- Email: [email protected]
Role of different B-cell subsets in the specific and polyclonal immune response to T-independent antigens type 2
Immunology Letters 88 (2003) 37 /42 www.elsevier.com/locate/
Role of different B-cell subsets in the specific and polyclonal immune response to T-independent antigens type 2 E.V. Sidorova a,*, Lu Li-Sheng b, B. Devlin b, I. Chernishova a, M. Gavrilova a a
Institute for Viral Preparations, Russian Academy of Medical Sciences, 1st Dubrovskaya Street 15, 115088 Moscow, Russia b Herzenberg Laboratory, Beckman Center B-007, Stanford, CA, USA Received 18 February 2003; accepted 25 February 2003
Abstract Role of different B-cell subsets in the immune response to T-independent antigen type 2 (TI-2) was studied. BALB/c and C57BL/6 mice were immunized by polyvinylpyrrolidon (PVP), and the numbers of antibody- and Ig-forming cells (AFC and IFC, respectively) were determined by ELISPOT method. The number of cells producing non-specific Ig (nIFC) was calculated as the difference between the number of IFC and AFC; the number of nIFC induced by PVP was calculated as the difference between the number of nIFC in immune and control splenocytes. Immunization by PVP induced not only the AFC appearance, but also the increase in the number of the antigen-induced nIFC. The treatment of splenocytes by anti-CD5 antibodies and guinea pig complement reduced the increase in the numbers of newly formed AFC and nIFC to /40% of control level. It means that CD5/ cells play an important role not only in the specific, but also in polyclonal immune response to non-self TI-2. To be sure that the decrease of AFC and nIFC numbers is due to depletion for CD5/ B-cells, but not CD5/ T-cells, splenocytes were separated to B-1 and B-2 subsets, and the numbers of AFC, IFC and nIFC were determined in each B-cell subpopulation separately. The overwhelming majority of newly formed AFC and nIFC was detected in B-1 subset. The numbers of AFC and nIFC in B-1 compartment was /10-fold greater than in B-2 cells. A close parallelism between AFC and nIFC formation was observed. It is concluded that specific and polyclonal immune response to non-self TI-2 */PVP */depends mainly on CD5/ B-1 subset. # 2003 Elsevier Science B.V. All rights reserved. Keywords: B-cells; Antibody; Immunoglobulins; T-independent antigens; CD5
1. Introduction The injection of the antigen (Ag) induces not only the biosynthesis of antibodies (Ab) and the appearance of the Ab-forming cells (AFC), but also greatly increases non-specific Ig (nIg) formation and the number of nIgforming cells (nIFC) [1]. The increase in nIFC number under the influence of T-dependent Ag (TD) is due mainly to the action of stimulatory T factors. The increase in nIFC number under the influence of Tindependent type 1 Ags (TI-1) depends on their mitogenic activity. The increase in the nIFC number under the influence of T-independent type 2 Ags (TI-2) lacking mitogenic activity is unexplained till now, and the role of
* Corresponding author. Tel.: /7-095-274-5501; fax: /7-095-2745710. E-mail address: [email protected] (E.V. Sidorova).
different B-cell subpopulations and the mechanisms responsible for this increase are not yet defined. B-cells are subdivided into CD5/ (B-1a), CD5/ (B1b) and CD5/ B-2 subsets [2,3]. It is hypothesized that B-1 cells can be positively selected by TI-2 [4 /7], and AFC arising in response to some TI-2 (TNP-Ficoll, Streptococcus pneumoniae , PVP) were shown to belong to B-1a cells [8,9]. At the same time, Ab to TI-2 could be produced by CD5/ B-cells [10,11], and in addition, CD5/ B-2 cells may become CD5/ B-cells after activation [12,13]. Thus, experiments have generated contradictory results, and the role of different B-cell subsets in the formation of specific Abs in response to TI-2 is not fully elucidated till now. Still less is known about the role of different B-cell subsets in nIg formation induced by TI-2 and about the mechanisms responsible for polyclonal B-cell activation in these cases. As a matter of fact, the cellular mechanisms of
0165-2478/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0165-2478(03)00055-5
38
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
B-cell activation by TI-2 as such are unclear. This research was done to study the role of CD5/ and CD5/ B-cells in the formation of Ab and nIg in response to non-self TI-2 */polyvinylpyrrolidon (PVP). Two different approaches were used. In the first one, splenocytes of mice immunized by PVP were depleted for CD5/ cells by the treatment with anti-CD5 Ab and complement (C), and the numbers of AFC and Igforming cells (IFC) in splenocyte suspensions were compared with those of control splenocytes. In the second approach, splenocytes from control and immunized mice were separated into CD5/ B-1 and CD5/ B-2 subsets, and the numbers of AFC, IFC and nIFC were determined in each of them.
2. Materials and methods 2.1. Animals, antigens and immunization BALB/c and C57Bl/6 female mice were obtained from Animal Facilities ‘‘Stolbovaya’’ (Russia) and Stanford University (USA). Polyvinylpyrrolidon 350 kDa (PVP, Serva) was used as TI-2. Mice were injected intravenously with 2 mg PVP in 0.2 ml of PBS. On the third or fourth day after immunization, spleens were taken, mononuclear suspensions were prepared and total splenocytes or CD5/ B-1 and CD5/ B-2 cells were used for the determination of AFC and IFC by ELISPOT method. Unmanipulated mice of the same age were used as control. Mice were killed either by cervical dislocation or by CO2 asphyxiation. 2.2. Depletion for CD5/ cells Single-cell suspensions (2 /106 cells/ml) were treated with C (pooled guinea pig serum, final dilution 1:10) and anti-CD5 Ab (Cederlain, Canada) according to manufacturer’s instruction (1 h at /4 8C and after addition of C 1 h at 37 8C). After incubation, cells were centrifuged (7 min, 1500 rpm), suspended in the RPMI 1640 medium with fetal calf serum (FCS) and antibiotics, and the number of viable cells was determined either with Trypan blue or with ethidium bromide/acridyl orange. 2.3. Flow cytometry mAb used here included Cascade Blue-labeled antiCD3e (2C11), anti-CD4 (GK1.5), anti-F4/80 and antiGr-1 (RB8-8C5), FITC-anti-CD21 (7G6), PE-antiCD23 (B3B4), Cy7-PE-anti-IgD (1126), Texas Red anti-IgM (331) and biotin anti-CD5 (53-7.8). mAb and fluorochrome conjugates were either prepared at Stanford [14,15] or obtained from Pharmingen (San Diego, CA). Streptavidin-Cy5PE from BD Pharmingen was
used as second-step reagent. Data were collected on HiD FACS (Cytomation/Becton Dickinson hybrid FACS) developed in FACS Facility at Stanford FACS/DESK as described [15]. Finally, data cataloged and stored with FACS/DESK (Stanford) were transferred to Flowjo (Tree Star, San Carlos, CA) for fluorescence compensation and analysis. 2.4. Cell sorting Single-cell suspensions from spleens were stained on ice with a combination of the anti-murine Abs in staining medium (biotin, flavin-deficient RPMI 1640 medium supplemented with 5% FCS and 0.02% sodium azide). Cells were washed with staining medium and incubated with Streptavidine-Cy5PE as second-step reagent for 15 min. Propidium iodide (PI) was added immediately before sorting at a final concentration of 0.25 mg/ml to discriminate viable cells. Cell sorting was performed by using a highly modified triple laser Hi-D FACS described in Refs. [14,15]. B-1 cells were identified as IgMhigh, IgDlow, CD23/, CD5/, and B-2 cells as IgMhigh, IgDhigh, CD23/, CD5/. Sorted cells were re-analysed after sorting and purities were /98%. 2.5. ELISPOT To determine the AFC and IFC numbers according to Ref. [16], nitrocellulose plates (MAHA 4550, Millipore Corp., Bedford, MA) were used. To determine the AFC number, nitrocellulose was treated by PVP (10 mg/well, 16 h at /4 8C, or 2 h at 37 8C), blocked with 1% BSA (Serva) (1.5 h at room temperature) and washed with PBS and RPMI 1640 medium. To determine IFC number, nitrocellulose coated by goat anti-mouse Ig (ICN) was used. To detect AFC 100/200/103 of total spleen cells per well, 40/80/103 of B-2 cells per well and 1.5 /2.0 /103 of B-1 cells per well suspended in RPMI 1640 medium with 3/5% of FCS were used in triplicate. To detect IFC 10 /20 /103 of total splenocytes or B-2 spleen cells and 1500 /2000/103 of B-1 cells per well were added. Cells were cultivated 18 /20 h at 37 8C in 5% of CO2. Cultivation being over, cells were washed out, and nitrocellulose wells were treated at first by rabbit antiserum specific to mouse IgM (anti-m serum) and then by horseradish peroxidase conjugated with sheep Ab to rabbit IgG. Membranes were dried in the dark, and the blue spots in each well were counted under microscope. The number of nIFC was calculated as the difference between the number of IFC and AFC; the number of Ag-induced nIFC was calculated as the difference in nIFC numbers in immune and normal splenocytes. All values for AFC, IFC and nIFC were expressed as the arithmetic mean9/S.D. per 106 of
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
pooled viable splenocytes or splenocytes of individual mice assayed separately.
3. Results In the first series of experiment, BALB/c mice were intravenously immunized by PVP (2 mg/mice), spleens were removed on the fourth day, and mononuclear suspensions were prepared. Splenocytes were treated with C or with anti-CD5 Ab plus C, and the numbers of AFC and IFC were determined by ELISPOT method according to Ref. [16]. The data obtained are shown in Table 1. It may be seen that PVP induces not only specific immune response, but also the increase in the number of cells producing nIg (nIFC), i.e. behaves as polyclonal B-cell activator. The treatment of splenocytes by anti-CD5 Ab and C markedly reduced the increase of AFC and nIFC in immune suspensions. The increase of AFC number diminished from /456/499 to /286 per 106 of viable splenocytes, and the increase of nIFC number from /593 /954 to /415 per 106 cells, i.e. for /40 and /36%, respectively. It means that during immune response to TI-2 CD5/ cells play role not only in AFC but also in nIFC formation as well. It must be noted that the treatment by anti-CD5 Ab of normal spleen cells practically did not influence the numbers of background AFC and nIFC. To determine if AFC and Ag-induced nIFC belong to B-1 subset in the next series of experiment, splenocytes from normal and immunized BALB/c and C57Bl/6 mice on the third day after immunization by PVP were separated into CD5/ B-1 and CD5/ B-2 subsets as described in Section 2. The FACS sorting of B-1 and B-2 cells from the spleen of BALB/c mice is shown in Fig. 1. It may be seen that the output of living cells was / 90.7% of input, the number of lymphocytes /70.4%, the number of total B-cells /24.6% and the numbers of B-1 and B-2 cells 14.8 and 68.5%, i.e. 0.45 and 10.6% of total spleen cells, respectively. The results obtained in experiments with 6 /8 individual BALB/c mice are represented in Table 2. It may be seen that immunization with PVP resulted in the increase in the number of cells producing Ab to PVP
39
and in a great increase in the number of IFC. The overwhelming majority of AFC and IFC were produced by CD5/ B-1 cells: 30179/1288/106 B-1 cells against 9399/317/106 cells and 104499/2168/106 B-1 cells against 25129/963/106 cells in control sample (respectively). The increase of PVP-induced nIFC number was equal to 42549/1777/106 B-1 cells. In immune B-2 subset, the numbers of AFC and IFC were also greater than in normal B-2 cells but nevertheless about 10 times less than in B-1 subset. It means that in both B-cell subsets PVP induces not only Ab formation but the formation of nIg as well. The relative increase of AFC and nIFC numbers in immune B-1 splenocytes was almost the same: 2.2 and 2.7, respectively. In B-2 subset the increase in nIFC number was much greater than in AFC content. However, these data were not beyond for doubts since the numbers of AFC and IFC in the wells could not be precisely estimated. Similar results were obtained in the experiments with C57B/6 mice, although immune response to PVP in C57Bl/6 mice was much less as compared to BALB/c mice. Thus, the number of AFC in control and immune C57Bl/6 B-1 subsets were 4749/190/106 cells and 7299/ 147/106 cells against 9399/317/106 cells and 30179/1288/ 106 cells in BALB/c B-1 cells, respectively, and the number of IFC in control and immune splenocytes were 18409/410/106 cells and 30999/699/106 cells against 25129/963/106 cells and 104499/2168/106 cells in BALB/c mice, respectively. But nevertheless, all values for AFC and IFC in B-1 subset of C57Bl/6 mice were again about 10-fold higher than in B-2 subset.
4. Discussion The phenotypic and functional relationships among various B-cell subset are of importance for better understanding of the immune system. In this report, the role of CD5/splenocytes and B-1 and B-2 murine cells in the formation of specific Ab and nIg under the influence of TI-2 was studied by two different approaches. In the first series of experiments, it was established that the depletion of immune splenocytes for CD5/ cells strongly reduced the increase in AFC
Table 1 AFC, IFC and nIFC formation in suspension of mouse splenocytes treated by anti-CD5 antibodies and complement (per 106 of viable cells)a Splenocytes
AFC
IFC
nIFC
AFC increase
nIFC increase
Normal Normal treated by C Normal treated by C and anti-CD5 Ab Immune Immune treated by C Immune treated by C and anti-CD5 Ab
2369/92 (11) 1899/79 (9) 2779/83 (8) 7359/134 (8) 6259/143 (5) 5649/116 (8)
22459/592 22339/577 22449/468 35619/460 36759/745 29459/594
22339/560 20969/526 19669/434 28269/450 30509/671 23819/635
4999/130 4569/106 2869/87
5939/204 9549/425 4159/166
a
The number of experiments with pooled splenocytes from 2 to 3 mice were given in parentheses.
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
40
Fig. 1. FACS sorting of B-1 and B-2 cells. SSC */side scatter; FSC */forward scatter.
and nIFC number (Table 1). It meant that CD5/ cells played an important role in specific and non-specific immune response to this Ag. (It must be pointed that the use of increasing anti-m serum in ELISPOT allows to determine the numbers of AFC and IFC with the same sensitivity.)
Unexpectedly, the treatment by anti-CD5 Ab and C of normal spleen cells did not result in any decrease in AFC or IFC/nIFC numbers (Table 1). It may depend either on the production of background anti-PVP and nIg by CD5/ B-2 cells or by B-1b cells. But in any case, it points to the participation in immune response to PVP
Table 2 The number of AFC, IFC and nIFC in B-1 and B-2 subsets of BALB/c mice immunized by PVP B subsets
B-1 B-2
AFC
IFC
Normal
Immune
Normal
Immune
9399/317 629/26
30179/1288 979/65
25129/963 2629/174
104499/2168 6059/375
* nIFCi, AFCi, nIFCi */IFC, AFC and nIFC induced by PVP, respectively.
nIFC
AFCi/AFC*
nIFCi/IFC*
42549/1777 4239/339
2.2 0.6
2.7 1.6
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
of B cells distinctive from ‘‘normal’’ background AFC/ IFC. Most of CD5/ splenocytes belong to T-lymphocytes. It means that the decrease in the AFC and IFC number after anti-CD5 Ab and C treatment might depend not on the depletion for CD5/ B-cells but on the depletion for T-cells (indirectly). To determine the role of B-1 and B-2 cells in AFC and nIFC formation under the influence of TI-2 splenocytes more directly in the second series of experiments, B-1 and B-2 subsets separated by FACS were used. In these experiments, it was proved that CD5/ B-1 cells were the main producers of the anti-PVP Ab and PVP-induced nIg. The numbers of AFC and nIFC in this subset exceeded those in B-2 subset about 10 times (Table 2). It is necessary to take into account, in addition, that the detection of small numbers of AFC and nIFC in B-2 subset may depend on the presence of CD5/ B-1b cells in this subpopulation, since FACS method used for cell separation did not allow to discriminate between B-1a and B-1b lymphocytes. Thus, a key role of CD5/ B-1 cells in the formation of AFC and nIFC under the influence of TI-2 was demonstrated by two different approaches. Our data are in agreement with the data of Whitmore et al. [8] who have found AFC to PVP in CD5/ B compartment, but differ from that of Smith et al. [10] who detected AFC to TNP-Ficoll in CD5/ B-cell fraction. It may be due either to the differences of TI2 used, or to the formation of anti-TNP-Ficoll Ab by CD5/ B-1b cells. It must be noted that CD5 expression per se is not a marker for B-1 cells in a functional sense. Therefore, our experiments did not rule out the formal possibility that CD5/ B-cells acquired CD5 expression after activation by TI-2 as it was suggested earlier [12,13]. Thus, although the possibility of change of B-2 phenotype to that of B-1 in our experiments seems to be extremely low it cannot, nevertheless, be totally excluded. To prove that TI-2 Ags induce the immune response in really preexisting CD5/ B-1 cells experiments with in vitro stimulation by TI-2 of CD5/ B-cells isolated from unimmunized mice must be done. Our study provides new insight into the functions of B-1 and B-2 cell subsets. Taken together they provide good evidence that: (1) PVP-non-self TI-2-induces marked immune response; (2) the majority of anti-PVP Ab are produced by CD5/ B-1 cells and (3) CD5/ B-1 cells are the main producers not only of Ab, but also of the Ag-induced nIg. It means that PVP may be considered as polyclonal B-1 cell activator. Earlier polyclonal activation of total B-cells by very high doses TI-2 was shown by Coutinho and Moller [17]. Polyclonal activation of B-1 cells by small immunizing doses of TI-2 is demonstrated for the first time. It was shown previously that simultaneous injection of TD and TI-2 resulted in the summation of total nIFC
41
number as compared with the quantities of these cells after injection TD or TI-2 alone, and it was suggested that B-cells responding to TD and TI-2 belonged to different B subsets [18]. Present experiments pointing to the formation of PVP-induced nIg mainly by B-1 cells confirm this suggestion. In our earlier experiments with TD and TI-2, the close parallelism between Ab and nIg formation was noted [1]. It concerned not only the dynamic and magnitude of specific and polyclonal immune responses, but also the Lyb5-phenotype of responding B-cells [19]. Similar correlations were found now in the experiments with CD5/ and B-1 cells. Immunization by PVP increased the numbers of AFC and nIFC in B-1 subset for 2 /3 times. The depletion of splenocytes for CD5/ cells reduced the increase in the number of cells producing Ab to PVP and the number of newly formed nIFC for /40%. Thus, the increase/decrease in the numbers of AFC and Ag-induced nIFC was similar. Mechanism responsible for such parallelism between AFC and Ag-induced nIFC formation remains mystery till now. The formation of nIg during immunization was described many years ago [20] and confirmed frequently. The capacity of TI-2 to induce nIg formation in B-1 cells representing the first line of defence points to that nIg formation is a general phenomenon and allows to suggest some important yet unknown functions of these proteins. In connection with the results obtained another very interesting question arises. It was shown [21] that the appearance of Ab to TI-2 depended on the presence of the most mature B-cell subpopulation, Lyb5/ B-cells. Later we have found that not only AFC, but also nIFC arising under the influence of TI-2 belong to Lyb5/ Bcells [19]. It is in accordance with the data about the late appearance of immune response to TI-2 in ontogenesis [22 /24]. However, according to modern point of view, namely B-1 cells selected by self Ags cross-reacting with bacterial poly- and lipopolysaccharides, are responsible for the defence against microbial infections. Thus, impaired responsiveness to TI-2 in neonates contradicts to the capacity of B-1 cells to respond to TI-2. Perhaps, this apparent paradox may be partially resolved if, according to recent data [25,26], the failure to respond to TI-2 in neonates depends not on the unresponsiveness of B-1 cells as such, but on some defect in accessory cells. This rises the questions about the mechanisms of B-1 cell activation and possible role in these processes of B, NK or macrophage/dendritic cells and their factors. Evidently, the contradiction between the capability of B1 cells to respond to TI-2 and the impaired responsiveness to these Ags in neonates remains open and waits its resolution.
42
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
Acknowledgements We are very grateful to Leonore and Leonard Herzenberg for the opportunity to perform experiments with B-1 and B-2 cells and for very interest and helpful discussion. Kind help of Ometa Herman in experiments is greatly appreciated. This research was supported (in part) by grant of Russian Foundation for Basic Researches.
References [1] E.V. Sidorova, Uspekhi Sovr. Biol. 113 (1993) 675 /701 (in Russian). [2] K. Hayakawa, R. Hardy, M. Honda, L. Herzenberg, A. Steinberg, L. Herzenberg, Proc. Nat. Acad. Sci. USA 81 (1984) 2494 / 2498. [3] A. Stall, S. Adams, L. Herzenberg, A.B. Kantor, Ann. NY Acad. Sci. 651 (1992) 33 /43. [4] A. Hayakawa, M. Asano, S. Shinton, M. Gui, D. Allman, C. Stewart, J. Silver, R. Hardy, Science 285 (1999) 113 /116. [5] H. Wortis, M. Teutsch, M. Higer, J. Zheng, D. Parker, Proc. Nat. Acad. Sci. USA 92 (1995) 3348 /3352. [6] G. Haughton, L. Arnold, A. Whitmore, S. Clarke, Immunol. Today 14 (1993) 84 /87. [7] L. Arnold, S. McCray, C. Tatu, S. Clarke, J. Immunol. 164 (2000) 2924 /2930. [8] A. Whitmore, G. Haughton, L. Arnold, Int. Immunol. 8 (1996) 533 /542.
[9] H. Masmoudi, T. Mota-Santos, F. Huetz, A. Coutinho, P. Cazenave, Int. Immunol. 2 (1990) 515 /520. [10] H. Smith, L. Yaffe, T. Chused, E. Raveche, D. Klinman, A. Steinberg, Cell. Immunol. 92 (1985) 190 /196. [11] H. Ohdan, K. Swenson, G. Kruger, Y. Yang, Y. Xu, A. Thall, M. Sykes, J. Immunol. 165 (2000) 5518 /5529. [12] Y. Cong, E. Rabin, H. Wortis, Int. Immunol. 3 (1991) 467 /476. [13] P. Youinou, Ch. Jamin, P. Lydyard, Immunol. Today 20 (1999) 312 /316. [14] M. Roederer, A. Kantor, D. Parks, L. Herzenberg, Cytometry 24 (1996) 191 /197. [15] M. Bigos, N. Baumgarth, G. Jager, O. Herman, T. Nozaki, R. Stovel, D. Parks, L.A. Herzenberg, Cytometry 36 (1999) 36 /45. [16] N. Logunova, M. Agadjanyan, E. Sidorova, Immunology 2 (1989) 77 /79 (in Russian). [17] A. Coutinho, G. Moller, Eur. J. Immunol. 3 (1973) 608 /613. [18] M. Agadjanyan, T. Megrabyan, E. Sidorova, Bull. Exp. Biol. Med. 8 (1981) 66 /68 (in Russian). [19] I. Chernishova, T. Kondratieva, E. Sidorova, Immunol. Lett. 62 (1998) 15 /18. [20] W. Boyd, H. Bernard, J. Immunol. 33 (1937) 111 /122. [21] A. Ahmed, S. Scher, S. Sharrow, A. Smith, D. Paul, J. Exp. Med. 145 (1977) 101 /111. [22] W. Timens, A. Boes, T. Rozeboom-Uiterwijk, S. Poppema, J. Immunol. 143 (1989) 32 /3206. [23] S. Smyk, H. Kohler, J. Immunol. 134 (1985) 2254 /2259. [24] R. Chelvarajan, N. Gilbert, S. Bondada, J. Immunol. 161 (1998) 3315 /3324. [25] J. Mond, Q. Vos, A. Lees, C. Snapper, Curr. Opin. Immunol. 7 (1995) 349 /354. [26] S. Bondada, H. Wu, D. Robertson, R. Chelvarajan, Vaccine 19 (2000) 557 /565.
Role of different B-cell subsets in the specific and polyclonal immune response to T-independent antigens type 2 E.V. Sidorova a,*, Lu Li-Sheng b, B. Devlin b, I. Chernishova a, M. Gavrilova a a
Institute for Viral Preparations, Russian Academy of Medical Sciences, 1st Dubrovskaya Street 15, 115088 Moscow, Russia b Herzenberg Laboratory, Beckman Center B-007, Stanford, CA, USA Received 18 February 2003; accepted 25 February 2003
Abstract Role of different B-cell subsets in the immune response to T-independent antigen type 2 (TI-2) was studied. BALB/c and C57BL/6 mice were immunized by polyvinylpyrrolidon (PVP), and the numbers of antibody- and Ig-forming cells (AFC and IFC, respectively) were determined by ELISPOT method. The number of cells producing non-specific Ig (nIFC) was calculated as the difference between the number of IFC and AFC; the number of nIFC induced by PVP was calculated as the difference between the number of nIFC in immune and control splenocytes. Immunization by PVP induced not only the AFC appearance, but also the increase in the number of the antigen-induced nIFC. The treatment of splenocytes by anti-CD5 antibodies and guinea pig complement reduced the increase in the numbers of newly formed AFC and nIFC to /40% of control level. It means that CD5/ cells play an important role not only in the specific, but also in polyclonal immune response to non-self TI-2. To be sure that the decrease of AFC and nIFC numbers is due to depletion for CD5/ B-cells, but not CD5/ T-cells, splenocytes were separated to B-1 and B-2 subsets, and the numbers of AFC, IFC and nIFC were determined in each B-cell subpopulation separately. The overwhelming majority of newly formed AFC and nIFC was detected in B-1 subset. The numbers of AFC and nIFC in B-1 compartment was /10-fold greater than in B-2 cells. A close parallelism between AFC and nIFC formation was observed. It is concluded that specific and polyclonal immune response to non-self TI-2 */PVP */depends mainly on CD5/ B-1 subset. # 2003 Elsevier Science B.V. All rights reserved. Keywords: B-cells; Antibody; Immunoglobulins; T-independent antigens; CD5
1. Introduction The injection of the antigen (Ag) induces not only the biosynthesis of antibodies (Ab) and the appearance of the Ab-forming cells (AFC), but also greatly increases non-specific Ig (nIg) formation and the number of nIgforming cells (nIFC) [1]. The increase in nIFC number under the influence of T-dependent Ag (TD) is due mainly to the action of stimulatory T factors. The increase in nIFC number under the influence of Tindependent type 1 Ags (TI-1) depends on their mitogenic activity. The increase in the nIFC number under the influence of T-independent type 2 Ags (TI-2) lacking mitogenic activity is unexplained till now, and the role of
* Corresponding author. Tel.: /7-095-274-5501; fax: /7-095-2745710. E-mail address: [email protected] (E.V. Sidorova).
different B-cell subpopulations and the mechanisms responsible for this increase are not yet defined. B-cells are subdivided into CD5/ (B-1a), CD5/ (B1b) and CD5/ B-2 subsets [2,3]. It is hypothesized that B-1 cells can be positively selected by TI-2 [4 /7], and AFC arising in response to some TI-2 (TNP-Ficoll, Streptococcus pneumoniae , PVP) were shown to belong to B-1a cells [8,9]. At the same time, Ab to TI-2 could be produced by CD5/ B-cells [10,11], and in addition, CD5/ B-2 cells may become CD5/ B-cells after activation [12,13]. Thus, experiments have generated contradictory results, and the role of different B-cell subsets in the formation of specific Abs in response to TI-2 is not fully elucidated till now. Still less is known about the role of different B-cell subsets in nIg formation induced by TI-2 and about the mechanisms responsible for polyclonal B-cell activation in these cases. As a matter of fact, the cellular mechanisms of
0165-2478/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0165-2478(03)00055-5
38
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
B-cell activation by TI-2 as such are unclear. This research was done to study the role of CD5/ and CD5/ B-cells in the formation of Ab and nIg in response to non-self TI-2 */polyvinylpyrrolidon (PVP). Two different approaches were used. In the first one, splenocytes of mice immunized by PVP were depleted for CD5/ cells by the treatment with anti-CD5 Ab and complement (C), and the numbers of AFC and Igforming cells (IFC) in splenocyte suspensions were compared with those of control splenocytes. In the second approach, splenocytes from control and immunized mice were separated into CD5/ B-1 and CD5/ B-2 subsets, and the numbers of AFC, IFC and nIFC were determined in each of them.
2. Materials and methods 2.1. Animals, antigens and immunization BALB/c and C57Bl/6 female mice were obtained from Animal Facilities ‘‘Stolbovaya’’ (Russia) and Stanford University (USA). Polyvinylpyrrolidon 350 kDa (PVP, Serva) was used as TI-2. Mice were injected intravenously with 2 mg PVP in 0.2 ml of PBS. On the third or fourth day after immunization, spleens were taken, mononuclear suspensions were prepared and total splenocytes or CD5/ B-1 and CD5/ B-2 cells were used for the determination of AFC and IFC by ELISPOT method. Unmanipulated mice of the same age were used as control. Mice were killed either by cervical dislocation or by CO2 asphyxiation. 2.2. Depletion for CD5/ cells Single-cell suspensions (2 /106 cells/ml) were treated with C (pooled guinea pig serum, final dilution 1:10) and anti-CD5 Ab (Cederlain, Canada) according to manufacturer’s instruction (1 h at /4 8C and after addition of C 1 h at 37 8C). After incubation, cells were centrifuged (7 min, 1500 rpm), suspended in the RPMI 1640 medium with fetal calf serum (FCS) and antibiotics, and the number of viable cells was determined either with Trypan blue or with ethidium bromide/acridyl orange. 2.3. Flow cytometry mAb used here included Cascade Blue-labeled antiCD3e (2C11), anti-CD4 (GK1.5), anti-F4/80 and antiGr-1 (RB8-8C5), FITC-anti-CD21 (7G6), PE-antiCD23 (B3B4), Cy7-PE-anti-IgD (1126), Texas Red anti-IgM (331) and biotin anti-CD5 (53-7.8). mAb and fluorochrome conjugates were either prepared at Stanford [14,15] or obtained from Pharmingen (San Diego, CA). Streptavidin-Cy5PE from BD Pharmingen was
used as second-step reagent. Data were collected on HiD FACS (Cytomation/Becton Dickinson hybrid FACS) developed in FACS Facility at Stanford FACS/DESK as described [15]. Finally, data cataloged and stored with FACS/DESK (Stanford) were transferred to Flowjo (Tree Star, San Carlos, CA) for fluorescence compensation and analysis. 2.4. Cell sorting Single-cell suspensions from spleens were stained on ice with a combination of the anti-murine Abs in staining medium (biotin, flavin-deficient RPMI 1640 medium supplemented with 5% FCS and 0.02% sodium azide). Cells were washed with staining medium and incubated with Streptavidine-Cy5PE as second-step reagent for 15 min. Propidium iodide (PI) was added immediately before sorting at a final concentration of 0.25 mg/ml to discriminate viable cells. Cell sorting was performed by using a highly modified triple laser Hi-D FACS described in Refs. [14,15]. B-1 cells were identified as IgMhigh, IgDlow, CD23/, CD5/, and B-2 cells as IgMhigh, IgDhigh, CD23/, CD5/. Sorted cells were re-analysed after sorting and purities were /98%. 2.5. ELISPOT To determine the AFC and IFC numbers according to Ref. [16], nitrocellulose plates (MAHA 4550, Millipore Corp., Bedford, MA) were used. To determine the AFC number, nitrocellulose was treated by PVP (10 mg/well, 16 h at /4 8C, or 2 h at 37 8C), blocked with 1% BSA (Serva) (1.5 h at room temperature) and washed with PBS and RPMI 1640 medium. To determine IFC number, nitrocellulose coated by goat anti-mouse Ig (ICN) was used. To detect AFC 100/200/103 of total spleen cells per well, 40/80/103 of B-2 cells per well and 1.5 /2.0 /103 of B-1 cells per well suspended in RPMI 1640 medium with 3/5% of FCS were used in triplicate. To detect IFC 10 /20 /103 of total splenocytes or B-2 spleen cells and 1500 /2000/103 of B-1 cells per well were added. Cells were cultivated 18 /20 h at 37 8C in 5% of CO2. Cultivation being over, cells were washed out, and nitrocellulose wells were treated at first by rabbit antiserum specific to mouse IgM (anti-m serum) and then by horseradish peroxidase conjugated with sheep Ab to rabbit IgG. Membranes were dried in the dark, and the blue spots in each well were counted under microscope. The number of nIFC was calculated as the difference between the number of IFC and AFC; the number of Ag-induced nIFC was calculated as the difference in nIFC numbers in immune and normal splenocytes. All values for AFC, IFC and nIFC were expressed as the arithmetic mean9/S.D. per 106 of
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
pooled viable splenocytes or splenocytes of individual mice assayed separately.
3. Results In the first series of experiment, BALB/c mice were intravenously immunized by PVP (2 mg/mice), spleens were removed on the fourth day, and mononuclear suspensions were prepared. Splenocytes were treated with C or with anti-CD5 Ab plus C, and the numbers of AFC and IFC were determined by ELISPOT method according to Ref. [16]. The data obtained are shown in Table 1. It may be seen that PVP induces not only specific immune response, but also the increase in the number of cells producing nIg (nIFC), i.e. behaves as polyclonal B-cell activator. The treatment of splenocytes by anti-CD5 Ab and C markedly reduced the increase of AFC and nIFC in immune suspensions. The increase of AFC number diminished from /456/499 to /286 per 106 of viable splenocytes, and the increase of nIFC number from /593 /954 to /415 per 106 cells, i.e. for /40 and /36%, respectively. It means that during immune response to TI-2 CD5/ cells play role not only in AFC but also in nIFC formation as well. It must be noted that the treatment by anti-CD5 Ab of normal spleen cells practically did not influence the numbers of background AFC and nIFC. To determine if AFC and Ag-induced nIFC belong to B-1 subset in the next series of experiment, splenocytes from normal and immunized BALB/c and C57Bl/6 mice on the third day after immunization by PVP were separated into CD5/ B-1 and CD5/ B-2 subsets as described in Section 2. The FACS sorting of B-1 and B-2 cells from the spleen of BALB/c mice is shown in Fig. 1. It may be seen that the output of living cells was / 90.7% of input, the number of lymphocytes /70.4%, the number of total B-cells /24.6% and the numbers of B-1 and B-2 cells 14.8 and 68.5%, i.e. 0.45 and 10.6% of total spleen cells, respectively. The results obtained in experiments with 6 /8 individual BALB/c mice are represented in Table 2. It may be seen that immunization with PVP resulted in the increase in the number of cells producing Ab to PVP
39
and in a great increase in the number of IFC. The overwhelming majority of AFC and IFC were produced by CD5/ B-1 cells: 30179/1288/106 B-1 cells against 9399/317/106 cells and 104499/2168/106 B-1 cells against 25129/963/106 cells in control sample (respectively). The increase of PVP-induced nIFC number was equal to 42549/1777/106 B-1 cells. In immune B-2 subset, the numbers of AFC and IFC were also greater than in normal B-2 cells but nevertheless about 10 times less than in B-1 subset. It means that in both B-cell subsets PVP induces not only Ab formation but the formation of nIg as well. The relative increase of AFC and nIFC numbers in immune B-1 splenocytes was almost the same: 2.2 and 2.7, respectively. In B-2 subset the increase in nIFC number was much greater than in AFC content. However, these data were not beyond for doubts since the numbers of AFC and IFC in the wells could not be precisely estimated. Similar results were obtained in the experiments with C57B/6 mice, although immune response to PVP in C57Bl/6 mice was much less as compared to BALB/c mice. Thus, the number of AFC in control and immune C57Bl/6 B-1 subsets were 4749/190/106 cells and 7299/ 147/106 cells against 9399/317/106 cells and 30179/1288/ 106 cells in BALB/c B-1 cells, respectively, and the number of IFC in control and immune splenocytes were 18409/410/106 cells and 30999/699/106 cells against 25129/963/106 cells and 104499/2168/106 cells in BALB/c mice, respectively. But nevertheless, all values for AFC and IFC in B-1 subset of C57Bl/6 mice were again about 10-fold higher than in B-2 subset.
4. Discussion The phenotypic and functional relationships among various B-cell subset are of importance for better understanding of the immune system. In this report, the role of CD5/splenocytes and B-1 and B-2 murine cells in the formation of specific Ab and nIg under the influence of TI-2 was studied by two different approaches. In the first series of experiments, it was established that the depletion of immune splenocytes for CD5/ cells strongly reduced the increase in AFC
Table 1 AFC, IFC and nIFC formation in suspension of mouse splenocytes treated by anti-CD5 antibodies and complement (per 106 of viable cells)a Splenocytes
AFC
IFC
nIFC
AFC increase
nIFC increase
Normal Normal treated by C Normal treated by C and anti-CD5 Ab Immune Immune treated by C Immune treated by C and anti-CD5 Ab
2369/92 (11) 1899/79 (9) 2779/83 (8) 7359/134 (8) 6259/143 (5) 5649/116 (8)
22459/592 22339/577 22449/468 35619/460 36759/745 29459/594
22339/560 20969/526 19669/434 28269/450 30509/671 23819/635
4999/130 4569/106 2869/87
5939/204 9549/425 4159/166
a
The number of experiments with pooled splenocytes from 2 to 3 mice were given in parentheses.
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
40
Fig. 1. FACS sorting of B-1 and B-2 cells. SSC */side scatter; FSC */forward scatter.
and nIFC number (Table 1). It meant that CD5/ cells played an important role in specific and non-specific immune response to this Ag. (It must be pointed that the use of increasing anti-m serum in ELISPOT allows to determine the numbers of AFC and IFC with the same sensitivity.)
Unexpectedly, the treatment by anti-CD5 Ab and C of normal spleen cells did not result in any decrease in AFC or IFC/nIFC numbers (Table 1). It may depend either on the production of background anti-PVP and nIg by CD5/ B-2 cells or by B-1b cells. But in any case, it points to the participation in immune response to PVP
Table 2 The number of AFC, IFC and nIFC in B-1 and B-2 subsets of BALB/c mice immunized by PVP B subsets
B-1 B-2
AFC
IFC
Normal
Immune
Normal
Immune
9399/317 629/26
30179/1288 979/65
25129/963 2629/174
104499/2168 6059/375
* nIFCi, AFCi, nIFCi */IFC, AFC and nIFC induced by PVP, respectively.
nIFC
AFCi/AFC*
nIFCi/IFC*
42549/1777 4239/339
2.2 0.6
2.7 1.6
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
of B cells distinctive from ‘‘normal’’ background AFC/ IFC. Most of CD5/ splenocytes belong to T-lymphocytes. It means that the decrease in the AFC and IFC number after anti-CD5 Ab and C treatment might depend not on the depletion for CD5/ B-cells but on the depletion for T-cells (indirectly). To determine the role of B-1 and B-2 cells in AFC and nIFC formation under the influence of TI-2 splenocytes more directly in the second series of experiments, B-1 and B-2 subsets separated by FACS were used. In these experiments, it was proved that CD5/ B-1 cells were the main producers of the anti-PVP Ab and PVP-induced nIg. The numbers of AFC and nIFC in this subset exceeded those in B-2 subset about 10 times (Table 2). It is necessary to take into account, in addition, that the detection of small numbers of AFC and nIFC in B-2 subset may depend on the presence of CD5/ B-1b cells in this subpopulation, since FACS method used for cell separation did not allow to discriminate between B-1a and B-1b lymphocytes. Thus, a key role of CD5/ B-1 cells in the formation of AFC and nIFC under the influence of TI-2 was demonstrated by two different approaches. Our data are in agreement with the data of Whitmore et al. [8] who have found AFC to PVP in CD5/ B compartment, but differ from that of Smith et al. [10] who detected AFC to TNP-Ficoll in CD5/ B-cell fraction. It may be due either to the differences of TI2 used, or to the formation of anti-TNP-Ficoll Ab by CD5/ B-1b cells. It must be noted that CD5 expression per se is not a marker for B-1 cells in a functional sense. Therefore, our experiments did not rule out the formal possibility that CD5/ B-cells acquired CD5 expression after activation by TI-2 as it was suggested earlier [12,13]. Thus, although the possibility of change of B-2 phenotype to that of B-1 in our experiments seems to be extremely low it cannot, nevertheless, be totally excluded. To prove that TI-2 Ags induce the immune response in really preexisting CD5/ B-1 cells experiments with in vitro stimulation by TI-2 of CD5/ B-cells isolated from unimmunized mice must be done. Our study provides new insight into the functions of B-1 and B-2 cell subsets. Taken together they provide good evidence that: (1) PVP-non-self TI-2-induces marked immune response; (2) the majority of anti-PVP Ab are produced by CD5/ B-1 cells and (3) CD5/ B-1 cells are the main producers not only of Ab, but also of the Ag-induced nIg. It means that PVP may be considered as polyclonal B-1 cell activator. Earlier polyclonal activation of total B-cells by very high doses TI-2 was shown by Coutinho and Moller [17]. Polyclonal activation of B-1 cells by small immunizing doses of TI-2 is demonstrated for the first time. It was shown previously that simultaneous injection of TD and TI-2 resulted in the summation of total nIFC
41
number as compared with the quantities of these cells after injection TD or TI-2 alone, and it was suggested that B-cells responding to TD and TI-2 belonged to different B subsets [18]. Present experiments pointing to the formation of PVP-induced nIg mainly by B-1 cells confirm this suggestion. In our earlier experiments with TD and TI-2, the close parallelism between Ab and nIg formation was noted [1]. It concerned not only the dynamic and magnitude of specific and polyclonal immune responses, but also the Lyb5-phenotype of responding B-cells [19]. Similar correlations were found now in the experiments with CD5/ and B-1 cells. Immunization by PVP increased the numbers of AFC and nIFC in B-1 subset for 2 /3 times. The depletion of splenocytes for CD5/ cells reduced the increase in the number of cells producing Ab to PVP and the number of newly formed nIFC for /40%. Thus, the increase/decrease in the numbers of AFC and Ag-induced nIFC was similar. Mechanism responsible for such parallelism between AFC and Ag-induced nIFC formation remains mystery till now. The formation of nIg during immunization was described many years ago [20] and confirmed frequently. The capacity of TI-2 to induce nIg formation in B-1 cells representing the first line of defence points to that nIg formation is a general phenomenon and allows to suggest some important yet unknown functions of these proteins. In connection with the results obtained another very interesting question arises. It was shown [21] that the appearance of Ab to TI-2 depended on the presence of the most mature B-cell subpopulation, Lyb5/ B-cells. Later we have found that not only AFC, but also nIFC arising under the influence of TI-2 belong to Lyb5/ Bcells [19]. It is in accordance with the data about the late appearance of immune response to TI-2 in ontogenesis [22 /24]. However, according to modern point of view, namely B-1 cells selected by self Ags cross-reacting with bacterial poly- and lipopolysaccharides, are responsible for the defence against microbial infections. Thus, impaired responsiveness to TI-2 in neonates contradicts to the capacity of B-1 cells to respond to TI-2. Perhaps, this apparent paradox may be partially resolved if, according to recent data [25,26], the failure to respond to TI-2 in neonates depends not on the unresponsiveness of B-1 cells as such, but on some defect in accessory cells. This rises the questions about the mechanisms of B-1 cell activation and possible role in these processes of B, NK or macrophage/dendritic cells and their factors. Evidently, the contradiction between the capability of B1 cells to respond to TI-2 and the impaired responsiveness to these Ags in neonates remains open and waits its resolution.
42
E.V. Sidorova et al. / Immunology Letters 88 (2003) 37 /42
Acknowledgements We are very grateful to Leonore and Leonard Herzenberg for the opportunity to perform experiments with B-1 and B-2 cells and for very interest and helpful discussion. Kind help of Ometa Herman in experiments is greatly appreciated. This research was supported (in part) by grant of Russian Foundation for Basic Researches.
References [1] E.V. Sidorova, Uspekhi Sovr. Biol. 113 (1993) 675 /701 (in Russian). [2] K. Hayakawa, R. Hardy, M. Honda, L. Herzenberg, A. Steinberg, L. Herzenberg, Proc. Nat. Acad. Sci. USA 81 (1984) 2494 / 2498. [3] A. Stall, S. Adams, L. Herzenberg, A.B. Kantor, Ann. NY Acad. Sci. 651 (1992) 33 /43. [4] A. Hayakawa, M. Asano, S. Shinton, M. Gui, D. Allman, C. Stewart, J. Silver, R. Hardy, Science 285 (1999) 113 /116. [5] H. Wortis, M. Teutsch, M. Higer, J. Zheng, D. Parker, Proc. Nat. Acad. Sci. USA 92 (1995) 3348 /3352. [6] G. Haughton, L. Arnold, A. Whitmore, S. Clarke, Immunol. Today 14 (1993) 84 /87. [7] L. Arnold, S. McCray, C. Tatu, S. Clarke, J. Immunol. 164 (2000) 2924 /2930. [8] A. Whitmore, G. Haughton, L. Arnold, Int. Immunol. 8 (1996) 533 /542.
[9] H. Masmoudi, T. Mota-Santos, F. Huetz, A. Coutinho, P. Cazenave, Int. Immunol. 2 (1990) 515 /520. [10] H. Smith, L. Yaffe, T. Chused, E. Raveche, D. Klinman, A. Steinberg, Cell. Immunol. 92 (1985) 190 /196. [11] H. Ohdan, K. Swenson, G. Kruger, Y. Yang, Y. Xu, A. Thall, M. Sykes, J. Immunol. 165 (2000) 5518 /5529. [12] Y. Cong, E. Rabin, H. Wortis, Int. Immunol. 3 (1991) 467 /476. [13] P. Youinou, Ch. Jamin, P. Lydyard, Immunol. Today 20 (1999) 312 /316. [14] M. Roederer, A. Kantor, D. Parks, L. Herzenberg, Cytometry 24 (1996) 191 /197. [15] M. Bigos, N. Baumgarth, G. Jager, O. Herman, T. Nozaki, R. Stovel, D. Parks, L.A. Herzenberg, Cytometry 36 (1999) 36 /45. [16] N. Logunova, M. Agadjanyan, E. Sidorova, Immunology 2 (1989) 77 /79 (in Russian). [17] A. Coutinho, G. Moller, Eur. J. Immunol. 3 (1973) 608 /613. [18] M. Agadjanyan, T. Megrabyan, E. Sidorova, Bull. Exp. Biol. Med. 8 (1981) 66 /68 (in Russian). [19] I. Chernishova, T. Kondratieva, E. Sidorova, Immunol. Lett. 62 (1998) 15 /18. [20] W. Boyd, H. Bernard, J. Immunol. 33 (1937) 111 /122. [21] A. Ahmed, S. Scher, S. Sharrow, A. Smith, D. Paul, J. Exp. Med. 145 (1977) 101 /111. [22] W. Timens, A. Boes, T. Rozeboom-Uiterwijk, S. Poppema, J. Immunol. 143 (1989) 32 /3206. [23] S. Smyk, H. Kohler, J. Immunol. 134 (1985) 2254 /2259. [24] R. Chelvarajan, N. Gilbert, S. Bondada, J. Immunol. 161 (1998) 3315 /3324. [25] J. Mond, Q. Vos, A. Lees, C. Snapper, Curr. Opin. Immunol. 7 (1995) 349 /354. [26] S. Bondada, H. Wu, D. Robertson, R. Chelvarajan, Vaccine 19 (2000) 557 /565.