Reprints Available Directly from the Publisher Photocopying Permitted by License Only Expression of Growth Hormone Receptors by Lymphocyte Subpopulations in the Human Tonsil

The ability of human tonsillar lymphoid cells to express growth hormone receptor (hGH-N-R) was analyzed by flow cytometry. FITC-coupled recombinant human growth hormone (hGH-N) was used to reveal the receptors, in combination with phenotype markers. Unlike T cells, tonsillar B cells constitutively express the hGH-N receptor. Quiescent cells separated from activated cells by Percoll-gradient centrifugation bear fewer receptors than activated ones. Activated T cells express hGH-N-R, but the typical germinal centre CD4+CD57 T cells do not. These latter thus appear not to be fully activated. Inside the lymph follicles, the germinal centre CD38 B-cell population and the mantle-zone CD39 B-cell population display similar levels of hGH-N-R expression, but receptor density is lower on dividing dark-zone CD38+CD10 B cells. Different lymphoid-cell populations thus differ markedly in their ability to express the growth hormone receptor, in relation notably to their activation status. This highlights the link between the neuroendocrine system and the active immune defense.


e-Peters and
ooghe, 1995; Hooghe et  al., 1996).This highly regulated ensemble forms a network of physiological reactions.Growth hormone produced by the pituitary, under the control of the hypothalamus, is one of the main actors in this network.

Growth hormone (hGH-N), in addition to its well- known metabolic effects, can modulate the immune system through its direct action on lymphocytes and their accessory cells.It is well established that human lymphoid cells express the growth hormone receptor *Corresponding author.Present address: Institute of Human Histology, University of Liege, 20 rue de Pitteurs, B-4020 Libge, Belgium.

(hGH-N-R) and even produce hGH-N ( Kiess and  Butenandt, 1985; Badolato et al., 1994; see for review Hooghe et al., 1996).The studies establishing these facts focused solely on crude lymphoid-cell populations, taking no account of the activation or matura- tion levels of subpopulations present.

Here we have examined expression of hGH-N-R by lymphocyte subpopulations freshly prepared from human tonsillar lymph follicles and by a mitogen- activated lymphocyte population.We have addressed the question: Does hGH-N-R expression by lympho- cytes correlate with th ir activation or maturation status?

Different lymphocyte subpopulations were ana- lyzed by flow cytometry.Mantle-zone B cells were studied by CD39 labeling, germinal center B cells by CD38 labeling, and dividing germinal center B cells by CD38+CD10 labeling (Gregory et al., 1987; Liu et  al., 1992; Hardie et al., 1993; Lagresle et al., 1993;  MacLennan, 1994; Kremmidiotis and Zola, 1995;  Lagresle et al, 1995; Dono et al., 1996).A peculiar population of T cells has been found inside the germinal centers; these cells are CD4+, most of them are CD57 (Velardi et al., 1986; Bouzahzah et al.,  1996a).On the basis of CD57 expression, we purified them by MACS and analyzed their hGH-N-R expression.Activated cells were separated from quiescent ones by Percoll-gradient centrifugation.

Expression of hGH-N-R was revealed by two different means: with mAb263 (mouse anti-hGH-N- R) and with FITC-labeled hGH-N.Since both meth- ods gave similar results (to be published), we present here the results obtained with hGH-N/FITC.B double labeling (anti-CD/PE and hGH-N/ FITC), we identified cell types and checked their hGH-N-R expression.


MATERIAL AND METHODS


Antibodies

Phycoerythrin (PE)-conjugated anti-CD3, anti-CD 10, anti-CD19, and unconjugated anti-CD19 and anti- CD8 monoclonal antibodies were provided by DAKO (Denmark).Simultest anti-CD3/FITC/CD 19/PE, anti- C

anti-CD57 m
bs were purchased from Becton-Dickinson (Rutherford, NJ).

Rabbit anti-human Ig coupled to polyacrylamide beads was produced by Biorad Laboratories (Rich- mond, VA) and used at 1/600 final dilution.Anti- CD39/PE mAb was provided by Pharmingen (USA) and streptavidinE by Boehringer Mannheim (Ge many).OKT3 and OKTll mAbs were kindly pro- vided by Thierry de France (Lyon, France) and mAb263 was a gracious gift from Ross Barnard (Queensland, Australia).


Isolation of T Cells

Tonsils from 1 to 8-year-old children were surgically removed and carried to the laboratory at 4C in a physiological solution containing 0.4% BSA (Sigma, USA).Lymphocyte populations were prepared by gently te

ing the tonsils.Free
ells were harvested and T cells were separated by rosetting using sheep red blood cells (SRBC), treated with 2-aminoethylth- iouronium, on a Ficoll Paque gradient (Pharmacia, Sweden).CD3 cells were selected by eliminating CD 19 cells from the rosetted population (Bouzahzah  et al., 1996b).For this, the cells were incubated with mouse anti-CD19 antibodies before rinsing and adding Dynabeads-coupled anti-mouse IgG (Dynal, Norway).Labeled B cells were eliminated in a magnetic field.The purity of the T cells usually exceeded 95%, as determined by cytofluorometry (FACScan, Becton-Dickinson, USA).

CD4 lymphocytes were purified similarly with anti-CD19 and anti-CD8 mAbs and Dynabeads.

CD4+CD57 T cells were prepared (Bouzahzah et al., 1995) by means of a magnetic cell sorter (MACS, Becton-Dickinson, USA): Briefly, CD4 cells were incubated for 30 min at 4C with anti-CD57 mAb, then rinsed and placed in contact with PE-conjugated avidin and biotinylated magnetic beads (Sanvertec, The Netherlands).The cells were transferred to a separation column in the magnetic field of a MACS apparatus.CD4+CD57 cells passed through the column, CD57 cells were retained by the magnetic field.The resulting CD57 and CD57-cells preparations were respectively 90% and 95% pure.

Bovine serum albumin (1 mg/ml BSA, Sigma) was conjugated to FITC by the same protocol and used as a negative control.


Isolation of B Cells

Of the nonrosetted cells, around 90% were B cells (CD19+).The concentration of B cells was im roved to 95% and 98% of total population by eliminating the contaminating T cells with OKT3 and OKTll and then Dynabea

TC (1 mg/ml) was dissolved
in dimethylsulfoxyde (DMSO, Sigma).A 250-/xl FITC solution was added to the dissolved hGH-N in 30-/xl aliquots dispensed at 15-sec intervals under constant stirring.The mixture was then incubated overnight in darkness at 4C. Free FITC was removed by three rounds of dialysis against 1000 volume of sodium bicarbonate buffer solution.This was performed at 4C in the dark.

The protein concentration and the fluorescein/ protein ratio (F/P) were determined by optical densi- tometry at 280 nm (protein) and 492 nm (FITC).

Fluorescein Labeling Binding of hGH-N/FITC was evaluated by incubating 10 6 lymphocytes with 5 / g hGH-N/FITC or BSA/ FITC in PBS, pH 7.4, for 1 hr at 4C in the dark.

BSA/FITC was used as a negative control.After 30 min, PE-conjugated mAbs (a

i-CD3, -CD 10, -CD 19
-CD38, or-CD39) were added at the adequate dilution.The cells were then washed twice in 1 ml cold PBS, resuspended in 750 /xl cold PBS, and analy ed with the FACScan.


RESULTS


Saturation Curve

To test the capacity of hGH-N/FITC to bind hGH-N receptors of living cells, we prepared tonsillar lymphoid cells, labeled them with anti-CD19/PE to stain B cells, and incubated them with increasing amounts of hGH-N/FIT

suring the percen
age of FITC-positive cells with the FACS.The results indicate that saturation level begins at approximately 5/xg/ml (Figure 1).BSA/FITC was used as a negative control (Rapaport et al., 1995).The binding speci- ficity of hGH-N/FITC has previously been proven by blocking with unlabeled hGH (Igout et al., 1995).The subsequent experiments were performed with 5 /xg hGH-N/FITC/ml, which enabled us to label a high percentage of hGHN-R while limiting non-specific binding.

Total Band T-Cell Populations Crude and cultured lymph-cell populations were double-labelled with hGH/N-FITC and anti-CD3/pE (for T lymphocytes) or anti-CD 19/PE (for B lympho- cytes) mAbs.The results (Table I and Figures 2 and 3) obtained from six

nsils show that practically no
T cell express hGHN-R wherease 60-95% of B cells express it.These percentages are underestimated,  since we used a nonsaturating concentration of hGH- N/FITC to stain the cells.Furthermore, experiments showed a strong rise in hGHN-R expression by T cells when these cells were activated by culturing for 24 or 48 hours in DMEM/F-12 with phycohemagglutinin L (PHA-L; 2 /xg/ml); hGHN-R expression also increased, but to a lesser extent, in lipopolysaccharide-activated B cells (10 /xg/ml lipopoly- saccharide; same culture conditions) (data not shown).In the tonsil, most T cells thus appear not to express hGHN-R, whereas nearly all B cells exhibit a basal expression level.To verify this pattern of hGHN-R expression, we prepared and studied iso- lated tonsillar Band T-cell subpopulations.T-Cell Subpopulations

We double-labeled subpopulations of T cells (CD4/, CD8+, or CD57+) in mixed B-, T-cell populations and purified CD3 or CD4+CD57 cells with hGH-N/ FITC.The results (Table II and Figure 2) show that in all tonsillar T-cell subpopulations, even for the CD57 subpopulation, less than 5% of the cell express hGHN-R.No difference was seen between the CD4 an

CD8 populations.Incuba
ion of these cells with PHA-L enhanced their hGH-N-R expression (data to be published).


B-Cell Subpopulations

Crude B-, T-cell populations or purified CD19 B cells were double-labeled with hGH-N/FITC and with anti-CD10, anti-CD19, anti-CD38, or anti-CD39/PE mAbs (Figure 4).The Mann-Whitney test was not applicable to either preparations because the results varied considerably from tonsil to tonsil (six tonsils, ce 5%) (Table III and Figure 3).To analyze differences between cell subpopulations, we thus examined each tonsil separately.In all six tonsils, CD38 and CD39 cells were found systematically to e s.Since this is probably due to the proliferation status of CD38/CD10 cells, we investigated hGH-N-R expression among high-density (nonactivated) and low-density (activated) tonsillar B and T lymphoid cells, separated on Percoll gradients.

T-Cell Gradients between tonsils, constantly show that more low- density T cells express hGHN-R than high-density T cells.The percentage of hGH-N-R positive cells is lower, nevertheless, among activated (low-density) T cells t an in the corresponding low-density B-cell population.

Unseparated lymph cells were centrifuged on Percoll gradients, yielding high-density (small, nonactivated) and low-density (large, activated) cells.Both populations were investigated by double-labeling with hGH- N/FITC and anti-CD3/PE, anti-CD4/PE, or anti-CD8/ PE mAbs.The results (Table IV), highly variable  N/FITC and anti-CD10/PE, anti-CD19/PE, anti- CD38/PE, or anti-CD39/PE mAbs.The results (Table V and Figure 5) show that low-density B cells express slightly more hGHN-R than high-density B cells and also more than the total B-cell population prior to separation.The hGH-N-R density varies on low- density B cells according to cell type: CD10 cells express less hGHN-R than other cell types.Here again, high variability between tonsils is apparent, but the results for each tonsil show a higher perce

age of labe
ed cells in the low-density population.In both the high-and low-density populations, fewer CD38/CD 10 cells than CD38/CD 10-or CD39 cell express the receptor.


DISCUSSION

Germinal centers develop in the secondary follicles of lymphoid organs by B-cell proliferation during a Tcell-dependent antibody response (MacLennan et al.,  1992; MacLennan, 1994).Each germinal center is divided into a dark and a light zone.Centroblasts proliferate in the dark zone, and, migrating to the light zone, mature in contact with antigen-presenting follicular dendritic cells (FDC).During their matura- tion, CD38/CD 10 B cells first lose CD 10, then CD38 at the end of maturation (Kremmidiotis and Zola,  1995), and finally CD39 after maturation has finished (Lagresle et al., 1993    Kiess and Butenandt (1985) have detected hGHN- R on mononuclear cells from peripheral venous blood.Here we have investigated the capacity o subpopulations of tonsillar B and T lymphocytes to express hGHN-R.Using fluorescein-labeled human growth hormone, cytometry, and cell-selection proce- dures, we have obtained new data on hGHN-R.Unlike tonsillar T cells, tonsillar B cells constitutively express these receptors.T cells do so transiently when activated.Interestingly, the peculiar CD4/CD57 germinal center T cells do not express hGH-N-R.This confirms data from other authors, indicating that these cells are only in a preactivated state (Bouzahzah et al.,  1995).

Among the follicular B cells, the dark- one CD38+CD 10 B cells possess few hGH-N-R.Perhaps proliferation causes B cells to reduce the density of hGH-N-R on their surface: During proliferation, activated cells express less hGH-N-R.These differences between B and T cells and between resting and proliferating cells were observed on both mixed (crude) and purified lymphoid-cell populations and after centrifugation in Percoll.The results are confirmed by in vitro studies on mitogen- activated lymphoid cell (in preparation).Contamina- tion by other cells bearing the CD of interest does not appear significant, since most of our data were obtained on highly purified Bor T-lymphocyte populations.

Since lymphocyte activation requires antigen rec- ognition and costimulatory signals given by accessory or lymphoid cells, we can now hypothesize that activated cells bear higher numbers of certain receptors rendering them particularly receptive to mes- sages, such as growth hormone, from the neuroendo- crine system.Quiescent B cells appear more sensitive to growth hormone signaling than T cells, but activation opens the latter to the influence of growth hormone.The sensitivity of t