Final Draft of Paper published in Toxicology,2000, Vol 143:125-140 IMMUNO-BE-LPT in Toxicology,143:125-140

DETECTION OF BERYLLIUM SENSITIVITY USING A FLOW CYTOMETRIC LYMPHOCYTE PROLIFERATION TEST: THE IMMUNO-Be-LPT

Georgia M. Farris1, Lee S. Newman2, Edward L. Frome3, Yulin Shou1, Elizabeth Barker2, Robert C. Habbersett1, Lisa Maier2, Hugh N. Smith1, and Babetta L. Marrone1*

1 Los Alamos National Laboratory, Life Sciences Division, Los Alamos, NM, 87545

2 National Jewish Medical and Research Center, Division of Environmental and Occupational Health Sciences, Denver, CO, 80206

3 Oak Ridge National Laboratory, Computer Science and Mathematics Division, Oak Ridge, TN, 37831

Abstract

Measurement of lymphocyte proliferation to detect hypersensitivity to beryllium (Be-LPT) in vitro is done presently using a method based on tritiated thymidine incorporation. Although this method is sensitive, it gives no information on cell viability or responding lymphocyte subsets. We have developed reliable and simple flow cytometric assays for lymphocyte proliferation testing (Immuno-Be-LPT) by combining immunophenotyping with bromodeoxyuridine (BrdU) incorporation or DNA content using propidium iodide (PI) or 4’6’-diimidazolin-2-phenylindole (DAPI). Evaluation of beryllium-induced lymphocyte proliferation in blood cells from 7 patients with Chronic Beryllium Disease (CBD) and 120 beryllium workers by both the Be-LPT and the Immuno-Be-LPT showed agreement between the tests. The Immuno-Be-LPT provided additional information about the specific type of lymphocytes responding. CD4+ lymphocytes proliferated in response to beryllium in blood samples from all 7 CBD individuals and CD8+ lymphocytes proliferated in 6 of the 7. Four beryllium workers without CBD had positive responses to beryllium in primarily the CD8+ cells. The use of the individual's own plasma supported a greater beryllium or tetanus-induced proliferation of CD4+ lymphocytes when compared to commercial human serum. The response of CD4+ lymphocytes measured in the Immuno-Be-LPT may provide a new marker for the diagnosis of CBD.

Keywords: Beryllium, lymphocyte proliferation test, flow cytometry, autologous plasma, CD4, CD8, immunophenotyping, hypersensitivity, BrdU, DAPI, PI, tetanus, Chronic Beryllium Disease

1. Introduction

The lymphocyte proliferation test (LPT) measures increased proliferation of blood lymphocytes in response to an antigen or mitogen in vitro. The LPT has been used to detect individual sensitization to a wide variety of antigens and toxins including: nickel (Rasanen and Tuomi, 1992) chromium (Rasanen et al., 1991), cow’s milk (Rasanen et al., 1992; Werfel et al., 1997)Error!Bookmarknotdefined., latex (Turjanmaa et al., 1989), and cytomegalovirus (Boland et al., 1990). The LPT has also been used clinically for detecting individual sensitivity to beryllium (Be-LPT) (Kreiss et al., 1993; Mroz et al., 1991; Yoshida et al., 1997)Error!Bookmarknotdefined.Error!Bookmarknotdefined. and as a predictor for development of Chronic Beryllium Disease (CBD), a lung granulomatous disease caused by a cell-mediated immune response to beryllium.

In the above studies, lymphocyte proliferation was determined by incorporation of [3H]-thymidine or [125I]-labeled iododioxyuridine following several days in culture with the antigen. Although the LPT is very sensitive, the bulk measurement of radioactivity by scintillation counting provides only a single value averaged over all of the cells (live or dead) in the culture at the time of harvest. Moreover, the use of the Be-LPT as part of medical surveillance programs to detect sensitivity in workers occupationally exposed to beryllium has been criticized because of reports of variability in results between labs performing the test (Stange et al., 1996). Improvements in the Be-LPT that would allow the elimination of dead cells in the culture from the analysis could play a key role in reducing inter-laboratory variability in Be-LPT results. In addition, the ability to measure lymphocyte proliferation within specific lymphocyte subsets would help to characterize the immune response within each individual. If proliferation within a specific subset of lymphocytes proves to have a strong correlation with CBD, then results from immunolabeling simultaneous with proliferation markers may facilitate diagnosis of disease or prediction of disease development. Both the CD4+ and CD8+ T lymphocytes appear to play critical roles in the development of lung granulomas (Suga et al., 1997) but proliferation of the lung CD4+ lymphocytes has been reported to be involved in Chronic Beryllium Disease (CBD) (Saltini et al., 1989).

In this report we describe two new ‘Immuno-Be-LPT’ assays that combine immunophenotyping analysis with a lymphocyte proliferation test. Fluorescent monoclonal antibodies (mAbs) and fluorescent DNA stains were used to evaluate DNA synthesis and DNA content in specific lymphocyte subsets. Lymphocytes were identified as helper/inducer and cytotoxic/suppressor T-lymphocytes by CD4 and CD8 mAbs, respectively. In the first Immuno-Be-LPT, cell proliferation was detected after several days in culture with tetanus toxoid or beryllium sulfate by measuring incorporation of bromodeoxyuridine (BrdU) in a manner analogous to the incorporation of thymidine analogs except that fluorescent tags were used instead of radiolabels. In the second Immuno-Be-LPT, cell proliferation was confirmed using DNA content analysis and cell cycle modeling after several days in culture with tetanus toxoid or beryllium sulfate and staining with 4’6’-diimidazolin-2-phenylindole (DAPI) or propidium iodide (PI). In addition, we evaluated an early marker of cell activation (CD69) to see if it correlated with the measurements of cell proliferation.

2. Materials and Methods

2.1. Blood samples

Blood samples from 7 individuals diagnosed with CBD were provided from NJMRC and a second source (requesting anonymity) for use in developing the Immuno-Be-LPT. All blood samples from CBD patients were analyzed by both the BrdU and DNA content Immuno-Be-LPT’s.

One hundred and twenty beryllium workers at Los Alamos National Laboratory (LANL) provided blood samples for this study and signed an informed consent approved by the LANL Institutional Review Board. These blood samples were split with 10 ml analyzed at LANL by the BrdU Immuno-Be-LPT and 30 ml of heparinized blood going to National Jewish Medical and Research Center (NJMRC) for analysis by the tritiated thymidine method (Be-LPT). Any LANL workers with an abnormal response to beryllium on the BrdU Immuno-Be-LPT then had a repeat sample analyzed by both the BrdU Immuno-Be-LPT and the DNA content Immuno-Be-LPT.

2.2. Immuno-Be-LPT

2.2.1. Blood mononuclear cell isolation and culture

Heparinized whole blood (8 ml) was mixed with 3 ml RPMI (Gibco, Grand Island, NY) and layered onto Histopaque-1077 (Sigma, Saint Louis, MO). The sample was then centrifuged for separation and collection of plasma and mononuclear cells (lymphocytes and monocytes). The diluted plasma (autologous) was collected off the top and the isolated mononuclear cells were washed twice with RPMI and reconstituted for culture in RPMI containing 10% pooled human serum, from AB+ males (Sigma) and RPMI containing 20% diluted autologous plasma. The specific lot of Sigma commercial human serum was chosen after evaluating several lots of serum from different sources. The lot of serum chosen supported the largest response to tetanus toxoid in 4 blood samples and also supported a response to beryllium sulfate with blood from a CBD patient. In a few cases (identified specifically in the results) mononuclear cells were also cultured in RPMI containing 10% autologous serum or 10% pooled human plasma from AB+ males (Biochemed Pharmacologicals, Winchester, VA).

The following three treatments were added to separate tubes of cells cultured with serum or plasma: no treatment (media control), 10 mM beryllium sulfate (Aldrich, Milwaukee, WI), and 1 Limit of Flocculation (LF) tetanus toxoid per ml (Pasteur Merieus Connaught, Swiftwater, PA. One LF tetanus toxoid (1 ml at 1,000 LF/ml) provided the optimal antigen response without toxicity. The set of samples for each type of proliferation assay and for each addition of serum or plasma included 2 media controls; 2 beryllium treated (10 mM); and 1 tetanus toxoid positive control. Additional culture tubes from individuals with CBD and LANL workers were treated with 100 mM beryllium sulfate until this treatment was discontinued. The final cell concentration was 0.6-1.0 x 106 cells per ml in 1 ml cultures within sterile Falcon 12x75 mm polystyrene tubes (Becton Dickinson). The cells were incubated at 37°C under 5% CO2 for 2 days for surface expression of CD69, and 6-7 days for proliferation analysis. This study started with analysis of proliferation after 7 days of culture based on the time course used in published methods (Frome et al., 1996; Mroz et al., 1991). Halfway through this study, the 6-day treatment period was determined as optimal based on evaluation of the proliferative response on days 5, 6, and 7.

2.2.2. BrdU incorporation and immunolabeling

Bromodeoxyuridine (BrdU) (Sigma) was added to the cultured cells at a final concentration of 40 µM, 16 hours before the end of the 6-day incubation period. At the end of the incubation, the cells were then washed with PBS/BSA and stained with CD8-TRI-COLOR (TC) and CD4-phycoerythrin (PE) antibodies and anti-BrdU-fluorescein isothiocyanate (FITC) as previously described (Carayon and Bord, 1992; Farris et al., 1997). After incubating the cells with surface antibodies the labeled cells were washed and fixed in 1% methanol-free formaldehyde (Electron Microscopy Sciences, Fort Washington, PA) with 0.05% Tween20 overnight. The cells were incubated with DNase (Sigma) in PBS containing Ca++ and Mg++ and washed with PBS containing 10% BSA and 0.5% Tween 20 before adding anti-BrdU-FITC (Becton-Dickinson). Although samples were analyzed immediately after washing off the anti-BrdU-FITC, it was observed that the staining appeared satisfactory for up to 48 hours at 4°C before analysis.

2.2.3. DNA staining

For DNA staining with 4’6’-diimidazolin-2-phenylindole (DAPI) (Sigma), cells were washed with PBS/BSA and incubated with CD3-TC, CD4-PE, and CD8-FITC for 20 min at room temperature. The cells were washed once with PBS/BSA, resuspended in 0.5% methanol-free formaldehyde and maintained at 4°C overnight or up to 2 weeks without adverse effects on the staining. One hour before evaluation by flow cytometry, the formaldehyde was washed off the cells using PBS and this was followed by staining with 1 µg/ml DAPI in PBS with 0.05% Tween 20. The DAPI-stained cells were kept dark at room temperature for 30 minutes before analysis.

For DNA staining with propidium idodide (PI) (Sigma), cells were washed with PBS/BSA and incubated with CD8-TC and CD4-FITC (Becton Dickinson, San Jose, CA) for 20 min at room temperature. These cells were washed once with PBS/BSA and cooled on ice, before staining in an ice-cold hypotonic PI solution (Brons et al., 1994). The samples were analyzed for DNA content and cell cycle stage after staining 30 minutes on ice, which was a modification from the published procedure.

2.2.4. Flow cytometry analysis

A majority of the debris and dead cells were excluded from the flow cytometry data acquisition by thresholding on forward light scatter. Intact lymphocytes were identified by forward and right angle (side) light-scatter. Cells being evaluated for immunolabeling, CD69 expression, or anti-BrdU were analyzed using a FACSCalibur model flow cytometer (Becton Dickinson) equipped with an argon-ion laser (488 nm). Emissions for FITC, PE and TC were collected at 515-545, 564-606, and >650 nm respectively. For each sample, 9000 lymphocytes were evaluated for CD69 expression and 12,000 for BrdU incorporation. Data was acquired using the Cell Quest software program (Becton-Dickinson) and then analyzed using the Winlist software program (Verity, Topsham, ME).

Cells being evaluated for immunolabeling and PI staining were analyzed using a FACSCalibur model flow cytometer equipped with an argon-ion laser (488 nm) and a red diode laser (635 nm). PI and CD4-FITC were excited by the 488 nm laser with collection of PI emissions at 564-606 nm. CD8-TC was excited by the red diode laser and captured by the FL4 detector (>650 nm). If only the FACSCalibur with the single 488 nm laser was available, then the samples were split and CD4-FITC and CD8-FITC were used in combination with PI in separate tubes. Cells being evaluated for immunolabeling and DAPI staining were analyzed using the Los Alamos Multiparameter Flow Cytometer (Steinkamp et al., 1991). DAPI was excited in the UV (351 nm), and emission was collected at 400-500nm. For each sample, 40,000 lymphocytes were collected for cell cycle analysis of DNA content distributions.

For analysis of proliferation by DNA content, intact lymphocytes were first selected in the bivariate light-scatter distribution. Cell doublets were gated out of the analysis in a bivariate distribution of DNA peak height and DNA peak area. Lymphocyte subsets were identified by TC, PE, and/or FITC fluorescence. The histogram of the DNA pulse area for each lymphocyte population was calculated by the Winlist software program (Verity) and the cell cycle position for each DNA pulse area histogram was then analyzed using the Modfit LT software program (Verity).

2.3. Tritiated thymidine Be-LPT

Mononuclear cells from whole blood were cultured in the presence or absence of 1, 10 or 100 uM beryllium sulfate in media containing 10% commercial human serum (Flow Laboratories, McLean, VA) as previously described (Mroz et al., 1991). Proliferation of mononuclear cells was evaluated on days 5 and 7 for the first half of the study and on days 4 and 6 for the second half of the study. The method for isolation of mononuclear cells, cell culture, pulse with tritiated thymidine and radiation counting has been described in detail (Frome et al., 1996).

2.4. CD69 labeling

The 48-hour cultured samples containing either cells alone, tetanus toxoid, or beryllium were washed with PBS containing 1% bovine serum albumin (Sigma, PBS/BSA). Monoclonal antibodies CD3-TC, CD4-PE or CD8-PE were added in addition to the mAb for CD69-FITC (Cal-Tag, Burlingame, CA) for 20 minutes at room temperature. The cells were washed with PBS/BSA and fixed with 0.5 % methanol-free formaldehyde (Electron Microscopy Services, Fort Washington, PA) for 30 minutes before analysis by flow cytometry.

2.5. Statistics

Determination of significant beryllium-induced proliferation using the Immuno-Be-LPT was based on the difference between the average of the two values for proliferation within the no-treatment cultures and the average of the two values for proliferation within the 10 uM beryllium-treated cultures. If this difference was greater than two times the standard deviation of the no-treatment proliferation values for all workers, then the response to beryllium was categorized as positive (abnormal). This formula was used for all lymphocytes, CD4+ lymphocytes, and CD8+ lymphocytes regardless of the serum or plasma source. Two standard deviations of the proliferation in the media control cultures was calculated several times during the LANL worker screening and consistently fell between 1 and 2% for each of the subsets of lymphocytes or total lymphocytes. Therefore, 2% proliferation above the media control was determined as the cut-point for all treatment effects and culture conditions.

Data from individuals with CBD were analyzed for statistical significance of beryllium-induced lymphocyte proliferation as compared to the no-treatment controls by a one-tailed, paired T-test using the Minitab software program (Minitab Inc., State College, PA). The percent proliferation values were transformed into natural log for stabilization of variance before this statistical analysis.

After the completion of the study a statistical approach was developed to determine a cut-point using outlier resistant methods. This approach used standard functions in the statistical package Splus (MathSoft, Inc., Cambridge, MA).

Determination of significant beryllium-induced proliferation using the Be-LPT (tritiated thymidine method) used the mean peak stimulation index (SI), which was the maximal observed at any concentration of beryllium on any day tested (Mroz et al., 1991). Two or more beryllium SI values must be above an individually calculated cut-point for the beryllium response to be considered positive.

3. Results

3.1. Light scatter

The bivariate light scatter plot gave an indication of lymphocyte proliferation after 6 days of culture with an antigen even without analysis of fluorescence tags (Figure 1). Cells cultured in media alone had a relatively tight cluster of lymphocytes representing a homogeneous population of cells. By contrast, most of the cell cultures treated with tetanus and all of the CBD patient cell cultures treated with beryllium clearly had an increased lymphocyte size (forward scatter) and increased cytoplasm granularity (side scatter) indicating lymphocyte activation.

A bivariate distribution of forward and side light scatter allowed identification of lymphocytes separate and apart from dead cells (high side scatter and low forward scatter) and debris (low forward and side scatter). Typically we found less than 8% of the cells appeared dead in the no-treatment, 10 mM beryllium-treated, or tetanus-treated cultures. Cell counts at the end of the culture period were equivalent in the 10 mM beryllium and no-treatment cell cultures. However, the highest concentration of beryllium (100 mM), used commonly in the Be-LPT, induced up to 60% cell death determined by light scatter and this was confirmed by cells counts that were decreased to about 40-60% of the no-treatment values. therefore, the 100 mM beryllium sulfate dose was discontinued as a treatment early in the study.

 

 

3.2. BrdU assay

Cellular proliferation determined by the incorporation of BrdU was evaluated in CD4+ and CD8+ cells by gating on both light scatter and positive mAb staining. Figure 2 illustrates lymphocytes from an individual with CBD where 20% of the CD4+ cells incorporated BrdU in response to 10 mM beryllium. A 16 hour incorporation of 40 mM BrdU gave a consistent proliferation labeling and did not induce cell death, as determined by light scatter and cell counts. Figure 2 also shows an increased intensity of CD4-PE fluorescence that was seen in the ‘activated’ and enlarged CD4+ lymphocytes. We also observed an increased CD8-FITC fluorescence when CD8+ cells were activated (not shown).

The optimal culture period for treating cells with beryllium or tetanus toxoid was 6 days. By day 7 the percent proliferation of lymphocytes began decreasing compared to day 6 in most cases. In some cases, the number of CD4+ lymphocytes increased to be the primary cell type in the culture by day 7 rather than maintaining the balance of cell populations seen up to day 6. For example, when lymphocytes from one individual were cultured with tetanus toxoid for 5, 6, or 7 days there was 7.6%, 11.4% and 11.0% proliferation of CD4+ cells, respectively. By day 7, the CD4+ cells were 71% of the total lymphocytes as compared to 37% on day 5 and 48% on day 6.

 

 

 

3.3. DNA content assay

An assay for DNA content analysis in CD4+ and CD8+ cells was developed using either DAPI or PI for DNA staining. Light scatter measurements showed a clear distinction between live and dead cells and showed the increased forward and side scatter of activated lymphocytes similar to that illustrated for the BrdU method. In addition, the surface immunolabeling for CD4+ and CD8+ cells gave bright fluorescence distinct from unlabelled cells. Using PI or DAPI to determine DNA content produced clear visual evidence of proliferation. Cell cycle analysis software provided percentages of cells in G1, S and G2/M phases of the cell cycle. Figure 3 illustrates cell cycle modeling of the DNA pulse area for lymphocytes when blood mononuclear cells from an individual with CBD were exposed to 10 mM beryllium.

Acquiring flow cytometry information on 40,000 lymphocytes was necessary for cell cycle analysis since the CD8+ cell population was as low as 10% of the total lymphocytes in some of the cultures. Analyzing 40,000 lymphocytes insured adequate CD8+ cells for accurate cell cycle determinations considering that the percentage of cells in S phase was close to zero for lymphocytes cultured without treatment.

3.4. Lymphocyte proliferation in response to beryllium

Lymphocyte proliferation in response to tetanus occurred in most individuals tested, and thus served as an excellent positive control in the Immuno-Be-LPT. By comparison, lymphocyte proliferation to beryllium was rarely observed, and should only occur in individuals who have been sensitized to beryllium through exposure. Evaluation of the Immuno-Be-LPT response in blood from individuals with CBD was important for development of the assay before it was used to screen workers. Figure 4 shows the average lymphocyte response to beryllium in the BrdU assay (a) and the DNA content assay (b) when cells from CBD patient samples were cultured either in autologous plasma or commercial human serum. Good agreement was observed between the BrdU assay and the cell cycle analysis. Statistical analysis showed significant beryllium-induced proliferation of CD4+ and CD8+ cells when cultured with either autologous plasma or commercial human serum when evaluated by BrdU or DNA content. The CD4+ response to beryllium was about double that of the CD8+ response in autologous plasma. In 1 of the 7 CBD patients, there was a strong CD4+ response only when autologous plasma was used in the culture (10% by BrdU, and 14.7% by DNA content) and no beryllium-induced proliferation of CD8+ lymphocytes when cultured with either plasma or commercial human serum.

Of the 120 LANL workers screened with the BrdU Immuno-Be-LPT, 4 had beryllium-induced lymphocyte proliferation. The mean lymphocyte proliferation values for these 4 are shown in Figure 5. In the BrdU assay (a) there was proliferation of CD8+ lymphocytes only when commercial human serum was used in the culture. In contrast, the means of the DNA content assay results (b) did not show a beryllium-induced proliferation of any lymphocyte subset. Three of the 4 workers had a beryllium-induced proliferation of only CD8+ cells, only on the BrdU assay, and only when commercial human serum was used. The other worker had an increased proliferation of CD8+ cells when evaluated by BrdU or DNA content and also had a slight increase in proliferation of CD4+ cells on the BrdU assay. Clinical pulmonary evaluations have been completed on 3 of these workers and they were found not to have CBD.

 

 

 

Figure 6 shows the percent proliferation distributions of the natural log values from the BrdU assay for all lymphocytes, CD4, and CD8+ T lymphocytes in media only, 10uM beryllium , or tetanus toxoid for 58 of the 120 LANL workers, including 3 with a positive response to beryllium. These last 58 screenings were done using the 6-day culture as compared to earlier data where the cultures were 7 days. The distributions for the beryllium-treated samples and the media controls for all lymphocytes, CD4+ and CD8+ T lymphocytes cultured with commercial human serum or autologous plasma were very similar. The shaded notches in the box plots for the beryllium-treated samples and controls overlap, indicating that the differences in the medians were not significant at an approximate 5% level (Chambers et al., 1983). Figure 6 also shows a strong increase in the proliferation of all lymphocytes, CD4, and CD8+ T lymphocytes in response to tetanus toxoid in both commercial serum and autologous plasma. Based on this data set, a statistical approach was developed to evaluate beryllium-induced proliferation detected by the Immuno-Be-LPT. Graphical analyses indicated that the natural log of the cell proliferation values for all lymphocytes, CD4+ cells, and CD8+ cells cultured in commercial human serum and autologous plasma were approximated well by a Gaussian distribution. The log ratios of treatment and no-treatment (media) results were therefore approximately normally distributed. The median (M) of the log ratios was used to estimate the location parameter and Sm = median absolute deviation (MAD) about M was used to estimate the scale parameter (Frome et al., 1996; Rousseeuw and Droux, 1993). A cut-point equal to M + 2.33*Sm was calculated, and individuals with log (ratios) that exceeded this value were identified for each distribution. Note that 2.33 is the 99th percentile of the standard normal distribution, so the probability of a log (ratio) exceeding this cut-point was 0.01. Using this statistical approach, we confirmed that the 3 positive LANL workers included in this data set were outliers of the beryllium treated groups.



 

Fig. 6. Boxplots showing lymphocyte proliferation based on the percentage of cells incorporating BrdU for 58 beryllium workers evaluated at LANL . Samples from all 58 workers were tested under identical treatment conditions. The percent proliferation (PP) for lymphocytes (All lymphocytes, CD4+, or CD8+) is shown on the right vertical axis, and has been plotted in natural logrithmic units shown on the left vertical axis. Results are shown for media only (M), 10mM beryllium sulfate (B) and tetanus toxoid (T). The ends of the boxes correspond to the 25th and 75th percentile, so fifty percent of the data are contained in the box for each group. The open circles indicate outliers. The shaded notches represent confidence intervals for the median of the log (PP).

The tritiated thymidine Be-LPT determined the same 4 LANL workers as positive and the other 116 workers as negative. Table 1 shows the agreement between results of the Be-LPT and the Immuno-Be-LPT for the 7 CBD patients and the 4 positive LANL workers. It is important to note that there are major differences in the analysis of the samples and the measurement of response when comparing Be-LPT and Immuno-Be-LPT results. The Be-LPT evaluates all cellular material in culture and the stimulation index is the measurement of total radioactivity count using an equation with the no-treatment value in the denominator. By comparison, the Immuno-Be-LPT evaluates live lymphocytes (not dead cells or cell debris) and the measurement is the percent cells in proliferation above the no-treatment control amount. For the closest comparison possible, the results of the Be-LPT are listed along side the results of the Immuno-Be-LPTs when all lymphocytes were evaluated and commercial human serum was used in the cultures. Our results found that these culture conditions and non-specific identification of cells did not give the best separation of individuals with CBD from workers that were non-CBD and +LPT. Therefore, we also listed in Table 1 the data for CD4+ lymphocytes using autologous plasma from the same blood samples in order to show the value of the additional data provided by the Immuno-Be-LPT.

 

Table 1. Proliferation responses in blood samples analyzed by the conventional Be-LPT and the Immuno-Be-LPT. Blood mononuclear cells were cultured in commercial human serum (serum) or autologous plasma (plasma) and cultured for 6-7 days with 10 mM beryllium sulfate, except where noted. Percent proliferation values have the corresponding media control values subtracted.

 

3[H]Thymidine

Stimulation Index (SI)

Total culture

serum

DNA content

% proliferation

All lymphocytes

serum

BrdU

% proliferation

All lymphocytes

serum

DNA content

% proliferation

CD4+ lymphocytes

plasma

BrdU

% proliferation

CD4+ lymphocytes

plasma

CBD

A

8.1 a

16.7

29.6

14.0

16.6

B

51.9

10.3

13.3

14.2

40.4

C

46.0

15.7

26.8

14.0

19.3

D

27.6

37.1

6.2

14.9

7.5

E

1.8

4.3

4.3

6.1

6.1

F

nd

17.0

10.5

10.4

7.9

G

nd

0.3

0.1

14.7

10.0

LPT + non-CBD

AA

4.4

1.6

2.5

0.4

0

BB

5.0

3.5

6.0

0.6

2.7

CC

9.1

1.1

1.9 b

0.3

0

DD

11.7 a

0.9

3.7

0.8

0

a The SI in response to 1 mM beryllium sulfate.

nd not determined on a split sample

b The response in CD8+ lymphocytes was 3.3%

 

3.5. Effect of Serum or Plasma

The lymphocyte samples from cultures using commercial serum appeared to have slightly more cell death; dimmer CD4-PE fluorescence; and a smaller percentage of CD4+ lymphocytes of the total lymphocytes as compared to cells cultured in autologous plasma (data not shown). Overall, cells cultured in autologous plasma gave a cleaner separation of populations by light scatter, surface immunolabeling, and anti-BrdU labeling than cells cultured with commercial human serum.

The proliferation response to tetanus toxoid or beryllium was influenced by the addition of commercial serum or autologous plasma to the cell culture. In the presence of commercial human serum, there was a greater tetanus-induced proliferation of CD8+ cells compared to the cultures containing autologous plasma (Figure 6). Commercial human serum consistently supported a beryllium-induced CD8+ lymphocyte proliferation in the 4 LANL workers (Figure 5a). Conversely, autologous plasma appeared to stimulate a greater proliferation of CD4+ cells as compared to commercial human serum. There was an average of 8.4% proliferation of CD4+ cells when lymphocytes from LANL workers were stimulated with tetanus toxoid and cultured with autologous plasma as compared to only 5.2% proliferation when cultured with commercial human serum. A similar pattern was seen with blood samples from CBD patients stimulated with beryllium (Figure 4). The individual with CBD labeled ‘G’ in Table 1 showed a CD4+ response to beryllium only in the presence of autologous plasma and no proliferation when commercial human serum was used in the culture.

In a pilot study evaluating the effects of various plasma and serum sources on lymphocyte proliferation, CD4+ responses to tetanus toxoid were greatest in the presence of autologous plasma, less by autologous serum, and lowest with commercial human serum. Figure 7 demonstrates a representative comparison from one LANL worker of tetanus toxoid-induced replication of CD4+ lymphocytes when autologous plasma, autologous serum, or commercial human serum were used in the culture. Collecting autologous serum required an additional tube of blood (clot tube), and because it was not as effective as


autologous plasma we did not pursue its use any further. Although the use of commercial pooled human plasma (from AB+ males) was considered initially to be a convenient source of plasma, it did not support antigen induced lymphocyte proliferation as well as autologous plasma (data not shown).

Fig. 7. An example of light scatter, CD4-PE fluorescence, and anti-BrdU-FITC fluorescence in lymphocytes stimulated with tetanus toxoid in cultures containing (a) autologous plasma, (b) autologous serum, or (c) commercial human serum.

 

3.6. CD69 Activation Assay

Assay of CD69 has been proposed as an indicator of cell activation that would correlate with cells in S phase several days later (Maino et al., 1995). Therefore, 48 hours after incubation with media, beryllium, or tetanus, lymphocytes were evaluated for expression of CD69. However, the percentage of CD69 positive cells did not consistently have a strong relationship with the eventual proliferation (on day 6 or 7) in response to tetanus toxoid when either commercial human serum or autologous plasma was used in the culture (data not shown). Therefore, the CD69 assay was not pursued further.

4. Discussion

We have developed two new multiparameter flow cytometry assays that combine lymphocyte proliferation with identification of lymphocyte populations to provide a more complete analysis of immune system hypersensitivity than the tritiated thymidine incorporation Be-LPT that is widely in use. We have termed our new method the ‘Immuno-Be-LPT’ because it combines immunophenotyping with a measure of antigen-induced lymphocyte proliferation in the same test. There are intrinsic advantages in replacing a radiochemical, bulk measurement assay such as the tritiated thymidine incorporation Be-LPT with a flow cytometry based assay. These include high-throughput, single cell measurement, information about cell viability, and proliferation measured in specific lymphocyte subsets. In addition, less blood is required, the fixed samples can be saved for 2-14 days before flow cytometry analysis, the data can be stored for reanalysis, fluorescent rather than radioactive reagents are used, and the reagent and labor cost is similar

to the tritiated thymidine incorporation test. Each sample is cultured, labeled with fluorescence, and analyzed on the flow cytometer without transferring the cells from the original culture tube. The flow cytometry analysis is straightforward and can be done on most clinical commercial flow cytometers. However, maintaining and operating a flow cytometer may be a disadvantage for some laboratories.

The analysis of lymphocyte proliferation by flow cytometry and BrdU or cell cycle analysis has been used to detect replication of lymphocytes in the study of various cancers, mitogen stimulation, or in vivo immunologic responses (Houck and Loken, 1985; Farris et al., 1997; Storek et al., 1992). However, this is the first report combining immunolabeling and cell proliferation markers for detection of antigen sensitization in in vitro cultures with the assays developed for large volume diagnostic testing. By optimizing the treatments, the amount of blood drawn from a worker for evaluation of hypersensitivity was minimal (8-10 ml of blood with heparin).

Although BrdU incorporation is a valuable tool for detecting proliferation, we recognized the limitation of using any nucleotide analogue as a quantitative measure of proliferation. Incorporation of thymidine analogues does not always correlate well with true increases in DNA content and cell proliferation (Neckers et al., 1995). Therefore, we also developed an Immuno-Be-LPT assay using DNA content and immunophenotyping for confirmation of antigen-stimulated proliferation. DAPI and PI each work well for determining the DNA content and are compatible with surface immunolabeling. The DAPI staining assay has the advantages of requiring a relatively weak dilution of mAbs for the immunophenotyping and the fixed cells can be stored up to 3 weeks before analysis without deterioration of the immunophenotyping results. Unfortunately, DAPI needs to be excited using a laser emitting light with an ultraviolet wavelength and this laser is not available on most of the clinical model flow cytometers. The PI staining assay requires a higher concentration of mAbs than the DAPI assay and data acquisition must be done within 3 hours after staining. However, the assay can be done on clinical model flow cytometers because the laser producing light at a 488 nm wavelength is present on a standard instrument.

Two assays were developed for detecting beryllium sensitivity. Up until now, beryllium workers who tested positive on the tritiated thymidine Be-LPT only had the option of having the same test repeated before deciding to undergo an extensive clinical pulmonary evaluation (often including invasive procedures) for CBD. We advocate use of a ‘Confirmatory’ DNA content Immuno-Be-LPT to follow up all positive beryllium responses on the BrdU ‘Screening’ Immuno-Be-LPT or on the conventional tritiated thymidine incorporation Be-LPT. The use of two different assays for evaluation of hypersensitivity will provide a double layer of certainty in our overall testing strategy and should minimize the number of individuals needing invasive lung testing.

In the CBD cases presented here, we observed a strong CD4+ response to beryllium in both the BrdU and DNA content assays. Of the 120 beryllium exposed workers tested, only 4 responded positively to beryllium in the Immuno-Be-LPT. By comparison to the CBD cases, this response was primarily in the CD8+ cells when cultured with commercial human serum. This difference in lymphocyte subset response to beryllium between the non-CBD and CBD individuals warrants further investigation from a larger cohort of current or former beryllium workers. Although we did not obtain a history from the LANL workers at the time of blood sampling it has been confirmed that these 4 workers with positive LPT results are all non-smokers. It has been previously demonstrated that cigarette smoking, age, gender, and race did not have an effect on the occurrence of CBD in beryllium workers (Kreiss et.al., 1977).

Commercial human serum and autologous plasma were compared to determine whether they influenced the proliferation response of the lymphocyte subsets in our assay. Commercial human serum has long been used as a nutrient source for lymphocyte stimulation cultures without considering the effect it may have on specific subsets, since total culture stimulation index is the typical endpoint. Although tetanus has been used experimentally to provoke antigen-stimulated lymphocyte proliferation (Allsopp et al., 1998), this is the first report showing that the amount of CD4+ and CD8+ cell proliferation in response to tetanus toxoid is dependent on the serum or plasma source in the culture. Commercial human serum enhanced the CD8+ lymphocyte response to antigens as compared to autologous plasma, and this enhancement was occasionally enough to make the total lymphocyte response positive in a few cases even though the CD4+ lymphocytes did not respond. This highlights the importance of being able to identify CD4+ and CD8+ lymphocytes and to characterize the proliferation response of each population. We suspected that the interaction of cells with commercial human serum was a possible source of variability reported in the Be-LPT as done by tritiated thymidine incorporation (Stange et al., 1996). By using autologous plasma we may learn more about the beryllium sensitivity and disease process and the roles of the components of autologous plasma that contribute to the reaction by comparing the responses of CD4+ and CD8+ lymphocytes. It has been reported that lymphocytes from lung aspirates of CBD patients respond to beryllium in vitro by clonal expansion of the CD4+ lymphocyte subset (Saltini et al., 1989). In particular, for evaluating sensitivity to beryllium, we would advocate adding autologous plasma to the cell cultures in order to optimize for CD4+ responses.

 

Two statistical approaches for determining positive responses on the Immuno-Be-LPT were used. The method used during the study considered a 2% increase in as indication of a positive response. The second method was developed at the end of the study to establish a more rigorous procedure based on a partial data set using the optimal flow cytometry protocol determined during the study. In future studies, an outlier resistant statistical approach similar to the second method will be developed to identify positive responders as results are collected.

5. Conclusion

We have developed more informative assays for clinical application to the in vitro testing of cell mediated immune responsiveness to beryllium. Our assays, based on analyses of cell proliferation by flow cytometry, have been optimized for medical surveillance to detect beryllium hypersensitivity in workers who have been occupationally exposed to beryllium. However, the assays may be readily adapted for testing sensitization to a variety of antigens. The Immuno-Be-LPT will help to define the roles of specific lymphocytes in the antigen sensitization process. Moreover, in the specific application to detection of beryllium sensitivity, it may improve the earlier prediction and diagnosis of CBD.

Acknowledgements

This work was performed at the Los Alamos National Laboratory, Los Alamos, NM, under the joint sponsorship of the United States Department of Energy (EH-61, and BER-72) and the National Flow Cytometry Resource (National Institutes of Health Grant No. P41-RR013150). The authors thank the medical technologists of the Occupational Medicine group at LANL for collecting the blood samples. Dr. Carlton Stewart provided valuable guidance on flow immunophenotyping during the initial phase of this work.

 

 

References

Allsopp, C.E.M., Nicholls, S.J., Langhorne, J., 1998. A flow cytometric method to assess antigen-specific proliferative responses of different subpopulations of fresh and cryopreserved human peripheral blood mononuclear cells. J. Immunol. Methods 214, 175-186.

Boland, G.J., DE Gast, G.C., Hené, R.J., Jambroes, G.J., Donckerwolcke, R., The, T.H., Mudde, G.C., 1990. Early detection of active cytomegalovirus (CMV) infection after heart and kidney trasplantation by testing for immediate early antigenemia and influence of cellular immunity on the occurrence of CMV infection. J. Clin. Microbiol. 28, 2069-2075.

Brons, P.P.T., Van Erp, P.E.J. & Pennings, A.H.M., 1994. Simultaneous DNA content and cell surface immunofluorescence analysis. In Methods in Cell Biology, Academic Press, pp. 95-102.

Carayon, P., Bord, A., 1992. Identification of DNA-replicating lymphocyte subsets using a new method to label the bromo-deoxyuridine incorporated into the DNA. J. Immunol. Methods 147, 225-230.

Chambers, J.M., Cleveland, W.S., Kleiner, B., Tukey, P.A., 1983. Graphical Methods for Data Analysis. Duxbury Press, Boston.

Farris, G.M., Robinson, S.N., Gaido, K.W., Wong, B.A., Wong, V.A., Hahn, W.P., Shah, R.S., 1997. Benzene-induced hematotoxicity and bone marrow compensation in B6C3F1 mice. Fund. Applied Toxicol. 36, 119-129.

Frome, E.L., Smith, M., Littlefield, G., Neubert, R., Colyer, S., 1996. Statistical methods for the blood beryllium lymphocyte proliferation test. Environ. Health Perspect. 105, 957-968.

Houck, D.W., Loken, M.R., 1985. Simultaneous analysis of cell surface antigens, bromodeoxyuridine incorporation and DNA content. Cytometry 6, 531-538.

Kreiss, K., Wasserman, S., Mroz, M.M., Newman, L.S., 1993. Beryllium disease screening in the ceramics industry. J. Occup. Med. 35, 267-274.

Kreiss, K., Mroz, M.M., Zhen, B., Wiedemann, H., and Barna, B., 1997. Risks of beryllium disease related to work processes at a metal, alloy, and oxide production plant. Occup. Environ. Med. 54:605-612.

Maino, V.C., Suni, M.A., Ruitenberg, J.J., 1995. Rapid flow cytometric method for measuring lymphocyte subset activation. Cytometry 20, 127-133.

Mroz, M.M., Kreiss, K., Lezotte, D.C., Campbell, P.A., Newman, L.S., 1991. Reexamination of the blood lymphocyte transformation test in the diagnosis of chronic beryllium disease. J. Allergy Clin. Immunol. 88, 54-60.

Neckers, L.M., Funkhouser, W.K., Trepel, J.B., Cossman, J., Gratzner, H.G., 1995. Significant non-s-phase DNA synthesis visualized by flow cytometry in activated and in malignant human lymphoid cells. Exp. Cell Res. 156, 429-438.

Rasanen, L., Lehto, M., Reunala, T., 1992. Diagnostic value of skin and laboratory tests in cow's milk allergy/intolerance. Clin. Exp. Allergy 22, 385-390.

Rasanen, L., Sainio, H., Lehto, M., Reunala, T., 1991. Lymphocyte proliferation test as a diagnostic aid in chromium contact sensitivity. Contact Dermatitis 25, 25-29.

Rasanen, L., Tuomi, M., 1992. Diagnostic value of the lymphocyte proliferation test in nickel contact allergy and provocation in occupational coin dermatitis. Contact Dermatitis 27, 250-254.

Rousseeuw, P.J., Droux, C., 1993. Alternatives to the median absolute deviation. J. Am. Statis. Assoc. 88, 1273-1283.

Saltini, C., Winestock, K., Kirby, M., Pinkston, P., Crystal, R.G., 1989. Maintenance of alveolitis in patients with chronic beryllium disease by beryllium-specific helper T cells. N. Engl. J. Med. 320, 1103-1109.

Stange, A.W., Furman, F.J., Hilmas, D.E., 1996. Rocky flats beryllium health surveillance. Environ. Health Perspect. 104, 981-986.

Steinkamp, J.A., Habbersett, R.C., Hiebert, R.D., 1991. Improved multilase/multiparameter flow cytometer for analysis and sorting of cells and particles. Rev. Sci. Instrum. 62, 2751-2764.

Storek, J., Schmid, I., Ferrara, S., Saxon, A., 1992. A novel B cell stimulation/proliferation assay using simultaneous flow cytometric detection of cell surface markers and DNA content. J. Immunol. Methods 151, 261-267.

Suga, M., Yamasaki, H., Nakagawa, K., Kohrogi, H., Ando, M., 1997. Mechanisms accounting for granulomatous responses in hypersensitivity pneumonitis. Sarcoidosis Vasculitis and Diffuse Lung Diseases 14, 131-138.

Turjanmaa, K., Rasanen, L., Lehto, M., Makinen-Kiljunen, S., Reunala, T, 1989. Basophil histamine release and lymphocyte proliferation tests in latex contact urticaria. Allergy 44, 181-186.

Werfel, T., Ahlers, G., Schmidt, P., Boeker, M., Kapp, A., Neumann, C., 1997. Milk-responsive atopic dermatitis is associated with a casein-specific lymphocyte response in adolescent and adult patients. J. Allergy Clin. Immunol. 99, 124-133.

Yoshida, T., Shima, S., Nagaoka, K., Taniwaki, H., Wada, A., Kurita, H., Morita, K., 1997. A study on the beryllium lymphocyte transformation test and the beryllium levels in working environment. Industrial Health 35, 374-379.