompatibility complex (MHC) class I
molecule bound to a unique peptide. A major goal in the study of
adaptive immune responses is to understand the developmental
progression of antigen-speci fi c T cells from naive precursors to
activated effector cells and long-lived memory cells ( 1, 2 ) . Prior to
1996, limiting dilution analysis was the standard method for esti-
mating the frequency of antigen-speci fi c T cells. The major limitation
of this approach is the requirement for exogenous stimulation and
expansion, introducing potential bias and signi fi cant inter-assay
variability. Notably, cloning ef fi ciency is typically <40%, suggesting
310C. Alanio et al.
that the assay necessarily underestimates precursor frequencies ( 3 ) .
Other techniques such as ELISpot or intracellular cytokine staining
(ICS) are based on the ability of antigen-speci fi c T cells to secrete
cytokines upon short in vitro restimulation with the cognate peptide
( 4, 5 ) . Such approaches identify lymphocytes possessing the capacity
to secrete a given cytokine at the time of the assay, however this
represents only a fraction of the antigen-speci fi c population(s).
The generation of MHC class I tetrameric or multimeric com-
plexes (referred to herein as MHC multimers), originally described
by Altman and Davis ( 6, 7 ) , represents a major technical advance
for the study of T cell repertoires. MHC multimers are reagents
that carry multiple MHC Class I/peptide (MHC-I/pep) com-
plexes, and thus have the ability to interact with multiple TCRs on
a single CD8 + T cell (Fig. 1 ). Fluorescent-labeling of MHC multi-
mers permits identi fi cation of antigen-speci fi c T lymphocytes based
on the avidity of their TCR, independent of their functional or dif-
ferentiation state. This technology has been recently reviewed by
Davis, Altman, and Newell ( 8 ) . Using MHC multimers, it is now
possible to directly track and quantitate antigen-speci fi c T cells
during the course of immunization ( 9 ) . And by co-staining with
antibodies directed against phenotypic cell surface proteins, one
can de fi ne subsets of cells of interest based on their activation or
differentiation state, or chemokine receptors expression ( 10 ) .
MHC multimer technology has also been successfully coupled to
conventional functional assays (e.g., CFSE dilution, ICS), and speci fi c
T cells can be sorted for ELISpot, cytotoxicity, gene expression
studies or for generating long-term cultures ( 11 ) . Multimers are
also widely used in the immune monitoring of T cell responses
Fig. 1. Schematic representation of puri fi ed biotinylated MHC Class I molecule ( left ) and multimer ( right ) (adapted from
Klenerman et al . ( 9 ] ).
31123 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
following therapeutic or prophylactic vaccination. Finally, the
recent availability of GMP-grade multimers is enabling the ex vivo
expansion of T cells for immunotherapy ( 12– 15 ) .
Several key improvements have been reported since the initial
description of multimer technology ( 8, 16, 17 ) . First, as multi-
merization is the key to overcoming the relatively low intrinsic
af fi nity of TCR/MHC interaction, MHC-I multimers now exist as
tetramers, pentamers, and dextramers (with the latter containing
>10 MHC I/pep complexes). Monomer production has also been
substantially optimized. Most notable is the work of Schumacher
and colleagues, who demonstrated the possibility to generate high-
throughput production of MHC I/pep complexes using a photo-
destructible peptide that permits an exchange reaction with peptides
of interest ( 18 ) . Additionally, the implementation of a dump chan-
nel, dual tetramer labeling and multiplexing have all helped establish
a robust foundation for translating this technology into the clinics
( 18– 21 ) . Nonetheless, there exist remaining technical limitations:
staining methods, analysis protocols, validation and data sharing
have to be standardized ( 22 ) ; and the limit of detection for standard
multimer assays is 10 −4 , which does not allows for direct detection of
rare antigen-speci fi c populations such as naive ones ( 23 ) .
In order to improve the limit of detection, MHC multimer
staining has recently been combined with magnetic bead enrich-
ment ( 24 ) , a concept initially developed in mice for assessment of
CD4 + T and CD8 + T cells ( 25– 27 ) . Following from these studies,
our lab, as well as others, developed a similar enrichment protocol
for human CD8 + T cells (Fig. 2 ) ( 23, 28 ) . Efforts have been made
to standardize the procedure—herein described in details—and to
optimize any details in order to achieve suf fi cient sensitivity to
allow detection of naive antigen-speci fi c T cells from human
peripheral blood. This protocol permits up to 100-fold increased
detection of antigen-speci fi c populations, allowing assessment of
populations with frequencies as low as 10 −6 . As such, it is now
Fig. 2. Schematic representation of the enrichment procedure. Starting from cellular suspension, cells are stained with
PE-labeled multimer, then incubated with antiPE-microbeads before loading on a MACS column. Flow through is the
Depleted fraction. By removing the column, you then have access to your Enriched fraction, containing increased numbers
of multimer-positive cells.
312C. Alanio et al.
possible to characterize the naive T cell repertoire, opening up new
opportunities for de fi ning how T cells are selected, as well as to
investigate aspects of their homeostasis ( 29 ) . These approaches
may also serve as powerful strategies for tracking rare antigen-
experienced self-, tumor-, transplant- or microbe-speci fi c T cells,
either in mice or in humans, in turn providing insight into param-
eters that shape immune T-cell responses. This unit describes our
method for labeling antigen-speci fi c CD8 + T cells obtained from
mice or human peripheral blood with MHC class I multimers, for
enriching and enumerating them, and eventually multiplexing the
assay and/or coupling it to ICS procedure. Future developments
in cytometric systems (e.g., mass spectroscopy-based cytometry)
and gene expression studies (e.g., single cell PCR) will further
extend these approaches and provide an unprecedented look at
the immune repertoire ( 8, 11 ) .
1. Fresh or frozen sample (for mice, see Subheading 3.1 ; for
humans, prepare PBMCs according to standard procedures
( see Note 1 )).
2. For humans studies only: fl uorescently labeled mAb speci fi c for
MHC class I molecules of interest, suitable for fl ow cytometry
(e.g., anti-human HLA A2 antibody, BD Biosciences), and
corresponding isotype ( see Note 2 ).
3. 60-mm Petri dish.
4. Falcon 15 mL tube.
5. 5 mL FACS tubes.
6. FcR blocking reagents.
7. Anti-PE microbeads ( see Note 3 ).
8. MACS separation columns, magnets, stands ( see Note 4 ).
9. BD Falcon Cell Strainer 70 μ m.
10. More than fi ve-color fl ow cytometer, ideally with possibility to
2. PBS–2% FCS.
3. Human Pulldown Buffer (HPB; ~50 mL for one enrichment
sample): PBS 1×, 5% of Human Serum Albumin 20% ( fi nal con-
centration 1%), 5% Citrate Dextrose Anticoagulant ( see Note 5 ).
4. Mice Pulldown Buffer (MPB): PBS 1×, 2% FCS, 0.001%
2.1. Common Reagents
313 23 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
5. Mice Pulldown Buffer (MPB) without azide.
6. Mice R-10 buffer: RMPI, 10% fetal calf serum, 10 mM HEPES,
1× nonessential amino acids, 1 mM sodium pyruvate, 60 nM
2-mercaptoethanol, 20 ng/mL gentamycin.
1. PE- and/or APC-labeled MHC class I multimers ( see Notes 6
and 7 ).
2. For multiplexing experiments (determination of multiple
speci fi cities in one single tube), biotinylated monomers ( see Note
8 ) and streptavidin coupled to fl uorochrome or reporter of
interest (PE-, APC-, PE-Cy7-, APC-Cy7-, Qdots-streptavidin,
3. Cocktail of fl uorescently labeled mAb that are known to be
expressed on cells you wish to exclude from analysis (e.g.,
monocytes, B cells, and NK cells) ( see Note 9 ). These mAb
should be coupled to a common fl uorochrome, for example
Paci fi c Blue, thus giving a positive signal in one fl uorescent
channel, which will be referred to as the “dump channel” in
our gating strategy ( see Subheading 3.5 and Fig. 5a ).
4. Viability marker that will speci fi cally stain dead cells (e.g.,
DAPI Nucleic Acid Stain, Invitrogen) ( see Note 10 ).
5. Fluorescently labeled mAbs including at least an anti-CD8
antibody. Others will be chosen depending on the desired phe-
notypic characterization of target T cells ( see Note 11 ).
6. For ICS in mice: CpG formulated with DOTAP, and speci fi c
peptide for in vivo restimulation.
7. For ICS, BD Cyto fi x/Cytoperm Fixation/Permeabilization
solution kit with BD GolgiPlug containing Brefeldin A (BD
8. For ICS, LIVE/DEAD fi xable dead cell stain kit such as Aqua
(Invitrogen) ( see Note 12 ).
Please note that we describe in this section both mice and human
protocols, and the reader should be careful to utilize the appropri-
ate reagents and buffers.
For human experimentation, you will start with HLA typing
(Subheading 3.2 ), then stain with multimer(s) (Subheading 3.3 ),
with the option to perform enrichment (Subheading 3.4 ), and
fi nally acquire your samples on fl ow cytometer (Subheading 3.5 )
and evaluate precursor frequency (Subheading 3.6 ).
For mice studies, you will start with mice dissection
(Subheading 3.1 ), then stain with multimer(s) (Subheading 3.3 ),
2.3. Flow Cytometry
314C. Alanio et al.
and optionally continue by enrichment (Subheading 3.4 ), and/or
ICS (Subheading 3.7 ). In all cases you will acquire your samples on
a fl ow cytometer (Subheading 3.5 ), and evaluate precursor fre-
quency (Subheading 3.6 ).
1. Harvest 15 lymph nodes (2 inguinal, 2 axillary, 2 brachial, 4
cervical-deep and super fi cial, 2 peri-aortic, and the mesenteric
chain) and the spleen in a 60 mm-Petri dish containing 2 mL
of Mice R-10 buffer.
2. Mash the organs and transfer the cells into a Falcon 15 mL
tube after fi ltering the cell suspension with a 70 μ m cell
3. Wash the well with 3 × 1 mL Mice R-10 to recover the maximum
4. Add 10 mL of Mouse Pulldown Buffer (MPB), count them, spin
down at 300 × g for 5 min at 4°C, and go to Subheading 3.3 .
1. Generic haplotyping of the sample can be easily achieved by
fl ow cytometry, and is suf fi cient for most multimer uses.
2. Count PBMCs and resuspend in PBS at 10 7 cells/mL.
3. Dispense 2 × 50 μ L of this solution into 5 mL FACS tubes. The
remaining cells will be spun down (300 × g , 5 min, 4°) and
used for multimer staining (Subheading 3.3 ).
4. Add either isotype or anti-HLA antibody titrated to the opti-
mal concentration to each FACS tube (optimal Cf = 1/400 in
our hands, meaning that you put 1 μ L of a solution diluted
1/8 in 50 μ L staining volume).
5. Incubate for 15 min at 4°C in the dark.
6. Wash cells once at 300 × g for 5 min at 4°C and resuspend in
100 μ L of PBS.
7. Acquire these 100 μ L in fl ow cytometry (Fig. 3 ).
1. Use cells prepared as detailed above, Subheadings 3.1 (mice)
or 3.2 (human). Resuspend cells in cold Pulldown Buffer
(MPB for mice or HBP for human, hereafter referred as PB),
and aliquot de fi ned numbers of cells in Falcon 15 mL tubes (one
for each speci fi city) ( see Note 13 ).
2. Wash once in PB (300 × g , 5 min, 4°) and resuspend each sample
in 90 μ L cold PB.
3. Add 10 μ L of FcR Blocking Reagent to each tube. Vortex.
4. Incubate 10 min at 4°C.
5. Add PE MHCI multimer and APC MHCI multimer of the
same speci fi city at the appropriate concentration ( see Notes 14
and 15 , and Fig. 4 ).
3.1. Mice Dissection
3.2. HLA Typing
3.3. Multimer Staining
315 23 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
6. If needed, it is possible to multiplex the experiment (i.e., deter-
mine multiple speci fi cities—up to 25—within one single tube)
by preparing each speci fi c multimer with a unique combina-
tion of two different colors ( 21 ) . In the case you want to simul-
taneously enrich your target populations with antiPE
microbeads, one of these two colors will have to be PE ( see
Note 16 and Fig. 6a ).
7. Vortex gently and incubate 30 min at 4°C ( see Note 17 ).
8. Wash once in 2 mL PB, spinning at 300 × g for 5 min at 4°C.
9. If you stop here, transfer your cells into 5 mL FACS tubes,
spin, resuspend in 90 μ L of PBS–2% FCS, and proceed directly
to fl ow cytometry analysis on Subheading 3.5 . Otherwise, you
can follow the procedure by enrichment (Subheading 3.4 ).
Fig. 3. Flow cytometry based-HLA typing (Human). PBMCs are prepared, then incubated with either isotype or anti-HLA
antibody of interest titrated to the optimal concentration. Histograms represent data obtained from one HLA-A2 positive
( left ) and one HLA-A2 negative ( right ) blood donors.
Fig. 4. Representative example of single and double multimer staining. PBMCs from one healthy donor were stained with
In fl uenza A-Matrix1 58-66 MHCI multimer labeled either in PE ( left ), APC ( middle ), or both ( right ). Plots are gated into global
CD8 + population using the gating strategy described in Subheading 3.5 and Fig. 5a .
316C. Alanio et al.
1. To start the enrichment protocol, resuspend labeled cells
obtained in Subheading 3.3 ( step 8 ) in 400 μ L PB.
2. Take a 10 μ L aliquot of labeled cells, and place it into 5 mL
FACS tubes. Complete with 90 μ L with PBS–2% FCS. This
gives you your “Pre-enriched” fraction.
3. To the cells used for enrichment, add 100 μ L of anti-PE micro-
beads ( see Note 18 ).
4. Vortex and incubate for 20 min at 4°C in the dark.
5. Wash twice in 2 mL cold PB, spinning cells at 300 × g for 5 min
6. During the washing step, prepare MACS columns (one per
Falcon 15 mL tube) on a magnet support ( see Note 4 ). Rinse
each column with PB (discard elution). Label Falcon 15 mL
tubes for collecting the fl ow through fraction.
7. Resuspend each sample in 1 mL PB, and load the column.
It is important to fi lter cells just prior to loading on the col-
umn in order to remove any clumped cells.
8. Wait until the sample has completely passed through the col-
9. Add 1 mL of PB to the initial Falcon 15 mL tube (wash step to
get every last cell).
10. Load column with this fraction.
11. Wait until the sample has completely passed through.
12. Collect fi rst fl ow-through fraction and load it on the same col-
umn a second time (again, an effort to capture all multimer-
13. Again, wait until the sample has completely passed through the
14. Wash the column with 3 × 1 or 2 × 3 mL of PB (for MS and LS
15. Wait until the sample has completely passed through: the
collective liquid in the collection tube ( fl ow through fraction)
is your “Depleted fraction.”
16. Remove one column at a time. Place it in a corresponding
labeled Falcon 15 mL tube. Add 2–5 mL (for MS and LS col-
umns, respectively) of PB to the upper fraction of the column.
Push the plunger using steady pressure.
17. Gently remove the plunger.
18. Add again 2–5 mL of PB to the upper fraction of the column.
19. Push the plunger. The collective liquid (total volume is
4–10 mL) is considered the “Enriched fraction.”
20. Spin Depleted and Enriched fractions at 300 × g for 5 min at 4°C.
31723 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
21. For the Depleted fraction ( see Note 19 ):
(a) Resuspend in 1 mL of PBS–2% FCS.
(b) Aliquot 90 μ L in one 5 mL FACS tube and add Ab mix.
(c) Incubate 20 min at 4°C in the dark.
(d) Wash in 3 mL PBS–2% FCS at 300 × g for 5 min at 4°C.
(e) Resuspend in 300 μ L PBS–2% FCS.
22. For the Enriched fraction, you can either continue with ICS
(proceed to Subheading 3.7 ) or prepare your samples for fl ow
cytometry analysis on Subheading 3.5 :
(a) Resuspend in 90 μ L PBS–2% FCS.
(b) Add your Ab mix directly into the Falcon 15 mL tube.
(c) Incubate 20 min at 4°C in the dark.
(d) Add 1 mL of PBS–2% FCS; transfer to 5 mL FACS
(e) Add 1 mL of PBS–2% FCS to the initial Falcon 15 mL
(f) Transfer the 1 mL to the same 5 mL FACS tubes.
(g) Spin at 300 × g for 5 min at 4°C.
(h) Resuspend in 300 μ L PBS–2% FCS.
(Examples for mice and human data are provided in Fig. 5b .)
1. If you came directly from Subheading 3.3 , add your mAb mix
and incubate 20 min at 4°C in the dark ( see Note 20).
Wash in 3 mL PBS–2% FCS at 300 × g for 5 min at 4°C.
Resuspend in 300 μ L PBS–2% FCS.
2. If you have pursued with enrichment on Subheading 3.4 , your
Depleted and Enriched fractions are now ready to be analyzed.
3. Add DAPI to each sample just prior to acquisition (Cf = 1/5,000;
3 μ L of solution 1/50 in 300 μ L of cells).
4. Set stopping gate at 2,000,000 events on Single cells
( SSC-A low /SSC-W low ).
5. Importantly, acquire all samples for Enriched fraction (add
PBS twice) ( see Note 21 ).
6. Gating strategy: SSC-A vs . SSC-W to exclude doublets; Dump
vs . CD3 to isolate viable pure CD3; CD3 vs . CD8 to gate on
CD3 + CD8 + ; Multimer-PE vs . CD8 gated on CD3 to have
background evaluation; Multimer-PE vs . CD8 gated on CD8
to have percentages; any further phenotypic analysis on
Multimer-PE + cells (Fig. 5a ).
7. If you enriched multiple speci fi cities, you will gate on CD8 + PE +
multimer positive cells, then discriminate antigen speci fi city
3.5. Flow Cytometry
318C. Alanio et al.
from another by gating on double positive T cells: PE + color-A +
will be T cells with speci fi city A, PE + color-B + will be T cells
with speci fi city B,… ( see Note 22 and Fig. 6b ).
1. To determine the size of the epitope-speci fi c populations
within each sample, we recommend a precise calculation, ini-
tially proposed by Moon et al. ( 26 ) .
2. The absolute number of total CD8 + T cells within any sample is
determined using the following equation: absolute number of
Fig. 5. Gating strategy and Enrichment. ( a ) Example of the gating strategy applied to an In fl uenza A-Matrix 1 58–66 enriched
human sample. SSC-A low /SSC-W low initial gating permit to exclude doublets, then Dump vs . CD3 contour plot permits to
isolate viable pure CD3 for further analysis. Upper line shows evaluation of background on a Multimer-PE vs. CD8 contour
plot still gated on the total CD3 population. Bottom line illustrates CD3 + CD8 + selection, then fi nally evaluation of
Multimer-PE + cells percentages within CD8 + T cells. ( b ) Left plots : PBMCs from a healthy donor have been incubated with
CMV, Flu or HCV MHCI multimers, then enriched as described in the protocol. Right plots : C57BL/6 mice were immunized
intradermally, either with male (HY model) or K bm1 mOva (Ova model) splenocytes. On day 11, the spleen and lymph nodes
were harvested and enrichment was performed as described in the protocol, using D b -UTY (HY model) or K b -SIINFEKL (Ova
319 23 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
CD8 + sT cells = (number of CD8 + T cells acquired in the pre-
enriched sample) × [(total number of PBMCs in the pre-
enriched sample)/(total number of cells acquired in the single
cell gate in the pre-enriched sample)].
3. The absolute number of multimer-positive T cells is the num-
ber of multimer-positive cells within the “single, live, non-
dump CD3 + CD8 + ” T-cell gate present in the enriched fraction.
( see Note 23 ).
4. The frequency of circulating multimer-positive cells is de fi ned
as the absolute number of multimer-positive T cells/absolute
number of CD8 + T cells.
1. Restimulation of cells is performed in vivo. Three hours prior
to leukocyte harvest, inject mice intravenously with 5 μ g of
CpG/DOTAP formulated as a mixture with 1 μ g speci fi c
peptide (e.g., SIINFEKL peptide in the Ovalbumin model).
2. Perform the staining and enrichment as described in
Subheadings 3.3 and 3.4 with the addition of BD GolgiPlug
containing Brefeldin A during multimer and beads incubation
steps ( fi nal concentration 1/1,000).
3. After elution from the column, resuspend enriched cells in
MPB without azide, add Aqua fl uorescent reactive dye ( fi nal
(Optimized for Mice)
Fig. 6. Multi-enrichment. ( a ) PBMCs from a healthy donor are incubated with a cocktail of MHCI multimers. Each speci fi city
(CMV, Flu, MART1) is labeled with PE and with another color (PE-Cy7, APC, APC-Cy7, respectively). Enrichment is performed
with anti-PE microbeads as described in Subheading 3.4 . ( b ) After applying the gating strategy described in Subheading 3.5
and Fig. 5a , CD8 + PE + cells are gated ( left plot ). Each speci fi city is then identi fi ed within CD8 + PE + population using the
second color readout ( middle and right plots ).
320C. Alanio et al.
concentration 1/1,000) to stain dead cells ( see Note 12 ), and
incubate 30 min at 4°C in the dark.
4. Spin 5 min at 300 × g at 4°C in 3 mL MPB without azide.
5. Resuspend cells in 100 μ L MPB without azide, add the mix of
antibodies for surface staining, and incubate 20 min at 4°C in
6. Wash with 3 mL of MPB without azide and resuspend thor-
oughly cells with 250 μ L of Cyto fi x/Cytoperm reagent. Vortex
and incubate 20 min at 4°C.
7. Wash with 1 mL of 1× Perm/wash buffer.
8. Incubate cells for 30 min at 4°C with anti-IFN γ antibody
diluted in Perm/Wash buffer.
9. Wash cells once with Perm/Wash buffer and once with MPB
10. Resuspend in 300 μ L PBS–2% FCS and acquire sample in fl ow
cytometry (Fig. 7 ).
1. Although the protocol described here focuses on antigen-
speci fi c T cells harvested from human peripheral blood and
from mice, similar procedures can be applied to other nonhuman
Fig. 7. Intracellular cytokine staining on mouse samples. C57BL/6 mice were immunized intradermally with K bm1 mOva
splenocytes. On day 12, mice were injected intravenously with CpG/DOTAP and SIINFEKL peptide. 3 h later, the spleen and
lymph nodes were harvested and Ova-speci fi c T cells were enriched with K b -SIINFEKL multimer, fi xed and stained intracel-
lularly for IFN γ as described in the protocol.
321 23 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
samples ( 25– 27 ) and to other tissues (e.g., tumor in fi ltrating
lymphocytes, TILs). For human peripheral blood, prepare
PBMCs using Ficoll separation. For tissue-based applications,
we recommend including a CD45 staining in one of the chan-
nels in order to segregate CD45-positive hematopoietic cells,
and decrease noise in the assay.
2. Although the protocol described here focuses on HLA–A2
individuals as example, it can be applied to any HLA speci fi city
without modi fi cation. Moreover, enrichment protocols would
also be applicable for CD4 + T cell, NK-T, and γ δ T cell popula-
tions, using respective multimer reagents.
3. Although the protocol described here is based on the combi-
nation of PE-labeled multimers and anti-PE microbeads, you
can similarly stain with APC-labeled multimers and enrich with
4. When establishing the assay on human samples, we found a
better recovery of rare speci fi c T cells when using MS columns,
regardless of the number of loaded PBMC used (1 × 10 7 −4 × 10 8
starting cell populations tested). Exceptions concern TILs, for
which you need to use LS columns in order to avoid clumps
and blockage of the column. Similarly, for mouse experiments,
LS columns are recommended due to potential of stromal tissue
from lymph nodes and spleen to clog the columns.
5. This recipe was chosen based on our experience in the lab.
Other conventional sorting buffers can be used, but may result in
slightly different background signals. Note that sodium azide
should be omitted if planning to cultivate the cells or perform
6. Concerning MHC multimers, there are two options.
Commercial vendors exist and will sell off-the-shelf reagents as
well as generate custom materials. Providers include Beckman
Coulter, ProImmune and Immudex. The alternative and rec-
ommended option is to prepare your own monomers, thus
facilitating multimerization with streptavidin coupled to your
desired fl uorescent tag. This approach allows you to work with
the same multimer labeled in different colors, thus improving
speci fi city of the assay ( see Note 14 ) and permits multiplexing
different speci fi cities in the same tube ( see Notes 16 and 22 ).
Note that regardless of the source, high-quality monomers are
important, with monomer purity impacting multimerization.
Moreover, the choice of streptavidin reagent is critical, and it is
recommended to purchase high-quality streptavidin conju-
gated to bright fl uorochromes.
7. When using MHC multimers, it is recommended to choose an
appropriate method to validate the speci fi city of tetramer-
stained cells. Positive controls will be multimers targeting
abundant populations of CD8 T cells. For human, EBV
322C. Alanio et al.
BMLF1 280–288 or In fl uenza A-M1 58–66 can be used as a positive
control. For mice, the strategy will be to stain splenocytes from
a TCR-transgenic mouse with the corresponding multimer
(OT-I CD8+ T cells that are speci fi c for H2-K b -SIINFEKL
complexes stained with H2-K b -SIINFEKL multimer for
instance). It will help you to establish the assay and ensure that
enrichment is suf fi cient for detection of rare cells. Concerning
naïve cells in humans, MART1 26–35(Leu27) is a good choice, as it
will also be a useful reference for establishing gating parameters
for naïve vs . memory populations ( 30 ) . Negative control
tetramers can be employed to help establish the assay, although
there is the potential to observe CD8 + T cells with the capacity
to bind any MHCI, including self-antigens. An alternative
option is the evaluation of background staining based on the
nonspeci fi c labeling of CD4 + T cells. That said, recent work
has suggested that even this interaction might be of physio-
logic relevance ( 28 ) . More de fi nitive controls are also impor-
tant, such as assessment of TCR CDR3-variable region usage
skewing, peptide-induced TCR downregulation and, after cell
sorting, TCR sequences analysis, or TCR genes transfer into
immortalized cell lines to show that the speci fi city can be
reconstituted ( 8 ) .
8. Biotinylated monomers may be stored for months at −80°C.
Stability testing is recommended. In contrast, multimers are
less stable, should be stored at 4°C and ideally should be used
within 4 weeks. Best is even to multimerize the amount you
will need for each experiment 1 day before.
9. The use of a “Dump channel” is essential as it excludes cells
that bind nonselectively to the MHC multimer reagents. Its
composition has to be reviewed in the context of the experi-
mental aims. For example, CD56 is useful for exclusion of NK
cells in human samples when evaluating naive cell repertoires,
but should be used with caution when studying human mem-
ory or activated T cell populations as some cells express CD56
and would thus be lost in the gating strategy. Similarly, it can
be useful to add anti-CD33 and anti-CD34 antibodies when
studying T cell populations present in bone-marrow popula-
tions. As indicated above, we deliberately do not included CD4
as we use the staining of this population as an assessment of
background, but this can be added when multiple-free chan-
nels are needed for complex multicolor experiments. In the
same way, the Dump channel must to be chosen carefully for
mouse experiments: some markers, such as CD11c, may in fact
upregulate on activated T cells.
10. Addition of a viability marker is necessary in order to avoid
nonspeci fi c background staining on dead cells. While this may
be omitted in some instances in which fresh blood is utilized,
32323 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
it should be noted that the enrichment columns have an af fi nity
for dead cells. Ideally, select a viability dye in the same channel
as the Dump channel—thus keeping the maximum number of
channels free for phenotypic characterization.
11. At least four fl uorescent channels are necessary for careful
assessment of enriched T cells: (a) a Dump channel; (b) CD3
staining for gating on T cells; (c) CD8 staining for gating of
CD8 + T cells; and (d) the multimer-conjugated label for the
speci fi city of interest. This can be extended when differently
labeled multimers are included in the same experiment (for
reducing nonspeci fi c binding in the assay and/or for multi-
plexing enrichment). Anti-CD3 can eventually be omitted if
you really need a maximum of free channels for phenotypic
characterization. Additional channels that are available will
depend upon the technical speci fi cations of the cytometer and
antibodies will be chosen depending on the experimental ques-
tions being evaluated. Note that it is crucial to stain with mul-
timers prior to washing and staining with other Abs, especially
for CD8 and CD4, as some clones have been shown to in fl uence
multimer staining ( 16 ) .
12. As cells will be fi xed to perform intracellular staining, DAPI
cannot be used as a viability marker. We therefore recommend
the use of a fi xable live/dead cell marker such as Aqua. Note
that Aqua labeling has to be done in azide-free buffer.
13. The starting number of cells is a critical point. It is required in
order to calculate precursor frequency; and in most instances it
is the determinant of the limit of sensitivity for the assay. For
example, if you suppose your population to be around 10 −6
(meaning 1 cell into 10 6 CD8), you will need to start from at
least 10 7 cells to maximize the possibility of achieving a well-
de fi ned multimer-positive population. Of note, during the
enrichment procedure, cell loss is in the range of 10–30%.
14. Staining cells with two multimers sharing the same speci fi city
but labeled in different colors permits a further decrease in the
nonspeci fi c binders ( 21 ) . We and others strongly recommend
to include dual labeling, especially if your aim is to detect ultra-
rare populations of cells of variable avidity. Of course, if you
want to pursue with enrichment with anti-PE microbeads, one
of the two colors need to be PE. Otherwise, you can use any
fl uorochrome combination (up to 25) ( 21 ) , as soon as you
titrate both streptavidin and multimers before use, and be cau-
tious with settings and compensations.
15. Concentration of multimers is another important parameter to
consider. Optimal concentrations must be de fi ned for each mul-
timer by titrating on speci fi c cell lines. In general, we recommend
working at high concentrations, i.e., 10–20 nM = 3–10 μ g/mL
324 C. Alanio et al.
fi nal concentration (NB: PE Tetramers ~ 500 kDa; APC
Tetramers ~ 350 kDa). Of note, the receptor density of TCR
on responding T cells is also critical. While not thoroughly
validated, it may be of interest to evaluate exposure to the pro-
tein tyrosine kinase inhibitor dasatinib prior to staining and/or
enrichment as a means of enhancing tetramer binding ( 16 ) .
16. For example, if you aim to stain and enrich T cells speci fi c for
CMV, In fl uenza A, and MART1-speci fi c populations at the
same time, you will put in the same tube, at the same incuba-
tion step, the following multimers: PE-CMV, PE-Cy7-CMV,
PE-Flu, APC-Flu, PE-MART1, APC-Cy7-MART1 ( see Note 22
and Fig. 6a ).
17. Temperature is a critical parameter in multimer staining. It is
advisable to assess the effects of temperature (and time) for
each individual system. In our assays, MHC I staining is
performed at 4° for 30 min. For staining at room temperature,
be cautious with respect to internalization of multimers—
which might interfere with the enrichment procedure.
18. Please note that the number of beads used here is not adjusted
to the number of targeted cells but rather fi xed at a high level,
suf fi cient for enrichment of rare or common populations.
19. We recommend to systematically analyze the Depleted frac-
tion, at least during assay optimization in order to evaluate cell
loss during the procedure (usually 10–30% in our hands).
20. All reagents should be titrated before use.
21. To ensure you can detect rare speci fi c T-cell populations, it is
important to start with suf fi cient number of cells ( see Note 13 ),
and acquire suf fi cient number of events ( 31 ) .
22. In the example provided above ( see Note 16 and Fig. 6 ), after
gating on CD8 + PE-multimer-positive T cells (which will con-
tain the four T cell speci fi cities), CMV-speci fi c T cells may be
segregated based on PE-Cy7 positivity; Flu-speci fi c cells will be
stained with APC; and MART1-speci fi c T cells will be APC-Cy7
labeled. With the increased number of available fl uorescent
channels (e.g., 18-parameter cytometer), one can theoretically
combine up to ten enriched speci fi cities at the same time,
the dif fi culty being the optimization of compensation settings.
The quality of the enrichment procedure, and capacity of
detecting rare events will in fact also depend on ones experi-
ence using the cytometer.
23. This calculation is based on the hypothesis that you recover all
epitope-speci fi c T cells while acquiring your Enriched sample
and should therefore be considered as the lower limit of pre-
cursor frequency. However, this requires rigorous adherence
to the protocol, and in particular the standardization of wash
32523 Tracking Antigen-Specifi c CD8+ T Cells Using MHC Class I Multimers
steps, and consistent acquisition of the Enriched sample.
Mixing studies with known input numbers of monoclonal
TCR transgenic cells in wild-type congenic mix, or T cell clones
into HLA-mismatched PBMCs may be used to evaluate
ef fi ciency and establish in-house criteria ( 23, 25 ) .
1. Arens R, Schoenberger SP (2010) Plasticity in
programming of effector and memory CD8
T-cell formation. Immunol Rev 235(1):190–
205. doi: 10.1111/j.0105-2896.2010.00899.x
2. Schumacher TNM, Gerlach C, Heijst JW
(2010) Mapping the life histories of T cells.
Nat Pub Group 10(9):621–631. doi: 10.1038/
3. Coulie PG, Karanikas V, Lurquin C, Colau D,
Connerotte T, Hanagiri T, Van Pel A, Lucas S,
Godelaine D, Lonchay C, Marchand M, Van
Baren N, Boon T (2002) Cytolytic T-cell
responses of cancer patients vaccinated with a
MAGE antigen. Immunol Rev 188:33–42
4. Lamoreaux L, Roederer M, Koup R (2006)
Intracellular cytokine optimization and standard
operating procedure. Nat Protoc 1(3):1507–
1516. doi: 10.1038/nprot.2006.268
5. Sun Y, Iglesias E, Samri A, Kamkamidze G,
Decoville T, Carcelain G, Autran B (2003) A
systematic comparison of methods to measure
HIV-1 speci fi c CD8 T cells. J Immunol
6. Altman JD, Moss PA, Goulder PJ, Barouch
DH, McHeyzer-Williams MG, Bell JI,
McMichael AJ, Davis MM (1996) Phenotypic
analysis of antigen-speci fi c T lymphocytes.
7. Altman JD, Davis MM (2003) MHC-peptide
tetramers to visualize antigen-speci fi c T cells.
Curr Protoc Immunol Chapter 17:Unit 17.13.
8. Davis MM, Altman JD, Newell EW (2011)
Interrogating the repertoire: broadening the
scope of peptide–MHC multimer analysis. Nat
Rev Immunol 11(8):551–558. doi: 10.1038/
9. Klenerman P, Cerundolo V, Dunbar PR (2002)
Tracking T cells with tetramers: new tales from
new tools. Nat Rev Immunol 2(4):263–272.
10. Romero P, Valmori D, Pittet MJ, Zippelius A,
Rimoldi D, Lévy F, Dutoit V, Ayyoub M,
Rubio-Godoy V, Michielin O, Guillaume P,
Batard P, Luescher IF, Lejeune F, Liénard D,
Rufer N, Dietrich P-Y, Speiser DE, Cerottini
J-C (2002) Antigenicity and immunogenicity
of Melan-A/MART-1 derived peptides as
targets for tumor reactive CTL in human mel-
anoma. Immunol Rev 188:81–96
11. Baitsch L, Baumgaertner P, Devevre E, Raghav
SK, Legat A, Barba L, Wieckowski S,
Bouzourene H, Deplancke B, Romero P, Rufer
N, Speiser DE (2011) Exhaustion of tumor-
speci fi c CD8+ T cells in metastases from mela-
noma patients. J Clin Invest 121(6):2350–2360.
12. Kanodia S, Wieder E, Lu S, Talpaz M, Alatrash
G, Clise-Dwyer K, Molldrem JJ (2010) PR1-
speci fi c T cells are associated with unmain-
tained cytogenetic remission of chronic
myelogenous leukemia after interferon with-
drawal. PLoS One 5(7):e11770. doi: 10.1371/
13. Rosenberg SA, Dudley ME, Restifo NP (2008)
Cancer immunotherapy. N Engl J Med
359(10):1072. doi: 10.1056/NEJMc081511
14. Appay V, Douek DC, Price DA (2008) CD8+
T cell ef fi cacy in vaccination and disease. Nat
Med 14(6):623–628. doi: 10.1038/nm.f.1774
15. Ma Q, Wang C, Jones D, Quintanilla KE, Li
D, Wang Y, Wieder ED, Clise-Dwyer K,
Alatrash G, Mj Y, Munsell MF, Lu S, Qazilbash
MH, Molldrem JJ (2010) Adoptive transfer of
PR1 cytotoxic T lymphocytes associated with
reduced leukemia burden in a mouse acute
Cytotherapy 12(8):1056–1062. doi: 10.3109/
16. Wooldridge L, Lissina A, Cole DK, van den Berg
HA, Price DA, Sewell AK (2009) Tricks with
tetramers: how to get the most from multimeric
peptide-MHC. Immunology 126(2):147–164.
17. Bakker AH, Schumacher TNM (2005) MHC
multimer technology: current status and future
prospects. Curr Opin Immunol 17(4):428–
433. doi: 10.1016/j.coi.2005.06.008
18. Toebes M, Coccoris M, Bins A, Rodenko B,
Gomez R, Nieuwkoop NJ, Van De Kasteele W,
Rimmelzwaan GF, Haanen JBAG, Ovaa H,
Schumacher TNM (2006) Design and use of
conditional MHC class I ligands. Nat Med
12(2):246–251. doi: 10.1038/nm1360
19. Hadrup SR, Toebes M, Rodenko B, Bakker
AH, Egan DA, Ovaa H, Schumacher TNM
326C. Alanio et al. Download full-text
(2009) High-throughput T-cell epitope dis-
covery through MHC peptide exchange.
Methods Mol Biol
20. Newell EW, Klein LO, Yu W, Davis MM
(2009) Simultaneous detection of many T-cell
speci fi cities using combinatorial tetramer
staining. Nat Pub Group 6(7):497–499.
21. Hadrup SR, Bakker AH, Shu CJ, Andersen
RS, van Veluw J, Hombrink P, Castermans E,
Thor Straten P, Blank C, Haanen JB,
Heemskerk MH, Schumacher TN (2009)
Parallel detection of antigen-speci fi c T-cell
responses by multidimensional encoding of
MHC multimers. Nat Methods 6(7):520–526.
22. Chattopadhyay PK, Melenhorst JJ, Ladell K,
Gostick E, Scheinberg P, Barrett AJ,
Wooldridge L, Roederer M, Sewell AK, Price
DA (2008) Techniques to improve the direct
ex vivo detection of low frequency antigen-
speci fi c CD8+ T cells with peptide-major
histocompatibility complex class I tetramers.
23. Alanio C, Lemaitre F, Law HKW, Hasan M,
Albert ML (2010) Enumeration of human
antigen-speci fi c naive CD8+ T cells reveals
24. Day CL (2003) Ex vivo analysis of human
memory CD4 T cells speci fi c for hepatitis C
virus using MHC class II tetramers. J Clin
25. Moon J, Chu H, Pepper M, Mcsorley S,
Jameson S, Kedl R, Jenkins M (2007) Naive
CD4+ T cell frequency varies for different
epitopes and predicts repertoire diversity and
response magnitude. Immunity 27(2):203–
213. doi: 10.1016/j.immuni.2007.07.007
26. Moon JJ, Chu HH, Hataye J, Pagán AJ, Pepper
M, Mclachlan JB, Zell T, Jenkins MK (2009)
Tracking epitope-speci fi c T cells. Nat Protoc
4(4):565–581. doi: 10.1038/nprot.2009.9
27. Obar J, Khanna K, Lefrancois L (2008)
Endogenous naive CD8+ T cell precursor fre-
quency regulates primary and memory
responses to infection. Immunity 28(6):859–
869. doi: 10.1016/j.immuni.2008.04.010
28. Legoux F, Debeaupuis E, Echasserieau K, De
La Salle H, Saulquin X, Bonneville M (2010)
Impact of TCR reactivity and HLA phenotype
on naive CD8 T cell frequency in humans.
J Immunol 184(12):6731–6738. doi: 10.4049/
29. Jenkins MK, Chu HH, Mclachlan JB, Moon JJ
(2010) On the composition of the preimmune
repertoire of T cells speci fi c for peptide–major
histocompatibility complex ligands. Annu Rev
30. Pittet MJ, Valmori D, Dunbar PR, Speiser DE,
Liénard D, Lejeune F, Fleischhauer K, Cerundolo
V, Cerottini JC, Romero P (1999) High fre-
quencies of naive Melan-A/MART-1-speci fi c
CD8(+) T cells in a large proportion of human
histocompatibility leukocyte antigen (HLA)-A2
individuals. J Exp Med 190(5):705–715
31. Roederer M (2008) How many events is
enough? Are you
73A(5):384–385. doi: 10.1002/cyto.a.20549