| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institut National de la Santé et de la Recherche Médicale Unité 25, and
Centre National de la Recherche Scientifique Unité MR8603, Hôpital Necker, Paris, France;
St. Bartholomew’s and Royal London School of Medicine and Dentistry, Bone and Joint Research Unit, London, United Kingdom;
Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Fundacion Campomar, Buenos Aires, Argentina; and
¶ Kennedy Institute of Rheumatology, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-galactosidase-transduced cells.
Protected mice showed a majority of islets (60%) presenting with
noninvasive peri-insulitis at variance with -galactosidase controls
that exhibited invasive/destructive insulitis. Importantly, in all
recipients, the transduced proteins were detected within islet
infiltrates. Infiltrating lymphocytes from recipients of
mIL-4-transduced cells produced high levels of mIL-4, as assessed by
ELISA. In recipients of -galactosidase-transduced cells, 
60% of
TCR
+ islet-infiltrating cells expressed
-galactosidase, as assessed by flow cytometry. The protection from
disease transfer is due to a direct effect of mIL-4 gene therapy on
immunoregulatory T cells rather than on diabetogenic cells.
mIL-4-transduced purified CD62L- effector cells or
transgenic BDC2.5 diabetogenic T cells still transferred disease
efficiently. Conversely, mIL-4 transduction up-regulated the capacity
of purified immunoregulatory CD62L+ cells to inhibit
disease transfer. These data open new perspectives for gene therapy in
insulin-dependent diabetes using T cells devoid of any intrinsic
diabetogenic potential. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initial data from our laboratory subsequently confirmed by other groups
have shown that in the nonobese diabetic
(NOD)3 mouse model of
spontaneous T cell-mediated autoimmune insulin-dependent diabetes, the
onset of disease is under the control of CD4+
immunoregulatory lymphocytes (ART, autoimmunity regulatory T cells
(9)), which appear in many respects similar to those
down-modulating natural autoreactivity (9, 10, 11, 12, 13, 14, 15). Much of
the phenomenology of this T cell-mediated immunoregulation has been
frequently linked to the concept of immune deviation that is based on
the existence of two main CD4 Th cell subsets Th1 and Th2 that
counterregulate each other (16). Thus, in the case of NOD
diabetes, in which pathogenic CD4 T cells chiefly belong to the
IFN-
-producing Th1 subset (14, 17, 18), the question is
posed to determine whether immunoregulatory T cells that control the
disease depend on Th2 cytokines for their differentiation, survival, or
mode of action. Systemic administration of IL-4 or overexpression of
IL-4 in islet cells prevents the occurrence of insulitis and
diabetes (19, 20). In addition, Th2 cells may indeed
down-regulate the islet-specific pathogenic response in NOD mice
notably after systemic administration of cell autoantigens
(21, 22, 23, 24).
In this study, we addressed this issue using a retrovirally based gene therapy approach in which diabetogenic or immunoregulatory T cells from NOD mice have been transduced with the murine IL-4 (mIL-4) gene and used in an adoptive transfer model we previously described. In this system, pathogenic T cells administered to immunoincompetent syngeneic adult recipients elicit disease, while the concomitant injection of immunoregulatory T cells efficiently protects from diabetes transfer. In this model, the L-selectin receptor (CD62L) was used as an instrumental marker to physically separate diabetogenic T cells, which lack CD62L, from immunoregulatory T cells that express it. After having shown that transduction with the mIL-4 gene of total spleen cells from diabetic mice inhibited their diabetogenic potential, we could evidence that the protection was mediated through a selective effect on immunoregulatory CD62L+ cells.
These results might open new perspectives for gene therapy in humans using as targets peripheral T cells devoid of any intrinsic diabetogenic potential and without the need for a previous cell immortalization step.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The retroviral vector pBABE-Neo containing the Mo-MLV long
terminal repeats, the SV40 promoter, and the neomycin gene was kindly
provided by Drs. Morgensten and Land (Imperial Cancer Research
Fund, London, U.K.) (25, 26). mIL-4 cDNA was cloned
into BamHI site of the vector pBABE Neo. The
-galactosidase gene was used as a reporter gene. The ecotropic
packaging line, GP+E86, was transfected with
pBABE Neo-IL-4 using calcium-phosphate precipitation. Transfected cells
were selected and maintained in medium with 1 mg/ml G418 geneticin
(Life Technologies, Gaithersburg, MD). Viral titers were at least
105 PFU/ml. Retroviral supernatants were stored
at -80°C.
Mice and adoptive cell transfer
NOD mice (Kd, I-Ag7,
Db) (Thy-1.2), congenic Thy-1.1, and NOD-scid
were bred in our animal facility under specific pathogen-free
conditions. NOD females developed insulitis by 4 wk of age, and
spontaneous diabetes appeared by 14 wk of age (95% incidence at 40
wk). Transgenic BDC2.5 NOD and NOD C-/- mice
were kindly provided by D. Mathis and C. Benoist (Institut de
Génétique et de Biologie Moléculaire et Cellulaire,
Strasbourg, France) and bred in our facility. These mice develop overt
diabetes by 4–6 wk of age. Colorimetric strips were used to monitor
glycosuria (Glukotest; Boehringer Mannheim, Mannheim, Germany) and
fasting glycemia (Haemoglukotest and Reflolux F; Boehringer
Mannheim).
Adult 6- to 8-wk-old NOD-scid mice were used as recipients. Depending on the experiment, the animals were transferred either with a single cell population or, in the case of cotransfer experiments, with a mixture of two distinct cell populations. The precise cell numbers used varied depending on the experiments and are detailed in Results.
Abs and FACS analysis
Abs to CD3 (145 2C11), CD4 (GK1.5), CD8 (53.6.7), and TCR
(H57-597) were purified in our laboratory. For flow cytometry, Abs were
fluoresceinated or biotinylated. MEL-14 (CD62L) was obtained from BD
PharMingen (San Diego, CA). PE-labeled streptavidin used to detect the
cell-bound biotinylated Abs was obtained from BD PharMingen. Briefly,
cells were stained in PBS containing 5% heat-inactivated FCS and
0.01% sodium azide and incubated for 30 min with the appropriate
concentration of biotin-labeled mAbs in 96-well round-bottom microplate
on ice. After washing, the cells were incubated for another 30 min with
FITC-labeled mAbs and PE-labeled streptavidin or PE-labeled mAbs.
Control stainings were performed using isotype-matched biotinylated,
FITC- or PE-labeled irrelevant Abs. Flow cytometry was performed on a
FACScan flow cytometer (BD Biosciences, San Jose, CA). A minimum of
104 events was acquired on a gate including
viable cells. For the acquisition and the analysis, the software used
was CellQuest (BD Biosciences).
Cell preparations and infection with retroviruses
Splenocytes were recovered by gentle disruption from pools of 7–10 female mice (overtly diabetic Thy-1.2 or Thy-1.1 NOD, prediabetic Thy-1.2 NOD) and cultured in vitro (24-well culture plaques) in presence of Con A at 2.5 µg/ml for 48 h at 37°C in a humidified atmosphere. The culture medium was RPMI 1640 (Life Technologies, Paisley, U.K.) supplemented with glutamax, 10% FCS (Life Technologies), 0.05 mM 2-ME, penicillin, and streptomycin. At the end of the culture, the cells were recovered, washed, and incubated at 37°C for 4 h at 5 x 106/ml in the viral supernatant previously warmed at 37°C and supplemented with DEAE dextran at 70 µg/ml. Before transfer, the cells were washed twice in saline.
When needed, splenocytes were purified on the basis of the CD62L expression using MACS (Miltenyi Biotech, Bergisch-Gladbach, Germany), as previously described (13). Briefly, splenocytes were incubated with the appropriate concentration of the biotinylated MEL-14 Ab, incubated with the streptavidin-coated paramagnetic beads, and passed through the magnetic column within the MACS device (Miltenyi Biotech) (13). The purity of the recovered cells was analyzed by flow cytometry after staining with PE-labeled streptavidin. Purity of the sorted cells was in all cases greater than 97%, and recovery ranged from 50 to 70%.
Isolation of islet-infiltrating cells
Pancreata from NOD-scid recipients were recovered 12–15 days after transfer, and infiltrating cells were isolated by conventional collagenase digestion (collagenase-P, 5 mg/ml; Boehringer Mannheim) at 37°C for 5–8 min and hand picking under binocular lenses (27, 28). Islets were then pressed through a 100-µm metal sieve and filtered through a 40-µm and a 20-µm nylon screen to recover infiltrating cells. Cell viability was determined by trypan blue exclusion.
In vitro cultures and cytokine production
Cells were cultured in RPMI 1640 supplemented with glutamax,
10% FCS, 0.05 mM 2-ME, penicillin, and streptomycin. For cytokine
production, lymphocytes were plated in triplicate (1 x
105/well; 200 µl final volume), in 96-well
round-bottom microplates (Nunc, Roskilde, Denmark) coated with 10
µg/ml of purified TCR Ab (H57-597). Supernatants were recovered
at 48 h of culture at 37° in a humidified atmosphere containing
5% CO2, and stored at -80°C until tested. In
control cultures, the TCR Ab coating was omitted.
mIL-4 in the supernatants was measured by ELISA using 11B11 (provided by W. Paul, National Institutes of Health, Bethesda, MD) as a capture Ab and biotinylated BVD6 (provided by A. O’Garra, DNAX, Palo Alto, CA) as the secondary Ab. Mouse rIL-4 was from R&D Systems (Minneapolis, MN). The sensitivity of the IL-4 assay was 25 pg/ml.
Histopathology
Paraffin sections of Bouin-fixed pancreata were used. Serial 2-µm sections were stained with hematoxylin and eosin to score mononuclear cell infiltration as follows: grade 0 = normal islets; grade 1 = focal or peripheral insulitis (lymphocytes surrounding the islet without destruction of endocrine cells); grade 2 = invasive destructive insulitis. For each group and experimental condition examined, a minimum of 100 islets/mouse was scored.
Quantitative RT-PCR
RNA was extracted using TRIzol (Life Technologies), followed by
isopropanol precipitation. Following reverse transcription, 50–200 ng
of reverse-transcription product was amplified using PCR for 30 cycles.
The RT-PCR quantification was done using a multispecific internal
standard (pMus) provided by D. Shire (Sanofi Recherche, Labège,
France). The primers used were (Eurobio, Bioprobe, Les Ulis, France):
IL-4, 5'-TCG GCA TTT TGA ACG AGG TC; IL-4, 3'-GAA AAG CCC GAA AGA GTC
TC; 2-microglobulin, 5'-TGA CCG GCT TGT ATG
CTA TC; 2-microglobulin, 3'-CAG TGT GAG CCA
GGA TAT AG. RT-PCR products were separated by 1.2% agarose gel
electrophoresis in TBE 1x containing 0.2 µg/ml of ethidium bromide
and visualized under UV light.
Detection of the neo gene
Whole pancreas from NOD and NOD-scid mice, Con A-stimulated
splenocytes from diabetic NOD, and islet-infiltrating cells from
NOD-scid recipients injected with -galactosidase-transduced cells
were homogenized using a commercial lysis solution (Wizard genomic DNA
purification; Promega, Madison, WI). The DNA was subsequently purified
following manufacturer’s recommendations. Genomic DNA (10 µg) was
digested with XbaI/EcoRI (Life Technologies), and
the fragments were separated on a 0.8% agarose gel. DNA was
transferred to Hybond (N+) membranes (Amersham,
Arlington Heights, IL). The membranes were then hybridized with a
32P-labeled bacterial neomycin phosphotransferase
gene (neor) probe.
-galactosidase gene expression
The fluorescent lipophilic -galactosidase substrate,
fluorescein di--D-galactopyranoside,
C12FDG (Imagene green; Molecular Probes, Leiden,
The Netherlands), was used. When the substrate is cleaved by the
enzyme, upon 2 h of incubation at 37°C, the green fluorescence
of cells expressing -galactosidase activity can be visualized by
flow cytometry.
Statistical analysis
The occurrence of diabetes in the different experimental groups
was plotted using the Kaplan-Meier method, i.e., nonparametric
cumulative survival plot. The statistical comparison between the curves
was performed using the logrank (Mantel-Cox) test, which provided the
corresponding 
2 values. In addition, when
needed, results were analyzed using the Student’s t
test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It has been well established that splenocytes from overtly diabetic NOD mice include pathogenic CD4 and CD8 T cells that adoptively transfer the disease into immunoincompetent recipients (12, 14, 29). As we have previously shown, the i.v. injection of 2–10 x 106 total splenocytes into NOD-scid recipients triggers overt diabetes within 4–6 wk (10, 11, 12, 13).
Using this experimental system, we investigated the effect of retrovirally mediated mIL-4 gene therapy on disease transfer.
To ensure ideal conditions for the integration of the retroviral gene,
which mostly occurs upon cell proliferation, all splenocyte populations
used were stimulated with the polyclonal activator Con A. In a
preliminary set of experiments, we confirmed that the diabetogenic
capacity of Con A-stimulated spleen cells from overtly diabetic NOD was
identical with that observed with unstimulated populations (data not
shown, see Fig. 1
legend). Mice
recipients of splenocytes infected with a -galactosidase retrovirus
were used as controls and exhibited an incidence and kinetics of
disease transfer superimposable to that observed in animals injected
with uninfected and unstimulated or Con A-stimulated splenocytes from
diabetic NOD mice (Fig. 1). The -galactosidase gene served both as a
control for the IL-4 gene and as a reporter gene to trace the homing to
islets of transduced cells by means of cell labeling using a
fluorescent lipophilic substrate (see below).
|
, diabetes transfer was very significantly and
reproducibly affected when mature cells, recovered from the spleen of
overtly diabetic NOD mice, were transduced with mIL-4. Thus, the data
from three independent experiments obtained at 8–12 wk after transfer
showed that overt disease appeared in only 0–20% of mice injected
with mIL-4-transduced T cells as compared with 80–100% of mice in the
control groups (Fig. 1).
At 24 h and 48 h from the infection, we quantified by
competitive RT-PCR the expression of specific mIL-4 mRNA within cells
infected with the mIL-4 or the -galactosidase retrovirus. As shown
on Fig. 2, at 24 h of in vitro
culture, a higher amount of specific mIL-4 mRNA was observed in
mIL-4-transduced cells as compared with uninfected control.
Moreover, at 48 h of culture, only in mIL-4-transduced cells was
the expression of mIL-4 mRNA detected (Fig. 2). These results suggest
the possibility of achieving a specific and relatively high retroviral
gene expression.
|
Upon transfer of diabetogenic cells, NOD-scid recipients showed a
progressive infiltration of the target tissue, namely the islets of
Langerhans (i.e., insulitis). Histological studies were performed in
mice receiving mIL-4-transduced cells as well as in controls. Pancreata
were recovered at 4 and 9 wk following transfer and processed as
already described according to conventional procedures
(13). Infiltration was present in all experimental groups
(Fig. 3). However, the topography of the
infiltration significantly varied between the experimental groups. As
shown on Fig. 3, in control mice that received
-galactosidase-transduced T cells and that developed a high disease
incidence, the majority of islets (90%) showed a pattern of
invasive/destructive insulitis. In clear contrast, in mice that were
protected from disease through the injection of mIL-4-transduced cells,
the vast majority of the islets were either intact (25%) or presented
with (60%) an infiltration that remained confined to the periphery of
the islets (noninvasive peri-insulitis). This pattern was stable over
time because identical results were scored at 9 wk after transfer.
|
-galactosidase
Different methods were used in parallel to assess the presence of
the retroviral gene and of the transduced proteins within
islet-infiltrating cells. Concerning the retroviral gene, Southern
blots were performed to evidence the neor gene.
In fact, the retroviral vectors were engineered to include, in
association with the mIL-4 or -galactosidase genes, the bacterial
neomycin phosphotransferase gene (neor). As shown
on Fig. 4, the digested fragment of 1200
bp containing the neor gene, indicating the
presence of the retroviral vector, was present in DNA samples prepared
using islet-infiltrating cells from NOD-scid recipients of
-galactosidase-transduced cells.
|
-galactosidase, was also evidenced within islet-infiltrating
lymphocytes from adoptively transferred mice. In recipients of
mIL-4-transduced cells, infiltrating lymphocytes, recovered at 12 days
from transfer, produced significant levels of mIL-4, upon polyclonal
stimulation using an immobilized TCR Ab (as compared with control
recipients of -galactosidase-transduced cells), as assessed by
specific ELISA (Fig. 5A). We
also investigated the expression of -galactosidase activity in
pancreatic infiltrating cells recovered at 15 days from transfer in
recipients of diabetogenic -galactosidase retrovirus-infected cells.
To this aim, we took advantage of the C12FDG
lipophilic substrate that is fluorescent upon cleavage by
-galactosidase. The green fluorescence of the cleaved substrate
could be easily measured by flow cytometry, and the results showed that
60% of TCR+ islet-infiltrating cells
expressed a significant -galactosidase activity (Fig. 5B).
|
C, mIL-4
levels were significantly higher in supernatants from TCR
Ab-stimulated cultures using splenocytes from recipients of
mIL-4-transduced cells as compared with control recipients of
-galactosidase-transduced cells.
We took advantage of mixing experiments in the adoptive transfer
system, using mixtures of cells from Thy-1.1 and Thy-1.2 congenic NOD
mice, to further prove that the mIL-4 production observed was
essentially derived from the mIL-4-transduced splenocytes. Thus,
adoptive transfer was performed using mixtures of uninfected
diabetogenic Thy-1.1 cells (from the spleen of diabetic Thy-1.1 NOD
mice) and purified mIL-4-transduced CD62L+
splenocytes from prediabetic Thy-1.2 congenic NODs. Recipients’ spleen
cells were recovered at 21 days from transfer, and the Thy-1.1 and
Thy-1.2 subsets were purified by means of magnetic beads sorting. These
purified subsets were then tested in vitro for their mIL-4-producing
ability. As shown on Fig. 5D, mIL-4 production almost
exclusively originated from the infected Thy-1.2 subset.
Some of these supernatants using splenocytes recovered from different
experimental groups were also tested for their levels of IFN- by
specific ELISA. A similar IFN- production (ranging from 42 to 68
ng/ml) was detected in bulk splenocyte cultures from mIL-4-transduced
or nontransduced CD62L+ cells and from
nontransduced CD62L- cells. Thus, the deduced
IFN-/IL-4 ratio was significantly decreased in supernatants from
mIL-4-transduced CD62L+ splenocytes when compared
with those from nontransduced cells (data not shown).
mIL-4 gene therapy has no effect on diabetogenic cells, but selectively potentiates the functional capacity of immunoregulatory lymphocytes
As we previously discussed, T cells present in the spleen of prediabetic and diabetic mice are functionally distinct. These include diabetogenic CD4+ and CD8+ T cells that transfer acute diabetes into immunoincompetent syngeneic recipients and a subset of immunoregulatory CD4+ T lymphocytes mediating "active tolerance" that control the pathogenic potential of diabetogenic effectors (11, 12, 13, 14). These immunoregulatory T cells are assessed through their capacity to inhibit adoptive transfer of disease by diabetogenic lymphocytes.
We therefore explored whether there was or not a selective T cell target (i.e., diabetogenic cells and/or immunoregulatory cells) responsible for the protective effect observed upon mIL-4 gene therapy.
In these experiments, we took advantage of the results we already discussed demonstrating that the L-selectin receptor (CD62L) is a unique marker to physically purify diabetogenic lymphocytes and separate them from immunoregulatory T cells (10). Thus, in overtly diabetic NOD mice, pathogenic splenocytes that transfer diabetes are exclusively included within the CD62L- T cell population. Conversely, CD62L+ T cells are totally devoided of any diabetogenic potential (10), and cotranfer experiments confirmed that even in overtly diabetic mice they do include significant proportions of immunoregulatory T cells (10, 11).
Selective mIL-4 gene therapy of diabetogenic effector cells
Retroviral mIL-4 infection of purified polyclonal
CD62L- cell splenocytes.
Purification of spleen T cell subsets was performed as already
described by means of magnetic beads cell sorting using the
CD62L-specific Ab MEL-14 (10, 11, 13). We first examined
whether mIL-4 transduction of purified CD62L-
cells could modulate their capacity to transfer diabetes. Results
clearly showed (Fig. 6A) that,
independently from the dose of cells (0.4 x
106, 1 x 106, 5
x 106), the diabetogenic capacity of
CD62L- T cells was not influenced upon
retroviral mIL-4 gene transfer (p = 0.241 for
the two cell number conditions detailed on Fig. 6A).
Recipients transferred with 5 x 106
purified diabetogenic CD62L- cells showed 100%
diabetes at 3 wk from transfer.
|
-/- transgenic mice expressing a
rearranged TCR from a CD4+ diabetogenic T
cell clone (17, 30). In BDC
2.5/C-/- transgenic mice, the proportion of
CD4+ T cells expressing CD62L is 50 to 70% lower
than that observed in wild-type NOD. In fact, the transgenic cells
preferentially express an activated/memory phenotype.
As shown on Fig. 6B, the incidence and the kinetics of
diabetes transfer were identical with infected or uninfected BDC
2.5/C-/- splenocytes.
Selective mIL-4 gene therapy of immunoregulatory cells
Adoptive cotransfers were performed according to a
modification of the method initially described by Boitard et al.
(12). NOD-scid recipients (4–5 wk of age) were injected
i.v. with mixtures of two distinct populations, namely uninfected
splenocytes from diabetic NOD mice (diabetogenic cells) and
mIL-4-transduced or nontransduced purified CD62L+
T cells from the spleen of prediabetic NOD mice (Fig. 6C).
In these experiments, mice injected with uninfected diabetogenic cells
alone were used as positive controls.
The results showed, as already described (11, 13), that
uninfected CD62L+ cells inhibit diabetes transfer
in a dose-dependent fashion. Moreover, at all dosages of
CD62L+ cells used, the retroviral mIL-4 infection
reproducibly improved their protective ability (Fig. 6C).
The differences observed in diabetes transfer between recipients of
nontransduced vs mIL-4-transduced CD62L+ cells
were statistically significant (cotransfer of 5 x
106 CD62L+,
p < 0.007; cotransfer of 20 x
106 CD62L+,
p < 0.004). Here again pancreas histology indicated
that in protected mice a majority of islets presented with peripheral
insulitis nondestructive (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4–6 wk. In a first step,
we have used these unseparated splenocytes to investigate the capacity
of retroviral mIL-4 gene therapy to modulate diabetes transfer.
Transduction experiments were conducted in parallel with the IL-4 or
the control -galactosidase gene. After the infection and before cell
transfer, a significant amount of specific IL-4 mRNA was detected in
cells infected with the mIL-4 retrovirus, thus confirming the effective
transcription and overexpression of the retrovirally transduced
gene.
Diabetes transfer was significantly and reproducibly reduced when
spleen cells from diabetic mice were IL-4 transduced. At variance,
recipients of -galactosidase-transduced cells exhibited an incidence
and kinetics of disease transfer superimposable to that observed in
animals injected with uninfected splenocytes. These results argue
against possible nonspecific bystander immune interference mediated
either by the viral vector or the local expression of a transgene.
Progressive insulitis, i.e., infiltration of the pancreatic islets was
detected in recipients of both -galactosidase- and IL-4-transduced
cells. However, the topography of the infiltration varied significantly
between the experimental groups. Mice receiving
-galactosidase-transduced T cells, which developed a high incidence
of disease, had invasive/destructive insulitis. In clear contrast, in
mice that were protected from disease by the injection of
IL-4-transduced cells, a vast majority of the islets were either intact
or had an infiltrate that remained confined to the periphery of the
islets (noninvasive peri-insulitis).
Several methods allowed us to assess the expression of the retroviral
genome and of the transduced proteins within these
islet-infiltrating cells. In Southern blots performed on DNA
preparations using pancreatic infiltrating cells from recipients of
transduced cells, the genomic bacterial neomycin phosphotransferase
gene band (neor), included within the retroviral
vectors, was detected. Importantly, in recipients of mIL-4-transduced
cells, pancreatic infiltrating lymphocytes, recovered 12 days after
transfer, produced significant levels of IL-4. The expression of
-galactosidase activity among pancreatic infiltrating cells, in
recipients of -galactosidase-transduced cells, was also investigated
using the C12FDG lipophilic substrate,
which becomes fluorescent upon cleavage by the enzyme. Results showed
that 60% of TCR+ islet-infiltrating
cells expressed significant -galactosidase activity.
As a whole, these ex vivo data confirm the reality of the IL-4 gene transduction and provide direct support, not available from previous studies, that the transduced cells home to the target organ, namely the islets, and produce locally a significant amount of the retrovirally transduced cytokine.
Our previous data have indicated that, even at an advanced stage of
disease, peripheral T cells from NOD mice include both diabetogenic T
cells and nondiabetogenic immunoregulatory CD4+ T
cells (11, 12, 13). We wondered whether the gene therapy
effect was concomitantly exerted on these two subsets or selectively in
one of them. The present results clearly showed that selective
transduction of purified effector cells did not inhibit their
diabetogenic potential when these cells were administered alone to
immunoincompetent hosts. We first took advantage of our demonstration,
discussed above, that L-selectin (CD62L) is absent from diabetogenic T
cells (10) to examine whether IL-4 transduction of
purified diabetogenic CD62L- cells could
modulate their capacity to transfer diabetes. This discrimination
afforded by CD62L appeared particularly useful in the context of these
studies, in which it was necessary to separate regulatory from effector
cells. CD25 has recently been described as another important marker of
ART cells in NOD mice (15), but it appeared less
attractive in this study because there is not yet definitive proof that
CD25- cells only include diabetogenic effectors
as CD62L- cells do. Results clearly showed that
the diabetogenic capacity of CD62L- T cells was
not modified upon retroviral IL-4 transduction. Importantly, this was
true independently from the number of cells transferred, which was as
high as 5 x 106 of IL-4-transduced or
nontransduced CD62L- T cells per recipient.
These results, obtained with polyclonal diabetogenic cells, were then
extended to transgenic (monoclonal) diabetogenic splenocytes from
BDC2.5 C-/- transgenic mice. These mice
express a rearranged TCR from a CD4+
diabetogenic T cell clone (30). The incidence and kinetics
of diabetes transfer were identical with infected or uninfected BDC2.5
C-/- splenocytes. This is well in keeping
with the data showing that, at variance with conventional NOD, cell
overexpression of IL-4 does not protect BDC2.5 TCR transgenic NOD mice
from diabetes (31). It was then suggested by the authors
that T cell diversity and expansion of what they defined
"nonpathogenic" specificities were required for the
counterregulating effect of diabetes induced by IL-4 (31).
This conclusion was in support of the hypothesis that locally delivered
IL-4 did not act directly on diabetogenic effectors, but rather
indirectly through a distinct target cell type (31).
It was thus not surprising to observe that selective transduction of the IL-4 gene in CD62L+ regulatory cells strongly potentiated the functional capacity of these cells. Cotransfers were performed as previously described (12). NOD-scid recipients were injected with mixtures of two distinct populations, namely uninfected diabetogenic splenocytes from diabetic NOD mice, and IL-4-transduced or nontransduced purified CD62L+ T cells from the spleen of prediabetic NOD mice. In these experiments, mice injected with uninfected diabetogenic cells alone were used as positive controls. As previously reported, uninfected CD62L+ cells inhibited diabetes transfer in a dose-dependent fashion. Using suboptimal doses of CD62L+ immunoregulatory cells, providing only partial protection, we could evidence that retroviral IL-4 infection reproducibly augmented their protective ability.
Two distinct nonmutually exclusive hypotheses may be proposed to
explain these results. The first is that of a selective effect of the
mIL-4 gene therapy on the subset of CD4+
CD62L+ T cells (ART immunoregulatory lymphocytes)
that control spontaneous diabetes onset. Alternatively, phenotypically
naive CD62L+ lymphocytes may differentiate, upon
mIL-4 transduction, into classical Th2-type cells involved in mediating
the protection. In favor of this conclusion is our present data showing
that the IFN-/IL-4 ratio is decreased in supernatants from
splenocytes including mIL-4-transduced CD62L+
cells that mediate the protection.
Another important and still pending question is whether IL-4 gene therapy will exert its effect in cis, the transduced cell being the effector of the protection, or in trans, through the recruitment of nontransduced effectors.
An interesting aspect of the described mIL-4 gene therapy is its
effectiveness on fully differentiated T cells recovered from NOD mice
at a late stage of the disease. This is at variance with most
previous data showing that IL-4-based therapies (systemic delivery of
the cytokine or transgenic cell overexpression) could prevent
disease in NOD mice only when applied at a very early stage (19, 20). Conversely, our results are in keeping with the data
showing that Lentivirus-mediated transduction of islet grafts with IL-4
protected them from autoimmune destruction once implanted into
diabetes-prone NOD mice (32).
At the therapeutic level, the interest of such a late therapeutic
effect is obvious because immunotherapy of human diabetes will involve
in the first place subjects with relatively advanced autoimmune disease
history. Cytokine-based gene therapy of autoimmune diabetes has to date
essentially involved either the transduction of somatic cells,
affording the systemic delivery of the cytokine (33, 34, 35),
or of the disease target, namely islet cells (32, 36).
More limited data are also available in diabetes and other autoimmune
diseases using transduced T cell clones or T cell hybridomas (37, 38). The strategy we used in this study circumvents first the
need for the transduction of cell that is hardly feasible in
humans, and second the usage of immortalized or long-term culture T
cells that may also be delicate to achieve in humans.
|
Acknowledgments |
|---|
-/- mice. We are also
indebted to Isabelle Cisse for managing the specific pathogen-free
mouse animal facility and to M. Netter and M. Kadouche for
the iconography. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Lucienne Chatenoud, Institut National de la Santé et de la Recherche Médicale Unité 25, Hôpital Necker, 161 Rue de Sèvres, 75743 Paris Cedex 15, France.
3 Abbreviations used in this paper: NOD, nonobese diabetic; ART, autoimmunity regulatory T cell; mIL, murine IL.
Received for publication September 27, 2000. Accepted for publication February 6, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
+CD4+ thymocytes expressing L-selectin mediate "active tolerance" in the nonobese diabetic mouse. J. Immunol. 161:2620.-cell antigens. Diabetes 46:758.[Abstract]
/IgG1 fusion protein. Gene Ther. 6:771.[Medline]
1 somatic gene therapy prevents autoimmune disease in nonobese diabetic mice. J. Immunol. 161:3950.This article has been cited by other articles:
|
|
 |
|
  T. D. Zorina, V. M. Subbotin, S. Bertera, A. M. Alexander, C. Haluszczak, B. Gambrell, R. Bottino, A. J. Styche, and M. Trucco Recovery of the Endogenous {beta} Cell Function in the NOD Model of Autoimmune Diabetes Stem Cells, July 1, 2003; 21(4): 377 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. von Herrath, B. Coon, T. Wolfe, and L. Chatenoud Nonmitogenic CD3 Antibody Reverses Virally Induced (Rat Insulin Promoter-Lymphocytic Choriomeningitis Virus) Autoimmune Diabetes Without Impeding Viral Clearance J. Immunol., January 15, 2002; 168(2): 933 - 941. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or
stimulated with an immobilized TCR
); B,
pools of islet-infiltrating cells were recovered from NOD-scid
recipients injected with 
and •) or
nontransduced (
). The figure describes the results of a
representative experiment of three independent ones performed that
provided identical results. The differences in diabetes transfer are
not significant (p = 0.241). In one of the
experiments,