Natural killer cells: from CD3
S
NKp46
+
to post-genomics
meta-analyses
Thierry Walzer
1,2,3
,Se
´
bastien Jaeger
1,2,3
, Julie Chaix
1,2,3
and Eric Vivier
1,2,3,4
The original definition of NK cells was based on their ‘natural’
cytolytic response against tumor cells and virus-infected cells
in the absence of specific immunization. However, the term
‘natural killer’ reflects neither the education/maturation
requirements before NK cells can kill nor the entirety of their
biological functions. In light of new functional assays, genetic
models and genomics analysis, we propose a more accurate
definition of NK cells. This definition includes the phenotypical
identification of NK cells as CD3
À
NKp46
+
cells across
mammalian species. In general, this attempt to redefine NK
cells also highlights the need to update the operational
definition of cell types in the post-genomic area.
Addresses
1
Centre d’Immunologie de Marseille-Luminy, Universite
´
de la
Mediterrane
´
e, Case 906, 13288 Marseille cedex 9, France
2
Institut National de la Sante
´
et de la Recherche Me
´
dicale (INSERM),
U631, Case 906, 13288 Marseille cedex 9, France
3
Centre National de la Recherche Scientifique, CNRS, UMR6102,
Case 906, 13288 Marseille cedex 9, France
4
Assistance Publique-Ho
ˆ
pitaux de Marseille, Ho
ˆ
pital de la
Conception, 147 bd Baille 13005 Marseille, France
Corresponding authors: Walzer, Thierry (walzer@ciml.univ-mrs.fr);
Vivier, Eric (vivier@ciml.univ-mrs.fr)
Current Opinion in Immunology 2007, 19:365–372
This review comes from a themed issue on
Lymphocyte effector functions
Edited by Hans-Gustaf Ljunggren and Yenan Bryceson
Available online 17th April 2007
0952-7915/$ – see front matter
# 2006 Elsevier Ltd. All rights reserved.
DOI 10.1016/j.coi.2007.04.004
Introduction
Natural killer (NK) cells were identified more than 30
years ago as ‘small lymphocytes of undefined nature
exerting spontaneous selective cytotoxic activity against
Moloney leukemia cells’ — a phenomenon designated as
to ‘natural cytotoxicity’ [1]. This definition was purely
functional and did not exclude the contribution of several
cell lineages to this cytolytic function. Later, several cell
surface markers were discovered, the expression of which
correlated with most of the observed natural cytotoxicity
against the classical tumor cell targets (e.g. YAC-1 in the
mouse and K562 in humans). This led to the proposal in
1986 that ‘NK cells formed a lineage of lymphocytes
which specialized function was the natural cytotoxicity’
[2]. In this article, we revisit the basics of the biology of
NK cells, including their phenotype, their transcriptional
signature and their effector function, in an attempt to
propose an updated and operational definition of this type
of innate lymphocytes.
NK cell phenotype
To date, the cell surface markers used to define NK cells in
human, mouse and rat have been quite different. In
humans, NK cells are classically defined as CD56
+
CD3
À
lymphocytes. In NK cells, CD56 corresponds to the
140 kDa isoform of the neural cell adhesion molecule
(N-CAM) [3]. However, N-CAM is also expressed by
T-cell subsets, muscle cells and neurons, but it is not
expressed by mouse NK cells. In the rat, NK cells express
the activating receptor NKR-P1A, but this molecule is also
expressed by T-cell subsets [4]. In the mouse, the widely
used PK136 antibody reacts with NK1.1, an epitope shared
by the activating receptor NKR-P1C in C57BL/6 mice and
the inhibitory receptor NKR-P1B in SJL mice [5]. How-
ever, NK cells from other mouse strains do not react with
the anti-NK1.1 antibody, because of allelic divergence of
the nkrp1b/c genes [5]. In NK1.1
À
mouse strains, identifi-
cation of NK cells is therefore based on the cell surface
expression of the integrin subunit CD49b that is recog-
nized by DX5 antibodies [6], despite its expression on
T-cell and myeloid subsets [7,8].
These various phenotypic NK cell definitions and the lack
of specific NK cell markers impede the comparison of data
across species and even across mouse strains. For example,
NK1.1
+
DX5
À
TRAIL
+
NK cells that can represent up to
50% of total liver NK cells in C57BL/6 mice [9] would not
be scored as NK cells in BALB/C mice, which do not react
with the NK1.1 antibody. Moreover, none of the markers
used to define NK cells are NK cell-specific but in fact are
expressed at various degrees by subsets of T cells [10]. In
an effort to identify a NK cell marker across mammalian
species, we recently focused our attention on the natural
cytotoxicity receptor (NCR) — NKp46 (also known as
NCR1). NKp46 is selectively expressed on all peripheral
blood NK cells in healthy individuals [11] as well as in
macaques, chimpanzee and all strains of mice tested (TW
et al., unpublished; see Update). Moreover, a NKP46/NCR1
promoter region, which is highly conserved in all mamma-
lian species from opossum to human, drives NK cell
selective expression both in vitro and in vivo in transgenic
mice (TW et al., unpublished; see Update). Thus, we
propose to use the cell surface expression of NKp46 as a
www.sciencedirect.com Current Opinion in Immunology 2007, 19:365–372