Biology Reference
In-Depth Information
donor to generate alloreactive NK cells
in vitro
could not be determined.
Only donors that generated alloreactive NK cells were selected as suitable
donors. Furthermore it is unknown if alloreactive donor NK cells were pres-
ent in the recipient in these subsequent studies. Two other models for pre-
dicting NK alloreactivity have also been studied. The KIR ligand absence
model uses recipient HLA typing to predict the potential for NK alloreac-
tivity. This model assumes that the respective inhibitory KIRs that would
recognize the missing ligand in the recipient are present in the donor. The
receptor ligand model uses donor KIR genotyping and recipient HLA typing
to predict NK cell alloreactivity. This model allows for the potential for NK
cells expressing a known inhibitory receptor to eliminate recipient cells that
lack the receptor for that KIR. This model also allows for NK cell alloreactivity
in HLA-identical transplants in which self-tolerant clones in the donor may
become alloreactive in the post-transplant setting. It is becoming increasing
clear that it is not merely sufficient to be able to predict NK cell alloreactiv-
ity but also to be able to demonstrate it, as not all donor-recipient pairs will
behave as predicted. This is most apparent with Bw4 and KIR3DL1. Rug-
geri and colleagues
[70]
used the KIR ligand incompatibility to predict that
31 donor-recipient pairs had potential Bw4 mismatches but were able to
detect Bw4-specific alloreactive NK clones in only 20 donors. Approximately
90% of individuals are KIR3DL1 positive, so a lack of KIR3DL1 may account
for a small proportion of donors without Bw4-specific clones. Furthermore
a common allele of KIR3DL1, KIR3DL1*004, is not expressed on the surface
[71]
, and differing levels of surface expression of KIR3DL1 alleles have also
been described that may influence generation of Bw4-specific clones; for
example, KIR3DL1*01502 is a stronger allele for Bw4 than KIR3DL1*007
[72]
. Also not all Bw4 alleles may interact with KIR3DL1 as predicted, and it
has been demonstrated that the Bw4 alleles HLA-B*1301 and HLA-B*1302
did not bind to KIR3DL1
[73]
. Furthermore HLA-A alleles that express the
Bw4 epitope have been examined for their ability to bind to KIR3DL1
[73]
and should also be included in analyzing potential Bw4 clones. The Bw4
epitope, which spans amino acids 77-83, can have either an isoleucine (Ile)
at position 80 or a threonine. In HIV, Bw4 alleles with Ile80 have been asso-
ciated with delayed progression to AIDS
[7]
and it could be predicted that
Bw4 alleles with an Ile80 may have stronger interactions with KIR3DL1 and
result in higher numbers of Bw4-specific clones.
340
Disparities with HLA-C mismatching predictions may also occur. Yawata
and colleagues
[72]
examined NK cell receptor repertoire expression and
function in 58 healthy individuals. They found that various combinations
of HLA-C alleles and KIR genes resulted in differing levels of NK cell func-
tion. There was no hierarchy observed between C2 HLA-C alleles and NK
cell function, though NK cells expressing KIR2DL1*004 were poor produc-
ers even when the donor expressed C2. Furthermore, Bari and colleagues
[74]
reported that KIR2DL1 alleles that expressed an arginine in position
245 are stronger alleles. Yawata and colleagues also demonstrated that
KIR2DL3-expressing NK cells from donors who were homozygous for the
C1 allele HLA-C*07 produced high levels of IFN-γ, those expressing the C1
alleles HLA-C*01, *03, *08, or *1403 produced moderate levels of IFN-γ, and
those donors homozygous for HLA-C*1402 were poor producers of IFN-γ.
KIR-HLA combinations that result in lower NK cell function may be due to
Search WWH ::
Custom Search