Increased expression of major histocompatibility complex (MHC) class I transplantation antigens in bovine trophoblast cells before fusion with maternal cells.

The mammalian fetus is potentially at risk from maternal immune attack because it can express paternally inherited polymorphic antigens, including those encoded by the major histocompatibility complex (MHC). The aim of this study was to investigate in more detail MHC class I upregulation by binucleate trophoblast cells in the bovine placenta. A method was developed to isolate binucleate cells by enzymatic disaggregation and density gradient centrifugation of bovine placental cotyledons. In cytospin preparations, 25-30% of purified binucleate cells stained positively with antibodies that recognize bovine MHC class I. The same antibodies were used to immunoprecipitate radiolabelled class I molecules from lysates of binucleate cells and fetal peripheral blood mononuclear cells. The protein species isolated from the two types of cell were similar in size and degree of glycosylation. PCR amplification of cDNA generated from binucleate cells and subsequent sequence analysis demonstrated transcription of MHC class I mRNA species similar to those found in fetal peripheral blood mononuclear cells, and there was no evidence of genetic imprinting of paternally inherited alleles. These results indicate that binucleate cells upregulate expression of MHC class I as they differentiate from MHC-negative uninucleate trophoblast cells. This finding has important implications for the immunological status of the fetus, as binucleate trophoblast cells are destined to cross to the maternal side of the placenta where they fuse with maternal cells. The immunological function of the resulting antigenically mixed fetomaternal hybrid minisyncytia is unknown.

Flow cytometric measurement of intracellular Th1 and Th2 cytokine production by human villous and extravillous cytotrophoblast.

A wide variety of cytokines are present at the maternal-fetal interface, but the extreme cellular complexity of the placenta has made it difficult to determine which cytokines are produced by which cells. Hence novel flow cytometric methods have been applied to determine intracellular cytokine production by specific cell-types in placental cell suspensions. Cell suspensions were prepared from first and third trimester chorionic villi and third trimester amniochorion by enzymatic digestion and Percoll density gradient centrifugation. After overnight incubation in the presence of monensin, cells were fixed, permeabilized and labelled with antibodies for villous cytotrophoblast (cytokeratin+, MHC class I-), extravillous cytotrophoblast (cytokeratin+, MHC class 1+) and leucocytes (CD45+). These cell types were further characterized by their expression of EGFR (proliferative cytotrophoblast) and c-erbB2 (invasive cytotrophoblast). Production of IL-4, IL-10, TNF-alpha, IFN-gamma and IL-12 was determined by simultaneous labelling with the appropriate monoclonal antibodies. Only IL-4 was detected consistently in all samples of cytotrophoblast. IL-10 was not detected but IL-10 mRNA was demonstrated in third trimester chorionic villus digests by RT-PCR. Although IL-4 secretion has not been demonstrated, these data suggest that, in vivo there may be a "Th2 type cytokine bias" orchestrated by the trophoblast. It is proposed that other cytokines (including IL-10 and TNF-alpha) are produced by decidual leukocytes, and not cytotrophoblast, at the maternal-fetal interface.

HLA-G suppresses proliferation of CD4(+) T-lymphocytes.

HLA-G is a non-classical MHC class 1 molecule, expressed primarily on human foetal trophoblast cells, which exhibits almost no genetic polymorphism. Because of these unusual features, HLA-G has been suggested to help prevent maternal immune attack of the semi-allogeneic foetus. The aim of these experiments was to investigate the effects of HLA-G on T-lymphocyte responses by using MHC class II-bearing HLA-G transfectants as stimulators of a mixed lymphocyte reaction. The presence of HLA-G, but not classical HLA class I, on the surface of stimulator cells markedly suppressed thymidine incorporation by peripheral blood mononuclear responder cells from a class I-similar, class II-dissimilar male. The suppressive effect of HLA-G on the mixed lymphocyte reaction persisted after depletion of phagocytes and CD8(+) T-cells from the responder population, but the mixed lymphocyte reaction was entirely abolished by depletion of CD4(+) T-cells. These results suggest that HLA-G exerts a direct suppressive effect on CD4(+) T-lymphocytes, even in the absence of the CD8(+) cells with which other human MHC class I molecules are thought to interact. Thus, HLA-G may allow the foetus to escape maternal immune attack by modulating CD4(+) T-cell activity.

Evolution of mammalian pregnancy in the presence of the maternal immune system.

Eutherian mammals have inherited a typical vertebrate immune system, which protects the body against infectious organisms by detecting and destroying foreign biological material. However, with the evolution of longer gestation periods, this protective mechanism became a potential threat to the 'semi-foreign' fetus and so eutherians have developed systems to prevent immune rejection of their developing fetuses. In many species, this is achieved by reducing placental expression of major histocompatibility complex (MHC) genes, the products of which are responsible for most transplantation rejection reactions. Unexpectedly, however, major histocompatibility complex expression is often re-established in the most invasive trophoblast cells. It is not known why transplantation antigen expression in the fetal cells most exposed to the maternal immune system is advantageous. It is possible that such expression aids the process of invasion or exerts an immunoprotective effect on the fetus. It may prove possible to identify the essential steps that all eutherian fetuses take to ensure their survival in the face of potential maternal immune attack by studying the common features of the placental immunology of different species.

The short forms of HLA-G are unlikely to play a role in pregnancy because they are not expressed at the cell surface.

HLA-G is a nonclassical class I MHC molecule of unknown function expressed on human invasive trophoblast. In trophoblast cells, HLA-G mRNA is alternatively spliced into a variety of forms which are predicted to encode a full length membrane-bound form, three short membrane-bound isoforms and two soluble isoforms. The aim of this study was to determine which of these protein isoforms are translated, which are expressed on the cell surface and which are secreted. Artificial cDNAs encoding the isoforms were generated by PCR mutagenesis, ligated to an epitope tag and transfected into a human cell line capable of expressing MHC class I. Protein products of appropriate sizes were detected in cells transfected with cDNAs encoding all membrane-bound forms, but surface biotinylation studies indicated that only full length membrane-bound HLA-G was present at the cell surface. Full length HLA-G was also detected by surface antibody binding and flow cytometry. Soluble HLA-G1 was detected in cells transfected with the appropriate cDNA only after treatment with monensin, which inhibits transport of glycoproteins through the Golgi apparatus. These results suggest that full length HLA-G, but not short HLA-G isoforms can be expressed on the surface of human cells and that soluble HLA-G is rapidly secreted. Thus, it is likely that the full length membrane-bound and soluble forms of HLA-G are the only biologically active forms to which the mother is exposed.

Little evidence of HLA-G mRNA polymorphism in Caucasian or Afro-Caribbean populations.

HLA-G is a nonclassical class I MHC molecule of unknown function expressed on human trophoblast. The level of polymorphism at the HLA-G locus is of considerable importance, since the paternally inherited gene product is exposed to the maternal immune system during pregnancy. However, previous studies of HLA-G polymorphism using genomic DNA samples have produced conflicting results. Our aim was to investigate polymorphism in trophoblast HLA-G mRNA from pregnancies in ten Caucasian and twelve Afro-Caribbean women by RT-PCR. A similar PCR protocol was also applied to umbilical cord blood genomic DNA from two Caucasian and two Afro-Caribbean neonates. Caucasian cDNA yielded only two different sequences: G*01011, and one containing a previously reported synonymous substitution. Afro-Caribbean samples yielded these sequences as well as one previously reported conservative (leucine-to-isoleucine) substitution. PCR amplification from genomic DNA samples from both populations using previously published primer pairs generated sequences containing multiple substitutions, many of which were nonsynonymous. More than two sequences were produced from genomic DNA from each individual. In contrast, amplification from the same genomic DNA using new primers complementary to exons of the HLA-G gene yielded the same few sequences generated from cDNA. These results suggest that polymorphism at the HLA-G locus is extremely limited in Caucasian and Afro-Caribbean populations. This suggests that spurious polymorphism has been reported in African Americans due to the use of intron-complementary PCR primers on genomic DNA samples. The monomorphic nature of HLA-G may allow trophoblast to carry out the immunological functions of class I-bearing tissues without compromising successful pregnancy.

Source and site of action of anti-luteolytic interferon in red deer (Cervus elaphus): possible involvement of extra-ovarian oxytocin secretion in maternal recognition of pregnancy.

Six conceptuses were collected from red deer hinds on day 22 after synchronization of oestrus with intravaginal progesterone-releasing devices (removal of device = day 0). Within 24 h of culture in vitro, the supernatant from five of six conceptuses showed detectable antiviral activity. Interferon alpha (IFN-alpha) receptors were identified by immunohistochemistry on the luminal surface of the endometrium, in the neurohypophysis and paraventricular hypothalamus, but not in the ovaries of the hinds from which the conceptuses were collected. Another 16 intact hinds were synchronized as above. Injection of 4 mg IFN i.m. twice a day on days 13-15 had no effect on cloprostenol-induced oxytocin secretion on day 15 and did not prevent cloprostenol-induced luteal regression. Sixteen ovariectomized hinds received a protocol of steroid treatment to mimic ovarian hormone secretion during the normal oestrous cycle. On day 16, hinds showed undulant oxytocin secretion that showed a degree of temporal association with uterine PGF2 alpha release. Treatment with 4 mg IFN-alpha I 1 twice a day on days 13-16 had no effect on this spontaneous oxytocin secretion, but reduced the magnitude of cloprostenol-induced oxytocin secretion on day 17 (P < 0.05). These results indicate that red deer conceptuses secrete an anti-luteolytic IFN to which the endometrium expresses a receptor during early pregnancy. The presence of IFN receptors in the hypothalamus and posterior pituitary and the IFN-induced suppression of extra-ovarian oxytocin secretion provides tentative evidence of an involvement of the central nervous system in maternal recognition of pregnancy in deer.

Successful in vitro fertilization of in vivo matured oocytes aspirated laparoscopically from red deer hinds (Cervus elaphus).

Most current protocols of in vitro fertilization in ruminants are based on in vitro maturation of oocytes derived from abattoir material. For application of IVF technology to captive endangered species, however, noninvasive techniques are required which allow repeated collection of oocytes from live females. The aim of this study was to develop a method for embryo production from mature oocytes collected laparoscopically from red deer hinds. Follicular development was synchronized in red deer hinds by the insertion of intravaginal progesterone-releasing devices for 10 d, and ovarian stimulation was induced with 1000 IU, i.m. PMSG 48 h before progesterone device removal. Oocytes were harvested by laparoscopy under xylazine/ketamine sedation 24 h after progesterone device removal and then co-incubated with frozen-thawed red deer spermatozoa for 24 h. In Experiment 1, oocytes and embryos were fixed and stained at different developmental timepoints. Their external morphological changes (cumulus expansion, extrusion of the second polar body and cytokinesis) paralleled their nuclear developmental changes (formation of the 2nd metaphase spindle of meiosis, pronuclear formation and nuclear division, respectively). In Experiment 2, embryos were maintained in vitro until they ceased to undergo cell division. A total of 39 aspiration procedures was carried out on 14 red deer hinds. Forty-four cumulus-oocyte complexes (COC) were aspirated from 95 large Graafian follicles; of these, 27 were classed as mature/nondegenerated on the basis of cumulus/cytoplasmic morphology. Seventeen oocytes cleaved following in vitro fertilization, yielding six 2-cell embryos, six 4-cell embryos, four 8-cell embryos and one 16-cell embryo. The results indicate that laparoscopic aspiration of mature oocytes from hormone-treated females offers a valuable source of genetic material for assisted deer breeding programs.

Potential of assisted breeding techniques for the conservation of endangered mammalian species in captivity: a review.

An alarming worldwide extinction of animal species is taking place as a result of the activities of the increasing global human population. The original ranges of many animal species are being reduced and fragmented and, in some cases, they have been reduced to perilously small relict populations. The adverse genetic consequences of these restrictions are becoming clear, as are possible methods for their alleviation. The concept of ex situ genetic management of small captive populations of endangered species with a view to re-introducing them into the wild is attracting increasing interest. Modern reproductive techniques will play an important role in such programmes, and it is likely that an increasing number of veterinarians will become involved. However, the literature describing the aims and methods of reproductive genetic management is scattered and often not readily available to interested veterinary surgeons. The aim of this review is to deal with this problem by describing some potential approaches to the captive breeding of endangered species.

Regulation of MHC class I gene expression is at transcriptional and post-transcriptional level in bovine placenta.

A previous study of MHC in cattle trophoblast demonstrated low or absent class I expression, using a broad specificity monoclonal antibody. The study reported here uses MHC-defined cattle and embryo transfer to ensure MHC incompatibility between dam and calf. Transcription and expression of defined class I genes was examined in placentomes taken at term, using monoclonal antibodies to bovine class I, a gene-specific DNA-based typing system, and in situ hybridisation. Results demonstrate intermediate levels of fetal MHC class I mRNA in trophoblast, but no detectable fetal class I protein. This suggests a level of transcriptional down-regulation, and a post-transcriptional block which might involve other gene products, such as beta2-microglobulin (beta 2m), or proteins involved in generation/transport of peptides.

Premature luteal regression induced by equine chorionic gonadotropin and estrogen is suppressed by administration of exogenous interferon in red deer (Cervus elaphus).

Superovulation of red deer hinds with eCG causes premature luteal regression by inducing follicular hypersecretion of estrogen that activates the luteolytic mechanism. Six groups of hinds (n = 8 per group) were treated with progesterone-impregnated intravaginal controlled internal drug-releasing (CIDR) devices for 14 days to synchronize estrus (CIDR device withdrawal = Day 0). Group 1 served as controls; group 2 received an i.m. injection of 1200 IU eCG at -72 h; group 3 received similar eCG treatment as well as i.m. injections of 0.25 mg estradiol benzoate (EDB) at 72, 84, 96, and 108 h; group 4 received twice-daily i.m. injections of 4 mg recombinant bovine interferon-alpha(I)1 (IFN) from Days 2 to 7; group 5 received IFN and eCG as above; group 6 received IFN, eCG, and EDB. Ovarian response was determined by laparoscopy on Days 14 and 15. Progesterone profiles were determined from thrice-weekly plasma samples from Days -14 to 28. Both the incidence of visible signs of luteal regression and the variation in the time of termination of the luteal phase (plasma progesterone < 1 ng/ml) were greater in eCG+EDB-treated hinds than in control, IFN-, IFN+eCG-, and IFN+eCG+EDB-treated hinds (p < 0.05). The ovulation rate in the eCG+EDB-treated hinds was less than that in the eCG-, IFN+eCG-, and IFN+eCG+EDB-treated hinds (p < 0.05). These results suggest that treatment with interferon, the putative embryonic pregnancy recognition signal, suppresses premature luteal regression induced by hypersecretion of estrogen following treatment with eCG.

Can repeated plasma donation by asymptomatic HIV-infected individuals delay the onset of AIDS?

Healthy HIV-positive regular donors of plasma in a programme of passive immunotherapy for AIDS patients were studied over a period of about two years. None developed symptoms of clinical progression; most seemed to make substantial gains of CD4 cells by comparison with asymptomatic individuals who were not donating. The effects of donation did not seem to diminish with repetition, and donor CD4 counts tended towards stabilizing within normal limits. Asymptomatic HIV-positive individuals were compared immunologically with 'normals' and people with AIDS, using a battery of 25 measurements on peripheral blood. The immunological profiles of donor and non-donor asymptomatics, indistinguishable at the start, became dissimilar: donors' profiles resembled AIDS less, non-donors became less like 'normal' and a few non-donor results could not be distinguished from AIDS. Improvement in the CD4 counts and amelioration of the immunological profile in donors provide prima facie evidence that plasmapheresis may be therapeutic for asymptomatic HIV-positive people. Further studies are justified.

Effect of pregnancy and exogenous interferon on synchronous pulsatile release of oxytocin and luteolytic prostaglandin F2 alpha in red deer (Cervus elaphus).

Three experiments were carried out to investigate the secretion of luteolytic hormones in red deer hinds during the oestrous cycle, early pregnancy and after administration of interferon, the putative pregnancy recognition signal. Three groups of hinds (n = 8-9 per group) were treated with progesterone-impregnated intravaginal controlled internal drug releasing (CIDR) devices for 13 days (device withdrawal = day 0). Group 1 (n = 9) served as controls; Group 2 (n = 8) received injections of 4 mg recombinant bovine interferon-alpha 1 l twice a day on days 13-18; Group 3 (n = 9) were run with a fertile stag on days 0-3. Plasma samples collected each day on days 16-23 were analysed for progesterone. Plasma samples collected each hour for 16 h on days 4, 10, 16, 18 were analysed for oxytocin and the prostaglandin F2 alpha (PGF2 alpha) metabolite (PGFM). Plasma progesterone concentrations declined to < 1 ng ml-1 between days 18 and 25 in control hinds indicating that luteolysis had occurred, whereas there was no endocrine evidence of luteolysis in interferon-treated or pregnant hinds. Control hinds (6/9) exhibited synchronous pulses of oxytocin and PGF2 alpha secretion on day 18, a greater proportion than on any other day in these hinds or on any day in the interferon-treated or mated hinds (P < 0.05). In a second experiment, close synchrony in secretion of oxytocin and PGF2 alpha pulses was evident in an unmated hind when samples were collected every 12 min on day 18. In a third experiment, oxytocin-induced PGF2 alpha secretion was potentiated by oxytocin administration at an interval of 1 h and inhibited by administration at a 6 h interval (P < 0.05). These results suggest that synchronous pulsatile secretion of oxytocin and PGF2 alpha induces luteolysis and is suppressed by pregnancy or administration of interferon. Oxytocin-induced alterations in uterine oxytocin sensitivity may underlie the pulsatile nature of luteolytic hormone secretion in red deer.

Exogenous interferon delays luteal regression in red deer hinds (Cervus elaphus) by suppressing steroid-induced endometrial oxytocin sensitivity.

Three groups of intact hinds (n = 10-18) and one group of ovariectomized hinds were treated with progesterone by mean, of Controlled Internal Drug Releasing (CIDR) devices for 13 days (device removal = Day 0). Group 1 served as controls; group 2 received injections of 4 mg recombinant bovine interferon-alpha,1 twice daily from Days 13 to 21; group 3 was run with a stag from Days 0 to 3, and all hinds were subsequently diagnosed pregnant; group 4 (ovariectomized) was treated with CIDR devices and estradiol to mimic steroid secretion during the estrous cycle. Progesterone profiles were determined from thrice-weekly plasma samples from Days -13 to 28. Rectal temperature was measured in a subset of groups 1 and 2 from Days 9 to 21. Oxytocin-induced prostaglandin F2 alpha release was measured in a subset of groups 1, 2, and 4 on Days 2, 4, 10, 16, and 18. Data are presented as means +/- SEM. Exogenous interferon delayed luteolysis (> or = 28 vs. 21.2 +/- 0.55 days, P < 0.0005) and induced transient pyrexia after the first injection (39.89 +/- 0.11 vs. 38.88 +/- 0.19 degrees C, p < 0.0005). Incidence of oxytocin-induced PGF2 alpha release in control hinds was greater on Days 2 and 18 than on Days 4 and 10 (8/8 and 7/8 vs. 3/8 and 0/8, respectively; p < 0.05) and was greater in control than in interferon-treated hinds on Days 16 and 18 (5/8 and 7/8 vs. 1/8 and 1/8, respectively; p < 0.05). Profiles of plasma progesterone concentration and oxytocin sensitivity in steroid-treated ovariectomized hinds did not differ from those in control hinds. These results suggest that steroid-controlled uterine oxytocin sensitivity is important in luteolysis and is suppressed by the administration of interferon, the putative embryonic pregnancy recognition signal in red deer.

External quality assurance for CD34 cell enumeration--results of a preliminary national trial. Royal Microscopical Society Clinical Flow Cytometry Group QA Schemes.

With the development of haematopoietic stem cell (HSC) mobilisation strategies and the associated technical expertise in leukapheresis has come the need for accurate and reproducible enumeration of HSC in the peripheral blood. Enumeration of HSC is not only required for timing of the harvest but is valuable in determining that the minimum number of HSC required for successful engraftment has been collected. In order to establish a minimum number of HSC required, results from multiple institutions performing such transplants need to be assessed. Clearly, to compare results from multiple centres requires confidence in the reproducibility of the assay. We have evaluated an established EQAS method which has already been proven in the external quality assurance of CD4 measurement in clinical samples to assess the inter- and intra-laboratory reproducibility of CD34 measurement. Fifteen laboratories participated in two distributions in which 28 samples were analysed. Standardised methods were not employed, laboratories using their routine methods. Participants reported their results in terms of '% positive of total leukocytes' and 'absolute number of CD34+/microliters'. A wide range of clinical samples was despatched and analysed with CD34 cell content ranging from 0.08-19.31% positive. The coefficients of variance (CV) associated with the estimations of relative proportions and absolute numbers were maximally 100.1 and 136.6%, respectively. This study highlights the need for external quality assurance and standardisation of the methodology of this assay.

Role of estrogen and prostaglandin F2 alpha in premature luteal regression in monovulatory and superovulated red deer (Cervus elaphus).

The superovulation of red deer hinds with eCG is commonly associated with premature luteal regression. This study was an investigation of the endocrine mechanisms regulating luteal function after superovulation. Four groups of hinds (n = 7-8 per group) were treated with progesterone-impregnated intravaginal controlled internal drug-releasing (CIDR) devices for 12 days to synchronize estrus (CIDR device withdrawal = Day 0). Group 1 served as controls; group 2 received an i.m. injection of 0.25 mg estradiol benzoate (EDB) at 72, 84, 96, and 108 h after removal of the device; group 3 received an i.m. injection of 1200 IU eCG at -72 h; group 4 received both EDB and eCG treatments. Oxytocin-induced prostaglandin F2 alpha (PGF2 alpha) release was assessed on Day 4 by oxytocin challenge. Ovarian response was determined by laparoscopy on Day 14. Plasma steroid profiles were determined from thrice-weekly plasma samples collected from Day -13 to Day 35 (progesterone) and Days 0 to 14 (estradiol). EDB increased the incidence of premature luteal regression in monovulatory and eCG-treated animals (p < 0.05) and reduced the number of CL (p < 0.05) in eCG-treated animals. EDB and eCG each elevated plasma concentrations of estradiol and increased the incidence of significant oxytocin-induced PGF2 alpha release. These results support the hypothesis that eCG causes premature luteal regression by inducing prolonged estrogen secretion that sensitizes the endometrium to oxytocin, thereby eliciting PGF2 alpha release during the early luteal phase.

Hoechst 33258: a fluorescent nuclear counterstain suitable for double-labelling immunofluorescence.

Counterstaining with the chromosomal dye Hoechst 33258 is a simple procedure that provides an excellent general purpose nuclear counterstain for immunofluorescent work. Benefits from its use include the ready identification and orientation of structures, differentiation of lymphoid and non-lymphoid cells and easy assessment of frequency in specifically labelled cellular subpopulations. It can be used at the same time as FITC and TRITC double-labelling immunofluorescence.

Elimination of allogeneic lymphocytes by mice.

Two major classes of response to allogeneic lymphocytes can be detected in mice in vivo, based on injecting them intravenously with 51Cr-labelled lymph node cells and examining them in a short term assay. A natural immunity discriminating between allogeneic and syngeneic lymphocytes is seen in the lymph nodes (and to a lesser extent, the spleen), which has such close similarities to natural cell-mediated responses of the NK class as thymic-independence and radioresistance. However, it has immunological specificity of a conventional kind, probably towards serologically determined K/D antigens. There is also an active immune response, produced by immunisation with dissociated lymphoid cells or allografting, which consists of three elements: an IgG opsonising alloantibody response, diverting circulating lymphocytes to the liver; an IgM opsonin, localising them to the spleen; and a cell-mediated serum-dependent elimination mechanism that destroys cells entering the lymph nodes and spleen. Dose-response curves for the primary response show evidence of high-dose paralysis of elimination. Dose-time-response results for the secondary show a variety of unique characteristics; evidence is presented to suggest that several aspects of the phenomenon betray primitive features retained from an earlier stage in the evolution of the immunological system.

Nomarski differential interference contrast studies of murine lymphocytes.

Nomarski differential interference contrast microscopy (DIC) of lymphocyte surface morphology was combined with immunofluorescence studies on T and B cell markers on the thymus, lymph nodes, spleen, peripheral blood lymphocytes and thoracic duct lymph of female CBA mice. DIC identified smooth cells and several categories of villous cells; more extreme forms were present in lymph. Most B cells seemed to belong to the smooth group and most peripheral T cells to the villous group. Thymus cells were almost entirely smooth, but treatment with cortisone increased the proportion of villous cells to 50%. The surface morphology of lymphocytes was highly labile preventing direct identification or separation of T and B cells. In vivo removal of T cells by adult thymectomy, lethal irradiation and bone marrow reconstitution caused the villous cells to decrease. During recovery from irradiation, T lymphocytes tended to parallel villous cells, B lymphocytes smooth cells, but were differences between the spleen and lymph nodes. Mice deprived of T1 cells by adult thymectomy showed a modest decrease of smooth cells in the spleen and blood; mice depleted of T2 cells by anti-lymphocyte serum, or which were naturally deficient in T2 cells, were markedly lacking in villous cells. Thoracic duct lymph, which is rich in T2 cells, had a high proportion of extremely villous lymphocytes. Exposure to lymph induced extreme villous features in lymph node cells, and it was found that the thoracic duct lymph was markedly hypertonic to serum, although varying in osmolarity throughout the day. It is suggested that the villous shape of T2 cells is a circulatory adaptation, necessitated by the peculiar character of the lymphatic system in mice.