Rabbit lymphocyte subpopulations. II. Separation and characterization of two Ig+ cell subpopulations.

Rabbit lymph node cells (Ig⁺Ig^?) were separated into Ig⁺ and Ig^? populations by rosette formation with anti-Ig antibody-coated erythrocytes and centrifugation on Ficoll-Hypaque. Subpopulations of Ig⁺ cells were obtained by treating rosetted cells with autologous serum which dissociated approximately half of the rosettes. The stable rosetted cells (Ig⁺ S) were separated from the labile unrosetted cells (Ig⁺L) by centrifugation on Ficoll-Hypaque. The Ig⁺S population contained most of the Ig-secreting cells and responded poorly to mitogens. The Ig⁺L population contained few Ig-secreting cells and responded well to mitogens. Approximately 50% of Ig⁺L cells became Ig⁺S when cultured with Ig^? cells but this transition did not occur if Ig⁺S cells were added to the culture at the start of the incubation period. Purified Ig⁺ L cells lost their ability to form rosettes when cultured by themselves but retained their ability to form rosettes when cultured wih Ig^? cells. The data indicate that the Ig⁺S and Ig⁺L populations are at different stages in the differentiation of Ig⁺ cells (B cells) and that the Ig⁺L cells are subject to the regulatory influences of both Ig^? and Ig⁺S cells.     

Rabbit lymphocyte subpopulations. I. Separation of Ig+ and Ig- cells and their interaction in cultures stimulated by mitogens.

Rabbit lymph node cells (Ig⁺Ig^?) were rosetted with anti-Ig antibody-coated erythrocytes and the rosetted Ig⁺ cells (B cells) were separated from unrosetted Ig^? cells (T cells) by centrifugation through Ficoll-Hypaque medium. The Ig^? cells were recovered from the top and the Ig⁺ cells from the bottom of the Ficoll-Hypaque layer. Some of the purified Ig⁺ cells lost their ability to form rosettes when cultured with the mitogen associated with streptolysin O. This suggested that the Ig⁺ population might contain two distinct subpopulations. The response of Ig⁺Ig^?, Ig⁺, and Ig^? cells to various mitogens was studied. The Ig^? cells incorporated more ³H-TdR when they were incubated by themselves than when they were cultured with Ig⁺ cells in an Ig⁺Ig^? culture. On the other hand, the Ig⁺ cells incorporated less ³H-TdR when they were incubated by themselves than when they were incubated with Ig^? cells in an Ig⁺Ig^? culture. Thus, Ig⁺ cells suppressed the response of Ig^? cells whereas Ig^? cells enhanced the response of Ig⁺ cells. We conclude that rabbit Ig⁺ cells (B cells) and Ig^? cells (T cells) interact with a feedback pattern of regulation.     

Enumeration and isolation of rabbit T and B lymphocytes by using antibody-coated erythrocytes.

Rosette formation with antibody-coated erythrocytes (Ab-E) was employed for the enumeration and isolation of rabbit B cells (Ig+T-) and T cells (Ig-T+). The cells bearing surface Ig (Ig+ cells) were enumerated by a direct immunocytoadhesion technique utilizing anti-rabbit IgG antibody-coated erythrocytes (Ab-E). To enumerate cells bearing thymus cell antigen (T+ cells), an indirect rosette technique was used in which lymphocytes were first sensitized with guinea pig anti-rabbit thymus cell antiserum and then rosetted with anti-guinea pig IgG Ab-E. To demonstrate the specificity of the anti-thymus cell antiserum, a 51Cr radioimmunoassay for counting rosettes was employed along with visual counting to enumerate Ig+ and T+ cells in lymph node cell populations. When Ig+ and T+ lymph node cells were rosetted simultaneously with sheep and human erythrocytes, no mixed rosettes (less than 1%) were observed. Ficoll-Hypaque gradient centrifugation was used to obtain purified Ig+T- and Ig-T+ cells by removing rosetted T+ and Ig+ cells, respectively. The purity of isolated Ig-T+ cells was indicated by 94 to 95% indirect rosetting with anti-thymus cell antiserum and by 0 to 3% direct rosetting with anti-rabbit IgG Ab-E. The purity of isolated Ig+T- cells was indicated by 90 t0 94% direct rosetting with anti-rabbit IgG Ab-E and by 2 to 3% indirect rosetting with anti-thymus cell antiserum. The percentage of Ig+T- and Ig-T+ cells were determined in peripheral blood and in various lymphoid organs. The isolated Ig+T- and Ig-T+ cells were also characterized by their responses to mitogens. Thus, nearly pure Ig+T- and Ig-T+ cells were isolated by "negative selection," which should minimize functional changes of the cells, and thereby facilitate the study of their biologic properties, e.g., their response to mitogens.     

Interaction of peanut agglutinin,a lectin specific for nonreducing terminal d-galactosyl residues,with embryonal carcinoma cells

Peanut agglutinin (PNA), a lectin specific for terminal d-galactosyl residues, was found to react with embryonal carcinoma cells, but not with their differentiated derivatives. Receptors for PNA were detectable at the surface of all cells of the quasinullipotent F9 line and on only 50% of the multipotent PCC3/A/1 line. The fraction of the PCC3/A/1 population which expresses the F9 antigen was found to be included in the subpopulation carrying the PNA receptors. PNA⁺ and PNA^? subpopulations of PCC3/A/1 were separated by a PNA-mediated reversible agglutination of PNA⁺ cells with rabbit erythrocytes. These subpopulations were essentially F9⁺ and F9^?, respectively.     

Primary in vitro antibody response of rabbit lymphoid cells and T-B cell collaboration in the absence of detectable mitogens

To investigate whether the antibody response and T-B-cell collaboration in vitro can be obtained in the absence of mitogens, a method of obtaining an in vitro primary anti-sheep red blood cell antibody response by rabbit spleen and lymph node cells was developed. We used Marbrook culture vessels and a specially prepared medium containing 10% autologous serum and maintained at pH 7.4–7.6. The system was shown to be devoid of any polyclonal mitogens as assessed by [³H]thymidine incorporation and by direct examination for blast cells in stained smears. The primary response increased continuously over the 5-day cultivation period and only IgM but not IgG plaque-forming cells (PFC) were detected. In over 20 experiments, the response ranged from 357 ± 17 to 4425 ± 110 PFC/10⁷ cultured cells with a median stimulation index of 52. The spleen cells required less antigen than the lymph node cells and 2-mercaptoethanol inhibited the response of the spleen cells but not that of the lymph node cells. Lymphocytes were separated into highly pure T- and B-cell populations by negative selection using antibody-coated human erythrocytes to rosette either T or B cells and Ficoll-Hypaque centrifugation to remove rosetted cells. Upon cultivation, B cells alone gave a low IgM response, whereas B cells reconstituted with T cells gave a response similar to that obtained with unseparated lymphoid cells. We concluded that: (a) optimal conditions for obtaining primary in vitro antibody responses using rabbit spleen and lymph node cells were established, (b) T-B-cell collaboration was demonstrated in the rabbit primary antibody response to sheep erythrocytes, and (c) the primary antibody response in vitro and T-B-cell collaboration may occur in the absence of detectable polyclonal mitogens.     

Immunoglobulin-positive cells from the gland of harder and bone marrow of the chicken

Cells were collected from the gland of Harder (GH) and bone marrow (BM) of 14-, 21- and 32-week-old birds and were incubated with an ¹²⁵I-labeled rabbit anti-chicken Ig (IgG and IgM) serum. At 14 weeks of age the percentage of Ig⁺ small lymphocytes (SL) in the GH and BM was similar. However, by 21 weeks of age Ig⁺ SL in BM had increased to approximately 19% of total lymphocytes while the Ig⁺ SL in the GH represented less than 1.7% of the lymphocyte pool. A marked drop in the number of Ig⁺ SL in BM occurred by 32 weeks of age. These data suggest that either the BM may be dependent on the bursa for maintenance of its Ig⁺ SL or it is unable to produce in situ or maintain Ig⁺ SL with age. In the GH the predominant cell was the plasma cell (PC). Labeled PC (> 20 grains) exceeded 80% of the total PC pool in the GH. These data contrast with the apparent deficiency of Ig receptors on murine PC. The maintenance of a large number of PC in the GH without the presence of Ig⁺ SL illustrates the uniqueness of this gland.     

Ly-6.2: A new lymphocyte specificity of peripheral T-cells

A new cell-membrane alloantigen determining locus, Ly-6, has recently been described, and the single specificity Ly-6.2 has been defined by the serum (BALB/c× A)F₁ anti-CXBD. Using both fluorescence and cytotoxicity, we found this specificity predominantly on peripheral (extrathymic) T cells, as tissues react thus: thymus, 0–5 percent; spleen, 25 percent; lymph nodes, 69 percent; bone marrow, 15 percent. These reactions agree with the proportion of (Thy⁺, Ig^–) cells present in these tissues. Cortisone-resistant thymus cells were positive. Absorption studies with thymus cells demonstrated the sparse or absent representation of Ly-6.2 on intrathymic T cells. Examination of spleen and lymph node cells from T cell-depleted C57BL/6 mice (after in vitro treatment with anti-Thy-1 serum or examination of tissues of C57BL/6-nu/nu mice) also showed a depletion of Ly-6.2⁺ cells. Conversely, removal of Ig⁺ B cells, which caused a relative increase in the number of T cells in the residual population, also increased the number of Ly-6.2⁺ cells. Additive effects of anti-Thy-1.2 and anti-Ly-6.2 could not be demonstrated, which suggests that the same population was Thy-1.2⁺, Ly-6.2⁺. However, additive effects could be shown with an anti-Ia serum and anti-Ly-6.2. The Ly-6.2 specificity is not found on red cells, liver, brain, or antibody-forming cells, but has been identified on a T-cell (but not B-cell) tumor and on kidney. Ly-6.2 can therefore be considered to be a marker for peripheral T cells, and it differs from the Thy-1 and the Ly-1,2,3, and 5 specificities in its relative absence from the thymus.     

Idiotypic regulation of mouse anti-H-2 antibody responses

Mouse-immunoglobulin (MIg) tolerant rabbits immunized with mouse H-2 antibodies produced anti-idiotype antisera, which were reactive towards specific B- and T-cell receptors. One such rabbit antiserum (from rabbit 5936) defines a family of idiotypes (Id) designated 5936-idiotypes (Rubin et al. 1979). The present experiments were performed in order to establish (1) the nature of 5936-Id⁺ serum molecules, (2) the specificity of 5936-Id⁺ serum molecules, (3) the association of the 5936-Id genes to allotype and/orH-2 genes and (4) the immunological role of 5936-Id⁺ serum molecules. A sensitive, radioimmunoassay employing¹²⁵I-labelled-F(ab)₂ fragments of B6 anti-B10.BR MIg pool, 5936 antiserum, and a sheep anti-rabbit immunoglobulin antiserum, was used.—The results suggested that 5936-Id⁺ serum molecules were exclusively MIg, and that they were mainly of the IgG₁ class. Such molecules were induced in B6 mice (H-2 ^b /Ig-1 ^b ) upon immunization with H-2^k but not with H-2^q alloantigen or conventional antigens. The 5936-Id were found to be associated with Ig-1^b allotypes and theH-2 ^b complex may contain immune response (Ir) genes which, in comparison withIr genes inH-2 ^d andH-2 ^s , favor the expression of 5936-Id.—Adsorption of 5936-Id⁺ B6 anti-CBA MIg preparations on CBA (IA^k) spleen cells demonstrated that CBA antibodies were 5936-Id^?. It is dicussed whether 5936-Id⁺, IgG₁ molecules in B6 anti-CBA sera are anti-(anti-CBA) antibodies or nonspecific antibodies, the production of which is augmented by immunization with IA^k alloantigen.     

Phagocytic Activity of Three Naegleria Strains in the Presence of Erythrocytes of Various Types1

The phagocytic activities of N. lovaniensis (Aq/9/1/45D) and N. gruberi (1518/1f and 1518/1e) were studied in the presence of erythrocytes of various species: chicken, rabbit, goat, and human (A⁺, B⁺, and AB⁺ were tested). The percentage of amoebae with ingested red cells, the phagocytic index (PhI), can be considered as an expression of phagocytic activity. Under given conditions (erythrocyte concentration, incubation time, age of amoebic cultures) each strain of Naegleria prefers one erythrocyte type. Thus, for 72-h cultures, N. lovaniensis ingested more A⁺ type erythrocytes than did N. gruberi strains but had very low affinity for rabbit red cells except when very high concentrations were tested. Naegleria gruberi 1f was the most active of the three strains towards rabbit and B⁺ and AB⁺ human erythrocytes, but very low PhIs were obtained with goat erythrocytes. Naegleria gruberi le exhibited high phagocytic activity for every erythrocyte type except for rabbit red cells.     

BCG-induced chronic pulmonary inflammation and splenomegaly in mice: suppression of PHA-induced proliferation, delayed hypersensitivity to sheep erythrocytes, and chronic pulmonary inflammation by soluble factors from adherent spleen cells

C57BL/6 (B6), but not CBA, mice develop intense chronic granulomatous inflammation (CGI) in the lungs and spleen in response to an iv injection with killed BCG in an oil-in-saline emulsion (BCG-E). Concomitant with the development of CGI, these mice show diminished responsiveness to PHA and LPS, as well as suppression of antibody synthesis and production of delayed hypersensitivity (DH) to sheep erythrocytes (SRBC). Suppression results from the development of adherent, Thy-1^?, Ig^? spleen cells. The present study shows that cells from inflamed spleens of BCG-E-treated B6 mice elaborate factors in vitro which (a) inhibit PHA-induced proliferation of both normal syngeneic and allogencic cells, (b) suppress DH to SRBC in B6 mice, and (c) diminish the intensity of BCG-E-induced CGI in the lungs and spleens of B6 mice. These factors are produced by adherent Thy-1^? cells in BCG-injected mice but not in similarly treated CBA mice. These factors may be important in understanding the control of immunologically mediated chronic inflammation.     

Cultured cells from renal cortex of hibernators and nonhibernators. Regulation of cell K+ at low temperature

Cells were grown as primary monolayer cultures from kidney cortex of guinea pigs (nonhibernators), hamsters and ground squirrels (both hibernating species). When plates of cells were placed at 5 °C, cells of guinea pigs lost 37% of their K⁺ in 2 h and those of the hibernator lost about 10%.Uptake of ⁴²K into the cells exhibited a simple, single exponential time course at both temperatures. Unidirectional efflux of K⁺ was equal to K⁺ influx in all cultures at 37 °C and, within limits of error, in hibernator cells at 5 °C. Efflux was 3- to 5-fold greater than influx in guinea pig cells at 5 °C.After 2 h in the cold the ouabain-sensitive K⁺ influx remaining (7–15% of that at 37 °C) was about the same in the cells of the 3 species. Cells from active hamsters and from hibernating ground squirrels, however, exhibited significantly greater pump activity after 45 min in the cold (19 and 14%, respectively). The stimulation of K⁺ influx by increasing [K⁺]_o did not show an increase in K_m⁺ at 5 °C in cells of guinea pigs and ground squirrels. Lowering [K⁺]_c and/or raising [Na⁺]_c by treatment in low- and high-K⁺ media caused only slight stimulation of K⁺ influx, except in cells of ground squirrels at 5 °C in which the stimulation was at least 11-times greater than at 37 °C or in cells of guinea pigs at either temperature.This altered kinetic response of K⁺ transport to cytoplasmic ion stimulation with cooling accounted for about one-third of the improved regulation of K⁺ at 5 °C in ground squirrel cells; the other two-thirds was attributable to a greater decrease in K⁺ leak with cooling. The inhibition of active transport by cold in all 3 species was much less severe than that previously seen in any (Na⁺ + K⁺)-ATPase of mammalian cells.     

Ly-4.2: a cell membrane alloantigen of murine B lymphocytes. III. In vitro studies

The alloantigenic specificity Ly-4.2 is present on a restricted population of murine lymphocytes which have previously been shown to have some of the properties generally ascribed to B lymphocytes, both with regard to distribution and function. In the study reported herein, the effect of anti-Ly-4.2 and anti-Thy-1.2 (θ) antisera have been examined in various in vitro systems. (a) T cell-mediated lysis of ⁵¹Cr-labeled P815-X2 target cells by immune allogeneic peritoneal exudate cells is inhibited by anti-Thy-1.2, but not affected by the anti-B (Ly-4.2) reagent. (b) Antibody-dependent lymphocyte-mediated lysis of ⁵¹Cr-labeled sheep red cells was only slightly inhibited by anti-Ly-4.2 and anti-Ig antisera, and not at all by anti-Thy-1.2 antisera, indicating that this type of cell lysis is mediated by neither T (Thy-l⁺) nor B (Ly-4.2⁺,Ig⁺) cells. (c) The response of lymph node lymphocytes to various mitogens was affected thus: PHA, completely inhibited by anti-Thy-1.2 but not by anti-Ly-4.2; Con A, largely inhibited by anti-Thy-1.2, and slightly by anti-Ly-4.2; PWM (pokeweed mitogen), partially inhibited by both antisera; E. coli endotoxin lipopolysaccharide, greatly inhibited by anti-Ly-4.2 but only slightly by anti-Thy-1.2. The findings demonstrate that anti-Thy-1.2 reacts predominantly with T cells and anti-Ly-4.2 with B cells.     

A human lymphoid cell line with receptors for both sheep red blood cells and complement

The human lymphoid cell line MOLT 4, from a patient with acute lymphocytic leukemia, was initially considered to be derived from T lymphocytes, on the basis of rosette formation with sheep erythrocytes (E). This cell line has now also been found to form rosettes with sheep erythrocytes sensitized with rabbit antibody and mouse complement (EAC). Evidence is presented that the formation of both E and EAC rosettes is due to two separate receptors on the MOLT cells: (a) EAC rosettes were formed more rapidly and were more stable than E rosettes; (b) preincubation of MOLT with an EAC membrane preparation inhibited resetting with EAC and not with E; (c) MOLT formed rosettes with EAC prepared from trypsinized E, but did not bind to trypsin-treated E alone. The implications of this finding, in regard to the derivation of this cell line, are discussed.     

Allelic exclusion of surface IgM allotypes on spleen and bursal B cells in the chicken

The expression of IgM-1a and IgM-1b (C) allotypes on the surface of splenic and bursal cells of B14 line chickens was examined by the triple-layer immunfluorescence technique. We have demonstrated that all surface Ig⁺ cells in both spleen and bursa expressed sIgM-1 which was subject to allelic exclusion in M-l^a/M-l^b heterozygous chickens. Furthermore, a phenotypic equilibrium in the proportion of cells carrying the products of the twoM-1 alleles was observed.Abbreviations used in this paper Ig immunoglobulin - cIg cytoplasmic immunoglobulin - sIg surface immunoglobulin - RAC rabbit anti-chicken Ig serum - PBS phosphate buffered saline - NCS normal chicken serum     

Rapid magnetic purification of rosette-forming lymphocytes.

High gradient magnetic separation, which as previously been shown effective in extracting erythrocytes from a flowing cell suspension, has been used to separate rosetted and unrosetted human peripheral blood lymphocytes. The hemoglobin in the sheep red cells used to form rosettes was first oxidizied to the paramagnetic methemoglobin form. Samples of 50 x 10(6) lymphocytes could be processed in 10 min under sterile conditions with greater than 90% purity of the rosetted cell fraction and maintenance of T cell function in mixed lymphocyte cultures.     

Phylogeny of lymphocyte heterogeneity. II. Differential effects of temperature on fish T-like and B-like cells.

Lymphocytes from the bluegill, a freshwater fish, were observed to undergo in vitro mitogenic responses to a variety of “classical” mitogens. Using cell fractionation approaches based upon surface markers and in vitro mitogenesis, bluegill lymphocytes could be divided into two populations. One population responded to PHA and Con A but not to LPS, contained surface antigens in common with bluegill brain, and did not form spontaneous rosettes with rabbit erythrocytes. The other population responded to LPS but not to PHA or Con A, did not appear to contain surface antigens in common with brain, and did form rosettes with rabbit erythrocytes. The former population responded to mitogenic stimulation very well at 32 °C, whereas the latter population responded better at 22 than at 32 °C. The pattern of mitogenic responses and brain antigen distribution coupled with the observation that mixed lymphocyte responses were obtained at 32 but not at 22 °C makes it likely that the 32 °C responsive population represents the fish equivalent of T cells. The other population may be B cells. These data suggest that the immunosuppressive effects of low temperatures on cold-blooded animals may be effects on the generation of functional T cells and not on B cells.     

A rapid method for the isolation of human peripheral null lymphocytes.

We describe here a new technique for the isolation of human peripheral null lymphocytes: cells lacking detectable surface immunoglobulin as well as receptors for sheep erythrocytes. The double rosette technique described employs a combined negative selection for Ig⁺ cells by specific anti-Ig-coated sheep erythrocytes and E rosette selection for T cells. This singlestep procedure facilitates the isolation of null cells from large cell populations and can be completed in less than 2 hr. The isolated null subpopulation represents less than 5% of the total peripheral mononuclear cells and is at least 85–90% pure by sensitive surface marker analysis. The null subpopulation is shown to be highly enriched for effectors of both antibody-dependent cytotoxicity against autologous lymphocytes and spontaneous cytotoxicity against the K562 leukemia blast cell line. Application of this technique will allow further characterization of the effector populations for ADCC and NK as well as differentiation of T and B cell precursors.     

Lysis of herpes simplex virus-infected targets: II. Nature of the effector cells

Peripheral blood lymphocytes (PBL) from patients with herpes simplex virus (HSV)-1 recurrences in the cornea only (Group I) exhibited reduced lysis of HSV-1-infected targets compared to PBL from patients with oral-facial and corneal HSV recurrences (Group II). The cytotoxic lymphocytes appeared to belong to a subpopulation of natural killer (NK-HSV) cells. Monoclonal antibodies to human lymphocyte differentiation antigens were used to define the surface phenotype of the NK-HSV cells. Most of the NK-HSV activity was mediated by lymphocytes expressing the surface markers Leu-7⁺ (HNK-1), OKT3⁺ (pan T), OKM1⁺ (myeloid and NK), Leu-2^? (cytotoxic/ suppressor T cell), and Leu-8^? (regulatory T cell). In contrast, lysis of K562 cells (NK-K562) was mediated by lymphocytes bearing the surface phenotype Leu-7⁺, OKT3^?, OKM1⁺, Leu-2^+/?, and Leu-8^?. The low level of NK-HSV activity in PBL from Group I donors appeared to be due to active suppression by suppressor T lymphocytes. Depletion of Leu-2⁺ cells from PBL of Group I donors resulted in significant augmentation of NK-HSV activity. Similar treatment of PBL from Group II donors either had no effect or slightly diminished the NK-HSV activity.     

Characterization of Effector Cells against B16 Melanoma in Mice Inoculated with Allogeneic Spleen Cells

Inbred C57BL/6 (B6) mice which had received an inoculation of allogeneic spleen cells showed remarkable antitumor activity against syngeneic tumor challenge with B16 melanoma cells 3 days after the allogeneic cell inoculation. This antitumor activity was not specific to the inoculated alloantigen, since the challenging B16 cells are syngeneic to B6 mice and since it was induced by BALB/c spleen cells as well as C3H/He spleen cells. The antitumor activity was sensitive to an in vivo treatment with anti-asialo GM1 (AGM1) antiserum or anti-Thy.1 monoclonal antibody (mAb) just before the tumor challenge and was resistant to an in vivo treatment with anti-CD8 (Ly. 2) mAb. These results suggest that AGM1+Thy.1+CD8– activated natural killer (NK) cells were generated by alloantigen inoculation and took an important part in the antitumor effect of the alloantigen inoculation.     

A system of alloantigens that selectively identifies lymph-node lymphocytes

The antiserum (BALB.I-H-2 ^j x SWR/7)F₁ anti-I.29 ascites cells, in reaction with B6 lymph-node cells (LNC) in the cytotoxicity assay, defines an alloantigen system called Lna-1* (lymph-node alloantigen-1) which in normal, untreated mice is expressed on the cells of lymph nodes but not of other lymphoid organs. The T- and B-cell populations of lymph nodes evidently include Lna-1u1) subpopulations representing 30–40 percent of the total population. The Lna-1¹ phenotype could be induced on cells of thymus and spleen but not of bone marrow. Congenitally asplenic +/Dh mice have no Lna-1⁺ cells in their lymph nodes, but their LNC can be induced to express Lna-1; this suggests that the spleen is normally required for the differentiation of Lna-1⁺ cells from Lna-1^– precursors. It is not yet known whether thymus is also required for the expression of Lna-1 in lymph nodes. It remains to be seen whether the existence of the Lna-1 + B-cell subpopulation of lymph nodes depends on Lna-1⁺ T cells, and whether the Lna-1⁺ phenotype of B cells may be acquired rather than intrinsic. One hypothesis which is the basis of further study is that there is a T-cell pathway in which noninducible bone-marrow cells become Lna-l-inducible in the thymus, then travel to the spleen, where they are induced to become Lna-1⁺, after which they reside in lymph nodes.Dr. Jan Klein reminded the authors that Lna was previously used in reference to la antigens (David et al. 1973) and was later abandoned.