CELLULAR 134, IMMUNOLOGY Long-Term 157-170 Tracking (1991) of Lymphocytes of PKH-Labeled in Vivo: The Migration Lymphocytes’ GARY F.TEARE,* PA~JL K. HORAN,PSUSAN E. SLEZAK,~. CHERYL-SMITH,* AND JOHN B. HAY*,+' Studies of in rwo cell migration using cell markers such as “Cr. “‘In, FITC. or XRITC have been limited to short time periods due to the elution. toxicity. or raped loss of label detectability. WC have labeled sheep lymphocytes w virro with PKH-2. a new fluorescent cell membrane label, and, after their intravenous injection back into donor sheep. have been able to detect them tn efferent lymph. using tlow cytometry. for longer than 38 days. The PKH-2-lahclcd lymphocytes migrated with similar kinetics. efficiency, and tissue specificity as lymphocytes labeled with cell markers used previously. PKH-24abeled cells mediated graft versus host reactions indistinguishable from those mediated by unlabeled cells. and cell surface antigens were equally detectable on the surface of labeled and unlabeled lymphocytes. According to the slow. consistent loss of fluorescence intensity of the labeled cells in r’,w, we predict that labeled lymphocytes could remain detectable by flow cytometry for greater than 7 weeks with the labeling protocol used in these experiments. h 1991 Academic Press. Ini INTRODUCTION Understanding the role and timing of the arrival and departure of specific cell types or subsets in the pathogenesis of disease states, or in the generation of beneficial immune responses, may lead to rational strategies for immunoregulation. While much has been learned about lymphocyte localization and migration in animals using various cell labeling technologies, these studies have been limited to short-term time frames due to the instability or toxicity of the cell markers that have been employed. A new fluorescent cell labeling molecule, PKH-2. recently became available. This cell marker integrates into the lipid bilayer of cell membranes and has been used by various laboratories to label several mammalian cell types as well as bacteria and other cells ( 1). Studies with this labeling technology have shown that PKH-2 incorporates stably into the cell membrane of rabbit erythrocytes such that labeled cells become intensely fluorescent, and remain so over long periods of time after reinfusion into the blood ’ Funding for this work was provided by the Medical Research Council of Canada. G. F. Team is a fellow of the Medical Research Council of Canada. ’ To whom correspondence should be addressed at Department of Immunology, Faculty of Medicine. University of Toronto. I Kings College Circle. Toronto. Ontario. Canada M5S I AX. 157 0008-8749/9 copyright I $3.00 (cl 1991 h) Academx Press. Inc All rights of reproductmn I” any form resenrd. 158 TEARE ET AL. circulation of the rabbits (2). Several other parameters of cell function after labeling with PKH label were previously examined. The cytotoxic function of mouse spleen natural killer cells and the target susceptibility of Yac- 1 cells in culture were unchanged when either of these cell types were labeled with PKH label (3) and PKH-2-labeled Yac- 1 cells grew in culture with identical kinetics to unlabeled controls. Labeled cells did not transfer their label to unlabeled cells. Interestingly, with each doubling of cell number the mean fluorescence intensity of the labeled population decreased by half. Thus the PKH label has the potential to allow retrospective analysis of cell division. (S. E. Slezak, P. IS. Horan, and B. D. Jensen, in preparation). Before this label could be used to study cell migrations and interactions in vivo, the migratory capability of labeled cells must be evaluated. Here we have compared the migratory function of PKH-Zlabeled lymphocytes with that of lymphocytes labeled with radiolabels and other fluorescent markers in previous studies, and have also determined that these PKH-2-labeled lymphocytes are easily detectable over many weeks in vivo. MATERIALS AND METHODS ANIMALS,SURGERY,ANDLYMPHCOLLECTION Randomly bred ewe lambs, 9- 12 months of age, were used in all experiments. During the course of the experiments they were kept in metabolism cages with free access to food and water. The efferent lymphatics draining lymph nodes of the mesentery and subcutaneous (prefemoral, prescapular) lymph nodes were cannulated under pentobarbital anesthesia as described previously (4-6). Lymph was collected into sterile polyethylene bottles containing 0.25- 1.O ml of normal saline containing 1000 IU/ml heparin sodium USP (Hepalean, Organon Canada Ltd., Toronto, Ontario) and 20,000 IU/ml penicillin-G potassium (Ayerst Laboratories, Montreal, Quebec) such that the final concentration of heparin and penicillin in the lymph collection was approximately 5 IU/ml and 100 IU/ml, respectively. Lymph volume was measured in sterile polypropylene graduated cylinders and lymphocyte concentration in each sample was determined by hemocytometer. All cells for PKH-2 labeling were kept under strict sterile conditions throughout the labeling procedure. Lymphocytes for labeling were collected from the efferent lymph draining mesenteric or subcutaneous lymph nodes. The cells were isolated from lymph by centrifugation, washed once with Hank’s balanced salt solution (HBSS, pH 7.3) (GIBCO, Grand Island, NY) and resuspended to a concentration of 4 X 107/ml in the labeling diluent “A” which is provided with the PKH-2 dye (Zynaxis Cell Science, Malvern, PA). To a volume of diluent “A” equal to that into which the cells were suspended, sufficient PKH-2 dye was added to make a solution twice the desired labeling concentration. This dye/diluent solution was added to the cell/diluent suspension and gently but thoroughly mixed by agitation. The final labeling conditions were thus 2 X lo7 cells/ ml in diluent “A” with either 2.5 or 10 PM PKH-2 depending on the experiment. After incubating the cells with label at room temperature for 2.5 min, the labeling reaction was stopped by addition of a doubling volume of M- 199 medium containing LONG-TERM LYMPHOCYTE TRACKING 159 Earle’s salts, 25 mM HEPES, L-glutamine (M- 199; Flow Laboratories Inc., McLean, VA) with 10% lamb serum (GIBCO). All media for washes and incubations were used at room temperature. The labeled lymphocytes were recovered by centrifugation, and were resuspended in serum-free M- 199 for immunofluorescence, or in M- 199/ 10% lamb serum for in vivo experiments. An aliquot of labeled cells was always set aside for verification of staining and viability testing. The aliquot of labeled lymphocytes was analyzed by FCM3 for green (PKH-2) fluorescence intensity, percentage of cells labeled, and for viability by propidium iodide (7) and/or trypan blue exclusion. Greater than 98% of the cells became labeled every time we stained lymphocytes with PKH-2. Lymphocyte populations used for recirculation experiments were at least 85% viable when injected into the venous circulation of the sheep. INDIRECT IMMUN~FLU~RESCENCE Lymphocytes were recovered from lymph by centrifugation and were washed twice with HBSS. Some cells were labeled in 2.5 FM PKH-2 while controls were left unlabeled. Unlabeled cells were kept at room temperature in M-199 with 10% lamb serum while the other cells were being labeled. Both PKH-2 labeled and unlabeled cells were (separately) washed twice with HBSS at 4°C and resuspended to a concentration of 2 X 10’ cells/ml in serum-free M- 199. Two million cells ( 100 ~1) were placed into wells of a round-bottom 96-well microtiter plate so that each monoclonal antibody and each control was tested in duplicate. Two well-characterized monoclonal antibodies, SBU-T4, and SBU-I, which specifically bind sheep CD-4, and a nonpolymorphic determinant of MHC class I (OLA I) cell membrane antigens, respectively, were used in this study (8,9). The cells were suspended in 100 ~1 of a I:20 dilution ofthe primary antibody for 30 min at 4°C washed three times with ice cold M-199, and incubated in 100 ~1 of a 1:2 dilution (0.5 mg/ml) of goat anti-mouse Ig conjugated to R-PE (Sigma Chemical Co., St. Louis, MO) for 30 min at 4°C. After the secondary incubation, the cells were washed three times and were fixed in cold 1% paraformaldehyde for 30 min. Fixed cells were washed three more times and were resuspended in a storage medium consisting of M-199 cell culture medium, 1% bovine serum albumin (Sigma Chemical Co.), and 0.1% sodium azide. Cells were analyzed by FCM within 2 days. ANALYSIS OF CELLS BY FLOW CYTOMETRY Recirculution Experiments Cells collected for analysis by FCM were isolated from lymph by centrifugation (600 g for 5 min at 20°C) washed once with phosphate-buffered saline (PBS, pH 7.3) and resuspended in cold 1% paraformaldehyde at a concentration of 5 X 1O’/ml. After 30 min of fixation, the cells were again pelleted, the paraformaldehyde was decanted off, and they were resuspended in storage medium. The samples were stored in the dark at 4°C until analyzed. ’ ilhhrevia~iom zr.wd~ FITC, fluoresceinisothiocyanate;XRITC, substitutedrhodamine isothiocyanate: FCM. flow cytometry;MHC, major h&compatibility complex;lg. immunoglobulin;R-PE,R-phycoerythrin: MTT, modal transit time. TEARE ET AL 160 Five hundred thousand cells were analyzed for each sample on an EPICS V flow cytometer (Coulter Electronics, Hialeah, FL) with an argon laser set at 500 mW power, and 488 nm wavelength for excitation of the PKH-2 label which emits maximally at 504 nm. Green fluorescence of the PKH-2 was collected through a 525/40 bandpass filter. Forward and right angle light scatter were used to exclude cell clumps and debris from the analysis. All cells were analyzed within 7 days of their collection from the animal. Paraformaldehyde-fixed, PKH-2-labeled lymphocytes could be stored at 4°C for up to 14 days without loss of fluorescence intensity. Even after 28 days of storage, the mean fluorescence intensity did not vary by more than four channels when compared to the same sample analyzed within 7 days of collection from the animal. Since green fluorescence intensity was compared over a long period of time in these experiments, it was crucial to ensure that the flow cytometer was carefully aligned and calibrated for fluorescence intensity immediately prior to each analysis session. Immuno-Check alignment fluorospheres (Coulter Electronics, Batch 505 1) were run such that exactly the same fluorescence intensity (mean log channel number) was recorded prior to starting the analysis of cells. At the end of each analysis session, the fluorescence calibration was rechecked with the fluorospheres and was always found to be within less than one channel of the starting value. Immunojluorescence Experiments Ten thousand cells were analyzed from each sample on an EPICS V flow cytometer with an argon laser run at 500 mW power at an excitation wavelength of 488 nm. Green signal from PKH-2 fluorescence was collected through a 525/40 bandpass filter while orange-red signal from R-PE was collected through a 575/30 bandpass filter into separate photomultiplier tubes (PMT). Due to the high intensity of the PKH label, it was necessary to adjust both the electronic pulse subtraction module and the high voltage to the green PMT in order to achieve the necessary amount of compensation for spectral overlap. FLUORESCENCEMICROSCOPY PKH-2-labeled cells were examined and photographed using a Zeiss photomicroscope provided with an epifluorescence attachment and a phase contrast condenser. STATISTICAL ANALYSIS The paired-sample Student’s t statistic was used for comparison of the mean skin thicknesses in the graft-versus-host reaction assay. RESULTS INTENSEFLUORESCENCELABELINGOFLYMPHOCYTES PKH-2 was able to label sheep lymphocytes very intensely. In order to determine the most suitable labeling conditions, a series of experiments were performed varying the cell concentration, the PKH-2 concentration, or the time of incubation in the cell labeling procedure. The cells labeled very rapidly, with greater than 98% of the cells LONG-TERM LYMPHOCYTE 161 TRACKING becoming intensely fluorescent within as little as 60 set of incubation with PKH-2. It was found that cellular fluorescence increased with increasing concentration of PKH2 but that at labeling concentrations greater than 10 pA4 significant cytotoxicity occurred. At cell concentrations greater than 2 X 107/ml cell loss also increased proportional to cell concentration even at PKH-2 concentrations less than 10 PM. As a result of these experiments, we selected labeling conditions of 2 X IO7 cells/ml in 10 pA.4 PKH-2 for 2.5 min at 22°C for our in viva cell tracking experiments, in order to label the lymphocytes as intensely as possible with a minimum of cytotoxic effect. Cells labeled in this manner were approximately 3000 times as bright as unlabeled cells and were easily detected by flow cytometry (FCM) (Fig. 1A) or fluorescence microscopy (Fig. 2). Fluorescence microscopy of freshly labeled lymphocytes demonstrates that this dye labels the cell membrane and intracellular structures such as the nuclear membrane (Fig. 2). PKH-2-labeled lymphocytes with the same staining pattern were also easily visible in frozen tissue sections of lymph node. spleen. and Peyers patches (not shown). The labeling was so intense that very small numbers of labeled cells could be unequivocally identified in lymph hours or weeks after injection into the circulation of sheep. In one experiment, the efferent lymph draining a subcutaneous lymph node 1 Mean channel I 122 : I Mean channel 115 Mean channel 15 i lo3 Green Fluorescence FIG. I. Single parameter histograms showing the fluorescent intensity of PKH-?-labeled lymphocytes immediately prior to intravenous injection (A), and from efferent lymph 22-25.5 hr (B) and 501.5-5053.5 hr postinjection (C). The channel numbers from the O-256 logarithmic fluorescence intensity scale of the EPICS V have been converted to linear fluorescence units to simplify the comparisons. The “mean channel number” given is in these linear units. For histogram (A). a total of 10’ cells were counted. of which >9YX were PKH+. For each of histograms (B) and (C), a total of 5 X IO5 ceils were counted. Unlabeled cells in histograms (B) and (C) constituting 98.94 and 99.7% respectively of the total cells counted are extremely dim relative to the PKH-labeled ceils. and have accumulated in the first few channels of the histograms. 162 TEARE ET AL. FIG.2. Appearance of freshly labeled PKH-2-labeled 40X oil-immersion objective. lymphocyte under fluorescence microscopy using collected during the fifth hour after intravenous injection of PIUI-2-labeled cells was found to contain 2 1 labeled cells per million when analyzed by FCM (data not shown). FUNCTIONAL ABILITIES OF PKH-2-LABELED LYMPHOCYTES Several parameters of lymphocyte migration were examined with respect to possible effects of the PKH-2 cell label on the normal functioning of lymphocytes. When lymphocytes collected from mesenteric lymph were labeled with PKH-2 and returned to the venous circulation of sheep, the cells migrated into lymph in a nonrandom, tissue-specific manner. In both experiments, PKH-labeled mesenteric efferent lymphocytes migrated preferentially into mesenteric efferent lymph (Fig. 3). The ratio of the concentration of labeled mesenteric efferent lymphocytes returning in mesenteric efferent lymph to their concentration in subcutaneous efferent lymph over the first 40 hr postinjection was 1.87 in one experiment and 1.3 1 in the other. As expected, lymphocytes collected from subcutaneous efferent lymph and labeled with PKH-2 migrated randomly into subcutaneous efferent lymph after intravenous injection. PKH-2-labeled lymphocytes migrated from blood to lymph with the same efficiency as lymphocytes labeled with other cell markers as previously determined in this laboratory (Table 1). Lymphocytes enter efferent lymph at the rate of approximately 3 X 10’ cells per gram of lymph node each hour ( 10, 11). Since more lymphocytes will traffic through a large lymph node in a given time period than through a small one, we standardize and normalize our recovery values to the percentage of injected (labeled) cells recovered per 10’ cells collected (labeled and unlabeled) over 40 hr starting at LONG-TERM LYMPHOCYTE time 163 TRACKING (hours ) FIG. 3. Concentration of labeled cells in the mesenteric (e) and subcutaneous (0) efferent lymph of a sheep after intravenous injection of 3.0 X IO9 lymphocytes collected from mesenteric lymph and labeled in IO PM PKH-2. The kinetics and tissue specificity demonstrated by the recirculating PKH-?-labeled lymphocytes are not significantly different from lymphocytes labeled with radioisotopes or other fluorescent labels. the time of injection. This normalization permits comparison of labeled cell recovery regardless of the specific lymph node cannulated. The recovery of PKH-2-labeled lymphocytes returning to the efferent lymph compartment from which they were isolated was determined in four experiments and the mean recovery rate +- standard error of the mean (SEM) was found to be 0.45 + 0.1 I %/ 10’ cells in the first 40 hr after injection. Another well-defined parameter of lymphocyte tralhc in sheep is the time at which the peak output of labeled lymphocytes occurs in lymph relative to the time of injection of the cells (“modal transit time”). In the four experiments reported here, the modal transit time + SEM of PKH-2-labeled lymphocytes was 34.25 -t 3. I2 hr. Intravenously TABLE Recovery of PKH-Labeled Label in Lymph Comparison Recovery” (%/lo9 cells/40 hr) (mean t SEM) used Indium-I I1 Chromium-5 FITC XRITC PKH-2 Lymphocytes I 1 0.30 0.27 0.61 0.56 0.45 i f f k f 0.04 0.03 0.10 0.06 0. I I to Lymphocytes with Other Labels n” I I 25 8 6 4 ’ Percentage of labeled cells recovered in lymph per IO9 lymph cells collected in the first 40 hr after injection of labeled cells into the venous circulation of sheep. Normalization of the data to recovery per lo9 cells allows comparison of recoveries through lymph nodes of different sizes and thus differing cell output rates. h See text for references. 164 TEARE ET AL. injected PKH-2-labeled lymphocytes appeared in small numbers in the lymph within a few hours. The concentration of these cells increased rapidly over the course of the first 24 hr, peaked after 30 hr, and declined to 60-70% of the peak concentration over the subsequent 60-120 hr. After this initial decline the concentration of labeled cells remained steady, declining very slowly over time (Fig. 3). As a further test of function of PIUS2-labeled lymphocytes we employed the graft versus host (GVH) response as measured by change in skin thickness. This response depends on the viability, immune responsiveness, and capacity to proliferate of the injected allogeneic lymphocytes ( 12). As can be seen in Fig. 4, in all three experiments there was no statistically discernible difference in the GVH response of either 10’ viable PKH-2-labeled or unlabeled cells. In two of these experiments, labeled and unlabeled cells irradiated with 5000 rad of y radiation were tested for their GVH response simultaneously with the unirradiated lymphocytes. Although the irradiated cells were viable when injected, they failed to mediate a GVH response, confirming the role of donor cell proliferation in the GVH reaction. PKH Sheep DAY ID# 0 5 5 406 0 I SR 5R 0 682 5 5 0 I SR 5R 0 5 5 0 48 I 0 2 4 skin thickness 6 8 (mm) FIG. 4. IO7 viable PKH-2-labeled (PKH+) and unlabeled (PKH-) allogeneic lymphocytes were injected intradermally into at least three sites each in the hairless skin of the groins of three sheep. Skin thickness measurements were taken before injection (Day 0) and on the fifth day after injection of the cells (Day 5). In two of the sheep, labeled (PKH+ Day 5R) and unlabeled (PKFDay 5R) cells which had been irradiated with 5000 rad of y radiation were also injected into separate sites. There was no statistically discernible difference (406 and 682: P % 0.3; 48: P z 0.09) between the skin thicknesses of GVH responses within each sheep whether or not the allogeneic cells were labeled with PKH-2. Irradiated cells, though viable by the trypan blue exclusion method when injected, were unable to mediate a GVH reaction. Lymphocytes were labeled in IO PM PKH-2. Error bars show standard deviation. LONG-TERM LYMPHOCYTE 165 TRACKING Since PKH-2 labels the cell membrane we wanted to determine whether it would interfere with immunofluorescent detection of cell surface antigens. Lymph cells from the same collection were either labeled with PKH-2 or left unlabeled and were subsequently phenotyped using the SBU-T4 and SBU-1 antibodies to the sheep CD4 molecule and MHC class I products, respectively (Fig. 5). All tests were done in du- 96.72 0.95 0.06 D c 2.33 1.0 0. 0.02 i 0.26 0.11 4.78 yj 1: 0.29 I I! 0.08 0. 0. 1 G H 8-1 99.86 1 0.17 !i 1 0.06 I 1’ I 0.3 r! I I I 199.53 FIG. 5. Two-color flow cytometric analysis of lymphocytes labeled with PKH-2 and by indirect immunofluorescence. Two million unlabeled (A. C. E) or PKH-2-labeled (B. D. F) lymphocytes were incubated with either SBU-T4 (A, B) or SBU-I (C. D). monoclonal antibodies which identify the sheep CD4 molecule and MHC class I products, respectively, or with a control murine monoclonal antibody produced by ATCC hybridoma HB-65 (E, F), which recognizes a determinant on influenza nucleoprotein. The secondary reagent for immunofluorescence was goat anti-mouse-lg conjugated with R-PE. Two color histograms of unlabeled and PKH-2.labeled cells which were not incubated with antibodies are also shown (G, H). Ten thousand cells were analyzed for each histogram, and the numbers at the comers of the histogram quadrants indicate the percentage of total cells contained within that quadrant, 166 TEARE ET AL. plicate. The mean percentage of CD4+ and OLA I+ cells in the PKH-labeled (PKH+) and unlabeled (PKH-) groups were CD4+ PKH+, 50.7%; CD4+ PKH-, 51.9%; OLA I+ PKH+, 94.1%; OLA I+ PKH-, 96.8%. The presence of the PKH fluorescent label did not interfere with the detection of either of these cell surface molecules. LONG-TERMLYMPHOCYTETRACKING The mean fluorescence intensity of the labeled cells returning in the lymph in the first 24 hr was not substantially different from the intensity of the injected cells (Figs. 1A and 1B). In each of three experiments in which the labeled cells returning to lymph were monitored for greater than 48 hr the mean fluorescence intensity decreased at a steady rate over the entire time course, with a half-life ranging from 7.7 to 9.8 days in the three experiments (Fig. 6). As can be seen from Fig. 6a, PKH-2-labeled lymphocytes were readily detectable for greater than 38 days (-9 17 hr) in viva and indeed many of the labeled cells could remain detectable for 7 weeks or more assuming that the rate of fluorescence loss remained constant. As can be seen from Fig. 1, when the fluorescence half-life = 8.3 days o! . . . . . . . . . . . . . . . . . ,......,..,._.,..,,. 0 b. c. 0 !... 0 168 336 504 672 ‘~~“~~“‘~~““~~~“““~‘~‘~~~~~“’ 168 336 504 672 ,. __._,... ..,.. 336 504 672 840 840 226 o! .__... 0 I. 168 ,_, ..,,._..., 840 Time (hours) FIG.6. Mean fluorescence intensity of recirculating lymphocytes labeled in 10 ~~ PKH-2. Lymphocytes collected by chronic lymphatic drain were fixed in paraformaldehyde and analyzed by FCM. The figure shows results from three experiments in two separate sheep; (a) sheep T04, first injection of labeled lymphocytes; (b) sheep TOClabeled lymphocytes injected 26 days after the first injection; (c) sheep T05. The half-life of fluorescence intensity is remarkably consistent between experiments, with a mean of 8.6 days. LONG-TERM LYMPHOCYTE TRACKING 167 lymphocytes are labeled with PKH-2 the distribution of logarithmic fluorescence intensities in the cell population is symmetric and Gaussian. However, as time passed, the distribution of fluorescence intensity in the population of labeled lymphocytes recovered from the lymph became increasingly skewed and broadened to the left. DISCUSSION In this paper we have described some of the functional properties of sheep lymphocytes labeled with a new fluorescent marker molecule, PKH-2, which incorporates into the lipid bilayer of cell membranes and permits long-term cell tracking in rive. The integrity of the cell membrane is of primary importance to the viability and functionality of all cells. Previous studies had demonstrated that rabbit erythrocytes labeled with a PKH fluorescent cell membrane label were indistinguishable from unlabeled erythrocytes in their susceptibility to volume changes and lysis due to hypoosmolar stress (2). Thus, incorporation of the PKH label into the cell membrane did not physically weaken the membrane. Lymphocytes depend on the integrity and plasticity of their cell membrane for all of their functions including immune recognition and effector functions, and migration for the dissemination and surveillance of immunity. Slezak and Horan demonstrated that mouse spleen cells and a mouse lymphoma line, Yac- 1, could be labeled with a PKH cell marker molecule for use in a flow cytometric cell-mediated cytotoxicity assay which was found to have excellent correlation to the chromium release assay (3). Labeled Yac- 1 targets were neither more nor less susceptible to lysis by NK cells in the effector spleen cell population than were unlabeled targets. Similarly, PKH-labeled and unlabeled effector spleen cells were equally capable of killing targets. This data indicates that in lymphocytes and a lymphoid cell line, the presence of the label in the cell membrane neither interferes with recognition, adhesion, and lytic events nor renders the target cell membrane more susceptible to lysis. In their studies, the cells were labeled with comparable amounts of PKH-label as were used in our studies of sheep lymphocytes. Lymphocyte migration involves a complex series of events all of which are dependent on the functional integrity of the cell membrane. Lymphocytes must first attach themselves to endothelium by means of strong, sometimes tissue-specific intercellular adhesive interactions, then must squeeze themselves between the endothelial cells and cross through the basement membrane into the underlying tissue (13). Within tissues they presumably interact with numerous cell types by means of molecules on their cell surface until they enter the lymph which returns them to the blood circulation to repeat the process. This lymphocyte migration process has been studied in sheep by our laboratory and others for almost 20 years using cell markers such as Na2”Cr04, ‘“In-oxine fluorescein, or substituted rhodamine isothiocyanate (FITC, XRITC) as they became available ( 14- 18). The large body of data regarding the efficiency and kinetics of labeled-lymphocyte migration from blood to lymph, and tissue-specific migration patterns that has accumulated over the years, is remarkably consistent and therefore probably physiologically sound. This body of data also provides a standard against which we were able to compare the migration of PKH-2-labeled lymphocytes. Sheep lymphocytes labeled with PKH-2 migrated from blood to lymph with similar efficiency (see Table l), kinetics, and tissue specificity as lymphocytes labeled with radiolabels or fluorescent cellular protein labels. 168 TEARE ET AL. The length of time taken by the majority of intravenously injected, labeled lymphocytes to enter the efferent lymph has been variously reported at between 21 and 30 hr in sheep. This “modal transit time” (MTT) has been found to be relatively consistent whether the label was 5’Cr ( 17, 19), ” ‘In ( 15) FITC ( 16, 18), or XRITC ( 18). However, some studies of lymphocyte recirculation in sheep have shown a later time for the MTT, between 30 and 50 hr postinjection of “Cr- or FITC-labeled lymphocytes (20,2 1). The MTT we report here for PKH-2-labeled lymphocytes is slightly higher than the upper end of the range typically reported for lymphocytes labeled with the other markers. This may be a real phenomenon due to the label or labeling procedure, or may be due to our relative inexperience with this new labeling technique, compared with the other more familiar labels. As further evidence that the presence of the reporter molecules in the lipid bilayer of the cell membrane did not interfere with major function of lymphocytes, PKH-2labeled allogeneic lymphocytes were found to mediate graft versus host responses indistinguishable from those generated by unlabeled lymphocytes. The detectability of cell surface antigens was unaffected by the label at the lower (2.5 PM) labeling concentration, while at the higher concentration (10 p.M) the amount of spectral overlap of the extremely intense PKH fluorescence (green) into the orange-red emission region of the R-PE conjugated to the secondary antibody made adequate spectral overflow subtraction impossible. Thus, detection of the surface antigens CD4 and OLA I on the more intensely labeled cells was not feasible in our study. Earlier experiments in our laboratory determined that after intravenous injection lymphocytes taken from mesenteric efferent lymph and labeled with 51Cr or I1‘In returned preferentially to mesenteric efferent lymph such that the ratio of the specific activity in mesenteric efferent to that in subcutaneous efferent lymph (i.e., mesenteric cpm/ 10’ cells/subcutaneous cpm/ 10’ cells) over the first 40 hr was (mean ? SD) 1.74 + 0.2 1 (18). The analogous ratio in this study was the ratio of the concentration of PKH-Zlabeled mesenteric efferent lymphocytes returning in mesenteric efferent lymph to that in subcutaneous efferent lymph over the first 40 hr postinjection. In the two experiments of this type which were performed, the ratios (% PKH+ in mesenteric lymph/% PKH+ in subcutaneous lymph) were 1.87 and 1.3 1. There is no apparent difference in the tissue-specific, nonrandom migratory behavior of lymphocytes labeled with PKH-2 compared to those labeled with other labels. PKH-2-labeled lymphocytes were easily detected by FCM in lymph samples from sheep over a period of greater than 5 weeks from the time they were injected intravenously. The half-life of the mean fluorescence intensity of PKH-2-labeled lymphocytes in vivo was remarkably consistent from one experiment to the next, being on average 9 days. Although we did not continue any of our experiments beyond 9 10 hr after intravenous injection of the labeled cells (>5 weeks), we predict, on the basis of the constant rate of fluorescence intensity decay and the fluorescence intensity distribution at 910 hr, that PKH-2-labeled lymphocytes would remain detectable in vivo for at least 7 weeks. Published results from other laboratories, and our own experience with lymphocytes labeled with FITC, RITC, or XRITC, indicate that detection of lymphocytes labeled with these labels is generally limited to less than 2 weeks and often to only a few days ( 16, 17,2 1,22). Intravenously injected FITC- or RITC-labeled murine lymph node cells can be recovered from syngeneic recipient mouse lymph nodes after up to 11 days in vivo and detected by FCM or fluorescent microscopy in LONG-TERM LYMPHOCYTE TRACKlNG 169 cell suspension (22). However, by 1 1 days after injection only half of the injected cells known to be in the lymph nodes could be detected, and these had a median fluorescence intensity of only 0.9% of their initial intensity. The PKH-2-labeled sheep lymphocytes, however, still retained 12% of their original intensity after 2 1 days in viva. The reasons for the loss of fluorescence intensity from the PKH-2-labeled lymphocytes are not entirely clear. Part of the loss is likely due to leaching of dye from the surface of the cells. Unlike what some of us (S.E.S., P.K.H.) found with Yac-1 cells, the PKH-2 dye did transfer from labeled to unlabeled sheep lymphocytes in vitro. When labeled lymphocytes were mixed with an equal number of unlabeled ones in the wells of a round-bottom microtiter plate, gently centrifuged to ensure close apposition of the cells, and incubated at 37°C for 3 hr in complete medium, the mean intensity of the unlabeled cells approached 6% of that of the labeled cells. Since the dye transferred in vitro, it is likely to do so in vivo as well. We saw evidence of increased “background” fluorescence on the unlabeled population of lymphocytes in many of the lymph samples we collected after PKH-2-labeled cells were injected intravenously. It is significant that the newer PKH-26 cell-labeling molecule. which can be exited at 488 nm and emits at 572 nm (red), does not transfer between labeled and unlabeled sheep lymphocytes (unpublished observations). This new red-fluorescing cell label will perhaps be even more useful than the green-fluorescing PKH-2, in that it can be used with FITC-conjugated antibodies for two color phenotypic analysis of recirculating lymphocytes. Since the chemistry of the two PKH molecules are very similar, we do not anticipate any difference between PKH-2- and PKH-26-labeled lymphocytes in terms of their capacity to recirculate. Studies of lymphocyte migration depend on marker molecules to enable tracking of the cells in the body. The various radiolabels and fluorochromes used in such studies each have their own particular advantages and disadvantages; however. none of the labels previously available enabled long-term tracking of lymphocytes. As we have shown, PKH-2 enables detection of labeled lymphocytes for several weeks to months. The observations of this study should be seen as early hndings in the application of this new cell labeling technology. We find that when labeled at the higher doses of PKH-2 (i.e., greater than 5 PM) there is considerable variability in the viability of sheep lymphocytes from one experiment to the next. The cell death is due to a concentration-dependent cytotoxic effect of the PKH-2 molecule (unpublished observations); however, we do not have an adequate explanation yet for the variable susceptibility of lymphocytes from different sheep to the toxic effect. The PKH label will likely be most useful for labeling lymphocytes at less than or equal to 5 PM concentrations in suspensions of lymphocytes ranging from 2 to 5 X lO’/ml. Along with recent information on the role of cytokines on the local recruitment or local detention of lymphocytes (24-26) this new cell labeling technology makes possible a new realm of experimentation and possibly the determination of the real lifespans and long-term migration patterns of lymphocytes and their subsets in normal and pathological conditions. REFERENCES 1, Slezak, 2. Slezak, 3. Slezak, S. E., and Horan. S. E., and Horan, S. E., and Horan. P. K.. h’uruw (London) 340, 167. 1989. P. K.. Blood 74, 2 172, 1989. P. K., J. Immunol. MethA 177, 205. 1989. 170 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 2 I. 22. 23. 24. 25. 26. TEARE ET AL. Lascelles, A. K., and Morris, B., Q. J. Exp. Physiol. 46, 199, 1961. Hall, J. G., Q. J. Exp. Physiol. 52, 200, 1967. Heitmann, H. H., Zentrolbl. Veterinarrned. (A) 17, 517, 1970. Krishan, A., J. Cell Biol. 66, 188, 1975. Maddox, J. F., MacKay, C. R., and Brandon, M. R., Immunology 55, 739, 1985. Gogolin-Ewens, K. J., Mackay, C. R., Mercer, W. R., and Brandon, M. R., Immunology 56,717, 1985. Hall, J. G., and Morris, B., J. Exp. Med. 121, 901, 1965. Hay, J. B., and Hobbs, B. B., J. Exp. Med. 145, 31, 1977. Jones, M. A. S., and Lafferty, K. J., Aust. J. Exp. Biol. Med. Sci. 47, 159, 1969. Schoefl, G. I., and Miles, R. E., J. Exp. Med. 136, 568, 1972. Cahill, R. N. P., Poskitt, D. C., Frost, H., and Trnka, Z., J. Exp. Med. 145,420, 1977. Issekutz, T., Chin, W., and Hay, J. B., Clin. Exp. Zmmunol. 39, 215, 1980. Schnorr, K. L., Pearson, L. D., Knisley, K. A., and DeMartini, J. C., Znt. Archs. Allergy Appl. Zmmunol. 72, 239, 1983. Chin, G. W., and Cahill, R. N. P., Immunology 52, 341, 1984. Abernethy, N. J., Chin, W., Lyons, H., and Hay, J. B., Cytometry 6,407, 1985. Chin, W., and Hay, J. B., Gastroenterology 79, 123 1, 1980. Frost, H., Cahill, R. N. P., and Trnka, Z., Eur. J. Zmmunol. 5, 839, 1975. Reynolds, J., Heron, I., Dudler, L., and Trnka, Z., Immunology 47, 4 15, 1982. Butcher, E. C., Scollay, R. G., and Weissman, I. L., J. Zmmunol. Methods 37, 109, 1980. Melnicoff, M. J., Morahan, P. S., Jensen, B. D., Breslin, E. W., and Horan, P. K., J. Leuk. Biol. 43, 387, 1988. Kaalaji, A. N., Abemethy, N. J., McCullough, K., and Hay, J. B., Reg. Zmmunol. 1, 56, 1988. Hein, W. R., and Supersaxo, A., Zmmunology64, 469, 1988. Kaalaji, A. N., McCullough, K., and Hay, J. B., Immunology Lett. 23, 143, 1989.