Ex Vivo Costimulatory Blockade to Generate Regulatory T Cells From Patients Awaiting Kidney Transplantation.

Short-term outcomes of kidney transplantation have improved dramatically, but chronic rejection and regimen-related toxicity continue to compromise overall patient outcomes. Development of regulatory T cells (Tregs) as a means to decrease alloresponsiveness and limit the need for pharmacologic immunosuppression is an active area of preclinical and clinical investigation. Nevertheless, the immunomodulatory effects of end-stage renal disease on the efficacy of various strategies to generate and expand recipient Tregs for kidney transplantation are incompletely characterized. In this study, we show that Tregs can be successfully generated from either freshly isolated or previously cryopreserved uremic recipient (responder) and healthy donor (stimulator) peripheral blood mononuclear cells using the strategy of ex vivo costimulatory blockade with belatacept during mixed lymphocyte culture. Moreover, these Tregs maintain a CD3(+) CD4(+) CD25(+) CD127(lo) surface phenotype, high levels of intracellular FOXP3 and significant demethylation of the FOXP3 Treg-specific demethylation region on allorestimulation with donor stimulator cells. These data support evaluation of this simple, brief Treg production strategy in clinical trials of mismatched kidney transplantation.

Parametric analysis of overcorrection duration effects. Is longer really better than shorter?

Positive practice overcorrection (PPOC) has long played a significant role in the behavioral treatment of serious self-stimulatory behavior. Three experiments comparing the effectiveness of 30-second, 2-minute, and 8-minute PPOC on reduction of stereotypic hand behavior of adults with severe to profound developmental disabilities were conducted to resolve inconsistencies in previously reported findings concerning the role of PPOC duration in response suppression. Experiment 1, which used an alternating treatments--multiple baseline design, suggested that the different durations were equally effective in reducing the stereotypic behaviors to near-zero levels. Experiment 2, which used a reversal design, supported the findings of Experiment 1. Experiment 3, which used a reversal design to test the shortest and longest durations, generally confirmed the results of the first two experiments. This study therefore failed to support the oft-claimed superiority of long-duration PPOC. The possible factors underlying these findings and their implications for future research and practice are discussed.

Efficient priming of CD8+ memory T cells specific for a subdominant epitope following Sendai virus infection.

The relationship between the primary effector CTL response to viral infection and the subsequent pool of memory CTL precursors (CTLp) is poorly understood. Here, we have analyzed the induction of both effector CTL and memory CTLp to dominant and subdominant epitopes following Sendai virus infection of C57BL/6 mice. A single peptide derived from the Sendai virus nucleoprotein (NP(324-332)) binds to both H-2 Kb and Db MHC class I molecules, generating both immunodominant (NP(324-332)/Kb) and subdominant (NP(324-332)/Db) epitopes. Following intranasal Sendai virus infection, NP(324-332)/Kb-specific CTL dominated the primary effector CTL response in the lung and were present at high frequency in the memory CTLp pool. In contrast, NP(324-332)/Db-specific CTL were not a detectable component of the effector response to primary Sendai virus infection. However, memory CTLp specific for this subdominant epitope were induced at frequencies approaching those of CTLp specific for the immunodominant epitope. These data indicate that memory CTLp specific for subdominant epitopes can be primed by Sendai virus infection in the absence of a detectable effector response. To determine whether CTLp memory to subdominant epitopes is functional in the context of Sendai virus infection, memory CTLp specific for a subdominant epitope were selectively primed by vaccination. These cells dominated the subsequent effector CTL response to Sendai virus infection, demonstrating that memory CTLp primed against subdominant epitopes can participate in an immune response and effectively compete with T cells specific for immunodominant epitopes. These data have implications for the development of vaccines designed to emphasize cellular immunity.

Major histocompatibility complex class II-associated peptides control the presentation of bacterial superantigens to T cells.

Recent studies have shown that only a subset of major histocompatibility complex (MHC) class II molecules are able to present bacterial superantigens to T cells, leading to the suggestion that class-II associated peptides may influence superantigen presentation. Here, we have assessed the potential role of peptides on superantigen presentation by (a) analyzing the ability of superantigens to block peptide-specific T cell responses and (b) analyzing the ability of individual peptides to promote superantigen presentation on I-Ab-expressing T2 cells that have a quantitative defect in antigen processing. A series of peptides is described that specifically promote either toxic shock syndrome toxin (TSST) 1 or staphylococcal enterotoxin A (SEA) presentation. Whereas some peptides promoted the presentation of TSST-1 (almost 5,000-fold in the case of one peptide), other peptides promoted the presentation of SEA. These data demonstrate that MHC class II-associated peptides differentially influence the presentation of bacterial superantigens to T cells.

Binding motifs predict major histocompatibility complex class II-restricted epitopes in the Sendai virus M protein.

Major histocompatibility complex (MHC) class I ligand motifs have been defined for a number of class I molecules and have been successfully used to identify class I-restricted cytotoxic T-cell epitopes. In contrast, the relative degeneracy of sequence motifs in naturally processed MHC class II ligands has suggested that they may be of more limited use. Here, we use a predicted I-Ab ligand motif to identify antigenic peptides in the Sendai virus Enders strain matrix (M) protein. The entire coding sequence of the M protein was derived, and seven peptide sequences that contained the predicted I-Ab motif were identified. Analysis of I-Ab-restricted M-specific T-cell hybridomas for reactivity to these synthetic peptides identified two distinct epitopes. These data demonstrate that MHC class II motifs can be valuable in predicting T-cell epitopes.

T cell recognition of the immunodominant Sendai virus NP324-332/Kb epitope is focused on the center of the peptide.

The CTL response to Sendai virus in C57BL/6 mice is directed almost exclusively to a single H-2Kb-restricted epitope derived from the virus nucleoprotein, NP324-332 (FAPGNYPAL). We have previously shown that the repertoire of T cells elicited by this epitope following primary Sendai virus infection is very diverse. The current experiments were undertaken to determine how a diverse array of TCR are able to interact with a single class I epitope. Crystallographic analysis of NP324-332 bound to Kb has shown that the side chains of peptide residues F1, G4, N5, and A8 protrude toward the solvent and are potentially available for recognition by the TCR. Notably, the N5 residue protrudes prominently from the peptide-binding site due to its localization on a bulge in the center of NP324-332. To determine the importance of these residues for T cell recognition, we analyzed the response of a large panel of hybridomas to NP324-332 analogues substituted at these four positions. The data suggested that there is dominant recognition of the central G4 and N5 residues at the center of the peptide. However, individual hybridomas exhibited distinct patterns of fine specificity for residues F1 and A8, in that they were dependent on one, both, or neither of these residues for recognition of NP324-332. These data are consistent with a critical role for the G4 and N5 residues in governing NP324-332/Kb recognition by T cells and may have implications for T cell recognition of class-I restricted epitopes in general.

Phytohemagglutinin-inducible p24 in peripheral blood mononuclear cells as a predictor of human immunodeficiency virus type 1 vertical transmission and infant clinical status.

We sought to determine whether the detectability of phytohemagglutinin-inducible p24 (PHA-p24) in short term cultures of peripheral blood mononuclear cells correlates with an increased risk of vertical transmission among human immunodeficiency virus type 1 (HIV-1)-infected pregnant women and more severe symptomatology among HIV-1-infected infants. The assay for PHA-p24 was performed on specimens obtained from HIV-1-infected women during their pregnancy and from infants during the first 6 months of life. Infants were followed prospectively to determine HIV-1 infection outcome and symptomatology. Among PHA-p24 positive women 9 of 19 (47.4%) gave birth to HIV-1-infected infants compared with 4 of 25 (16.0%) of PHA-p24-negative women (P = 0.02). Among women who tested PHA-p24-positive and had a CD4+ lymphocyte count < 500 cells/mm3, 8 of 15 (53.3%) gave birth to HIV-1-infected infants compared with 4 of 26 (15.4%) not meeting these conditions (P = 0.01). Among HIV-1-infected infants 4 of 5 (80%) of those testing PHA-p24-positive by one month of age developed an opportunistic infection or encephalopathy by 12 months of age, compared with none of the 11 infants testing PHA-p24-negative (P = 0.003). We conclude that PHA-p24 may be a useful in vitro measure for increased risk of vertical transmission among HIV-1-infected pregnant women and increased risk for rapid development of severe disease among HIV-1-infected infants.

Analysis of the primary T-cell response to Sendai virus infection in C57BL/6 mice: CD4+ T-cell recognition is directed predominantly to the hemagglutinin-neuraminidase glycoprotein.

Sendai virus infection of C57BL/6 mice elicits a strong CD4+ and CD8+ T-cell response in the respiratory tract. To investigate the specificity of the CD4+ T-cell response, a panel of hybridomas was generated from cells recovered from the respiratory tracts of infected mice. Using vaccinia virus recombinants expressing individual Sendai virus proteins, we found that the majority of these hybridomas (34 of 37) were specific for the hemagglutinin-neuraminidase (HN) glycoprotein. The hybridomas were then analyzed for reactivity to a set of overlapping peptides spanning the entire length of the hemagglutinin-neuraminidase glycoprotein. At least five H-2 I-Ab-restricted epitopes were defined in HN. The strong bias toward recognition of class II epitopes derived from a single viral protein contrasts with T-cell recognition of epitopes of several proteins in influenza A virus as found previously by others.

The MHC class I-restricted T cell response to Sendai virus infection in C57BL/6 mice: a single immunodominant epitope elicits an extremely diverse repertoire of T cells.

We have used Sendai virus infection of C57BL/6 mice as a model with which to study the T cell response to a single MHC class I epitope. Cells taken from the bronchoalveolar lavage or restimulated in vitro from the mediastinal lymph nodes of virus-infected mice were strongly cytotoxic for a single nucleoprotein epitope, NP324-332/Kb. To correlate TCR usage with specificity for the immunodominant epitope, we generated T cell hybridomas from the bronchoalveolar lavage and mediastinal lymph node cells of C57BL/6 mice at the peak of infection. Altogether, 20 hybridomas were identified that specifically secreted IL-2 in response to NP324-332-pulsed L929-Kb cells. TCR usage in this panel of hybridomas was extremely diverse. Over half of the available J beta and V beta elements present in the C57BL/6 strain of mouse were represented in the hybridomas. Similarly, V alpha usage was also diverse and all 12 of the alpha chains sequenced used distinct J alpha elements. The only relatively conserved feature of the TCR in these hybridomas was the presence of an arginine residue in the junctions of 70% of the beta chains. These data demonstrate that a diverse repertoire of TCR is able to recognize a single MHC class I epitope. Moreover, the data demonstrate that mice make use of this potential diversity in the primary response to a natural viral infection.

Effect of 2',3'-didehydro-3'-deoxythymidine in an in vitro hollow-fiber pharmacodynamic model system correlates with results of dose-ranging clinical studies.

We sought to validate an in vitro system which could predict the minimal effect dose of antiretroviral agents. Mixtures of uninfected CEM cells and CEM cells chronically infected with human immunodeficiency virus (HIV) type 1 MN were exposed to 2',3'-didehydro-3'-deoxythymidine (D4T) in vitro in a hollow-fiber model which simulates the plasma concentration-time profile of D4T in patients. Drug concentration was adjusted to simulate continuous intravenous infusion, or an intravenous bolus administered twice daily. The effect of the dosing regimen was measured with viral infectivity, p24 antigen, and reverse transcriptase or PCR for unintegrated HIV DNA. Dose deescalation studies on a twice-daily dosing schedule predicted a minimum effect dose of 0.5 mg/kg of body weight per day which correlated with the results of a clinical trial. Antiviral effect was demonstrated to be independent of schedule for every 12-h dosing versus continuous infusion. Finally, at or near the minimal effect dose, efficacy appeared to depend on the viral load. The ability of this in vitro pharmacodynamic model to assess the response of HIV-infected cells to different doses and schedules of antiviral agents may be useful in the design of optimal dosing regimens for clinical trials but requires validation with other types of antiretroviral agents.

Virus-specific CD8+ T-cell responses in mice transgenic for a T-cell receptor beta chain selected at random.

The consequences of severely limiting the T-cell receptor (TCR) repertoire available for the response to intranasal infection with an influenza A virus or with Sendai virus have been analyzed by using H-2k mice (TG8.1) transgenic for a TCR beta-chain gene (V beta 8.1D beta 2J beta 2.3C beta 2). Analyzing the prevalence of V beta 8.1+ CD8+ T cells in lymph node cultures from nontransgenic (non-TG) H-2k controls primed with either virus and then stimulated in vitro with the homologous virus or with anti-CD3 epsilon showed that this TCR is not normally selected from the CD8+ T-cell repertoire during these infections. However, the TG8.1 mice cleared both viruses and generated virus-specific effector cytotoxic T lymphocytes (CTL) and memory CTL precursors, though the responses were delayed compared with the non-TG controls. Depletion of the CD4+ T-cell subset had little effect on the course of influenza virus infection but substantially slowed the development of the Sendai virus-specific CTL response and virus elimination in both the TG8.1 and non-TG mice, indicating that CD4+ helpers are promoting the CD8+ T-cell response in the Sendai virus model. Even so, restricting the available T-cell repertoire to lymphocytes expressing a single TCR beta chain still allows sufficient TCR diversity for CD8+ T cells (acting in the presence or absence of the CD4+ subset) to limit infection with an influenza A virus and a parainfluenza type 1 virus.

Cross-protection against lymphocytic choriomeningitis virus mediated by a CD4+ T-cell clone specific for an envelope glycoprotein epitope of Lassa virus.

Recombinant vaccinia virus expressing the Lassa virus (LV) envelope glycoprotein precursor, V-LSGPC, was used to study the basis of LV-induced cross-protective immunity against the closely related arenavirus lymphocytic choriomeningitis virus (LCMV). C3H/HeJ mice primed with V-LSGPC developed neither circulating antibodies nor CD8+ cytotoxic T cells specific for LCMV, yet they resisted a normally lethal LCMV challenge. Spleen cells from such mice gave a proliferative response to LCMV in vitro that was inhibitable by anti-CD4 antibody. Synthetic peptides corresponding to predicted T-cell sites common to the envelope glycoprotein precursor (GP-C) of LV and that of LCMV were used to map the specificity of the proliferative response to an epitope located between amino acids 403 and 417 of LV GP-C. Several CD4+ T-cell clones specific for the 403-417 peptide were isolated and found to produce gamma interferon in response to both the peptide and LCMV. One of these clones, C9, was selected for further study. C9 lysed I-AK-bearing target cells, and when adoptively transferred to C3H/HeJ mice, it was capable of mediating both a peptide-specific delayed hypersensitivity reaction and resistance to lethal LCMV challenge. These collective findings demonstrate, for the first time, that CD4+ T cells can play a major role in arenavirus-specific cross-protective immunity.

Inhibition of HIV-1 IIIb replication in AA-2 and MT-2 cells in culture by two ligands of poly (ADP-ribose) polymerase: 6-amino-1,2-benzopyrone and 5-iodo-6-amino-1,2-benzopyrone.

The effects of two adenosine diphosphoribose transferase (ADPRT) enzyme inhibitory ligands, 6-amino-1,2-benzopyrone and its 5-iodo-derivative, were determined in AA-2 and MT-2 cell cultures on the replication of HIV-1 IIIb, assayed by an immunochemical test for the HIV protein p24, and syncytium formation, characteristic of HIV-infected cells. Intracellular concentrations of both drugs were sufficient to inhibit poly(ADP-ribose) polymerase activity within the intact cell. Both drugs inhibited HIV replication parallel to their inhibitory potency on ADPRT, but distinct differences were ascertained between the two cell lines. In AA-2 cells both p24 and syncytium formation were depressed simultaneously, whereas in MT-2 cells only syncytium formation was inhibited by the drugs, and the p24 production, which remained unchanged during viral growth, was unaffected. Both drugs only moderately depressed the growth rate of the AA-2 and MT-2 cells and there was no detectable cellular toxicity. Results suggest the feasibility of the development of a new line of ADPRT ligand anti-HIV drugs that fundamentally differ in their mode of action from currently used chemotherapeutics.

Rejection of allogeneic tumor is not determined by host responses to MHC class I molecules and is mediated by CD4-CD8+ T lymphocytes that are not lytic for the tumor.

In previous studies, the murine SaI (A/J derived, KkDd) sarcoma was transfected with the allogeneic MHC class I H-2Kb gene, and expressed high levels of H-2Kb antigen. Contrary to expectations, the tumor cells expressing the alloantigen (SKB3.1M tumor cells) were not rejected by autologous A/J mice. Because these results contradict the laws of transplantation immunology, the present studies were undertaken to examine the immunogenicity of SKB3.1M and SaI cells in allogeneic hosts. Similar to SKB3.1M, SaI cells are lethal in some allogeneic strains, despite tumor-host MHC class I incompatibilities. Tumor challenges of SKB3.1M and SaI cells, however induce MHC class I-specific antibodies and CTL in both tumor-resistant and -susceptible hosts. Although the tumors induce specific CTL, tumor cells are not lysed in vitro by these CTL, suggesting that the tumor cells are resistant to CTL-mediated lysis. Since growth of these tumors does not follow the classical rules of allograft transplantation, and because the tumor is not susceptible to CTL-mediated lysis, we have used Winn assays to identify the effector lymphocyte(s) responsible for SaI rejection. Depletion studies demonstrate that the effector cell is a CD4-CD8+ T lymphocyte. Collectively these studies suggest that the host's response to MHC class I alloantigens of SKB3.1M and SaI cells does not determine tumor rejection, and that effector cells other than classically defined CTL, but with the CD4-CD8+ phenotype, can mediate tumor-specific immunity.

Tumor-specific immunity can be enhanced by transfection of tumor cells with syngeneic MHC-class-II genes or allogeneic MHC-class-I genes.

Mouse Sal sarcoma cells are lethal in the autologous A/J (KkDd) host. In order to improve the immune response to the Sal tumor, Sal cells have been transfected with syngeneic MHC-class-II or allogeneic MHC-class-I genes. MHC-class-II transfectants are uniformly rejected by the autologous host and immunization with them protects against subsequent Sal challenge. The improved immunity is probably the result of enhanced generation of tumor-specific Th cells. We hypothesize that class-II tumor cells trigger an improved Th-cell response because they directly present Sal tumor antigens in the context of class-II molecules to Th cells, by-passing professional APC. Studies by others have demonstrated that antigen presentation requires an intracellular signal transmitted by the cytoplasmic domain of the APC class-II molecule. Sal cells expressing class-II antigens with truncated cytoplasmic domains are as malignant as wild-type Sal cells. These experiments therefore support the role of tumor-cell class-II molecules as antigen presentation elements, and demonstrate the requirement for intact class-II molecules for tumor protection. Sal cells have also been transfected with allogeneic MHC-class-I genes. Although Kb-transfected cells are not rejected by A/J mice, Db-transfected Sal cells and Kb- plus Db-transfected cells are rejected. The Db transfectants effectively immunize A/J mice against subsequent Sal challenge. These experiments demonstrate that expression of certain allogeneic MHC-class-I genes can lead to tumor-specific immunity, and that such transfectants can protect against challenges of wild-type tumor cells. Transfection of tumor cells with syngeneic MHC-class-II or allogeneic MHC-class-I genes may therefore be a potential strategy for improving tumor-specific immunity in the autologous host.

In vitro inhibition of HIV-1 infectivity by human salivas.

Inhibitory factors to human immunodeficiency virus type 1 (HIV-1) in saliva may be responsible for the infrequent isolation of virus from saliva and also may account for the marked infrequency of salivary and/or oral transmission of HIV-1. Incubation of HIV-1 with human saliva followed by addition of the mixture to susceptible cells leads to partial or complete suppression of viral replication in vitro. We investigated the inhibitory effects of whole saliva and specific glandular salivas on HIV-1 infectivity as measured by viral-induced cytopathic effects in susceptible cells. Whole saliva contained marked inhibitory activity to HIV-1, strain HTLV-IIIB, and to virus infected cells. Submandibular saliva contained inhibitory activity, but of lesser quantity. Parotid saliva demonstrated no HIV-inhibitory activity. Whole saliva also appeared to contain filterable components that were inhibitory to lymphocyte growth. Passage through a .45 micron pore-size filter eliminated the viral inhibitory activity of submandibular saliva and some of the activity in whole saliva. All salivas except parotid incubated with HIV-1 followed by filtration were inhibitory suggesting that complexing of virus with high molecular weight, submandibular mucins may play a role in viral inhibition.

Transfection and expression of syngeneic H-2 genes does not reduce malignancy of H-2 negative teratocarcinoma cells in the autologous host.

Rejection of the MHC class I negative 402AX teratocarcinoma is accompanied by induction of tumor cell-encoded H-2K and H-2D antigens by the genetically resistant host. To determine whether MHC antigen expression is required for 402AX rejection, we have prepared H-2Db-transfected 402AX cells (402AX/Db). Transfectants express high levels of H-2Db, most of which is not associated with beta 2-microglobulin. MHC syngeneic and allogeneic mice susceptible to 402AX are resistant to 402AX/Db, suggesting that MHC class I antigen expression is required for tumor rejection. Autologous 129 hosts, however, are susceptible to 402AX/Db. 402AX cells transfected with the H-2Kb gene (402AX/Kb) are also lethal in the autologous 129/J host, but rejected by MHC syngeneic and allogeneic mice. Non-129 strain 402AX-susceptible mice pre-immunized with 402AX/Db or simultaneously challenged with 402AX/Db plus 402AX are immune to 402AX. Mice immunized with 402AX/Db produce MHC class I induction factor. 402AX/Db and 402AX cells are lysed equally by natural killer cells, indicating that in 402AX cells the expression of class I antigens is unrelated to NK susceptibility. These studies confirm the requirement for class I expression in 402AX immunity, but demonstrate that in the autologous host immunity requires additional factors beyond class I antigen expression.

Transfection of major histocompatibility complex class I and class II genes causes tumour rejection.

Many human and mouse tumours do not express MHC class II antigens and have reduced levels of class I antigens. Because of the requirement for class I and/or class II antigen for antigen presentation to Th and Tc cells, these phenotypes may enable tumour cells to 'escape' the host's immune response. Experiments presented here are designed to assess the role of MHC class I and class II antigens in tumour immunity, and to overcome the MHC class I- or class II-negative phenotype. When transfected with the syngeneic H-2Db gene, the MHC antigen-negative 402AX teratocarcinoma expresses high levels of H-2Db antigen. 402AX/Db cells are rejected by MHC allogeneic and some MHC syngeneic 402AX-susceptible mice, however the fully syngeneic strain of origin (129) remains tumour-susceptible. Induction of MHC class I gene products on class I antigen-negative embryonal carcinoma cells therefore increases tumour immunogenicity in some hosts, but not in the fully syngeneic mouse. In an attempt to enhance antigen presentation of tumour-associated antigens to Th cells, MHC class I antigen-positive SaI (KkDd) sarcoma cells were transfected with syngeneic A alpha k and A beta k genes to generate Iak-expressing tumour cells. SaI/Ak cells are efficiently rejected by syngeneic A/J (KkDd) mice, while untransfected SaI cells are lethal. Induction of MHC class II antigen expression on the class I antigen-positive SaI sarcoma therefore completely abrogates malignancy.

Developmental profile of chromogranin, hormonal peptides, and 5-hydroxytryptamine in gastrointestinal endocrine cells.

In this immunocytochemical study, we have analyzed the developmental profile and phenotypic expression of the endocrine cell antigens chromogranin, 5-hydroxytryptamine, gastrin/cholecystokinin, cholecystokinin (9-20), somatostatin, somatostatin 28 (1-14), somatostatin cryptic peptide, glucagon, glucagonlike peptides 1 and 2, glicentin, peptide YY, glucose-dependent insulinotropic peptide, secretin, neurotensin, and substance P in human fetal stomach and intestine. All currently identifiable endocrine cell types were detected by 10 wk of gestation. Immunostaining for the endocrine cell marker chromogranin revealed abundant endocrine cells in the earliest specimens (8 wk of gestation) with a relatively higher frequency in both proximal duodenum and distal colon/rectum compared with other areas. Quantification of endocrine cells showed an increase with age that was roughly parallel to the growth of the gut as a whole. These studies show that the diversity of the endocrine component of the gut appears to be established by 10 wk of gestation and that gut activity is preceded by the development of a fully differentiated endocrine component, which may subserve or even initiate the onset of functional maturity.

Stage-selective inhibition of rodent malaria by cyclosporine.

The relative susceptibility of different developmental stages of Plasmodium berghei to cyclosporine was investigated in vivo. Within 12 h of receiving a single 25-mg/kg (body weight) dose of cyclosporine, mice with patent P. berghei infections uniformly exhibited a rapid fall in asexual parasite stages. Initially, ring forms and mature schizonts disappeared. Subsequently, trophozoites disappeared between 21 and 24 h, whereas gametocytes persisted for 36 h. In contrast, when cyclosporine was administered to mice 1 day before inoculation (100 mg/kg) with P. berghei sporozoites and for 2 consecutive days after inoculation (25 mg/kg), infections developed normally. When mice with patent infections were placed on prolonged cyclosporine therapy (25 mg/kg per day), parasitemia initially disappeared but often recrudesced. Recrudescent parasites were frequently resistant to cyclosporine (Csr). The Csr phenotype remained stable after serial passage of parasites in mice and after transmission through Anopheles stephensi mosquitoes, in which the capacity to produce oocysts was reduced. When infections of untreated mice were initiated with equal numbers of Csr and cyclosporine-susceptible (Css) parasites and then carried through two serial cycles of mosquito-to-mouse transmission without cyclosporine treatment, the Csr phenotype was lost. The results indicate that cyclosporine selectively inhibits asexual blood stages of P. berghei and favors the emergence of Csr parasites with diminished infectivity for mosquitoes.