A Murine Locus on Chromosome 18

Controls NKT Cell Homeostasis and

This information is current as
of November 21, 2009
Th Cell Differentiation
Feng Zhang, Zhiyan Liang, Naoto Matsuki, Luc Van
Kaer, Sebastian Joyce, Edward K. Wakeland and Thomas
M. Aune

J. Immunol. 2003;171;4613-4620
http://www.jimmunol.org/cgi/content/full/171/9/4613

References This article cites 37 articles, 16 of which can be accessed free at:
http://www.jimmunol.org/cgi/content/full/171/9/4613#BIBL

2 online articles that cite this article can be accessed at:

http://www.jimmunol.org/cgi/content/full/171/9/4613#otherarticl
es

Downloaded from www.jimmunol.org on November 21, 2009

Subscriptions Information about subscribing to The Journal of Immunology is
online at http://www.jimmunol.org/subscriptions/

Permissions Submit copyright permission requests at

http://www.aai.org/ji/copyright.html

Email Alerts Receive free email alerts when new articles cite this article. Sign
up at http://www.jimmunol.org/subscriptions/etoc.shtml

The Journal of Immunology is published twice each month by The American Association
of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994.
Copyright ©2003 by The American Association of Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
 The Journal of Immunology

A Murine Locus on Chromosome 18 Controls NKT Cell

Homeostasis and Th Cell Differentiation1

Feng Zhang,*† Zhiyan Liang,‡§ Naoto Matsuki,† Luc Van Kaer,† Sebastian Joyce,†
Edward K. Wakeland,‡§ and Thomas M. Aune2*†
Th cell differentiation is a critical event in the adaptive immune response. C57BL strains develop predominant Th1 responses while
BALB/c develops a predominant Th2 response. To identify quantitative trait loci controlling this variation, we performed Th1/Th2
differentiation assays of F1 ⴛ BALB/c progeny. A single strong quantitative trait locus was identified on chromosome 18, with
weaker effects detectable on chromosomes 5, 12, and 14. By preparing a congenic BALB.B10.D2c18 strain, we were able to
demonstrate that this single locus was sufficient to “repolarize” spleen cell cultures. This difference was not due to intrinsic
differences in CD4ⴙ T cells. Rather, introgression of the chromosome 18 locus into BALB/c disrupted Va14Ja18 NKT cell homeostasis
resulting in the almost complete absence of this T cell subset. Taken together, these data indicate that genes within chromosome 18
control strain-dependent development of Va14Ja18 NKT cells. The Journal of Immunology, 2003, 171: 4613– 4620.

ifferentiation of CD4⫹ Th cells into Th1 or Th2 subsets

Downloaded from www.jimmunol.org on November 21, 2009

is often polarized and either Th1 or Th2 responses tend
to dominate a given adaptive immune response (1– 4).
Preferential development of either Th1 or Th2 responses has pro-
found effects on the types of immune responses that develop in a
host. Examples include development of effective or ineffective im-
munity against a given pathogen or onset of destructive or protec-
tive immunity to a self-Ag (5, 6). Although type and dose of Ag
(environmental factors) can elicit a predominant Th1 or Th2 re-
sponse (7), genetic factors also contribute significantly to Th1/Th2
polarization (8 –10). Genetic factors that govern Th differentiation
may be broadly classified as those either intrinsic or extrinsic to the
cell. Intrinsic factors would be those that directly govern the prop-
erties of individual CD4⫹ T cells and include genetic loci on chro-
mosomes 11 (8, 9) and 16 (10). The tpm1 locus on chromosome 11
controls maintenance of IL-12 responsiveness and promotes Th1
differentiation. Conversely, the locus on chromosome 16 controls
early commitment of CD4⫹ T cells to the Th2 lineage. Extrinsic
factors that control Th differentiation may include levels of polar-
izing cytokines, IL-4 and IL-12 (1– 4), as well as other cytokines,
such as IL-6, that stimulate Th2 differentiation and inhibit Th1
differentiation (11, 12), costimulatory factors (13), or the type or
properties of APCs (14). In early immune responses, NKT cells
and monocytes/macrophages can be a major source of IL-4 and
IL-12 (1– 4), respectively, suggesting that intrinsic regulation of
these cells may potentially polarize Th cells.
*Division of Rheumatology, Department of Medicine, and †Department of Microbi-
ology and Immunology, Vanderbilt University School of Medicine, Nashville, TN
37232; ‡Simmon’s Arthritis Research Center, Department of Medicine, and §Center
for Immunology, University of Texas Southwestern Medical Center, Dallas, TX
75235
Received for publication May 21, 2003. Accepted for publication August 26, 2003.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by Grants from the National Institutes of Health DK 58765
(to T.M.A.), AI 44924 (to T.M.A.), and AI 42284 (to S.J.), the Arthritis Foundation
(to T.M.A.), and the Juvenile Diabetes Research Foundation (to S.J.).
2
Address correspondence and reprint requests to Dr. Thomas M. Aune, Division of
Rheumatology, Department of Medicine, Medical Center North T3219, Vanderbilt
University Medical Center, 1161 21st Avenue South, Nashville, TN 37232. E-mail
address: Thomas.Aune@vanderbilt.edu
Classical examples of strains of mice that develop highly po-
larized Th1 or Th2 responses, in vitro and in vivo, are C57BL
strains and the BALB/c strains, respectively (5, 6). These mice
exhibit markedly different responses to certain pathogens as well
as to self-Ags. It is generally thought that differences in responses
to pathogens or self-Ags by these two strains result from differ-
ences in Th cell polarization. Linkage studies have identified ge-
netic loci that control responses by these two strains to bacterial
pathogens and these loci do not overlap with loci that control in-
trinsic CD4⫹ T cell differentiation (15, 16). Thus, we hypothesized
that genetic loci that regulate Th differentiation through extrinsic
pathways may also contribute to differences in immune responses
to bacterial pathogens or self-Ags by C57BL and BALB/c strains.
Genetic loci that control extrinsic pathways may be detected in
cultures of heterogeneous mixtures of cells (spleen cells) instead of
cultures of highly purified cells used to identify intrinsic factors
(CD4⫹ T cells).
We asked three questions to test our hypothesis. First, what is
the degree of Th1/Th2 polarization in B10.D2 and BALB/c spleen
cell cultures primed and restimulated with anti-CD3 mAb? Sec-
ond, can linkage analysis identify loci on chromosomes 11 or 16 or
other regions of the genome that govern these responses? Third,
should we find loci separate from those previously identified on
chromosomes 11 and 16, do they regulate Th differentiation
through extrinsic or intrinsic factors? First, B10.D2 and
(B10.D2 ⫻ BALB/c)F1 cultures produced highly polarized Th1
responses whereas BALB/c cultures produced highly polarized
Th2 responses. Thus, the B10.D2 Th1 phenotype was the dominant
phenotype. Second, we failed to identify linkage to chromosomes
11 or 16 in our assay systems, but rather identified a predominant
locus that mapped to the middle of chromosome 18. Third, rather
than affecting intrinsic properties of CD4⫹ T cells, our analysis
demonstrates that introgression of the B10.D2 chromosome 18 lo-
cus into BALB/c causes an almost complete absence of NKT cells
in thymus, spleen, and liver. Thus, genetic factors that influence
extrinsic events predominate over intrinsic events to control Th
polarization in cultures of heterogeneous mixtures of lymphocytes.
Taken together, these data argue that multiple genetic loci are re-
sponsible for differences in Th differentiation and immune re-
sponses in C57BL and BALB/c strains and localize the position of

Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00

4614
FIGURE 1. Th1 differentiation is the dominant response. Cultures of B10.D2, BALB/c, and (B10.D2 ⫻ BALB/c)F1 splenocytes were stimulated with
anti-CD3 (1 ␮g) and 100 U/ml human recombinant IL-2. After 5 days, cultures were restimulated with plate-bound anti-CD3. Culture fluids were harvested
after 48 h and analyzed for levels of Th2 cytokines: IL-4 (A), IL-5 (B), IL-10 (C), and Th1 cytokines: IFN-␥ (D).
a potent genetic polymorphism affecting NKT cell development
and Th differentiation to a specific congenic interval.
Materials and Methods
Mice
B10.D2, BALB/c, or DBA2 mice were obtained from The Jackson Laboratory
(Bar Harbor, ME) and were used between 6–10 wk of age. (B10.D2 ⫻ BALB/
c)F1, F1 ⫻ BALB/c (back-crossed (BC)1), BALB.B10.D2c18 congenic,
(BALB ⫻ BALB.B10.D2c18)F1, (B10 ⫻ BALB.B10.D2c18)F1, and BALB/c
Cd1d⫺/⫺ mice (17) were bred in our animal facility at Vanderbilt University
(Nashville, TN) and used between 6 and 10 wk of age.
Lymphocyte cultures
Spleen cells (5 ⫻ 106 cells/ml) were suspended in RPMI 1640 medium
with 10% FCS, Penn/Strep, L-glutamine, and 2-ME (5 ⫻ 10⫺5 M) and were
stimulated with anti-CD3 mAb (1 ␮g/ml 2C11; American Type Culture
Collection, Rockville, MD) and 100 U/ml human IL-2 (Hoffmann-La-
Roche, Nutley, NJ). After 5 days, cultures were washed, resuspended in
fresh medium, and equivalent numbers of viable cells were restimulated
with plate-bound anti-CD3 mAb (10 ␮g/ml coating solution in 0.1 M
NaHCO3). Culture fluids were harvested after 48 h and analyzed for cy-
tokine levels by ELISA using mAbs from BD PharMingen (San Diego,
CA) according to their recommended protocols. Sensitivity of ELISAs was
10 –20 pg/ml. Alternatively, naive CD4⫹ T cells (CD4⫹, CD62L⫹, 2.5 ⫻
105 cells/ml), purified by both negative (CD8⫺, Ia⫺) and positive
FIGURE 2. Th1/Th2 phenotype of individual BC1 progeny. Individual
BC1 progeny were analyzed for production of IL-4, IL-5, IL-10, and IFN-␥
and assigned either a Th1 (high IFN-␥ levels and low Th2 cytokine levels)
or Th2 (high Th2 cytokine levels and low IFN-␥) phenotype.
NKT AND Th CELLS AND QTL ON CHROMOSOME 18
(CD62L⫹) selection using MACS magnetic beads (Miltenyi Biotec, Au-
burn, CA) and were cultured with or without IL-4 (5 ng/ml) for 5 days
before restimulation with plate-bound anti-CD3.
Genetic analysis
Downloaded from www.jimmunol.org on November 21, 2009
Genomic DNA was prepared from kidney tissue using standard techniques
(18). Primers of simple sequence repeat markers were obtained from Re-
search Genetics (Huntsville, AL) as MapPairs. Genomic DNA was ampli-
fied as previously described (19). Samples were either resolved on 4%
agarose gels (size differences ⬎8 bp) or on 8% polyacrylamide plus urea
gels (size differences ⬍8 bp). Linkage relationships between the 73 poly-
morphic markers were determined by analyzing their segregation patterns
among the 83 BC1 progeny representing the phenotypic extremes using the
MAPMAKER-EXP (version 3) (20, 21) computer package as described.
Generation of BALB/c congenic strains
Congenic strains were derived using a combination of marker assisted se-
lection protocols (22) and screening progeny for presence or absence of the
B10.D2 Th1 trait (low/high IL-4 production by PBLs in tissue culture
assays). We prepared three F1 generations from three independent BALB/c
and B10.D2 founders. Each founder line was used to generate separate
progeny by back-crossing to independent BALB/c mice. At each subse-
quent generation, we screened individual progeny for the B10.D2 trait. We
also performed genome-wide genotyping using simple sequence repeat
markers. At each generation, progeny that contained the B10.D2 trait (low
IL-4) and the least amount of contaminating B10.D2 genome were selected
as parents for the next generation. Initial average genome spacing of simple
sequence repeat markers across the genome was ⬃20 cM. Where we found
recombination events in selected progeny, we performed more detailed fine
mapping across this region in the next generation to closely monitor elim-
ination of this contaminating genomic region. Primers were obtained from
Research Genetics. PCR products of tail DNA samples were analyzed on
4% agarose gels. BALB.B10.D2c18 congenic mice used for experiments
were from the N9 and N10 generations.
Stimulation of Va14Ja18 NKT cells in vivo
Mice were injected in the tail vein with ␣-galactosylceramide (␣-GalCer)3
(10 ␮g/mouse in 100 ␮l of 0.25% Tween 20 in PBS) or with vehicle alone.
After 90 min, sera and spleens were harvested. Single spleen cell suspen-
sions were prepared in complete medium and after 90 min of culture,
supernatant fluids were harvested. IL-4 levels in sera and culture fluids
were measured by ELISA. Kirin Brewery (Gunma, Japan) kindly provided
synthetic ␣-GalCer.
Detection of NKT cells by flow cytometry
Single cell suspensions of thymus, spleen, and liver were prepared by gen-
tle homogenization. Hepatocytes were centrifuged at 1200 ⫻ g for 5 min,
resuspended in RPMI 1640 (with 10% FCS) and underlaid below a Ficoll-
Hypaque solution. Cells were centrifuged for 20 min at 2000 ⫻ g. Intra-
hepatic mononuclear cells were collected from the interface. Cells were
subsequently washed once with RPMI 1640. One million cells were used
3
Abbreviations used in this paper: ␣-GalCer, ␣-galactosylceramide; QTL, quantita-
tive trait loci; SSR, simple sequence repeats; LOD, logarithm of odds.
The Journal of Immunology 4615

Table I. Linkage analysis of QTL associated with the Th1/Th2 responsea

LOD/p Value

Chromosome Distance, cM Locus IFN-␥ IL-4 IL-5 IL-10

18 11.0 D18mit68317 2.0/0.002 1.4/0.013

1.0 D18mit173149 2.1/0.002 1.4/0.019
18.4 D18mit1493207 2.4/0.003 1.5/0.012
16.4 D18mit2073184 2.0/0.003 1.5/0.012
5 12.2 D5mit387381
35.9 D5mit813314 1.1/0.20
28.3 D5mit314331 1.2/0.022
12 6.5 D12mit1363285
19.9 D12mit2853277 1.6/0.010
14 15.7 D14mit1333142
13.8 D14mit1423158 1.2/0.051
29.2 D14mit1583195 1.3/0.017
a
Distances are in centimorgans (Kosambi function) between two consecutive markers as calculated by MAPMAKER.EXP.
Markers are arranged from centromeric to telomeric. The LOD likelihood ratio scores were analyzed by MAPMAKER.QTL and
p values by ANOVA analysis.

for staining with APC-conjugated CD1d1-␣-GalCer tetramer (CD1 tetram- total of 83 BC1 progeny selected from the phenotypic extremes in
ers) and PerCP-conjugated anti-CD3⑀. PE-conjugated anti-HSA and anti- cytokine production were genotyped with a panel of 73 poly-

Downloaded from www.jimmunol.org on November 21, 2009

CD8 were used to gate out immature cells and CD8⫹ cells in thymus. morphic SSR marker loci. Linkage groups were identified using
PE– conjugated anti-B220 and anti-CD8 were used to gate out B cells
and CD8⫹ cells in spleen and liver. Labeled cells were analyzed with a MAPMAKER-EXP as previously described (20, 21). To minimize
FACStarPlus flow cytometer using the CellQuest program (BD Biosciences, experimental variability, groups of 30 – 40 age-matched BC1 were
San Jose, CA). analyzed in individual experiments. Several of these experiments
were grouped for analysis of cytokine levels. Cytokine assays were
Results performed at least twice.
B10.D2 and BALB/c spleen cells exhibit extremes in production The positions of quantitative trait loci (QTL) affecting levels of
of Th1/Th2 cytokines individual cytokine levels were identified using MAPMAKER-
Spleen cells from B10.D2, BALB/c, or F1 mice were stimulated QTL (Table I). A single locus was identified in the middle of
with anti-CD3 (2C11) and 100 U/ml IL-2. After 5 days, cultures chromosome 18, most closely linked to D18mit149 and D18mit207
were harvested and equivalent numbers of viable cells were re- (log of odds (LOD) 2.4 for IFN-␥). Linkage to levels of IL-4, IL-5,
stimulated with plate-bound anti-CD3. Culture fluids were har- and IL-10 was less evident (IL-4, LOD 0.04; IL-5, 0.9; IL-10, 1.5).
vested 48 h later for analysis of cytokine levels. Secondary B10.D2 In addition to this QTL for IFN-␥ levels, three additional intervals
cultures contained high levels of IFN-␥ but low levels of IL-4, showed weaker linkage to cytokine levels. A region on chromo-
IL-5, and IL-10 when compared with BALB/c cultures (Fig. 1). some 12 also showed linkage to IFN-␥ levels (D12mit285 to
The F1 cultures and B10.D2 cultures contained comparable levels D12mit277; LOD 1.6) and a region on chromosome 14,
of IFN-␥, IL-4, IL-5, and IL-10. These data demonstrate that the D14mit142 to D14mit195, showed weak linkage to levels of IL-5
B10.D2 phenotype of high IFN-␥, low IL-4, IL-5, and IL-10 is the (LOD 1.2). A region on chromosome 5 also showed weak linkage
dominant phenotype. to levels of IL-5 (D5mit314 to D5mit31; LOD 1.2). Surprisingly,
To identify the number of B10.D2 loci that control this pheno-
typic trait, we crossed F1 mice to the recessive BALB/c strain to
generate a cohort of back-crossed mice (BC1). Phenotypic analysis
was performed by culturing spleen cells as outlined in Fig. 1. An
example of representative data is shown (Fig. 2). Generally, anal-
ysis of individual BC1 progeny revealed that levels of Th1 cyto-
kines (IFN-␥) and Th2 cytokines (IL-4, IL-5, and IL-10) exhibited
a reciprocal phenotype in individual mice. Extremes in levels of
cytokine production were greatest for IL-4; and achieved over 100-
fold differences among the progeny. Other cytokine levels did not
differ to this same degree. For example, average differences in
extremes of IFN-␥ levels were closer to 10-fold and even less for
IL-5 and IL-10. Although not shown, BC1 progeny did not always
yield parallel values of the Th2 cytokines, IL-4, IL-5, and IL-10.
For example, some progeny produced high levels of IL-4 but low
levels of IL-5 or IL-10. Thus, it is possible that multiple loci con-
tribute to the regulation of production of certain cytokines gener-
ally classified as either Th1 or Th2 cytokines. FIGURE 3. BALB.B10.D2c18 congenic mice exhibit a Th1 rather than
a Th2 phenotype. BALB.B10.D2c18 congenic mice were developed by
Linkage analysis marker assisted selection protocols as outlined in Materials and Methods
and in the study. Th1/Th2 differentiation assays were performed as in Fig.
A genome-wide mapping analysis using microsatellite markers 1. Results are expressed as the percentage of the BALB/c response. The
(simple sequence repeat (SSR) markers) was performed to identify actual cytokine levels produced by splenocytes from either BALB/c or
loci that controlled levels of expression of individual cytokines. A BALB.B10.D2c18 mice (n ⫽ 5 each) are shown above the figure.
4616
FIGURE 4. The introgressed region on chromosome
18 contains both B10 and DBA/2 loci. The indicated
SSR markers were used for the analysis of a series of
backcrossed nine-generation congenic BALB.B10.D2c18
mice. Actual PCR sizes (bp) of different SSR markers in
BALB, C57BL10, and DBA/2 strains and the PCR re-
sults from the BALB.B10.D2c18 mice are shown. Num-
bers in bold show contaminating DBA/2 segment of
chromosome 18. Expected and observed results for
BALB, B6, and DBA strains were equal. The location of
the introgressed B10 (white), and DBA2 (gray) regions
on chromosome 18 of the BALB.B10.D2c18 strain are
indicated. LOD scores from Table I are aligned inMb
across chromosome 18.
we were unable to identify strong linkage groups for levels of IL-4
even though levels of IL-4 exhibited the strongest phenotypic dif-
ferences among the BC1 progeny. We did not detect significant
linkage to regions on chromosomes 11 or 16.
BALB/c congenic strains heterozygous for B10.D2 chromosome 18
Because the previous linkage analysis including 83 progeny may
be underpowered, we considered screening additional BC1 prog-
eny. Concurrently, we initiated attempts to produce a BALB/c con-
genic line that carried the B10.D2 Th1 trait. Our analysis of BC2
(BC1 ⫻ BALB/c) and BC3 (BC2 ⫻ BALB/c) progeny indicated
we could identify BALB/c progeny with the B10.D2 Th1 trait.
Therefore, we elected to devote further efforts to production of this
congenic line rather than the screening additional BC1 progeny. To
assess the effect of the predominant QTL on chromosome 18 on
Th1/Th2 responses in the BALB/c strain, we prepared several in-
dependent congenic lines using marker assisted selection proto-
cols. These strains, BALB.B10.D2c18, contained ⬍1% contami-
nating B10.D2 genome outside the introgressed region
(chromosome 18; distances 11– 41 cM). Each of these congenic
lines exhibited similar properties in our assays as further described.
Congenic BALB/c mice were compared with wild-type BALB/c
mice to determine whether the introgressed region from B10.D2
would alter cytokine profiles in tissue culture assays. The BALB/c
congenic strain containing a single copy of the introgressed
B10.D2 chromosome 18 region exhibited a cytokine profile com-
parable to B10.D2; it had suppressed IL-4 cytokine production and
enhanced IFN-␥ production (Fig. 3). Levels of IL-5 and IL-10
were also depressed in cultures from the congenic strain compared
with BALB/c. However, levels of these cytokines were substan-
tially higher than observed in the B10.D2 or (BALB/c ⫻
NKT AND Th CELLS AND QTL ON CHROMOSOME 18
B10.D2)F1 strain (see Fig. 1). This suggests that additional alleles
from B10.D2 either separately, or together with the interval on
chromosome 18, regulate levels of these two cytokines. The weak
QTL that mapped to regions of chromosomes 5 or 14 may contain
Downloaded from www.jimmunol.org on November 21, 2009
one or more genes that regulate IL-5 and IL-10 levels.
Fine mapping of the B10.D2c18 interval
To perform fine mapping studies, a series of back-crossed nine
generations congenic BALB.B10.D2c18 mice were produced and
analyzed for recombination events across the chromosome 18 in-
terval. These mice were also analyzed for cytokine profiles (IL-4
and IFN-␥) in tissue culture assays as previously outlined and as-
signed either a BALB/c or a B10.D2 phenotype. We used the in-
dicated SSR markers for the analysis (Fig. 4). Upon analysis of the
congenic interval, primers for a series of the SSR markers
(D18mit53, D18mit55, D18mit123, D18mit124, and D18mit52)
did not produce PCR products of the expected size for BALB/c
and B6 strains. B10.D2 was derived from a cross between B10 and
DBA2 strains and back-crossed four generations to B10. There-
fore, we suspected that the unexpected typing results were from
DBA2 contamination. To test this possibility, we rechecked the
size of PCR products and compared them to the size of PCR prod-
ucts from BALB/c, B10, DBA2, and BALB.B10.D2c18 strains.
We identified five microsatellite markers that were clearly from
DBA2, D18mit53, D18mit55, D18mit123, D18mit124, and
D18mit52 (Fig. 4).
Comparison of the results of fine mapping and the cell culture
assays clearly demonstrated that the locus on chromosome 18 that
controls reciprocal Th1/Th2 differentiation is located between 54.1
and 57.8 Mb from the centromere (Fig. 5). This map position is
FIGURE 5. Fine mapping of the chromosome 18 lo-
cus that regulates Th1/Th2 differentiation in the BALB.
B10.D2c18 strain. A cohort of (BALB.B10.D2c18 ⫻
BALB/c)F1 mice were screened for recombination events
using the following SSR markers: D18mit11, D18mit17,
D18mit149, D18mit150, D18mit124, D18mit152, and
D18mit209. Genotypes heterozygous (he) (both B10 and
BALB/c alleles) or homozygous (ho) (only BALB/c al-
lele) at each marker are shown as well as the distance
(cM) and position (Mb) of each marker. Cultures were
performed and analyzed as in Fig. 1, and levels of IL-4
and IFN-␥ were measured after secondary stimulation
(ng/ml ⫾ SD of triplicate cultures; data are pooled from
all mice with a given genotype). Number of mice (n)
analyzed with each genotype is shown.
The Journal of Immunology
similar to the map position identified by the linkage analysis (Ta-
ble I). Mice that had recombination events between 54.1 and 57.8
Mb were easily segregated into those that had the BALB/c phe-
notype (high IL-4, low IFN-␥) and those that had the B10.D2
phenotype (low IL-4, high IFN-␥). Additional fine mapping within
this region should help further localize this trait.
Absence of intrinsic defects in Th2 polarization by
BALB.B10.D2c18 CD4⫹ T cells
The strongest phenotypic effect of the B10.D2c18 QTL on Th
differentiation was its reciprocal influence on IL-4 and IFN-␥ pro-
duction in the secondary spleen cell cultures compared with
BALB/c. Therefore, we wanted to probe this mechanism in more
detail. First, we determined whether BALB/c splenocytes required
IL-4 to generate polarized Th2 responses by adding anti-IL4 to
primary cultures. Addition of anti-IL4 mAb to primary cultures
diminished IL-4 production in secondary BALB/c cultures by
⬎90% (64 ⫾ 5 ng/ml to 4 ⫾ 1 ng/ml). Second, we asked whether
there were differences in polarization when highly purified
BALB/c or BALB.B10.D2c18 CD4⫹, CD62L high, naive CD4⫹ T
cells were cultured with or without IL-4. Equivalent levels of IL-4
were produced in primary cultures of T cells from both strains. In
addition, equivalent levels of IL-4 were detected in secondary cul-
tures of primed naive CD4⫹ T cells from both strains if primed
with or without IL-4 (Fig. 6). These results argue against an in-
trinsic difference in CD4⫹ T cells between the two strains and
hence, do not account for marked differences in Th2 polarization
by splenocyte cultures.
Absence of NKT cells in BALB.B10.D2c18 mice
In splenocyte cultures, NKT cells may produce sufficient quantities
of IL-4 to polarize Th2 responses by conventional CD4⫹ T cells
(17, 23–26). Therefore, we wanted to directly compare the func-
tion and presence of NKT cells between BALB/c and
BALB.B10.D2c18 strains. First, we tested the ability of ␣-GalCer,
a specific ligand for Va14Ja18 NKT cells, to induce IL-4 produc-
tion, in vivo, in both strains. In the BALB/c mice, we detected
substantial amounts of IL-4 produced in serum 90 min after ␣-Gal-
Cer injection (Fig. 7) and after 90 min of culture of splenocytes, in
vitro (data not shown). In contrast, IL-4 was not detected in serum
(Fig. 7) nor splenocyte cultures (data not shown) from BALB.
B10.D2c18 mice.
FIGURE 6. BALB.B10.D2c18 naive CD4⫹ T cells do not have an in-
trinsic defect in Th2 differentiation. Naive CD4⫹, CD62L⫹ T cells were
purified from BALB.B10.D2c18 and BALB/c mice and stimulated with
anti-CD3, anti-CD28 with or without IL-4. After 5 days, equivalent num-
bers of viable cells were restimulated with anti-CD3. Culture fluids were
harvested 48 h after primary or secondary stimulation. IL-4 was measured
by ELISA.
4617
FIGURE 7. Defective function of BALB.B10.D2c18 NKT cells in vivo.
Mice of the indicated strains were injected in the tail vein with ␣-GalCer
(10 ␮g/mouse). Sera were harvested after 90 min. IL-4 was measured by
ELISA. Data are pooled from three separate experiments using three mice
per strain in each experiment. Error bars represent SD (BALB.B10.D2c18
is shown but is too small to see). Statistical analysis (ANOVA) for BALB/c
vs BALB.B10.D2c18, p ⬍ 0.0001.
Defects in NKT cell IL-4 production may result from the absence
or dysfunction of NKT cells in the congenic BALB.B10.D2c18 strain.
Single cell suspensions of thymus, spleen, and liver were labeled with
Downloaded from www.jimmunol.org on November 21, 2009
an Ab against CD3 and CD1 tetramers and analyzed by flow cy-
tometry. Va14Ja18 NKT cells were almost undetectable in
BALB.B10.D2c18 mice (Fig. 8A). In contrast, similar numbers of
NKT cells were detected in BALB, B10D2, and DBA2 strains. Sim-
ilar results were obtained from all three organs. We also directly mea-
sured numbers of V␤8.1, -8.2, CD1 tetramer double positive cells in
thymus, spleen, and liver by flow cytometry and obtained similar
results: almost complete absence of NKT cells (data not shown). To
confirm these results, we measured levels of rearranged Va14Ja18
mRNA in thymus by RT-PCR. We were unable to detect Va14Ja18
mRNA in BALB.B10.D2c18 thymus but Va14Ja18 mRNA was
readily detected in BALB/c thymus (data not shown).
We have initiated fine-mapping studies to determine whether the
absence of NKT cells in the congenic BALB.B10.D2c18 strain is
also localized to the chromosome 18 introgressed region. Although
these studies are incomplete, our initial recombinant analysis in-
dicates that we can localize this trait between D18mit55 and
D18mit124. These data support the notion that the traits of low
IL-4 and low NKT cell numbers may colocalize.
Absence of NKT cells in (BALB/c ⫻ BALB.B10D2c18)F1 and
(B10 ⫻ BALB.B10D2c18)F1 mice
Although the BALB.B10D2c18 strain is ⬎99% BALB/c outside
the chromosome 18 interval, we wanted to further exclude the
possibility that the ⬍1% contamination by the B10.D2 genome
may cause the absence of NKT cells. Several (BALB/c ⫻
BALB.B10D2c18)F1 and (B10 ⫻ BALB.B10.D2c18)F1 mice
were analyzed for numbers of NKT cells. Thymic, splenic, and
hepatic NKT cells were undetectable in F1 mice (Fig. 8B). These
results further argue that the B10.D2 chromosome 18 interval ac-
counts for the almost complete absence of NKT cells observed in
the congenic strain.
BALB/c Cd1d⫺/⫺ and BALB.B10.D2c18 splenocytes exhibit
equivalent defects in IL-4 production
Cd1d⫺/⫺ mice are also deficient in NKT cells. Therefore, we
wanted to determine whether absence of NKT cells induced by a
completely independent method would result in the same defect in
Th2 polarization as observed in BALB.B10.D2c18 splenocyte cul-
tures. Splenocytes from both strains were cultured as previously
outlined and cultures were analyzed for IL-4 levels. Cultures from
both strains exhibited extreme defects in IL-4 production when
4618 NKT AND Th CELLS AND QTL ON CHROMOSOME 18

Downloaded from www.jimmunol.org on November 21, 2009

FIGURE 8. The introgressed chromosome 18 locus alters NKT cell homeostasis. A, Single cell suspensions of thymus, spleen, and liver prepared from
the indicated strains (8-wk-old mice) were stained with APC-conjugated CD1 tetramer and PerCP-conjugated anti-CD3⑀. PE-conjugated anti-HSA and
anti-CD8 were used to gate out immature cells and CD8⫹ cells in thymus. PE-conjugated anti-B220 and anti-CD8 were used to gate out B cells and CD8⫹
cells in spleen and liver. One representative of three separate experiments is shown. B, (BALB ⫻ BALB.B10.D2c18)F1 and (B10 ⫻ BALB.B10.D2c18)F1
mice also have altered NKT cell homeostasis. Thymus, spleen, and liver were harvested from the indicated mice and stained for NKT cells as outlined in
(A). One representative experiment of a total of three experiments is shown. All mice were age matched (6 to 8 wk old).

compared with BALB/c (Fig. 9). These data demonstrate that Th2
differentiation in spleen cell cultures is strongly regulated by the
level of NKT cells. We also directly measured CD1d levels in
thymus, spleen, and liver by flow cytometry. BALB/c and
BALB.B10.D2.c18 expressed similar levels of CD1d (data not
shown). Taken together, these results indicate that defects in NKT
cell homeostasis in the BALB.B10.D2c18 strain probably account
for defects in IL-4 production and Th2 polarization between the
congenic strain and the parent BALB/c strain. These results also
suggest that there are strain-dependent variations in NKT cell de-
velopment or homeostasis that are controlled by genes within
chromosome 18.
FIGURE 9. BALB.B10.D2c18 and Cd1d⫺/⫺ splenocyte cultures ex-
hibit comparable defects in Th2 differentiation. Cultures of wild-type
BALB/c, BALB/c Cd1d⫺/⫺, and BALB.B10.D2c18 splenocytes were
stimulated with anti-CD3 (1 ␮g) and 100 U/ml human recombinant IL-2.
After 5 days, cultures were restimulated with plate-bound anti-CD3. Cul-
ture fluids were harvested after 48 h and analyzed for levels of IL-4 cyto-
kines by ELISA.
Discussion
We have identified a single locus within mouse chromosome 18
that is responsible for the regulation of production of both Th1
(IFN-␥) and Th2 (IL-4, IL-5, and IL-10) cytokines in B10.D2 and
BALB/c strains in our assay systems. Mapping data did not show
strong linkage to this region for all cytokines. However, analysis of
the congenic BALB/c strain that contains one copy of the middle
region of B10.D2 chromosome 18 clearly demonstrates that this
single B10.D2 locus is sufficient to change polarization of spleen
cell cultures from the BALB/c phenotype (Th2) to the B10 phe-
notype (Th1). Analysis of congenic mice for recombination events
maps this locus between 54.1 and 57.8 Mb from the P terminus of
chromosome 18. Our mapping data also suggest that additional
loci may exist on chromosomes 5, 12, and 14 that may contribute
to this phenotype. These loci may cooperate with the chromosome
18 locus to suppress the Th2 response further than that already
observed in the chromosome 18 single congenic mice.
Taken together, our data support the notion that the B10.D2c18
locus, when introgressed into the BALB/c strain, regulates NKT
cell development and/or homeostasis resulting in the virtual ab-
sence of NKT cells in the congenic strain. Absence of NKT cells
causes the shift in the predominant Th2 cytokine profile to a pre-
dominant Th1 profile in splenocyte cultures. First, we were unable
to detect any intrinsic difference in the ability of purified naive
BALB/c or BALB.B10.D2c18 T cells to differentiate into Th1/Th2
cells. Second, IL-4 was required to polarize BALB/c splenocyte
cultures to the Th2 phenotype and IL-4 levels were much higher in
BALB/c splenocyte cultures than in purified T cell cultures. Third,
it has been clearly demonstrated that NKT cells can produce suf-
ficient quantities of IL-4 in splenocyte cultures to direct the po-
larization of splenic T cells to the Th2 lineage (26). Fourth,
Cd1d⫺/⫺ mice showed the exact same defect in Th2 differentiation
The Journal of Immunology
as congenic mice. Fifth, our additional recombinant analyses in-
dicate that genes that cause reduced levels of NKT cells found in
the BALB.B10.D2c18 strain are found between D18mit55 and
D18mit124. Therefore, we predict that altered NKT cell develop-
ment and/or homeostasis observed in the BALB.B10.D2c18 and
altered Th1/Th2 development will be localized to identical
genomic regions on chromosome 18. Further fine-mapping exper-
iments are in progress to directly test this.
BALB/c, B10, B10.D2, and DBA/2 all contain similar levels of
NKT cells in thymus, spleen, and liver. In contrast, NKT cells are
almost undetectable in BALB.B10.D2c18 thymus, spleen, and
liver and we know that this is due to introgression of the
B10.D2c18 locus. One possible explanation for this is that the
introgressed chromosome 18 locus has a unique dominant muta-
tion in a gene critical for NKT development and this mutation is
not present in the parent B10.D2 strain. An alternative explanation
is that this is actually a result of selectively breeding the
B10.D2c18 allele onto the BALB/c background. If so, a genetic
model to account for these results is more complex than a single
locus model. BALB congenic strains that are either homozygous or
heterozygous for the introgressed chromosome 18 region exhibit
similar disruptions in NKT cell number so the chromosome 18
locus must be a dominant trait. The fact that the same effects are
not observed in B10.D2 or in DBA2 argues that there must be a
second locus in these strains that suppresses the effects of the chro-
mosome 18 locus on NKT development and/or homeostasis. We
hypothesize that this second locus must be absent in BALB. To
provide genetic evidence for this chromosome 18-modifying locus,
we crossed BALB.B10.D2c18 mice with B10 or BALB/c. Both F1
mice lack NKT cells. The fact that F1 mice also lack NKT cells
argues that the activity of this hypothesized chromosome 18-mod-
ifying locus must represent a recessive trait. These results bear
certain similarities to the epistatic modifiers of lupus development
described in the NZB/NZW/NZM strains (27). New Zealand Black
(NZB) and New Zealand White (NZW) strains are relatively lupus
resistant, whereas (NZB ⫻ NZW)F1 mice develop fatal lupus ne-
phritis. The F1 strain does not develop lupus because of the con-
tribution of dominant lupus-causing alleles from both strains.
Rather, the major lupus causing loci are found in NZW. However,
NZW also contains lupus-suppressing loci that are recessive (27).
The F1 strain loses the suppressive effect of these recessive loci
that are absent in NZB and the result is severe fatal lupus. This
model predicts that recessive epistatic modifiers exist in both B10
and DBA2 that suppress the activity of the chromosome 18 locus
and these epistatic modifiers are absent in BALB. We are in the
process of testing this model by back-crossing BALB.B10.D2c18
mice with DBA2 mice and screening for progeny that lack NKT
cells.
It is reasonable to expect that QTL controlling NKT cell func-
tion and Th cell differentiation may also play a role in immune-
mediated diseases. Loci associated with susceptibility to murine
systemic lupus erythematosus and murine experimental autoim-
mune encephalomyelitis are also localized to chromosome 18 in-
tervals that overlap this QTL (28, 29). This region of mouse chro-
mosome 18 is also syntenic with regions of human chromosomes
5 and 18 (30). Linkage studies have identified autoimmune disease
susceptibility loci in the middle of human chromosome 18 for a
variety of autoimmune diseases including insulin dependent dia-
betes mellitus, rheumatoid arthritis, systemic lupus erythematosus,
and Graves disease (31–34). In addition, linkage studies of human
asthma, dermatitis, and conjunctivitis have identified loci in the
region of 5q31–33 that are syntenic with this same region on
mouse chromosome 18 (35– 40). Human 5q31–33 also contains a
Th2 cytokine locus and Cd14. It is proposed that both of these
4619
groups of genes may be linked to atopy. The murine Th2 cytokine
locus is not found on chromosome 18 but murine Cd14 maps to
murine chromosome 18 (31 cM).
Acknowledgment
We thank Kirin Brewery (Gunma, Japan) for providing synthetic
␣-GalCer.
References
1. Paul, W. E., and R. A. Seder. 1994. Lymphocyte responses and cytokines. Cell
76:241.
2. Abbas, A. K., K. M. Murphy, and A. Sher. 1996. Functional diversity of helper
T lymphocytes. Nature 383:787.
3. Mosmann, T. R., and R. L. Coffman. 1989. TH1 and TH2 cells: different patterns
of lymphokine secretion lead to different functional properties. Annu. Rev. Im-
munol. 7:145.
4. O’Garra, A. 1998. Cytokines induce the development of functionally heteroge-
neous T helper cell subsets. Immunity 8:275.
5. Sher, A., and R. L. Coffman. 1992. Regulation of immunity to parasites by T cells
and T cell-derived cytokines. Annu. Rev. Immunol. 10:385.
6. Scott, B., R. Liblau, S. Degermann, L. A. Marconi, L. Ogata, A. J. Caton,
H. O. McDevitt, and D. Lo. 1994. A role for non-MHC genetic polymorphism in
susceptibility to spontaneous autoimmunity. Immunity 1:73.
7. Parish, C. R., and F. Y. Liew. 1972. Immune response to chemically modified
flagellin. III. Enhanced cell-mediated immunity during high and low zone anti-
body tolerance to flagellin. J. Exp. Med. 135:298.
Downloaded from www.jimmunol.org on November 21, 2009
8. Guler, M. L., J. D. Gorham, C. S. Hsieh, A. J. Mackey, R. G. Steen,
W. F. Dietrich, and K. M. Murphy. 1996. Genetic susceptibility to Leishmania:
IL-12 responsiveness in TH1 cell development. Science 271:984.
9. Gorham, J. D., M. L. Guler, R. G. Steen, A. J. Mackey, M. J. Daly, K. Frederick,
W. F. Dietrich, and K. M. Murphy. 1996. Genetic mapping of a murine locus
controlling development of T helper 1/T helper 2 type responses. Proc. Natl.
Acad. Sci. USA 93:12467.
10. Bix, M., Z. E. Wang, B. Thiel, N. J. Schork, and R. M. Locksley. 1998. Genetic
regulation of commitment to interleukin 4 production by a CD4⫹ T cell-intrinsic
mechanism. J. Exp. Med. 188:2289.
11. Rincon, M., J. Anguita, T. Nakamura, E. Fikrig, and R. A. Flavell. 1997. Inter-
leukin (IL)-6 directs the differentiation of IL-4-producing CD4⫹ T cells. J. Exp.
Med. 185:461.
12. Diehl, S., J. Anguita, A. Hoffmeyer, T. Zapton, J. N. Ihle, E. Fikrig, and
M. Rincon. 2000. Inhibition of Th1 differentiation by IL-6 is mediated by
SOCS1. Immunity 13:805.
13. Lenschow, D. J., K. C. Herold, L. Rhee, B. Patel, A. Koons, H. Y. Qin, E. Fuchs,
B. Singh, C. B. Thompson, and J. A. Bluestone. 1996. CD28/B7 regulation of
Th1 and Th2 subsets in the development of autoimmune diabetes. Immunity
5:285.
14. Rissoan, M. C., V. Soumelis, N. Kadowaki, G. Grouard, F. Briere,
R. de Waal Malefyt, and Y. J. Liu. 1999. Reciprocal control of T helper cell and
dendritic cell differentiation. Science 283:1183.
15. Beebe, A. M., S. Mauze, N. J. Schork, and R. L. Coffman. 1997. Serial backcross
mapping of multiple loci associated with resistance to Leishmania major in mice.
Immunity 6:551.
16. Roberts, L. J., T. M. Baldwin, J. M. Curtis, E. Handman, and S. J. Foote. 1997.
Resistance to Leishmania major is linked to the H2 region on chromosome 17 and
to chromosome 9. J. Exp. Med. 185:1705.
17. Mendiratta, S. K., W. D. Martin, S. Hong, A. Boesteanu, S. Joyce, and
L. Van Kaer. 1997. Cd1d1 mutant mice are deficient in natural T cells that
promptly produce IL-4. Immunity 6:469.
18. Treco, D. A., W. M. Strauss, K. Wilson, M. Finney, F. M. Ausubel, R. Brent,
R. E. Kingston, D. D. Moore, J. G. Seidman, and K. Struhl, eds. 1987. Prepa-
ration and analysis of DNA. In Current Protocols in Molecular Biology. Wiley
Interscience, New York.
19. Wu, J., J. A. Longmate, G. Adamus, P. A. Hargrave, and E. K. Wakeland. 1996.
Interval mapping of quantitative trait loci controlling humoral immunity to ex-
ogenous antigens: evidence that non-MHC immune response genes may also
influence susceptibility to autoimmunity. J. Immunol. 157:2498.
20. Lander, E. S., and D. Botstein. 1989. Mapping mendelian factors underlying
quantitative traits using RFLP linkage maps. Genetics 121:185.
21. Lander, E. S., P. Green, J. Abrahamson, A. Barlow, M. J. Daly, S. E. Lincoln, and
L. L. Newburg. 1987. MAPMAKER: an interactive computer package for con-
structing primary.
22. Wakeland, E., L. Morel, M. Achey, M. Yui, and J. A. Longmate. 1997. Speed
congenics: a classic technique moves into the fast lane (relatively speaking).
Immunol. Today 18:472.
23. Moodycliffe, A. M., D. Nghiem, G. Clydesdale, and S. E. Ullrich. 2000. Immune
suppression and skin cancer development: regulation by NKT cells. Nat. Immun.
1:521.
24. Terabe, M., S. Matsui, N. Noben-Trauth, H. Chen, C. Watson, D. D. Donaldson,
D. P. Carbone, W. E. Paul, and J. A. Berzofsky. 2000. NKT cell-mediated re-
pression of tumor immunosurveillance by IL-13 and the IL-4R-STAT6 pathway.
Nat. Immun. 1:515.
25. Lan, F., D. Zeng, M. Higuchi, P. Huie, J. P. Higgins, and S. Strober. 2001.
Predominance of NK1.1⫹TCR ␣␤⫹ or DX5⫹TCR ␣␤⫹ T cells in mice condi-
tioned with fractionated lymphoid irradiation protects against graft-versus-host
disease: “natural suppressor” cells. J. Immunol. 167:2087.
4620
26. Chen, H., and W. E. Paul. 1997. Cultured NK1.1⫹ CD4⫹ T cells produce large
amounts of IL-4 and IFN-␥ upon activation by anti-CD3 or CD1. J. Immunol.
159:2240.
27. Morel, L., X. H. Tian, B. P. Croker, and E. K. Wakeland. 1999. Epistatic mod-
ifiers of autoimmunity in a murine model of lupus nephritis. Immunity 11:131.
28. Vyse, T. J., and B. L. Kotzin. 1998. Genetic susceptibility to systemic lupus
erythematosus. Annu. Rev. Immunol. 16:261.
29. Baker, D., O. A. Rosenwasser, J. K. O’Neill, and J. L. Turk. 1995. Genetic
analysis of experimental allergic encephalomyelitis in mice. J. Immunol.
155:4046.
30. Blake, J. A., J. T. Eppig, J. E. Richardson, C. J. Bult, and J. A. Kadin. 2001. The
Mouse Genome Database (MGD): integration nexus for the laboratory mouse.
Nucleic Acids Res. 29:91.
31. Merriman, T. R., I. A. Eaves, R. C. Twells, M. E. Merriman, P. A. Danoy,
C. E. Muxworthy, K. M. Hunter, R. D. Cox, F. Cucca, P. A. McKinney, et al.
1998. Transmission of haplotypes of microsatellite markers rather than single
marker alleles in the mapping of a putative type 1 diabetes susceptibility gene
(IDDM6). Hum. Mol. Genet. 7:517.
32. Cornelis, F., S. Faure, M. Martinez, J. F. Prud’homme, P. Fritz, C. Dib, H. Alves,
P. Barrera, N. de Vries, A. Balsa, et al. 1998. New susceptibility locus for rheu-
matoid arthritis suggested by a genome-wide linkage study. Proc. Natl. Acad. Sci.
USA 95:10746.
33. Shai, R., F. P. Quismorio, Jr., L. Li, O. J. Kwon, J. Morrison, D. J. Wallace,
C. M. Neuwelt, C. Brautbar, W. J. Gauderman, and C. O. Jacob. 1999. Genome-
NKT AND Th CELLS AND QTL ON CHROMOSOME 18
wide screen for systemic lupus erythematosus susceptibility genes in multiplex
families. Hum. Mol. Genet. 8:639.
34. Vaidya, B., H. Imrie, P. Perros, E. T. Young, W. F. Kelly, D. Carr, D. M. Large,
A. D. Toft, P. Kendall-Taylor, and S. H. Pearce. 2000. Evidence for a new Graves
disease susceptibility locus at chromosome 18q21. Am. J. Hum. Genet. 66:1710.
35. Ono, S. J. 2000. Molecular genetics of allergic diseases. Annu. Rev. Immunol.
18:347.
36. Soriano, J. B., R. de Cid, X. Estivill, J. M. Anto, J. Sunyer, D. Otero, J. Roca,
R. Rodriguez-Roisin, F. Morell, M. J. Rodrigo, G. Ercilla, T. H. Beaty, and
C. Lazaro. 2000. Association study of proposed candidate genes/regions in a
population of Spanish asthmatics. Eur. J. Epidemiol. 16:745.
37. Strauch, K., R. Bogdanow, R. Fimmers, M. P. Baur, and T. F. Wienker. 2001.
Linkage analysis of asthma and atopy including models with genomic imprinting.
Genet. Epidemiol. 21:S204.
38. Torres-Galvan, M. J., J. Quiralte, J. J. Pestano, N. Ortega, C. Blanco, R. Castillo,
T. Carrillo, P. Perez-Aciego, and F. Sanchez-Garcia. 2001. IL4 –R1 (5q31– q33)
and Fc⑀RI-␤ca (11q13) markers and atopy: a case/control study in a spanish
population. Allergy 56:159.
39. Beyer, K., R. Nickel, L. Freidhoff, B. Bjorksten, S. K. Huang, K. C. Barnes,
S. MacDonald, J. Forster, F. Zepp, V. Wahn, T. H. Beaty, D. G. Marsh, and
U. Wahn. 2000. Association and linkage of atopic dermatitis with chromosome
13q12–14 and 5q31–33 markers. J. Invest. Dermatol. 115:906.
40. Nishimura, A., R. S. Campbell-Meltzer, K. Chute, J. Orrell, and S. J. Ono. 2001.
Genetics of allergic disease: evidence for organ-specific susceptibility genes. Int.
Arch. Allergy Immunol. 124:197.
Downloaded from www.jimmunol.org on November 21, 2009