Hindawi Publishing Corporation
AIDS Research and Treatment
Volume 2010, Article ID 164586, 9 pages
Expressionand Functionof theChemokine,CXCL13, andIts
Receptor,CXCR5, inAids-Associated Non-Hodgkin’s Lymphoma
Jeffrey B.Smith,7Roger Detels,3,5,8and OtonielMart´ ınez-Maza1,3,5,8,9,10
1Department of Obstetrics & Gynecology, David Geffen School of Medicine at UCLA, CA 90095-1740, USA
2Department of Pathology and Laboratory Medicine, David Geffen School of Medicine at UCLA, CA 90095-1732, USA
3UCLA AIDS Institute and Jonsson Comprehensive Cancer Center, University of California, Los Angeles, CA 90095, USA
4Department of Psychiatry & Biobehavioral Sciences, David Geffen School of Medicine at UCLA, CA 90095-6967, USA
5Los Angeles Center of the Multicenter AIDS Cohort Study (MACS), University of California, Los Angeles, CA 90095, USA
6Department of Biostatistics, School of Public Health, University of California, Los Angeles, CA 90095, USA
7Department of Pediatrics, David Geffen School of Medicine at UCLA, CA 90095-1752, USA
8Department of Epidemiology, School of Public Health, University of California, Los Angeles, CA 90095, USA
9Department of Microbiology, Immunology & Molecular Genetics, David Geffen School of Medicine at UCLA, CA 90095-7364, USA
10UCLA AIDS Institute, BSRB Room 173, 615 Charles Young Drive South, Los Angeles, CA 90095-7363, USA
Correspondence should be addressed to Otoniel Mart´ ınez-Maza, email@example.com
Received 21 April 2010; Accepted 23 June 2010
Academic Editor: Guido Poli
Copyright © 2010 Daniel P. Widney et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
Background. The homeostatic chemokine, CXCL13 (BLC, BCA-1), helps direct the recirculation of mature, resting B cells, which
express its receptor, CXCR5. CXCL13/CXCR5 are expressed, and may play a role, in some non-AIDS-associated B cell tumors.
Objective. To determine if CXCL13/CXCR5 are associated with AIDS-related non-Hodgkin’s lymphoma (AIDS-NHL). Methods.
Serum CXCL13 levels were measured by ELISA in 46 subjects who developed AIDS-NHL in the Multicenter AIDS Cohort Study
and in controls. The expression or function of CXCL13 and CXCR5 was examined on primary AIDS-NHL specimens or AIDS-
AIDS-NHL specimens showed CXCR5 expression and most also showed CXCL13 expression. AIDS-NHL cell lines expressed
be involved in its biology. CXCL13 may have potential as a biomarker for AIDS-NHL.
HIV infection is associated with a greatly increased risk of
incidence of NHL in HIV-infected persons is believed to
be an end result of excess B cell activation that is seen
during HIV infection, as well as of loss of immunoregulatory
control of Epstein-Barr virus (EBV) infection . AIDS-
NHL tumors are generally of high grade and can be broadly
placed into one of several subtypes, which include Burkitt
DLBCL), and primary CNS lymphoma [4, 5]. Levels of B
cell-stimulatory cytokines, such as IL-6, are known to be
elevated during HIV infection and are further elevated in
those HIV(+) persons who go on to develop NHL [1, 6–8].
Thus, cytokines may be important in the development and
growth of these tumors.
Although it has been shown that cytokines probably
play an important role in the biology of AIDS-NHL,
little work has been done on chemokines, cytokines that
are most known for their ability to direct chemotaxis of
immune cells . In recent years, it has been shown that
chemokines can play important roles in a variety of cancers,
including cancers of the breast, prostate, lung, pancreas,
2AIDS Research and Treatment
liver, and skin, colorectal cancer, and cancers of the immune
system, including some lymphomas [10–18]. Chemokines
can potentially promote tumor progression by a number of
means, including by directly promoting tumor cell growth,
by inhibiting tumor cell apoptosis, and by promoting tumor
cell migration and metastasis [10–12, 19].
Although, in general, not much is known about the role
that chemokines could be playing in AIDS-NHL, several
studies are extant in the literature that cover aspects of this
issue. Specifically, Rabkin et al. have reported that HIV(+)
subjects who are at least heterozygous for a genetic dele-
tion in the chemokine receptor, CCR5, are at significantly
decreased risk for developing NHL. Conversely, HIV(+)
individuals who carry particular polymorphisms in the gene
for the chemokine, SDF-1, have a substantially increased risk
of developing NHL . Sei et al. have reported increased
expression of SDF-1 in children who had AIDS-NHL .
Sharma et al. reported that several AIDS-NHL cell lines
secreted the chemokines, IL-8, IL-16, and MIP-1α [22–24].
Some cell lines also expressed the receptor for MIP-1α,
it seems likely that chemokines can play a role in the biology
CXCL13, are substantially elevated during HIV infection
. CXCL13 is a homeostatic chemokine that helps direct
the normal trafficking of B cells in the body . It is
expressed by T follicular helper cells, dendritic cells, and
receptor, CXCR5, into these zones [27–31]. B cells become
activated in these zones if given appropriate help, and then
. Two other reports indicate that there are abnormalities
in the CXCR5/CXCL13 system during HIV infection: first,
mature B cells lose expression of CXCR5 during HIV infec-
tion; in contrast, these cells nearly uniformly express CXCR5
in healthy, uninfected, individuals . Second, Cagigi et al.
have recently shown that recirculating B cells may express
CXCL13 in HIV-infected subjects, in contrast to B cells from
healthy, uninfected, subjects, which are uniformly negative
forexpression of thismolecule.Theseobservations raise
the possibility that abnormalities in the CXCR5/CXCL13
system are contributing to the abnormalities that are seen in
be involved in the genesis of AIDS-NHL.
It should be noted that CXCR5 and/or CXCL13 have
been shown to be associated with several non-AIDS-related
B cell lymphomas, including extragastric lymphoma of
mucosa-associated lymphoid tissue (MALT) and follicular
lymphoma [13–17, 35]. In several of these cancers, tumor
cells were shown to express functional CXCR5 and to
migrate towards CXCL13 . In some of these cancers,
tumor cells were shown to express both CXCL13 and
CXCR5, suggesting a possibility for autocrine interactions,
or CXCL13 was shown to be expressed in tissue surrounding
tumors, raising the possibility that it could be directing the
metastasis of tumor cells [16, 17, 35, 36]. In this report, we
define the expression and function of CXCL13 and CXCR5
obtained from the Los Angeles (UCLA) center of the
Multicenter AIDS Cohort Study (MACS), an ongoing
natural history study of HIV in adult homosexual men
established in 1984 with 1637 original participants [37, 38].
In the MACS, participants are examined semiannually, and
blood samples are archived at each visit. The Institutional
Review Board at UCLA approved this study. Our study
assessed CXCL13 levels in sera collected from four groups
of adult homosexual men: Group I (AIDS-NHL) consisted
of 46 subjects who developed AIDS-lymphoma. For each
subject, archival serum was used from the closest available
visit prior to lymphoma diagnosis. The mean time to
diagnosis was 8.2 months (SD = 7.9). Of the 46 subjects in
Group I, 15 subjects had AIDS-NHL of the Burkitt subtype,
29 had AIDS-NHL of the diffuse large B cell lymphoma
(DLBCL) subtype, and 2 had AIDS-NHL of unspecified
subtype. Group II (AIDS, non-lymphoma) consisted of
41 subjects with AIDS , but no reported malignancy.
The mean CD4 T cell counts of Group I (199cells/mm3,
SD = 198) and Group II (182cells/mm3, SD = 162) were
similar. Group III subjects (n = 43) were HIV(+), but had
no AIDS-defining condition and were selected to have CD4
T cell counts greater than 500 (mean = 684cells/mm3, SD =
181). Group IV subjects (n = 40) were HIV seronegative. All
subjects were na¨ ıve for highly active antiretroviral therapy
Subjects. Archival storedserumspecimenswere
2.2. Determination of Levels of CXCL13 and Other Markers.
CXCL13 levels in sera, and in culture supernatants of cell
lines, were determined by ELISA (R&D Systems, Minneapo-
lis, MN; lower detection limit = 16pg/mL). Levels of soluble
CD44std (sCD44) in sera were determined by ELISA in
earlier studies using the same sera as in the current study
in correlational analyses.
2.3. Immunohistochemistry Studies. Tissue arrays containing
formalin-fixed, paraffin-embedded sections from AIDS-
lymphomas were obtained from the AIDS and Cancer
Specimen Resource (ACSR) of the National Cancer Institute.
who provided the serum samples for the CXCL13 serum
studies described above, since tumor specimens were not
available for most AIDS-NHL cases in the MACS that were
used in the serum studies. Conversely, prediagnosis serum
specimens were not available for the cases obtained from the
ACSR. Of the AIDS-NHL tumors represented in the tissue
arrays, about 1/3 were located in lymph nodes, while the
remaining 2/3 were of extranodal origin.
Initially, one set of array slides was stained with stan-
dard hematoxylin and eosin (H&E) stain to allow general
visualization of tissues and tumors. Tumors/tumor cells were
generally easily identified in the arrays.
AIDS Research and Treatment3
For immunohistochemical detection of specific markers,
sections were initially deparaffinized in xylene and rehy-
6.0) for 25 minutes at 90◦C (CXCR5), or heated in a pressure
cooker in TRIS (tris(hydroxymethyl)aminomethane) buffer
(pH 9.0) for 2 minutes at 115◦C (CXCL13). Endogenous
peroxidase activity was then quenched with 0.3% H2O2in
70% methanol for 10 minutes. The wash buffer in this
study wasPBS(phospho-buffered saline);allantibodies were
For CXCR5, a rat antihuman CXCR5 antibody (clone
RF8B2, R&D Systems) or IgG2bisotype control (clone KLH/
G2b-1-2, Southern Biotechnology Associates, Birmingham,
AL) was used (2μg/mL, 2 hours). Next, rabbit antirat
polyclonal antibody (DAKO, Carpinteria, CA) and then
horseradish peroxidase (HRP)-conjugated antirabbit EnVi-
sion reagent (DAKO) were applied, and color was developed
using Fast Red (Innogenex, San Ramon, CA).
For CXCL13, a specific polyclonal goat IgG (R&D
Systems) was used (10μg/mL, 30 minutes); the control was
normal goat IgG (Santa Cruz Biotechnology, Santa Cruz,
CA). Next, a biotinylated secondary antibody (Vectastain
Elite ABC kit, Vector Laboratories, Burlingame, CA) was
applied, followed by Vectastain Elite ABC reagent. Color
was developed using diaminobenzidine (Sigma-Aldrich, St.
Individual array sections were scored by a pathologist as
follows: 0 = no staining, 1+ = <30% of tumor cells positive,
2+ = 30%–60% of cells positive, 3+ = >60% of cells positive.
2.4. Cell Lines. Two AIDS-NHL cell lines, 2F7 (Burkitt
subtype) and R (diffuse large B cell lymphoma subtype),
and the Epstein-Barr virus (EBV)-lymphoblastoid B cell line,
40102, were used in these studies [41–43]. All cell lines were
cultured in standard media as previously described .
Both AIDS-NHL cell lines are positive for the Epstein-Barr
virus (EBV) and are not infected with HIV [41–43].
2.5. Flow Cytometry. Surface CXCR5 expression on cell
lines was examined by flow cytometry using a previously
described indirect staining protocol . The primary
anti-CXCR5 antibody was Clone RF8B2 (rat IgG2b, BD
Biosciences, San Diego, CA). The isotype control antibody
was clone KLH/Gb-1-2 (rat IgG2b, Southern Biotechnology
Associates, Inc., Birmingham, AL). The secondary antibody
was phycoerythrin (PE)-conjugated goat anti-rat IgG (Jack-
son ImmunoResearch Laboratories, Inc., West Grove, PA).
Flow cytometry was performed using a Becton Dickinson
LSR machine. Data was analyzed using CellQuest Pro 5.1
software (Becton Dickinson).
2.6. Cell Migration Assays. Cell migration assays were con-
ducted on the 2F7 and R AIDS-NHL cell lines using the
methods described by Kitchen et al. . Briefly, cells were
fluorescently prelabeled by culturing them for 30 minutes at
37◦in the presence of 5μg/mL of Calcein-AM (Molecular
Probes, Eugene, OR). Cells were then placed in RPMI 1610
medium without phenol red (Irvine Scientific, Irvine, CA)
at a concentration of 3.6 × 106cells/mL. For determination
of chemotaxis, cells (90,000; 25μL) were then loaded on top
of the filter of ChemoTX chemotaxis plates (NeuroProbe,
Gaithersburg, MD; filter pore size = 8-μm). Recombinant
human CXCL13 (R&D Systems) was then loaded into the
bottom well, and plates were incubated for 2 hours at
37◦C. Plate fluorescence was read using a Beckman Coulter
(Fullerton, CA) D7X880 Multimode Detector (bottom read
position; excitation, 485nm; emission, 530nm). A range of
0 (media only control) to 1000ng/mL of CXCL13 was tested.
each experiment was performed at least three times. As an
additional control, CXCL13 was loaded into both the lower
and the upper wells at some locations on the plates, to test
2.7. Statistical Analysis. Statistical analysis was performed
using the SAS system, version 9.1 (SAS Institute, Cary, NC).
In the study of serum CXCL13 levels, a moderate number
of subjects had values below the lower limit of detection
of the ELISA used (16pg/mL). Therefore, the significance
of differences in serum CXCL13 levels between groups was
determined using a variant of the two-sample t test that
of detection of the assay), as previously detailed by us in
an earlier publication . As we noted in that previous
reference, values less than 16pg/mL were considered left-
censored. We assumed that CXCL13 levels were distributed
log-normal and used the method of maximum likelihood to
estimate means and variance components of interest .
Although analysis was performed on the logarithmic scale,
for ease of interpretation, means and standard deviations
(SD) for CXCL13 are reported on the original scale using
the delta method. A two-sample z test that accounted for the
censoring was used to test for significance.
The degree of association between serum CXCL13 and
serum sCD44 was quantified using a bivariate linear regres-
sion analysis that again accounted for the left-censoring of
CXCL13. Pearson correlation coefficients are used to report
results of the regression analysis. For migration (chemotaxis)
assays, differences between conditions were analyzed using
the Wilcoxon rank sum test, as previously described .
3.1. Serum CXCL13 Levels Are Elevated Prior to the Develop-
ment of AIDS-NHL. In these studies, we examined CXCL13
levels in AIDS-NHL subjects, using serum obtained at a
mean time of 8.2 months prior to diagnosis. The mean
CXCL13 level seen in the AIDS-NHL group (158pg/mL, SD
= 153) was ∼50% higher than in the AIDS control group
(98.4pg/mL, SD = 70.9, P = .02, Figure 1). Furthermore,
CXCL13 levels correlated with sCD44 levels in the AIDS-
NHL group (R = 0.31, P = .04), but not in the AIDS control
group (R = 0.07; P = .7, data not shown); we previously
showed that sCD44 levels are elevated prior to AIDS-NHL
development . CXCL13 levels in the AIDS-NHL group
were also ∼2.5× greater than levels in the HIV(+) group,
4 AIDS Research and Treatment
and ∼7× greater than levels in the HIV(−) group; these
results were statistically significant (Figure 1). No significant
difference in CXCL13 levels was noted when comparing
the Burkitt’s and diffuse large B cell lymphoma (DLBCL)
may be a result of the relatively small number of subjects
tested in this study.
3.2. All Primary AIDS-NHL Specimens Showed CXCR5
Expression, and Most Showed CXCL13 Expression. For these
studies, we examined CXCR5 and CXCL13 expression in
tissue arrays containing sections of primary specimens of
AIDS-NHL from different subjects, using immunohisto-
chemistry. It was not technically feasible to examine both
markers simultaneously using double staining. We therefore
evaluated each marker using separate array slides containing
sections from the same tumors. All AIDS-NHL tumors
expressed CXCL13. Figure 2 shows representative staining
for the two major AIDS-NHL subtypes, Burkitt lymphoma
(BL), and diffuse large cell lymphoma (DLBCL). For AIDS-
BL, the mean score for CXCR5 expression (n = 9) was 1.7
(range = 1 to 3); the mean score for CXCL13 expression (n
= 7) was 1.0 (range = 0 to 3; data not shown). AIDS-DLBCL
showed higher expression of both CXCR5 and, particularly,
CXCL13; the mean score for CXCR5 (n = 19) was 2.2 (range
= 1 to 3); the mean score for CXCL13 (n = 17) was 1.9
(range = 1 to 3; not shown). For most specimens, most
expression of CXCR5 and/or CXCL13 appeared to be on
tumor cells themselves, as most tumors did not appear to
have many infiltrating cells (not shown). Eleven tumors had
high numbers of cells expressing CXCR5 (i.e., scoring 2+
or above), and high numbers of cells expressing CXCL13,
suggesting that some tumor cells coexpressed both markers.
Indeed, this was very likely the case for several of the
tumors, as virtually 100% of cells expressed CXCR5, and
virtually 100% expressed CXCL13 (not shown). Sections of
noncancerous lymph nodes were also included on the arrays
as controls—these typically showed definite, but relatively
low, expression of both molecules (a score of about 1;
not shown), which likely reflects the previously reported
observation that CXCL13 and CXCR5 expression in normal
lymph nodes is generally confined to cells located in or near
germinal centers [27, 29, 34].
3.3. AIDS-NHL Cell Lines Showed Surface Expression of
CXCR5, and Sometimes Also Secreted Low Levels of CXCL13.
In these studies, we examined expression of CXCR5 and
CXCL13 on the AIDS-NHL cell lines, 2F7 (Burkitt sub-
type) and R (DLBCL subtype). Cell surface CXCR5 was
detected using flow cytometry, and CXCL13 in cell culture
supernatants was detected by ELISA. Both cell lines clearly
expressed surface CXCR5; there was a large increase in mean
fluorescence intensity (MFI) when cells were stained for
CXCR5 compared to the isotype control (Figure 3). The two
cell lines showed somewhat different patterns of CXCR5
expression, with cells of the R cell line uniformly showing
high levels of CXCR5 expression. In contrast, cells of the 2F7
cell line showed a bimodal pattern of expression, with a large
Serum CXCL13 (pg/mL)
Figure 1: Serum CXCL13 levels are elevated prior to development
of AIDS-NHL. Serum from each person in the AIDS-NHL group
(n = 46) was obtained for one time point in the 2.5 years prior
to the development of lymphoma; subjects in the AIDS control
group (n = 41) had AIDS, but no malignancy. Subjects in the
HIV(+) group (n = 43) were HIV-seropositive, non-AIDS, and
had CD4 T cell numbers >500cells/mm3. Serum CXCL13 levels
were determined by ELISA. Subjects in the HIV(−) group (n = 40)
were HIV-seronegative. Analysis was performed on the logarithmic
scale, but we report means and standard deviations for CXCL13 on
the original scale using the delta method, as previously reported
. Bars represent mean values; error bars represent standard
deviations, (∗P < .05, and∗∗P < .001). Although not indicated in
the figure, all other comparisons (HIV+ versus HIV−, AIDS versus
HIV+, HIV− versus AIDS, and AIDS-NHL versus HIV−) were also
significant, P < .05.
subpopulation expressing low levelsof CXCR5, and a smaller
subpopulation expressing higher levels of CXCR5 (Figure 3).
Expression of CXCL13 was mixed, with culture supernatant
of the R cell line not containing detectable levels of CXCL13
(<16pg/mL; not shown). In contrast, culture supernatants
of new cultures of 2F7 did not contain detectable levels
of CXCL13, but culture supernatants of cultures of 2F7
that were several months old contained detectable CXCL13,
albeit at low levels (19ng/mL; data not shown). Culture
supernatants of the EBV-lymphoblastoid cell line, 40102,
which was used as a positive control, contained considerably
higher levels of CXCL13 (42pg/mL; not shown).
3.4. AIDS-NHL Cell Lines Showed Chemotaxis Towards
CXCL13 In Vitro. The ability of the AIDS-NHL cell lines to
respond to CXCL13 was examined using standard migration
assays. Both cell lines demonstrated statistically significant
(P < .05) migration towards CXCL13 (Figure 4). For the 2F7
cell line, 100ng/mL of CXCL13 was optimal; 50ng/mL was
optimal for the R cell line. At its optimal concentration of
CXCL13, each cell line always showed increased migration
in each of 3-4 replicate experiments compared to the media
only (spontaneous migration) control; the percent increase
AIDS Research and Treatment5
Figure 2: Representative expression of CXCR5 and CXCL13 protein in AIDS-NHL. Tissue arrays containing sections from numerous
AIDS-NHLs were examined for expression of CXCR5 (a) or CXCL13 (b) by immunohistochemistry, as noted in Section 2. For CXCR5,
an HRP/Fast Red system was used for color development (red); for CXCL13, Vectastain Elite ABC reagent and DAB were used (brown).
Arrays were counterstained with hematoxylin. Sections representative of typical CXCR5 and CXCL13 staining patterns in AIDS-associated
Burkitt lymphoma (AIDS-BL, indicated as “Burkitt’s”) and AIDS-associated diffuse large B cell lymphoma (AIDS-DLBCL, indicated as
“LCL”) are shown. Both AIDS-BL and AIDS-DLBCL show strong expression (3+) of CXCR5; AIDS-DLBCL shows strong expression (3+)
of CXCL13, whereas AIDS-BL shows more moderate expression (2+). The sections shown representing CXCR5 and CXCL13 expression in
AIDS-DLBCL came from the same tumor, an AIDS-DLBCL in the maxillary sinus. Normal sinus tissue (∗) is unstained. For each tumor
section stained for CXCR5 or CXCL13 expression, a negative control using an isotype-specific, non-cross-reactive, antibody is shown on
the right. All sections are shown at x100 original magnification, except that the Burkitt lymphoma in panel B is shown at x200 original
magnification. Pictures were taken using an Olympus DP11 camera attached to an Olympus BX51 bright field microscope, and recorded on
a Smart Media digital card. The 10x and 20x objective lenses (UPlan Apo, Japan) had apertures of 0.40 and 0.70, respectively. Pictures were
edited for publication using Adobe Photoshop 6.0 and Canvas 5.0.3 (ACD Systems of America, Inc.).
in cells migrated ranged from ∼30% to ∼90%. For both cell
lines, the addition of CXCL13 to the upper well abrogated
the response, indicating that chemotaxis, as opposed to
chemokinesis, was occurring (not shown).
These studies clearly demonstrate that there is an association
between CXCL13 expression and AIDS-NHL. First, in the
serum studies, serum CXCL13 levels were significantly ele-
vated prior to lymphoma diagnosis in the AIDS-NHL group,
compared to the AIDS control group (Figure 1). CXCL13
levels in the AIDS-NHL group were even more highly
elevated whencompared to the HIV(+)/non-AIDS and HIV-
seronegative groups. These results suggest that assessment of
CXCL13 levels could potentially be useful in the detection
and diagnosis of AIDS-NHL. Elevated serum CXCL13 levels
chronic lymphocytic leukemia . Second, many primary
AIDS-NHL specimens, of both the Burkitt and DLBCL
subtypes, showed expression of CXCL13 (Figure 2). Third,
as noted in (Section 3), the AIDS-Burkitt cell line, 2F7,
expressed low levels of CXCL13 after it had been growing
in culture for several months (not shown). Thus, these
combined data indicate that CXCL13 is frequently expressed
in AIDS-NHL. These results are not dissimilar to results
obtained in studies on several types of B cell lymphoma
not associated with HIV infection, including non-AIDS-
related primary central nervous system lymphoma (PCNSL)
[15, 16, 35, 36]. The frequent expression of CXCL13 in
the AIDS-NHL tumor specimens (Section 3, Figure 2) raises
the possibility that the developing tumors themselves are a
source of the elevated serum CXCL13 levels seen preceding
diagnosis in the AIDS-NHL group (Figure 1). If so, CXCL13
could potentially prove to be a useful biomarker for early
detection of AIDS-NHL, for determining burden of disease,
and for determining prognosis following chemotherapy.
The receptor for CXCL13, CXCR5, appears to be com-
monly expressed in AIDS-NHL, as well. In the immuno-
histochemistry studies, all AIDS-NHL specimens showed
expression of CXCR5 (Figure 2), as did both AIDS-NHL
cell lines (Figure 3). Additionally, the CXCR5 on the AIDS-
NHL cell lines appeared to be functional, as both cell lines
results are not dissimilar to other previously reported results
for several types of B cell lymphoma not associated with
HIV infection [13–17, 35, 36]. The fact that the optimal
amount of CXCL13 in chemotaxis studies on the R cell line
(50ng/mL; Section 3) was lower than the optimal amount
for the 2F7 cell line (100ng/mL; Section 3) is consistent with
the results in Figure 3, which indicate that the R cell line
expresses higher levels of CXCR5 than the 2F7 cell line.
Although our studies show that both CXCL13 and
CXCR5 are commonly expressed in AIDS-NHL, future
studies will be needed to determine more exactly what roles
these molecules may be playing in these cancers. Several
mechanisms, which are not necessarily mutually exclusive,
appear possible: first, it seems plausible that CXCR5 and
CXCL13 could be playing a role in the initial development
of AIDS-NHL. As noted in Section 1, serum CXCL13 levels
6 AIDS Research and Treatment
Figure 3: AIDS-NHL cell lines express CXCR5, as shown by flow cytometry. The AIDS-BL cell line, 2F7 (a), and the AIDS-DLBCL cell
line, R (b), were stained for CXCR5 expression using an indirect staining protocol, as noted in Section 2. First, cells were stained with a rat
IgG2banti-CXCR5 antibody (dotted lines). As a control, some cells were stained with a rat IgG2bisotype control antibody (solid lines). Cells
were then stained with a PE-conjugated goat anti-rat IgG secondary antibody, and examined by flow cytometry. During the flow cytometry
acquisition stage, at least 5,000cells/events were acquired per tube/condition. During analysis, dead cells were excluded using forward- and
appear to be elevated during HIV infection, and at least
some circulating B cells appear to (abnormally) express
CXCL13, and to lose expression of CXCR5 [25, 33, 34]. As
CXCL13/CXCR5 are important in guiding recirculating B
cells into B cell zones of secondary lymphoid tissues [26, 27],
if expression of these molecules is disturbed, it is possible
that B cells will not home normally to these tissues. This
could lead to inappropriate homing to other tissues and/or
inappropriate activation, promoting lymphomagenesis. In
support of this possibility is the fact that a large proportion
B cell lymphomas not associated with HIV infection . It
could be that the continued expression of CXCL13/CXCR5
in developed AIDS-NHLs, as shown in the current studies, is
a reflection of processes that occurred during lymphomage-
nesis. While these molecules might continue to have an effect
is merely an epiphenomenon that is reflective of the earlier
role they played in the initial development of these tumors.
A second possibility is that the CXCL13 and/or CXCR5
could be directly promoting AIDS-NHL tumor growth, or
performing some other tumor-promoting functions, such
as inhibiting apoptosis or enhancing angiogenesis. Other
chemokines, such as SDF-1, have been shown to directly
induce tumor cell growth in non-AIDS-related cancers [11,
12], and it is possible that CXCL13 could be doing so,
also. This promotion of cell growth could be coming from
high levels of circulating CXCL13 produced elsewhere in
the body, or from CXCL13 produced locally to the tumor,
either by tumor cells themselves or by local nontumor
cells. Our studies do suggest that autocrine interactions
could be occurring in AIDS-NHL, as some tumor cells
appeared to express both CXCR5 and CXCL13 in both
the immunohistochemistry and AIDS-NHL cell line studies
Third, it seems possible that CXCL13 could be directing
movement and/or metastasis of AIDS-NHL cells in the body.
As noted earlier (Figure 4), both AIDS-NHL cell lines used
in our studies demonstrated chemotaxis towards CXCL13.
Although CXCL13 has often described as being produced
in secondary lymphoid tissues, it is now clear that it can
be produced in other tissues as well, including brain, liver,
and lung, particularly in the context of inflammation [48–
50]. Perhaps CXL13 produced in such organs chemoattracts
AIDS-NHL tumor cells to metastasize to these locations.
It should be noted that at least aspects of these poten-
tial mechanisms involving CXCR5 and CXCL13 could be
operating in some non-AIDS-related B cell lymphomas,
as well, given the similarities noted above between the
expression/function of these molecules in AIDS-NHL in
the current study, and their previously reported expres-
sion/function in several types of non-AIDS-related B cell
lymphoma [13–17, 35, 36]. However, it also seems possible
that some of these mechanisms could prove to be unique to
AIDS-NHL, given our previous finding that HIV infection
is associated with an overproduction of CXCL13, even in
subjects who do not develop lymphoma . Such an
overproduction of CXCL13 may not necessarily be occurring
prior to the initiation of tumor development in otherwise
healthy subjects who develop B cell lymphoma outside of the
context of HIV infection. Thus, further study of CXCR5 and
CXCL13 expression and functionin AIDS-NHL,and in non-
AIDS-associated NHL, could ultimately reveal some unique
differences between the pathogenesis of AIDS-related and
non-AIDS-related B cell lymphomas.
In summary, we have demonstrated an association
between expression of the chemokine, CXCL13, and its
receptor, CXCR5, and AIDS-NHL. CXCL13 has potential as
a biomarker for AIDS-NHL. Future studies will be needed
to more fully determine the feasibility of using CXCL13 as
AIDS Research and Treatment7
Normalised migration (%)
Figure 4: AIDS-NHL cell lines demonstrate chemotaxis towards
CXCL13 in vitro. As noted in Section 2, cells of the AIDS-NHL
cell lines, 2F7 (AIDS-BL) or R (AIDS-DLBCL), were first prelabeled
with Calcein-AM, and then placed on top of 96-well chemotaxis
chambers and permitted to migrate for 2 hours in response to
CXCL13 or media alone placed in the lower wells. To create a
standard curve, preset numbers of cells were added to some wells.
Fluorescence in the bottom well was read in a plate reader, and this
data was converted to number of cells migrated using the standard
curve. As noted in Section 3, 100ng/mL of CXCL13 was used for
the 2F7 cell line, and 50ng/mL was used for the R cell line. Each
experiment was replicated at least 3 times, and, in each experiment,
6–12 replicate wells were used for each experimental condition.
The figure shows data from a representative experiment for each
cell line. Results are shown as average percent of cells migrating
compared to the media only controls (which are set at 100%).
Solid bars show results for media alone controls; hatched bars show
results after addition of CXCL13. Bars represent the SEM.
a biomarker for AIDS-NHL, as well as to define the function
of CXCL13 and CXCR5 in the pathogenesis of AIDS-NHL.
This work was supported in part by grants from the Califor-
nia Universitywide AIDS Research Program (UARP) (ID04-
LA-039), the Leukemia and Lymphoma Society (6155-03),
and the National Institutes of Health (NIH) (CA57152,
CA73475, and AI28697). The Multicenter AIDS Cohort
Study (MACS) is funded by the National Institute of Allergy
and Infectious Diseases, with additional supplemental fund-
ing from the National Cancer Institute (UO1-AI-35042, 5-
MO1-RR-00722 (GCRC), UO1-AI-35043, UO1-AI-37984,
UO1-AI-35039, UO1-AI-35040, UO1-AI-37613, and UO1-
AI-35041). The AIDS-NHL tissue arrays were provided by
the AIDS and Cancer Specimen Resource (ACSR) of the
National Cancer Institute (NCI). Flow cytometry was per-
formed in the UCLA Jonsson Comprehensive Cancer Center
(JCCC) and Center for AIDS Research Flow Cytometry Core
Facility that is supported by National Institutes of Health
awards CA16042 and AI28697, and by the JCCC, the UCLA
AIDS Institute, and the David Geffen School of Medicine
at UCLA. This work was carried out in the facilities of the
UCLA AIDS Institute, which were supported, in part, by
funds from the James B. Pendleton Charitable Trust and the
McCarthy Family Foundation. The authors would like to
thank the UCLA-MACS participants, who have made this
and many other studies of HIV disease possible. They also
thank Max Hechter, Kevin Barrett, Cindy Chang, and Najib
Aziz for assistance with MACS serum specimens and data,
and Benjamin Bonavida for assistance with the AIDS-NHL
 O.Mart´ ınez-MazaandE.C.Breen,“B-cellactivationandlym-
phoma in patients with HIV,” Current Opinion in Oncology,
vol. 14, no. 5, pp. 528–532, 2002.
 R. J. Biggar, “AIDS-related cancers in the era of highly active
antiretroviral therapy,” Oncology, vol. 15, no. 4, pp. 439–448,
 M. H. Gail, J. M. Pluda, C. S. Rabkin et al., “Projections of
immunodeficiency syndrome,” Journal of the National Cancer
Institute, vol. 83, no. 10, pp. 695–701, 1991.
 K. L. Grogg, R. F. Miller, and A. Dogan, “HIV infection and
lymphoma,” Journal of Clinical Pathology, vol. 60, no. 12, pp.
 A. Carbone, A. Gloghini, D. Capello, and G. Gaidano,
“Genetic pathways and histogenetic models of AIDS-related
lymphomas,” European Journal of Cancer, vol. 37, no. 10, pp.
 E. C. Breen, A. R. Rezai, K. Nakajima et al., “Infection with
HIV is associated with elevated IL-6 levels and production,”
Journal of Immunology, vol. 144, no. 2, pp. 480–484, 1990.
 J. M. Pluda, D. J. Venzon, G. Tosato et al., “Parameters
affecting the development of non-Hodgkin’s lymphoma in
patients with severe human immunodeficiency virus infection
receiving antiretroviral therapy,” Journal of Clinical Oncology,
vol. 11, no. 6, pp. 1099–1107, 1993.
 R. Masood, Y. Zhang, M. W. Bond et al., “Interleukin-10 is
an autocrine growth factor for acquired immunodeficiency
 A. Zlotnik and O. Yoshie, “Chemokines: a new classification
system and their role in immunity,” Immunity, vol. 12, no. 2,
pp. 121–127, 2000.
 A. Zlotnik, “Chemokines and cancer,” International Journal of
Cancer, vol. 119, no. 9, pp. 2026–2029, 2006.
 H. Kulbe, N. R. Levinson, F. Balkwill, and J. L. Wilson, “The
movement,” International Journal of Developmental Biology,
vol. 48, no. 5-6, pp. 489–496, 2004.
 F. Bertolini, C. Dell’Agnola, P. Mancuso et al., “CXCR4 neu-
tralization, a novel therapeutic approach for non-Hodgkin’s
8 AIDS Research and Treatment
lymphoma,” Cancer Research, vol. 62, no. 11, pp. 3106–3112,
 A. J. A. Deutsch, A. Aigelsreiter, E. Steinbauer et al., “Distinct
signatures of B-cell homeostatic and activation-dependent
chemokine receptors in the development and progression of
extragastricMALTlymphomas,” JournalofPathology, vol.215,
no. 4, pp. 431–444, 2008.
 A. V. Kurtova, A. T. Tamayo, R. J. Ford, and J. A. Burger,
“Mantle cell lymphoma cells express high levels of CXCR4,
CXCR5, and VLA-4 (CD49d): importance for interactions
with the stromal microenvironment and specific targeting,”
Blood, vol. 113, no. 19, pp. 4604–4613, 2009.
 H. W. Tun, D. Personett, K. A. Baskerville et al., “Pathway
analysis of primary central nervous system lymphoma,” Blood,
vol. 111, no. 6, pp. 3200–3210, 2008.
 A. Brunn, M. Montesinos-Rongen, A. Strack et al., “Expres-
sion pattern and cellular sources of chemokines in primary
central nervous system lymphoma,” Acta Neuropathologica,
vol. 114, no. 3, pp. 271–276, 2007.
 L. Trentin, A. Cabrelle, M. Facco et al., “Homeostatic
chemokines drive migration of malignant B cells in patients
with non-Hodgkin lymphomas,” Blood, vol. 104, no. 2, pp.
 R. J. Phillips, M. D. Burdick, M. Lutz, J. A. Belperio, M.
P. Keane, and R. M. Strieter, “The stromal derived factor-
1/CXCL12-CXC chemokine receptor 4 biological axis in
non-small cell lung cancer metastases,” American Journal of
Respiratory and Critical Care Medicine, vol. 167, no. 12, pp.
 T. C. Walser and A. M. Fulton, “The role of chemokines in the
biology and therapy of breast cancer,” Breast Disease, vol. 20,
pp. 137–143, 2004.
 C. S. Rabkin, Q.-E. Yang, J. J. Goedert, G. Nguyen, H.
Mitsuya, and S. Sei, “Chemokine and chemokine receptor
gene variants and risk of non-Hodgkin’s lymphoma in human
immunodeficiency virus-1-infected individuals,” Blood, vol.
93, no. 6, pp. 1838–1842, 1999.
 S. Sei, D. P. O’Neill, S. K. Stewart et al., “Increased level
of stromal cell-derived factor-1 mRNA in peripheral blood
mononuclear cells from children with AIDS-related lym-
 V. Sharma and L. Zhang, “Interleukin-8 expression in aids-
associated lymphoma B-cell lines,” Biochemical and Biophys-
ical Research Communications, vol. 282, no. 2, pp. 369–375,
 V. Sharma, J. L. Sparks, and J. D. Vail, “Human B-cell lines
constitutively express and secrete interleukin-16,” Immunol-
ogy, vol. 99, no. 2, pp. 266–271, 2000.
 V. Sharma, D. Walper, and R. Deckert, “Modulation of
macrophage inflammatory protein-1α and its receptors in
human B-cell lines derived from patients with acquired
immunodeficiency syndrome and Burkitt’s lymphoma,” Bio-
chemical and Biophysical Research Communications, vol. 235,
no. 3, pp. 576–581, 1997.
 D. P. Widney, E. C. Breen, W. J. Boscardin et al., “Serum levels
of the homeostatic B cell chemokine, CXCL13, are elevated
during HIV infection,” Journal of Interferon and Cytokine
Research, vol. 25, no. 11, pp. 702–706, 2005.
 G. M¨ uller, U. E. H¨ opken, H. Stein, and M. Lipp, “Systemic
immunoregulatory and pathogenic functions of homeostatic
1, pp. 1–8, 2002.
 J. G. Cyster, K. M. Ansel, K. Reif et al., “Follicular stromal cells
and lymphocyte homing to follicles,” Immunological Reviews,
vol. 176, pp. 181–193, 2000.
 R. F¨ orster, A. E. Mattis, E. Kremmer, E. Wolf, G. Brem, and
M. Lipp, “A putative chemokine receptor, BLR1, directs B cell
migration to defined lymphoid organs and specific anatomic
compartments of the spleen,” Cell, vol. 87, no. 6, pp. 1037–
 K. Reif, E. H. Ekland, L. Ohl et al., “Balanced responsiveness
to chemoattractants from adjacent zones determines B-cell
position,” Nature, vol. 416, no. 6876, pp. 94–99, 2002.
 M. D. Gunn, V. N. Ngo, K. M. Ansel, E. H. Ekland, J. G.
Cyster, and L. T. Williams, “A B-cell-homing chemokine made
Nature, vol. 391, no. 6669, pp. 799–803, 1998.
 C. H. Kim, H. W. Lim, J. R. Kim, L. Rott, P. Hillsamer, and
E. C. Butcher, “Unique gene expression program of human
 C. D. C. Allen, T. Okada, and J. G. Cyster, “Germinal-center
organization and cellular dynamics,” Immunity, vol. 27, no. 2,
pp. 190–202, 2007.
 R. F¨ orster, G. Schweigard, S. Johann et al., “Abnormal expres-
sion of the B-cell homing chemokine receptor BLR1 during
the progression of acquired immunodeficiency syndrome,”
Blood, vol. 90, no. 2, pp. 520–525, 1997.
 A. Cagigi, F. Mowafi, L. V. P. Dang et al., “Altered expression
of the receptor-ligand pair CXCR5/CXCL13 in B cells during
chronic HIV-1 infection,” Blood, vol. 112, no. 12, pp. 4401–
 J. R. Smith, R. M. Braziel, S. Paoletti, M. Lipp, M. Uguc-
cioni, and J. T. Rosenbaum, “Expression of B-cell-attracting
chemokine 1 (CXCL13) by malignant lymphocytes and
vascular endothelium in primary central nervous system
lymphoma,” Blood, vol. 101, no. 3, pp. 815–821, 2003.
and chemokines, BLC and SDF-1, in the eyes of patients with
primary intraocular lymphoma,” Ophthalmology, vol. 110, no.
2, pp. 421–426, 2003.
 R. A. Kaslow, D. G. Ostrow, R. Detels, J. P. Phair, B. F.
Polk, and C. R. Rinaldo Jr., “The Multicenter AIDS Cohort
Study: rationale, organization, and selected characteristics of
2, pp. 310–318, 1987.
 R. Detels, J. P. Phair, A. J. Saah, et al., “Recent scientific con-
tributions to understanding HIV/AIDS from the Multicenter
AIDS Cohort Study,” Journal of Epidemiology, vol. 2, pp. S11–
 Centers for Disease Control, “1993 revised classification
system for HIV infection and expanded surveillance case
definition for AIDS among adolescents and adults,” Morbidity
and Mortality Weekly Report. Recommendations and Reports,
vol. 41, no. 17, pp. 1–19, 1992.
 E. C. Breen, M. Epeldegui, W. J. Boscardin, D. P. Widney,
R. Detels, and O. Mart´ ınez-Maza, “Elevated levels of soluble
CD44 precede the development of AIDS-associated non-
Hodgkin’s B-cell lymphoma,” AIDS, vol. 19, no. 15, pp. 1711–
 V. L. Ng, M. H. Hurt, C. L. Fein et al., “IgMs produced by
two acquired immune deficiency syndrome lymphoma cell
lines: Ig binding specificity and V(H)-gene putative somatic
mutation analysis,” Blood, vol. 83, no. 4, pp. 1067–1078, 1994.
AIDS Research and Treatment9
 D. Widney, W. J. Boscardin, A. Kasravi, and O. Mart´ ınez-
Maza, “Expression and function of CD28 on Epstein-Barr
virus-positive B cell lines and AIDS-associated non-Hodgkin’s
lymphoma cell lines,” Tumor Biology, vol. 24, no. 2, pp. 82–93,
 A. Demidem, D. Lam, X. H. Gan, A. Levine, Z. Salahuddin,
and B. Bonavida, “Endogenous IL-10 is a protective factor in
drug-resistant AIDS-related lymphoma (ARL),” Proceedings of
 S. G. Kitchen, S. LaForge, V. P. Patel, C. M. Kitchen,
M. C. Miceli, and J. A. Zack, “Activation of CD8 T cells
induces expression of CD4, which functions as a chemotactic
receptor,” Blood, vol. 99, no. 1, pp. 207–212, 2002.
 M. Di Mascio, G. Dornadula, H. Zhang et al., “In a subset
of subjects on highly active antiretroviral therapy, human
of Virology, vol. 77, no. 3, pp. 2271–2275, 2003.
 A. B¨ urkle, M. Niedermeier, A. Schmitt-Gr¨ aff, W. G. Wierda,
M. J. Keating, and J. A. Burger, “Overexpression of the CXCR5
chemokine receptor, and its ligand, CXCL13 in B-cell chronic
lymphocytic leukemia,” Blood, vol. 110, no. 9, pp. 3316–3325,
 A. Tulpule and A. Levine, “AIDS-related lymphoma,” Blood
Reviews, vol. 13, no. 3, pp. 147–150, 1999.
 F. Aloisi, S. Columba-Cabezas, D. Franciotta et al., “Lymphoid
chemokines in chronic neuroinflammation,” Journal of Neu-
roimmunology, vol. 198, no. 1-2, pp. 106–112, 2008.
 D. F. Legler, M. Loetscher, R. S. Roos, I. Clark-Lewis, M.
Baggiolini, and B. Moser, “B cell-attracting chemokine 1,
a human CXC chemokine expressed in lymphoid tissues,
selectively attracts B lymphocytes via BLR1/CXCR5,” Journal
of Experimental Medicine, vol. 187, no. 4, pp. 655–660, 1998.
 P. C. Fulkerson, N. Zimmermann, L. M. Hassman, F. D.
Finkelman, and M. E. Rothenberg, “Pulmonary chemokine
expression is coordinately regulated by STAT1, STAT6, and
IFN-γ,” Journal of Immunology, vol. 173, no. 12, pp. 7565–