Natural Killer Cells Kill Virally
Infected Cells
Thus far, we have
dealt with situations that deal primarily with infectious agents that reside in
the extracellular space. But what if an infectious agent manages to enter cells
of the host, where they are protected from the attentions of the soluble PRRs (e.g.,
complement) and are also shielded from phagocytosis by macrophages and
neutrophils? To deal with this situation, another type of immune cell has
evolved – the natural killer (NK) cell, which is endowed with the ability to
inspect host cells for signs of abnormal patterns of protein expression that
may indicate that such cells might be harboring a virus. NK cells are also
capable of killing cells that have suffered mutations and are on the way to
malignant transformation into tumors. Note that although NK cells constitute a
component of the innate response, under certain circumstances they exhibit
immunological memory, a feature usually confined to adaptive responses.
Natural killer (NK) cells kill host cells that appear
abnormal
NK cells are large granular
leukocytes with a characteristic morphology. They choose their victims on the
basis of two major criteria. The first, termed “missing self,”
relates to the fact that practically all nucleated cells in the body express
molecules on their surface called major histocompatibility complex (MHC) proteins.
The latter molecules have a very important role in activating cells of the
adaptive immune system, which we will deal with later in this chapter, but for
now, it is sufficient to know that a cell lacking MHC molecules is not a good
proposition from the perspective of the immune system. NK cells exist as a countermeasure
to such an eventuality and cells lacking the normal pattern of expression of
MHC molecules are swiftly recognized and killed by NK cells. As we saw in the
previous section dealing with interferons, one way in which the expression of
MHC molecules can be reduced is as a consequence of interferon‐responsive gene
products that can interfere with protein translation within cells infected by
viruses, or in the vicinity of such cells.
In addition to reduced
or absent MHC expression, NK cells are also capable of inspecting cells for the
expression of MHC‐related molecules (called nonclassical MHC molecules) and
other proteins that are not normally expressed on cells, but become so in
response to certain stresses such as DNA damage. This scenario represents “altered
self” and also results in such cells being singled out for the
attentions of NK cells, culminating in swift execution. NK receptors have also
been found to be capable of detecting certain viral proteins directly, such as
hemagglutinin from the influenza virus, that qualifies such receptors as
another class of PRRs. There are additional receptors on the surfaces of NK
cells that enable these cells to recognize infected or transformed cells that we will
discuss in Chapter 4. Clearly an NK is not a cell to get on the wrong side of.
NK cells kill target cells via two different pathways
Upon recognition of a
target cell, through either of the mechanisms mentioned in the preceding
section, the NK cell has two main weapons at its disposal, either of which is
sufficient to kill a target cell within a matter of 30–60 minutes (see Video clip 3). In both cases the target
cell dies through switching on its own cell death machinery as a result of
encounter with the NK cell; thus, NK killing represents a type of assisted
cellular suicide. During NK‐mediated killing, killer and target are brought
into close apposition (Figure 1.39) as a result of detection of either missing
self or altered self on the target cell. This can engage either the death
receptor pathway or the granule‐dependent pathway to
apoptosis (Figure 1.40). We shall consider these in turn, although the outcomes
are very similar.
Death receptor‐dependent cell killing
Death receptors are a
subset of the TNF receptor superfamily, which includes the receptors for Fas,
TNF, and TRAIL, and these molecules derive their name from the observation that
ligation of such receptors with the appropriate ligand can result in death of
the cell bearing the receptor (Figure 1.40). When this observation was first made,
it was a fairly astonishing proposition as it suggested that a cell could be
killed through the simple expedient of tickling a membrane receptor in the
correct way. Clearly, this is a very different type of killing compared with
that seen upon exposure of a cell to a toxic chemical or physical stress that
can kill through disruption of normal cellular processes. Here we have a
physiological receptor/ligand system that exists for the purpose of killing
cells on demand– something it has to be said that the immune system does a lot
of. Naturally, this sparked a lot of investigation directed towards
understanding how ligation of Fas, TNF, and related receptors culminates in
cell death and this is now understood in fine detail as a consequence.
Engagement of Fas or TNF receptors with their trimeric ligands results in the
recruitment of a protease, called caspase‐8, to the receptor
complex that becomes activated as a result of receptor‐induced aggregation of
this protease that now undergoes auto activation (Figure 1.41). Activation of caspase‐8 at the receptor then
results in propagation of the signaling cascade in two possible ways, either
via proteolysis of Bid, which routes the signal through mitochondria, or by
direct processing of other effector caspases (caspases‐3 and ‐7) downstream. In each case,
activation of the effector caspases culminates in death of the cell via
apoptosis, which, as we mentioned earlier in this chapter, represents a
programmed mode of cell death. NK cells can kill target cells in a Fas ligand dependent
manner, but ca also kill through the related TNF ligand to some extent.
Granule‐dependent cell killing
NK cells also possess
cytotoxic granules that contain a battery of serine proteases, called granzymes,
as well as a poreforming protein called perforin. Activation of
the NK cell leads to polarization of granules between nucleus and target within
minutes, and extracellular release of their contents into the space between the
two cells followed by target cell death. Polarization of the granules towards
the target cell takes place as a result of the formation of a synapse between
the killer and target that is composed of an adhesion molecule called LFA‐1 and
its cognate receptor ICAM‐1.
Perforin bears some
structural homology to C9; it is like that protein, but without any help other
than from Ca2+ it can insert itself into the membrane of the target,
apparently by binding to phosphorylcholine through its central amphipathic
domain. It then polymerizes to form a transmembrane pore with an annular
structure, comparable to the complement membrane attack complex (Figure 1.41).
This pore then facilitates entry of the additional cytotoxic granule
constituents, the granzymes, which do the actual killing. Perforin deficient
animals are severely compromised in terms of their ability to kill target
cells, as the granule‐dependent pathway no longer functions in the absence of a
mechanism to deliver the granzymes into the target.
Granzymes kill through
proteolysis of a variety of proteins within the target cell. Most of the
killing potential resides in granzymes A and B, with the function of several
additional granzymes (H, K, and M in humans) still unclear. The mode of action of granzyme B is
particularly well understood and it has been found that this protease in
essence mimicks the action of caspase‐8 in the death receptor pathway to
apoptosis, as described above. Thus, upon entry into the target cell, granzyme
B can initiate apoptosis by cleaving Bid or through directly processing and
activating the downstream effector caspases (Figure 1.41). Both routes result
in the activation of the effector caspases that coordinate the dismantling of
the cell through restricted proteolysis of hundreds
of key cellular proteins.
NK cell activity can be enhanced by PAMPs as well as
type I interferons
NK cells also express
a subset of the TLRs that are focused towards detecting PAMPs, such as
double‐stranded RNA, that are typically associated with viruses. TLR3, TLR7,
and TLR8 all appear to be functional in NK cells and upon engagement of these
receptors, NK cells become activated and their killing potential is enhanced. Interferon‐α
and interferon‐β are also important activators of NK cells, the effects
of which can increase the killing activity of such cells by up to 100‐fold
(Figure 1.42). Recall from our earlier discussion of PRRs, especially those
that detect intracellular infections such as the cytoplasmic DNA sensor, STING,
and the viral RNA sensors within RIG‐I‐like receptor family (Figure 1.22 and
Figure 1.23), that activation of these PRRs induces the expression of Type I interferons,
such as IFN‐α and IFN‐β. This is an excellent example of cooperation between
cells of the innate immune system, where cytokines produced by macrophages or
other cells upon detection of a pathogen results in the activation of other
cells, NK cells in the present context, that may be better adapted to dealing
with the infectious threat.
Activated NK cells can amplify immune responses
through production of IFNγ
Another consequence of
the activation of NK cells is the production of another type of interferon,
IFNγ, a very important cytokine that has a set of activities distinct from that
of IFNα and IFNβ. Macrophages respond to IFNγ by greatly enhancing their
microbicidal activities and also by producing other cytokines (such as IL‐12)
that shape the nature of the ensuing immune response by T‐cells within the adaptive immune system (Figure
1.42). Another effect of IFNγ is to enhance the antigen presentation function
of dendritic cells, which is also important for activation of the adaptive
immune system. This cytokine can also influence the type of adaptive immune
response that is mounted by helping to polarize T‐cells towards a particular response pattern; we shall discuss
this issue at length in Chapter 8.
