CXCR3/CCR5 pathways in metastatic melanoma patients treated with adoptive therapy and interleukin-2.

BACKGROUND: Adoptive therapy with tumour-infiltrating lymphocytes (TILs) induces durable complete responses (CR) in ∼20% of patients with metastatic melanoma. The recruitment of T cells through CXCR3/CCR5 chemokine ligands is critical for immune-mediated rejection. We postulated that polymorphisms and/or expression of CXCR3/CCR5 in TILs and the expression of their ligands in tumour influence the migration of TILs to tumours and tumour regression. METHODS: Tumour-infiltrating lymphocytes from 142 metastatic melanoma patients enrolled in adoptive therapy trials were genotyped for CXCR3 rs2280964 and CCR5-Δ32 deletion, which encodes a protein not expressed on the cell surface. Expression of CXCR3/CCR5 in TILs and CXCR3/CCR5 and ligand genes in 113 available parental tumours was also assessed. Tumour-infiltrating lymphocyte data were validated by flow cytometry (N=50). RESULTS: The full gene expression/polymorphism model, which includes CXCR3 and CCR5 expression data, CCR5-Δ32 polymorphism data and their interaction, was significantly associated with both CR and overall response (OR; P=0.0009, and P=0.007, respectively). More in detail, the predicted underexpression of both CXCR3 and CCR5 according to gene expression and polymorphism data (protein prediction model, PPM) was associated with response to therapy (odds ratio=6.16 and 2.32, for CR and OR, respectively). Flow cytometric analysis confirmed the PPM. Coordinate upregulation of CXCL9, CXCL10, CXCL11, and CCL5 in pretreatment tumour biopsies was associated with OR. CONCLUSION: Coordinate overexpression of CXCL9, CXCL10, CXCL11, and CCL5 in pretreatment tumours was associated with responsiveness to treatment. Conversely, CCR5-Δ32 polymorphism and CXCR3/CCR5 underexpression influence downregulation of the corresponding receptors in TILs and were associated with likelihood and degree of response.

Biased epitope selection by recombinant vaccinia-virus (rVV)-infected mature or immature dendritic cells.

Recombinant expression vectors represent a powerful way to deliver whole antigens (Ags) for immunization. Sustained Ag expression in vector-infected dendritic cells (DC) combines Ag-specific stimulation with powerful costimulation and, simultaneously, through 'self-selection' of ad hoc epitopes broadens the scope of immunization beyond restrictions posed by individual patients' human leukocyte antigen (HLA) phenotype. In this study, therefore, we evaluated the efficiency of a recombinant vaccinia virus encoding the gp100/PMel17 melanoma Ag (rVV-gp100) to infect immature (iDC) or mature dendritic cells (mDC) derived from circulating mononuclear cells and the effect of infection on their status of maturation. In addition, we tested the ability of rVV-gp100-infected iDC and mDC to present the HLA-A*0201-associated gp100:209-217 epitope (g209). Irrespective of status of maturation, rVV-gp100 infection induced gp100 expression while only partially reversing the expression of some maturation markers. However, endogenous presentation of the wild-type g209 epitope was inefficient. The low efficiency was epitope-specific since infection of DC with rVV encoding a gp100 construct containing the modified gp100:209-217 (210M) (g209-2M) epitope characterized by high binding affinity for HLA-A*0201 restored efficient Ag presentation. Presentation of an HLA-class II-associated epitope and cytokine release by DC was not altered by rVV infection. Thus, Ag expression driven by rVV may be an efficient strategy for whole Ag delivery. However, since the effectiveness of Ag processing and presentation is subject to stringent HLA/epitope pairing, and for other yet undefined rules, the assumption that whole Ag delivery may circumvent HLA restriction is incorrect and recombinant expression vectors encoding well-characterized polyepitopic constructs may prove more effective.

Efficient chromosomal mapping of a methylcholanthrene-induced tumor antigen by CTL immunoselection.

It has been difficult to genetically map the genes encoding tumor Ags because they arise as a consequence of somatic mutational events. CTL-mediated immunoselection can impose potent immunoselective pressure against tumor cells, resulting in the survival of rare tumor Ag-loss variants. We subjected a heterozygous 3-methylcholanthrene-induced murine sarcoma cell line to CTL immunoselection, selecting for the loss of a tumor-specific Ag, recognized antigen from MCA-induced tumor 1 (Ram1). Several variants eluded CTL recognition by genetic loss of the hemizygously expressed tumor-specific Ag epitope. A frequently observed genetic escape mechanism was spontaneous mitotic recombination resulting in loss of heterozygosity on chromosome 4. Higher density genetic analyses along with functional confirmation with an independently produced chromosome 4 loss of heterozygosity variant positioned the Ram1 locus to a distal 7.1 cM interval on chromosome 4. This region of the mouse genome is rich in tumor-modifier genes and this positioning of Ram1 may thus provide insight into the genetic basis of 3-methycholanthrene-induced tumor Ags.

Adoptive transfer of cloned melanoma-reactive T lymphocytes for the treatment of patients with metastatic melanoma.

This report describes a phase I study of the adoptive transfer of cloned melanoma antigen-specific T lymphocytes for therapy of patients with advanced melanoma. Clones were derived from peripheral blood lymphocytes or tumor-infiltrating lymphocytes of patients who had received prior immunization with the melanoma-associated antigen, gpl00. In response to its cognate antigen, each clone used for treatment secreted large amounts of interferon-gamma and granulocyte-macrophage colony-stimulating factor, lesser amounts of interleukin (IL)-2 and tumor necrosis factor-alpha, and little or no IL-4 and IL-10. Clones also demonstrated recognition of human leukocyte antigen-matched melanomas using cytokine secretion and lysis assays. Twelve patients received 2 cycles of cells alone; 11 patients received additional cycles of cells and were randomized between two schedules of IL-2 (125,000 IU/kg subcutaneously daily for 12 days versus 720,000 IU/kg intravenously every 8 h for 4 days). A total of 51 cycles of cells were administered, with an average of 1 x 10(10) cells per cycle. Peripheral blood samples were analyzed for persistence of transferred cells by T-cell receptor-specific polymerase chain reaction. Transferred cells reached a maximum level at 1 h after transfer but rapidly declined to undetectable levels by 2 weeks. One minor response and one mixed response were observed (both in the high-dose IL-2 arm). This report demonstrates the safety and feasibility of cloned T-cell transfer as a therapy for patients with cancer. The lack of clinical effectiveness of this protocol suggests that transfer of different or additional cell types or that modulation of the recipient host environment is required for successful therapy.

Kinetics of TCR use in response to repeated epitope-specific immunization.

Selection of T cell-directed immunization strategies is based extensively on discordant information derived from preclinical models. We characterized the kinetics of T cell selection in response to repeated antigenic challenge. By enumerating with epitope/HLA tetrameric complexes (tHLA) vaccine-elicited T cell precursor frequencies (Tc-pf) in melanoma patients exposed to the modified gp100 epitope gp100:209-217 (g209-2M) we observed in most patients that the Tc-pf increased with number of immunizations. One patient's kinetics were further characterized. Dissociation kinetics of g209-2M/tHLA suggested enrichment of T cell effector populations expressing TCR with progressively higher affinity. Furthermore, vaccine-elicited T cells maintained the ability to express IFN-gamma ex vivo and proliferate in vitro. Thus, repeated exposure to immunogenic peptides benefited immune competence. These results provide a rationale for immunization strategies.

TCR stimulation protects CD8+ T cells from CD95 mediated apoptosis.

Activation of T cells through the T-cell receptor (TCR) induces the expression of Fas Ligand (CD95L). In turn, CD95L binds to the Fas receptor (CD95) and rapidly induces apoptosis in cycling cells. This interaction is involved in the elimination of reactive lymphocytes during an immune response. However, TCR activation cannot always trigger apoptosis because an effective immune response would then be compromised. Here we show that a short (2 to 3 h) activation of T cells through the TCR simultaneously induces an increase in CD95L mRNA and a dramatic decrease in caspase-8 mRNA levels and proteolytic activity in human CD8(+) T cells. In addition, there is a small reduction in CD95 mRNA and CD95 levels on the cell surface. We found that preactivation of T cells protected them from apoptosis induced by either religation of the TCR or direct exposure to CD95L. These results suggest a mechanism by which cycling CD95-sensitive peripheral T cells, become protected from CD95 mediated deletion when actively engaged in the specific recognition of target cells.

T-cell clones from melanoma patients immunized against an anchor-modified gp100 peptide display discordant effector phenotypes.

PURPOSE: The modified peptide epitope gp100:209-217 (210M), referred to as g209-2M, of the gp100 melanocyte differentiation protein, when administered to melanoma patients by subcutaneous injection in incomplete Freund's adjuvant, is capable of generating HLA-A2-restricted CD8+ lymphocytes that specifically recognize the native gp100:209-217 (g209) peptide as well as gp100-expressing tumor cells. To evaluate the suitability of cloned lymphocytes from immunized patients for use in adoptive transfer therapy protocols, the functional and phenotypic variation of individual CD8+ T cell clones comprising the antitumor immune response was evaluated. METHODS: T-cell clones from melanoma patients who received g209-2M immunization were isolated and expanded, and their specific antitumor functional phenotypes were characterized. RESULTS: g209-specific CD8+ lymphocytes that specifically recognized gp100-expressing tumor cells were readily obtained from g209-2M-immunized patients. There was substantial variation in the absolute levels of cytokine secretion and target cell lysis by g209-specific clones from each patient. Furthermore, individual clones demonstrated discordant secretion of different proinflammatory cytokines. These clonal phenotypes were stable, even after large expansions in cell number. DISCUSSION: These results indicate that g209-2M peptide immunization of melanoma patients leads to a functionally diverse population of T cells, many of which are capable of expansion ex vivo to cell numbers appropriate for adoptive immunotherapy. However, the selection of a particular antigen-specific T-cell clone for treatment should be based on multiple functional criteria.

Expansion and characterization of T cells transduced with a chimeric receptor against ovarian cancer.

Adoptive immunotherapy with genetically modified T lymphocytes is being utilized in clinical trials for the treatment of a broad range of diseases including cancer and HIV infection. To improve on these treatments, and to better understand their mechanisms of action, it is necessary to develop techniques to generate large numbers of cells and characterize the functional heterogeneity of the cells produced. In this study, patient peripheral blood lymphocytes were transduced with a chimeric antigen receptor (MOv-gamma) derived from a mouse monoclonal antibody against folate-binding protein, which is overexpressed on many ovarian cancers. Thus, irrespective of their original specificity, normal human T lymphocytes were redirected to react against ovarian cancer cells. Lymphocytes from five patients were transduced and grown to large numbers, with a median expansion of more than 7000-fold. When proliferation was inadequate, the cells were expanded by stimulation utilizing anti-CD3, IL-2, and irradiated allogeneic PBMCs. The cells maintained their functional ability to recognize ovarian cancer over several months. Cloning of transduced cells was undertaken to determine the level of gene expression and function of individual cells making up the bulk population. Transduced CD4(+) and CD8(+) cell clones were isolated from the bulk and demonstrated antitumor activity. These clones had a diverse repertoire with respect to secretion of cytokines, and individual clones maintained their cytokine profile on subsequent expansion. These studies establish the feasibility of consistently generating large numbers of gene-modified tumor-reactive lymphocytes, with a stable and diverse cytokine repertoire, that could be utilized for patient treatment.

Cutting edge: CD4+ T cell control of CD8+ T cell reactivity to a model tumor antigen.

Neoantigens resulting from the inherent genomic instability of tumor cells generally do not trigger immune recognition. Similarly, transfection of tumors with model Ags often fails to elicit CD8+ T cell responses or alter a tumor's growth rate or lethality. We report here that the adoptive transfer of activated Th1-type CD4+ T cells specific for a model tumor Ag results in the de novo generation of CD8+ T cells with specificity to that Ag and concomitant tumor destruction. The anti-tumor effects of the CD4+ T cells required the presence of both MHC class I and class II on host cells, as evidenced by experiments in knockout mice, suggesting that CD4+ T cells enhanced the ability of host APC to activate endogenous CD8+ T cells. These results indicate that the apparent inability of tumor cells expressing highly immunogenic epitopes to activate tumor-specific CD8+ T cells can be altered by activated CD4+ T cells.

Antitumor immunization with a minimal peptide epitope (G9-209-2M) leads to a functionally heterogeneous CTL response.

The goal of experimental clinical protocols using peptide antigen for active vaccination and treatment of patients with metastatic cancer is to induce a vigorous cytotoxic T lymphocyte (CTL) response against the immunizing antigen, and thereby against tumor cells expressing the antigen. However, the magnitude and breadth of human CTL responses induced by peptide immunization, and in particular against antigens expressed by normal tissues as well as tumors, is not well characterized. This issue was examined by characterizing CTL cloids derived from peripheral blood mononuclear cells of three patients who received peptide immunization as treatment for metastatic melanoma. All patients received G9-209-2M peptide, a modified epitope of the gp100 melanoma-associated antigen. The results indicated that the CTL response induced by this peptide antigen was highly heterogeneous both in terms of avidity toward the peptide antigen and recognition of tumor cell lines. Furthermore, avidity of each CTL cloid for the native peptide was highly predictive of tumor reactivity. These results are discussed in terms of their implications for peptide vaccination and adoptive tumor immunotherapy.

High avidity CTLs for two self-antigens demonstrate superior in vitro and in vivo antitumor efficacy.

A majority of the human tumor-associated Ags characterized to date are derived from nonmutated "self"-proteins. Little is currently understood about the nature of the self-reactive lymphocytes that recognize these Ags. We recently characterized two nonmutated tumor-associated Ags for the B16 murine melanoma: tyrosinase-related protein-2 (TRP-2) and the endogenous retroviral envelope protein, p15E. We previously reported that both TRP-2 and p15E reactive CTL could be detected in the spleens of naive animals after a single in vitro stimulation using 10(-5)-10(-6) M of the appropriate Kb-binding 9-amino acid epitope. In this report we show that the CTL found in naive animals are low avidity lymphocytes, that respond only to high concentrations of peptide in vitro. We demonstrate that titration of in vitro-stimulating peptide to limiting concentrations distinguishes qualitative differences in the lymphocyte reactivity to these two Ags between vaccinated and unvaccinated animals. We further demonstrate that in vitro expansion of CTL in either high or low concentrations of stimulating peptide generated CTL cultures with different avidities for the relevant epitopes. CTL expanded in low concentrations demonstrated higher avidity for peptide-pulsed targets and better tumor recognition, when compared to CTL generated in the presence of high concentrations of Ag. More importantly, high avidity CTL demonstrated superior in vivo antitumor activity. These results demonstrate that qualitative differences in the CTL that recognize these two self-Ags are critically important to their in vitro and in vivo anti-tumor efficacy.

Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma.

The cloning of the genes encoding cancer antigens has opened new possibilities for the treatment of patients with cancer. In this study, immunodominant peptides from the gp100 melanoma-associated antigen were identified, and a synthetic peptide, designed to increase binding to HLA-A2 molecules, was used as a cancer vaccine to treat patients with metastatic melanoma. On the basis of immunologic assays, 91% of patients could be successfully immunized with this synthetic peptide, and 13 of 31 patients (42%) receiving the peptide vaccine plus IL-2 had objective cancer responses, and four additional patients had mixed or minor responses. Synthetic peptide vaccines based on the genes encoding cancer antigens hold promise for the development of novel cancer immunotherapies.

Loss of a unique tumor antigen by cytotoxic T lymphocyte immunoselection from a 3-methylcholanthrene-induced mouse sarcoma reveals secondary unique and shared antigens.

Most chemically induced tumors of mice express unique antigens that can be recognized by cytotoxic T lymphocytes (CTL) and thereby mediate tumor rejection. The number of different antigens expressed by a single tumor and their interplay during immunization and rejection are largely unexplored. We used CTL clones specific to individual tumor antigens to examine the number and distribution of CTL antigens expressed by cell lines derived from 3-methylcholanthrene-induced sarcomas of (C57BL/6J X SPRET/Ei)F1 mice. Each tumor cell line expressed one or more antigens that were unique, that is, not detected on cell lines from independent sarcomas. Immunoselection against an immunodominant antigen produced both major histocompatibility complex class I antigen and unique tumor antigen loss variants. Immunization of mice with antigen-negative immunoselected variants resulted in CTL that recognized additional antigens that were also expressed by the progenitor tumor. Some CTL recognized additional unique tumor antigen(s); other CTL recognized a shared antigen expressed not only by the immunizing cell line, but also by independent sarcoma cell lines and untransformed myoblastoid cell lines. CTL that recognized the shared antigen were also recovered from mice immunized in vivo with an untransformed myoblastoid cell line. These findings support a model of immunodominance among chemically induced tumor antigens in which shared antigens are masked by unique immunodominant antigens.

MHC class I-mediated antigen presentation and induction of CD8+ cytotoxic T-cell responses in autoimmune diabetes-prone NOD mice.

The common class I alleles (e.g., Kd and Db) within the H2g7 major histocompatibility complex (MHC) clearly contribute to autoimmune IDDM in NOD mice, but the mechanism by which this occurs has been controversial. One laboratory has reported that the peptide transporter encoded by the Tap1 gene within H2g7 is defective, and this contributes to IDDM by impairing MHC class I-mediated antigen presentation. If true, defective MHC class I-mediated antigen presentation should segregate with the H2g7 haplotype. NOD mice, related congenic stocks, and other control strains were used to test this hypothesis. H2g7-positive strains did not differ from those expressing other MHC haplotypes in ability to present MHC class I-restricted H3aa or H3ab minor histocompatibility (H) antigens to cytotoxic T-lymphocytes (CTL). The H2g7 haplotype was found to have a reduced capacity to mediate MHC class I-restricted presentation of the H47a minor H antigen. However, MHC class I-restricted presentation of H47a was found to be Tap independent. NOD mice and control strains also did not differ in ability to activate adenovirus-specific MHC class I restricted CTL. Thus, the H2g7 haplotype is not characterized by a Tap gene defect that only impairs the inductive phase of the immune response. In addition, MHC class I-restricted presentation of either minor H or adenoviral antigens was equivalent in male and female NOD mice. Therefore, while the class I alleles of the H2g7 haplotype exert diabetogenic functions in NOD mice, this is not elicited through a Tap gene defect. The absence of female-specific Tap gene defects also indicates this cannot account for the reduced male incidence of IDDM in some NOD mouse colonies.

Frequency and histological appearance of adenomas in multiple intestinal neoplasia mice are unaffected by severe combined immunodeficiency (scid) mutation.

Well-characterized murine mutations are powerful analytical tools for the genetic analysis of tumorigenesis. We crossed the multiple intestinal neoplasia (Min) allele of adenomatous polyposis coli (Apc), which produces a profound pre-disposition to intestinal neoplasia, with the severe combined immunodeficiency (scid) mutation, which causes defective double-strand DNA repair and severe immunodeficiency, on the common C57BL/6J genetic background to assay for any combined effect on intestinal tumorigenesis. Several phenotypic traits were exacerbated in an apparently additive manner in the double mutant mice, including reduced immunoglobulin levels, reduced body weight and increased morbidity. However, quantitation and histological evaluation of polyp phenotype indicated that these mutations did not interact to affect either polyp frequency or progression. Thus, neither genome instability nor lack of immunosurveillance conferred by scid contributes to intestinal polyps in this model.

Motif-primed polymerase chain reaction-based allelotype of sarcomas induced by 3-methylcholanthrene in interspecific hybrid mice.

Loss of heterozygosity (LOH) occurs commonly in some human tumors and is thought to reflect the selective pressure exerted by tumor suppressor genes. The purpose of this study was to examine the genomic location and relative frequency of LOH of chemically induced tumors in an established mouse model for chemical carcinogenesis. Twenty-four tumors were induced by subcutaneous injection of 3-methylcholanthrene into inter-subspecific hybrid (C57BL/6J x SPRET/Ei)F1 mice. Tumors were diagnosed as poorly differentiated, spindle cell sarcomas. DNA isolated from cell cultures derived from the sarcomas was analysed for LOH using a multi-locus genome scanning technique, motif-primed PCR, followed by conventional simple sequence length polymorphism (SSLP) PCR markers. Ninety-one motif-primed PCR products and seven SSLP markers were evaluated, covering segments of 39 out of 41 chromosomes. LOH was extremely rare; average fractional allelic loss per tumor was less than 0.01 suggesting a remarkable genetic stability of the tumors and that genetic events leading to transformation are only rarely revealed as LOH. However, some regions exhibiting LOH were identified encompassing both known and hypothesized tumor suppressor loci such as Trp53. These data indicate that motif-primed PCR is an efficient method of scanning the genome of interspecific sarcomas for rare mutational events generating LOH during tumor induction. Cell lines derived from the sarcomas promise to be useful for identifying genes whose inactivation causes increased malignancy during tumor progression.

The therapeutic efficacy of murine anti-tumor T cells: freshly isolated T cells are more therapeutic than T cells expanded in vitro.

Adoptive immunotherapy (AIT) involving transfer of tumor-sensitized T lymphocytes in combination with cyclophosphamide (CY)-injection results in the eradication of the C57BL/6J (B6) rhabdomyosarcoma, 76-9 and is associated with the accumulation of a large number of tumor-infiltrating lymphocytes (TIL). Using immune spleen cells (ISC) from B6 and congenic B6. PL. Thy-1a mice, it was shown that most (> or = 97%) of the TIL were donor-derived. This in situ increase in donor-derived T cells was confirmed by using positively-selected Thy- 1.1+ and Thy- 1.2+ TIL for AIT after isolating them from regressing tumors and expanding them in rIL-2. The extent of CD8+ TIL expansion in vivo correlated with the numbers of TIL adoptively transferred and this in turn determined the degree of anti-tumor effects. It was evident, however, that these in vitro-expanded TIL expressing mRNA for TNF alpha and IFN gamma were qualitatively different and therapeutically less efficacious than the T cells associated with ISC or with freshly-isolated TIL. Unlike freshly isolated TIL that expressed specific cytotoxicity towards the 76-9 targets in vitro, IL-2 expanded TIL killed 76-9 cells and unrelated tumor targets to the same extent. A cytotoxic CD8+ T cell line derived from ISC and selected for activity against the 76-9 tumor cells showed no therapeutic efficacy. The data suggest that, in this tumor model, expansion of CD8+ T cells in vitro selects against anti-tumor efficacy.

Gene mapping in a murine cell line by immunoselection with cytotoxic T lymphocytes.

Minor histocompatibility (H) loci encode alloantigens that are recognized by cytotoxic T (Tc) lymphocytes. A (C57BL/10 x 129)F1-derived transformed lymphocyte cell line was immunoselected in vitro with cloned Tc cells that were specific for H-3aa, a Chromosome 2-encoded minor H antigen. This cell line is heterozygous at H-3a (former symbol, Cd-1) and other loci. Three groups of antigen-loss variants were identified. One group contained mutations affecting only the antigen-encoding gene. Another group probably arose through a single homologous interchromosomal exchange, resulting in extensive regions of loss of heterozygosity (LOH). The third group of variants contained an interstitial LOH, one of which was shown to be a significant deletion. Several deletion boundaries were identified, one of which ordered the closely linked H-3a and beta 2-microglobulin (B2m) genes. We suggest that Tc immunoselection against minor H antigens is a promising approach for targeting negative selection to specified chromosomal regions and can provide high-resolution genetic map information.

A non-nodulating alfalfa mutant displays neither root hair curling nor early cell division in response to Rhizobium meliloti.

The early events in the alfalfa-Rhizobium meliloti symbiosis include deformation of epidermal root hairs and the approximately concurrent stimulation of cell dedifferentiation and cell division in the root inner cortex. These early steps have been studied previously by analysis of R. meliloti mutants. Bacterial strains mutated in nodABC, for example, fail to stimulate either root hair curling or cell division events in the plant host, whereas exopolysaccharide (exo) mutants of R. meliloti stimulate host cell division but the resulting nodules are uninfected. As a further approach to understanding early symbiotic interactions, we have investigated the phenotype of a non-nodulating alfalfa mutant, MnNC-1008 (NN) (referred to as MN-1008). Nodulating and non-nodulating plants were inoculated with wild-type R. meliloti and scored for root hair curling and cell divisions. MN-1008 was found to be defective in both responses. Mutant plants inoculated with Exo- bacteria also showed no cell division response. Therefore, the genetic function mutated in MN-1008 is required for both root hair curling and cell division, as is true for the R. meliloti nodABC genes. These observations support the model that the distinct cellular processes of root hair curling and cell division are triggered by related mechanisms or components, or are causally linked.

Microscopic studies of cell divisions induced in alfalfa roots by Rhizobium meliloti.

We have used spot-inoculation and new cytological procedures to observe the earliest events stimulated in alfalfa (Medicago sativa L.) roots by Rhizobium meliloti. Roots were inoculated with 1-10 nl of concentrated bacteria, fixed in paraformaldehyde, and after embedding and sectioning stained with a combination of acridine orange and DAPI (4'-6-diamidino-2-phenylindole hydrochloride). Normal R. meliloti provoke cell dedifferentiation and mitosis in the inner cortex of the root within 21-24 h after inoculation. This activation of root cells spreads progressively, leading to nodule formation. In contrast, the R. meliloti nodA and nodC mutants do not stimulate any activation or mitosis. Thus the primary and earliest effect of Rhizobium nod gene action is plant cellular activation. A rapid, whole-mount visualization by lactic acid shows that the pattern of nodule form varies widely. Some R. meliloti strains were found to be capable of stimulating on alfalfa roots both normal nodules and a "hybrid" structure intermediate between a nodule and a lateral root.