| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Chemistry, Stanford University, Stanford, CA 94305
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
14 min. Evidence is presented suggesting that at
steady state <1% of the total cell surface I-Ek is active and that a
significant fraction of these active molecules originates from
intracellular pools as well as reactivation of inactive cell surface
I-Ek. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several laboratories have confirmed the existence of two functionally distinct empty class II MHC isomers in solution (6, 7, 8), an active or peptide-receptive isomer that binds peptide rapidly, and an inactive isomer that does not bind peptide. The formation of active MHC and its interconversion to the inactive isomer have been described by the following scheme:
|
|
Here p is peptide, p/M is peptide/MHC complex; koff and kon are the dissociation and association rate constants; Ma is the active form of MHC, Mi, the inactive form; and kac and kin are the activation and inactivation rate constants.
Rabinowitz et al. (6) provided indirect evidence for the existence of the two I-Ek isomers on fixed CH27 B cells. However, it remains unclear whether the behavior and kinetic properties measured for soluble I-Ek are representative of I-Ek expressed on the plasma membrane. This study is aimed at understanding the source of active class II MHC molecules, quantifying their amount and determining their kinetic properties on the surface of cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All peptides were synthesized using standard 9-fluorenylmethoxycarbonyl chemistry, purified by HPLC, and characterized by mass spectroscopy. The peptide sequences used were the following: invariant chain 85-99 mutant peptide (Ii 85-99 M90L M98L) (KPVSQLRMATPLLLR); invariant chain 85-99 peptide (Ii 85-99) (KPVSQMRMATPLLMR), and biotinylated moth cytochrome c 82-103 peptide (B-MCC) (FAGLKKANERADLIAYLKQATK). Biotin-labeling of Ii 85-99 and MCC 82-103 was conducted while still on resin by reaction of the free amino terminus with biotin N-hydroxysuccinimide ester (Pierce, Rockford, IL) in DMSO with a catalytic amount of diisopropylethylamine.
Cell lines
Chinese hamster ovary cells transfected with native I-Ek (CHO-Ek) (9) cells and CH27 B lymphomas were maintained by passage every 2 days in complete RPMI 1640 medium (Life Technologies, Grand Island, NY) containing 10% FBS (HyClone, Logan, UT), 25 mM HEPES, L-glutamine, 20 µM 2-ME, 900 U/ml penicillin-streptomycin (Life Technologies), and 9 µg/ml gentamicin (Life Technologies) at 37°C and 5% CO2. A subclone of Chinese hamster ovary cells (CHO-K1) was grown as described above and used as negative controls.
Abs and dyes
R-phycoerythrin 14.4.4S Ab was purchased from PharMingen (San Diego, CA) and used to determine the total amount of I-Ek on the cell surface. Briefly, cells were suspended in 100 µl PBS/2% FBS at pH 7.2 and incubated with R-PE 14.4.4S Ab (0.008 ng/cell) for 20 min at 4°C. Cells were washed twice with PBS/2% FBS, resuspended in 500 µl PBS/2% FBS, and immediately analyzed by FACS. Streptavidin-R-PE conjugate was purchased from both Sigma (St. Louis, MO) and PharMingen.
Preloading procedure
CHO-Ek cells were first incubated overnight with 200 µM Ii
85-99 M90L M98L in complete RPMI 1640 at pH 7.2 to generate Ii 85-99
M90L M98L/I-Ek complexes. Unbound Ii 85-99 M90L M98L was then removed
by aspirating the supernatant. Next, the Ii 85-99 M90L M98L/I-Ek
complexes were allowed to dissociate in the presence of 200 µM Ii
85-99 for a period of 2.5–3 h (2 half-lives). In this manner, we
replaced the Ii 85-99 M90L M98L/I-Ek complexes by Ii 85-99/I-Ek, which
dissociates at least 20 times faster from I-Ek. This faster
dissociation is essential for the experiments that require the rapid
generation active MHC.
Flow cytometry
In all experiments, cells were plated on six-well plates at a density of 250,000 cells/well in 2 ml RPMI 1640 and incubated overnight at 37°C and 5% CO2 in the presence or absence of a preloading peptide. For experiments measuring active I-Ek, cells were exposed to 50 µM B-MCC for varying time periods. For experiments determining the stability of I-Ek, cells were preloaded overnight as described above. The I-Ek/Ii 85-99 complexes were allowed to dissociate for varying time periods, and active I-Ek was measured by exposure to 50 µM B-MCC for 8 min. For peptide association experiments, cells were exposed to varying concentrations of B-MCC for 8 min.
In all experiments, cells were washed once with RPMI 1640 medium following exposure to B-MCC and released from the plate with a cold EDTA solution (0.526 mM). The cells were then centrifuged at 1200 rpm for 5 min, resuspended in 100 µl of PBS/2% FBS, pH 7.2, and stained for 20 min at 4°C with streptavidin-R-PE (0.36 µg). Cells were washed twice with PBS/2% FBS, resuspended in 500 µl PBS/2% FBS, and immediately analyzed.
For peptide dissociation experiments, cells were exposed overnight to 200 µM biotinylated invariant chain 85-99 peptide (B-Ii 85-99) at pH 7.2, 37°C and 5% CO2. Cells were released from the plate with EDTA and resuspended in 300 µl PBS/2% FBS. All handling was carried out at 4°C to avoid peptide dissociation. B-Ii 85-99/I-Ek complexes were allowed to dissociate at pH 7.2, 37°C for varied times in the presence of 1 µM MCC 95-103 as a competitor peptide. Then, cells were stained with streptavidin-R-PE as described above. Quantitation of fluorescence intensity was determined with a Quantum R-PE kit (Flow Cytometry Standards, San Juan, PR). Flow cytometry analysis was conducted on Coulter EPICS 753, Altra, and Elite XL flow cytometers. Data were analyzed using either Elite or EXPO Software (Beckman Coulter, Fullerton, CA).
Inhibitors
All inhibitors were purchased from Sigma and used at the following concentrations: brefeldin A, 20 µg/ml; sodium azide, 5 µM; cycloheximide, 5 µg/ml; and chloroquine, 5.2 µg/ml. Experiments with inhibitors involved incubation of cells with the above blockers for 1 h in RPMI 1640 at 37°C, pH 7.2 and 5% CO2. Cells were then washed with RPMI 1640 and either exposed to B-MCC to determine active MHC, or lifted off the plate and incubated with R-PE 14.4.4S Ab to determine total MHC.
Inactivation rate constant
The inactivation rate constant of surface-active I-Ek was
estimated by fitting the data in Fig. 3
B to the analytical
solution of the system of differential equations described by
Rabinowitz et al. (6). Briefly, the following differential
equations were derived from scheme I:
 | (1) |
 | (2) |
 | (3) |
|
which
arises from endogenous peptide dissociation, then we can derive the
following analytical solution to the system described above:
 | (4) |
is 0.3 from Fig. 3
B. Then,
[p/M]0 and
kin were estimated by obtaining the
best data fit to Equation 4. Apparent association rate constant
The apparent association rate constant for B-MCC binding to I-Ek
for preloaded cells was calculated by using the following simplified
version of scheme I:
|
|
Here p represents Ii 85-99, p/M is Ii 85-99/I-Ek complex, p* is B-MCC, p*/M is B-MCC/I-Ek complex, and kon is the B-MCC association rate constant. Our simplifying assumptions are that the reactivation of Mi is negligible on the time scale of our experiments because of its slow activation rate; that reassociation of unlabeled peptide p does not occur due to its low concentration; and that association of B-MCC is irreversible because of its small off-rate.
The concentration of B-MCC is essentially constant because it is in large excess with respect to cell surface I-Ek (µM range for B-MCC vs pM for I-Ek).
Thus we have the following system of differential equations:
 | (5) |
 | (6) |
 | (7) |
 | (8) |
 |
 | (9) |
A to this equation
using koff = 0.2
min-1 for Ii 85-99, an incubation time of 8 min
and kin = 0.049
min-1 ± 0.01 min-1 from
Fig. 3B.
Equation 9 is not applicable to the calculation of the association rate
constant for untreated cells. Instead, an apparent
kon value was calculated by fitting
the data in Fig. 3A to a single exponential curve.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
850,000
molecules/cell by comparing the fluorescence of known concentrations of
R-PE with the fluorescence of R-PE 14.4.4S Ab bound to cell surface
I-Ek (data not shown). B-MCC was used because of its slow dissociation
rate from I-Ek (t1/2 at pH 7.0
and 37°C > 200 h) (10). Peptide binding to
class II MHC was quantified using FACS together with streptavidin-R-PE
conjugate as a marker.
Specificity of the binding of B-MCC to I-Ek was determined using CHO-Ek
and CHO-K1 cells. Cells were incubated for 8 min with concentrations of
B-MCC ranging from 10 to 400 µM. CHO-K1 cells do not contain I-Ek on
their surface and do not exhibit significant peptide binding even at
high B-MCC concentrations (Fig. 1). In
contrast, CHO-Ek cells show a large increase in fluorescence after
peptide incubation, confirming that B-MCC binding is specific to I-Ek
molecules on the cell surface.
|
The dissociation kinetics of various peptides from solubilized I-Ek have been reported previously (11). To compare the dissociation rate of peptides from solubilized I-Ek to the dissociation rate from cell surface I-Ek, we exposed CHO-Ek cells to 200 µM B-Ii 85-99 at pH 7.2 and 37°C for 20 h. The cells were released from the plate with cold EDTA and resuspended in PBS/2% FBS in the presence of 1 µM competitor peptide (MCC 95-103). The B-Ii 85-99/I-Ek complexes were allowed to dissociate at 37°C for different time periods. Cells were stained with streptavidin-R-PE at 4°C.
Fig. 2 shows the dissociation of B-Ii
85-99 from I-Ek as measured by FACS. A dissociation half-life of 3.5
min for B-Ii 85-99 was calculated by fitting the data to a single
exponential curve. This value is slightly lower than that obtained for
the dissociation of carboxyfluorescein-labeled Ii 85-99 from
water-soluble I-Ek (t1/2 = 9
min at pH 7.0 and 37°C) (Peter Kasson and J.V., unpublished
results).
|
Rabinowitz et al. (6) found in experiments with purified soluble I-Ek that dissociation of a preformed peptide/MHC complex resulted in the formation of active I-Ek. The rate of peptide binding to MHC was approximately equal to the rate of dissociation of the preformed complex, suggesting that peptide associated rapidly when the active MHC isomer was formed. Based on these findings, we examined whether preloading CHO-Ek cells with fast dissociating peptides would result in an enhancement of peptide binding to cell surface I-Ek.
Preliminary experiments showed that incubating CHO-Ek cells
overnight with either 200 µM Ii 85-99 (dissociation
t1/2 = 3.5 min at pH 7.2 and 37°C),
or 200 µM Ii 85-99 M90L M98L (dissociation
t1/2 = 90 min at pH 7.0 and 37°C in
solution) (6) increased B-MCC binding to I-Ek by a factor
of 2 and 3, respectively, after a 30-min incubation when compared with
untreated cells (data not shown). It was also found that Ii 85-99 M90L
M98L bound 10 times more to cell surface I-Ek than Ii 85-99 (data
not shown), reflecting its greater affinity for I-Ek (11).
On the basis of these results and the need to generate the largest
possible amount of active I-Ek quickly, we used the two-step preloading
procedure described in Materials and Methods. This procedure
was found to be useful in terms of preparing a 5-fold enhanced
concentration of I-Ek loaded with Ii 85-99 as compared with no
preloading.
The physical chemistry of the two-step preloading procedure is
complex. The overall reaction stoichiometry (not mechanism) during the
overnight incubation is Mi +
p 
p/M. The peptide
equilibrium binding constants for this reaction are rather small. The
binding constants can be estimated as follows. The rate of inactivation
for Ma 
Mi is
0.049 min-1, as discussed later, and the rate activation
for Mi Ma is
0.00044 min-1 (6). From these values it can
be seen that the equilibrium constants for the reaction
Mi + p
p/M are 2 orders of magnitude smaller
than for Ma+ p
p/M. The data given later in this paper
enable one to estimate the equilibrium constants for the reaction
Ma + p
p/M at pH 7.0 to be 3 x
104 M-1 for Ii 85-99 and 8 x
105 M-1 for Ii 85-99 M90L M98L. Thus, the
binding of neither peptide is saturated at the concentration used (200
µM). During the second exchange step, where Ii 85-99 M90L M98L is
replaced by Ii 85-99, the system approaches equilibrium by a kinetic
pathway distinct from that compared with loading with Ii 85-99 alone.
This likely accounts for the modest relative enhancement of Ii 85-99
binding in the two-step procedure.
Fig. 3A shows an 5-fold
enhancement in B-MCC binding following the two-step preloading
procedure, whereas total levels of I-Ek (see Materials and
Methods) remained constant (data not shown), suggesting that the
dissociation of the prebound peptide resulted in the formation of
active I-Ek.
To determine whether the kinetic properties of active I-Ek on the cell
surface are similar to those in solution, we investigated the stability
of I-Ek generated by the preloading procedure. Preloaded CHO-Ek cells
were incubated in the absence of peptide for several time periods.
Cells were then exposed to 50 µM B-MCC for 8 min to estimate the
amount of active I-Ek present. Fig. 3B shows that active
I-Ek decreases rapidly when cells are incubated in the absence of
exogenous peptide. This decrease in B-MCC binding indicates that active
I-Ek rapidly inactivates following peptide dissociation. By fitting the
data in Fig. 3B to Equation 4 derived from scheme I
(6) (see Materials and Methods) with a
koff = 0.2
min-1 for Ii 85-99
(t1/2 = 3.5 min), one can
estimate an inactivation rate constant
kin = 0.049
min-1 ± 0.01 min-1
(t1/2 = 14 min).
Apparent association rate of MCC to I-Ek
The apparent association rate constant for B-MCC binding to I-Ek
on CHO-Ek cells was determined by measuring the concentration
dependence of peptide binding following an 8-min incubation.
Experiments were performed with both preloaded and untreated cells. The
association rate of B-MCC to I-Ek for preloaded cells was calculated by
fitting the data in Fig. 3A to Equation 9 (see
Materials and Methods), with t = 8 min and
peptide concentration variable. Because the rate of dissociation of
endogenous peptides from I-Ek is unknown, an apparent association rate
constant of B-MCC to I-Ek for untreated cells was calculated by fitting
the data to a single exponential curve. As a control, the data for
preloaded cells were also fit to a single exponential curve. Fig. 3A shows that although the saturation levels are higher for
preloaded cells, the apparent association rate constants obtained from
the curve fits are comparable, with approximate values of 25
M-1 s-1 for untreated
cells, and 35 M-1 s-1
(from single exponential curve) to 60 M-1
s-1 (from Equation 9) for preloaded cells.
Number of active MHC molecules present on the surface of CHO-Ek cells
To determine the number of active I-Ek molecules on CHO-Ek cells, cells were exposed to 50 µM B-MCC at pH 7.2 and 37°C for several short time periods. The goal was to expose cells to a sufficient concentration of peptide to quickly bind all active I-Ek molecules. The fluorescence intensity was calibrated as the number of molecules of equivalent soluble fluorochrome (MESF) using a Quantum R-PE kit from Flow Cytometry Standards. The MESF was used as a crude measure of the number of active I-Ek molecules present on the cell surface.
Fig. 4 shows the average of two
experiments performed simultaneously (to avoid small day-to-day
variations in the levels of active I-Ek). Each data point reflects the
number of active I-Ek molecules present. For untreated cells, the data
can be fit to a line with a slope giving the rate of formation of
active molecules, and an intercept at t = 0 equivalent
to the number of active I-Ek molecules at steady state. We estimate a
minimum of 2600 active I-Ek molecules present at steady state and a
rate of formation of 440 active molecules per minute.
|
Fig. 4 also shows the effect of the preloading procedure on the amount
of active I-Ek on the cell surface. For preloaded cells the half-time
of B-MCC binding is approximately equal to the half-time of Ii 85-99
dissociation. For this reason, the data were fit to a straight line
plus a single exponential curve that reflects the dissociation of
preformed complex. The binding rate constant derived from this fit
(0.18 min-1) is approximately equal to the Ii
85-99 dissociation constant obtained from Fig. 2 (0.2
min-1). When this binding rate constant is
multiplied by the B-MCC concentration used (50 µM), one obtains an
apparent association rate of 60 M-1
s-1 which is identical with that derived from
Fig. 3A.
Effect of transport and Ag processing inhibitors on I-Ek
A series of transport and Ag processing inhibitors were used to determine the origin of the active pool of surface MHC on cells. Brefeldin A reversibly disassembles the Golgi complex (13) preventing transport of Golgi-derived vesicles. Sodium azide inhibits mitochondrial respiration preventing energy-requiring processes (14). Chloroquine inhibits Ag processing and presentation of class II molecules (15) by raising the lysosomal pH, and cycloheximide is a protein synthesis inhibitor (14). We anticipated that each of the inhibitors would decrease the amount of active I-Ek on the cell surface.
CHO-Ek cells were exposed to inhibitors at pH 7.2, 37°C for 1 h.
The amount of active I-Ek was quantified by incubation with 40 µM
B-MCC for 10 min, and the total surface I-Ek expression was measured by
exposure to R-PE 14.4.4S Ab. Fig. 5 shows
that although the amount of active I-Ek on the cell surface decreased
by an average of 40% following treatment with inhibitors, the total
amount of I-Ek remained unchanged. (Similar results were obtained with
CH27 B lymphomas, where active I-Ek was undetectable after treatment
with inhibitors, whereas the total amount of I-Ek remained constant;
data not shown). These results show that MHC molecules are still
present at the cell surface after treatment with inhibitors and that a
significant part of the active MHC derives from newly synthesized or
recycled intracellular MHC. Assuming that the binding of peptide to
inhibitor-treated CHO-Ek cells follows a line similar to that in Fig. 4
(open circles), except with a lower slope, we infer that this source of
active I-Ek would arise from release of endogenous peptide, or
reactivation. Rough calculations indicate that reactivation dominates
this source of active I-Ek for inhibitor-treated cells. (We have
neglected reactivation in the analysis of experiments involving
preloading because this source of active I-Ek is rather small).
|
, B-MCC
bound I-Ek normally. The effect of the inhibitors was not detectable
when relatively large amounts of active I-Ek were generated by
preloading, suggesting that the steady-state number of active I-Ek
molecules is a small fraction of the total I-Ek on the cell
surface. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The dissociation rates for Ii 85-99/I-Ek in solution and on the cell surface differ by a factor of 3. The dissociation of B-Ii 85-99 follows a single exponential decay with a rate constant koff of 0.2 min-1 from I-Ek on the cell surface, compared with a koff of 0.077 min-1 for water-soluble I-Ek.
The apparent association rate constants of B-MCC to I-Ek on the cell
surface are 25 M-1 s-1
for untreated cells and 60 M-1
s-1 for preloaded cells. Due to the errors
involved in the inactivation and dissociation rate constants used to
determine kon, the values of these
apparent association rate constants should be taken as rough estimates.
Because the kon values for untreated
and preloaded cells are comparable, we infer that there is only one
form of active I-Ek present on CHO-Ek cells. The cell surface apparent
association rate constants are approximately 1 order of magnitude
slower than that obtained for water-soluble I-Ek and MCC peptide at pH
7.0 (6, 11).
Analogous to solution studies, we find that in the absence of peptide, active I-Ek is unstable, decaying to the inactive isomer with an approximate half-life of 14 min. One possible role of inactivation might be to maintain a steady but low number of peptide-receptive MHC molecules, facilitating the generation of a diverse repertoire of peptide/MHC complexes for T cell presentation while preventing the binding of a large number of self-peptides for which tolerance is weak or not established.
Consistent with studies indicating that B cells and macrophages express
little peptide-receptive MHC on their plasma membrane
(12), we find that at steady state 2600 I-Ek molecules
are active, representing <1% of the total MHC present in CHO-Ek
cells. A substantial fraction of these active I-Ek molecules must
derive from intracellular pools as well as reactivation. From the
steady-state number of active I-Ek molecules one estimates the minimum
number of peptide/MHC complexes required to trigger T cell activation.
Using the cytosensor microphysiometer as a measure of early T cell
activation, Beeson et al. (4) demonstrated that a 90-s
exposure to 10 nM MCC 88-103 was sufficient to induce a detectable acid
release in a mixture of 5CC7 T cells and CHO-Ek cells. However, a 1
µM MCC 88-103 concentration was required for maximal acidification
response. From these data and an association rate for B-MCC to I-Ek of
102 M-1
s-1, we estimate that 0.3 peptide/MHC
complexes per APC are required for initiation of detectable signaling
and 30 peptide/MHC complexes are required for maximal early
response. Our results agree with previous reports indicating that few
peptide/MHC complexes are required to trigger T cell activation
(16, 17, 18, 19, 20, 21).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Harden M. McConnell, Department of Chemistry, Stanford University, Stanford, CA 94305. E-mail address: harden{at}leland.stanford.edu
3 Abbreviations used in this paper: CHO-Ek, Chinese hamster ovary cells transfected with native I-Ek; CHO-K1, subclone of Chinese hamster ovary cells; B-MCC, biotinylated moth cytochrome c 82-103 peptide; Ii 85-99, invariant chain 85-99 peptide; Ii 85-99 M90L M98L, invariant chain 85-99 mutant peptide; B-Ii 85-99, biotinylated invariant chain 85-99 peptide; MESF, molecules of equivalent soluble fluorochrome; R-PE, R-phycoerythrin.
Received for publication December 26, 2000. Accepted for publication March 20, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Potolicchio, S. Chitta, X. Xu, D. Fonseca, G. Crisi, V. Horejsi, J. L. Strominger, L. J. Stern, G. Raposo, and L. Santambrogio Conformational Variation of Surface Class II MHC Proteins during Myeloid Dendritic Cell Differentiation Accompanies Structural Changes in Lysosomal MIIC J. Immunol., October 15, 2005; 175(8): 4935 - 4947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Marin-Esteban, K. Falk, and O. Rotzschke "Chemical Analogues" of HLA-DM Can Induce a Peptide-receptive State in HLA-DR Molecules J. Biol. Chem., December 3, 2004; 279(49): 50684 - 50690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Potolicchio, L. Santambrogio, and J. L. Strominger Molecular Interaction and Enzymatic Activity of Macrophage Migration Inhibitory Factor with Immunorelevant Peptides J. Biol. Chem., August 15, 2003; 278(33): 30889 - 30895. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
