PR1-specific Cytotoxic Lymphocytes Research Articles

Cathepsin G Is Expressed By B Cell Acute Lymphoblastic Leukemia and Is an Effective Immunotherapeutic Target

The azurophil granule proteases neutrophil elastase (NE) and proteinase 3 (P3) have been established as effective immunotherapeutic targets in hematological malignancies. Expression of PR1, an HLA-A2 restricted nonameric peptide derived from NE and P3, has been demonstrated in acute myeloid leukemia (AML). We have successfully targeted PR1 in AML patients using PR1-peptide vaccine, anti-PR1/HLA-A2 antibody and PR1-specific cytotoxic T-lymphocytes (PR1-CTL). We have also identified the uptake and cross-presentation of NE and P3 by breast cancer, lung cancer and melanoma, which underlines their potential as immunotherapeutic targets in non-myeloid malignancies.Cathepsin G is also an azurophil granule serine protease that is highly expressed by AML blasts and leukemic stem cells. Our previous studies showed a high level of efficacy in targeting AML blasts that express CG1 (FLLPTGAEA) using CG1-specific CTL (CG1-CTL), while sparing normal hematopoietic cells. However, two previous reports (Fujiwara et al., Clinical Cancer Research, 2005 and Gorodkiewicz et al., Analytical Biochemistry, 2012) and our published data (Alatrash G, Leukemia, 2017) have reported CG expression in lymphoid leukemia. We therefore hypothesized that CG is an effective immunotherapeutic target in lymphoid malignancies, specifically acute lymphoid leukemia (ALL).We first analyzed CG expression in ALL patient samples. Using reverse phase protein array (RPPA), we analyzed CG expression in ALL blasts from 345 adult ALL patient samples. We detected CG in 345 patient samples at levels comparable to G-CSF stimulated myeloid progenitors. The differential expression of CG in sorted ALL blasts ranged from -1.8030 to 5.4390, with a median value of 0.3506. In estimating CG levels, we found that more than 15% of our ALL samples expressed Log2 concentrations of CG between 0.3 and 0.4 (Figure 1). This was comparable to what we have previously reported in AML. We then wanted to determine whether the ALL-associated CG was due to uptake or endogenous expression. We performed PCR and western blotting and confirmed CG expression in 4 of 9 and 9 of 9 patient ALL samples, respectively (Figure 2). To assess whether CG is also taken up by ALL, we examined CG uptake by the B-ALL cell lines SB, RS4, Nalm6 and HMY (B-ALL). ALL cell lines were incubated with CG protein at increasing doses and durations. Intracellular staining of CG followed by flow cytometry analysis was used to determine uptake of CG by ALL cell lines. Our data show that all of the ALL cell lines take up CG; uptake correlated directly with increasing doses and time of incubation up to 24 hours.Having characterized the expression of CG by ALL, we sought to determine its potential as an immunotherapeutic target, as we have previously shown for AML. In these experiments, CG1-specific CTLs were generated using standard dendritic cells methodology, interleukin-7 (IL-7), IL-2 and CG1-peptide pulsing. Expanded CG1-CTL were confirmed by staining cells with CG1 tetramer followed by flow cytometry analysis. We employed standard calcein-AM cytotoxicity assays to determine specific lysis. In these experiments, patient ALL sample UPN8 was pulsed with calcein-AM and then co-cultured with CG1-CTL for 4 hours. Our results indicated that UPN8 was effectively lysed by CG1-CTL, reaching a maximum specific lysis of ~30% at the highest effector:target ratio (Figure 3).Overall, our analysis confirms CG endogenous expression and uptake by ALL. Furthermore, we conclude that CG is an effective target for immunotherapy in B-ALL. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

PR1-specific cytotoxic T lymphocytes are relatively frequent in umbilical cord blood and can be effectively expanded to target myeloid leukemia

Background aimsPR1 is an HLA-A2 restricted leukemia-associated antigen derived from neutrophil elastase and proteinase 3, both of which are normally stored in the azurophil granules of myeloid cells but overexpressed in myeloid leukemic cells. PR1-specific cytotoxic lymphocytes (PR1-CTLs) have activity against primary myeloid leukemia in vitro and in vivo and thus could have great potential in the setting of adoptive cellular therapy (ACT). Adult peripheral blood–derived PR1-CTLs are infrequent but preferentially lyse myeloid leukemia cells. We sought to examine PR1-CTLs in umbilical cord blood (UCB) because UCB units provide a rapidly available cell source and a lower risk of graft-versus-host disease, even in the setting of mismatched human leukocyte antigen (HLA) loci. MethodsWe first determined the frequency of PR1-CTLs in HLA-A2+ UCB units and then successfully expanded them ex vivo using repeated stimulation with PR1 peptide-pulsed antigen-presenting cells (APCs). After expansion, we assessed the PR1-CTL phenotype (naive, effector, memory) and function against PR1-expressing target cells. ResultsPR1-CTLs are detected at an average frequency of 0.14% within the CD8+ population of fresh UCB units, which is 45 times higher than in healthy adult peripheral blood. UCB PR1-CTLs are phenotypically naive, consistent with the UCB CD8+ population as a whole. In addition, the cells can be expanded by stimulation with PR1 peptide-pulsed APCs. Expansion results in an increased frequency of PR1-CTLs, up to 4.56%, with an average 20-fold increase in total number. After expansion, UCB PR1-CTLs express markers consistent with effector memory T cells. Expanded UCB PR1-CTLs are functional in vitro as they are able to produce cytokines and lyse PR1-expressing leukemia cell lines. ConclusionsThis study is the first report to show that T cells specific for a leukemia-associated antigen are found at a significantly higher frequency in UCB than adult blood. Our results also demonstrate specific cytotoxicity of expanded UCB-derived PR1-CTLs against PR1-expressing targets. Together, our data suggest that UCB PR1-CTLs could be useful to prevent or treat leukemia relapse in myeloid leukemia patients.

Research on PR1 specific cytotoxic T lymphocytes of HLA-A0201 positive chronic myelogenous leukemia

Objective To evaluate the relationship between proportion of PR1 specific cytotoxic T lymphocytes (CTLs) in the peripheral blood and prognosis and curative effect in patients with HLA-A0201 positive chronic myelogenous leukemia (CML), and to discuss whether PR1 peptide could be used as the following immune therapeutic method for patients who had achieved the standard of stop treatment. Methods The soluble HLA-A0201/PR1 tetramer and flow cytometry were applied to determine the proportion and the frequency of PR1 specific CTLs in peripheral blood from 28 HLA-A0201 positive CML patients. The proportions were compared among different phases of patients. The correlations between the proportion of PR1 specific CTLs and clinical parameters were analyzed. Results There was a negative correlation between PR1 specific CTLs and PCR (bcr-abl/abl)IS (r =-0.658, P 0.05). For patients treated with IM 400 mg qd, the frequency of PR1 specific CTLs in high-risk group was lower than that in low-risk or intermediate-risk groups. Conclusion PR1 specific CTLs can be detected in patients who achieved good curative effect, and is correlated with tumor burden, which indicates that PR1 specific CTLs may be related to the action of resisting leukemia and provide the evidence for PR1 peptide as a potential immune therapeutic schedule in patients who have achieved stable MR4.5 and MR5.0. Key words: Leukemia, myelogenous, chronic; PR1; Cytotoxic T lymphocytes; Tumor immunity

PR1 Is Cross-Presented By Multiple Myeloma Cells in Patients and Renders Multiple Myeloma Susceptible to PR1-CTL and Anti-PR1/HLA-A2 Antibody

PR1 is an HLA-A2-retricted, nonameric peptide that is derived from the azurophil granule proteases neutrophil elastase (NE) and proteinase 3 (P3). PR1 has been targeted successfully in acute (AML) and chronic (CML) myeloid leukemia using anti-PR1/HLA-A2 antibody (8F4), PR1-peptide vaccine and PR1-specific cytotoxic T lymphocytes (PR1-CTL). We have previously reported that NE and P3 are cross-presented by normal B cells and dendritic cells (DC), leading to PR1 expression by HLA-A2. Since multiple myeloma (MM) is a B cell malignancy, we investigated whether MM cells can cross-present PR1 as a possible target for immunotherapy.To study whether PR1 is presented by MM cells, patient bone marrow (BM) was stained with 8F4 antibody and then imaged using confocal microscopy. PR1/HLA-A2 was detected on the surface of CD138+ MM cells in BM samples from 3 of 6 HLA-A2+ MM patients (Fig. 1). We then investigated whether cellular immunity to PR1 is detected in peripheral blood (PB) from patients with MM. PB samples from MM patients who had undergone allogeneic (allo) (n=9) and autologous (auto) (n=2) stem cell transplantation (SCT) were stained with PR1/HLA-A2 dextramer in addition to standard lineage markers. PR1-specific CD8+CTL were detected in PB of 10 of 11 patients (range=0.02%-2.9%).Because P3 and NE expression is limited to myeloid cells, we sought to determine the mechanism of PR1 presentation in MM. We performed RT-PCR and western blotting on seven MM cell lines, including U266, ARK, ARP-1, OPM-2, LP-1, IM-9 and RPMI 8226. Neither NE nor P3 were detected in the MM cell lines studied at either the mRNA or proteins levels. We then investigated whether MM cells took up NE and P3. We cultured MM cells for 30 hours with 10 ug/mL of soluble NE and P3 or irradiated HLA-A2 negative PMN, the latter as a source for cell-associated NE and P3. Cells were then stained intracellularly for NE and P3 at different time points. Flow cytometry analysis showed that all the cell lines analyzed took up NE and P3. Uptake was seen as early as 1 hour after co-culture with soluble NE and P3 and was higher for soluble P3. Additionally, more uptake was seen in the cells that were co-cultured with irradiated PMN in comparison with soluble NE and P3.We then investigated whether PR1 expression in MM was through NE and P3 cross-presentation. We focused our studies on the HLA-A2+ U266 MM cell line. U266 cells were co-cultured with soluble NE, P3 or irradiated PMN, as described in the previous section, and then surface stained with 8F4. We detected PR1/HLA-A2 on the surface of U266 cells by flow cytometry as early as 6 and 24 hours after co-culture with soluble NE and P3, respectively. Cross-presentation was also seen in the cells that were co-cultured with irradiated PMN to a similar extent in comparison with the cells that were cultured with soluble NE and P3, however, cross-presentation occurred at an earlier time point (1 hour) in the cells that were cultured with PMNs. Cross-presentation was abrogated by lactacystin, a proteasome inhibitor, and brefeldin A, an ER/Golgi transport inhibitor, indicating that NE and P3 cross-presentation occurs through conventional cross-presentation mechanisms. Furthermore, because immune modulatory drugs and proteasome inhibitors are part of the standard of care therapy for MM, and since both have been shown to affect cross-presentation by DC, we tested whether lenalidomide and bortezomib altered PR1 cross-presentation by U266 and normal DC. U266 and DC were cultured with irradiated PMN in the presence of increasing doses of bortezomib or lenalidomide and then stained with PR1/HLA-A2. Flow-cytometry analysis showed a significant inhibition of PR1-cross-presentation by U266 after addition of bortezomib, but not lenalidomide. PR1 cross-presentation by DC was not affected by either drug.Finally, we tested whether PR1 cross-presentation caused MM cells to become susceptible to PR1-targeting immunotherapies. U266 cells were cultured with NE or P3 for 24 hours, the time point that showed the maximal cross-presentation, followed by addition of 8F4 antibody or co-culture with PR1-CTL. Using calcein AM cytotoxicity assays, we showed dose dependent killing of U266 by 8F4 and PR1-CTL following PR1 cross-presentation (Fig. 2).Together our data show that PR1 is cross-presented by primary MM cells and cell lines. These findings lay the foundation for the future applications of PR1-targeting immunotherapies in MM. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Interferon-Alpha (IFN-a) Is Able To Maintain Complete Molecular Response (CMR) In Chronic Myeloid Leukemia (CML) Patients Who Maintain Undetectable disease After Imatinib (IM) Discontinuation

Cytotoxic lymphocytes (CTLs) which recognize distinct antigenic peptides on CML cells, contributed greatly to the eradication of leukemic cells after allografting. Previously, BCR-ABL specific peptides and the PR1 non-apeptide, were found to elicit a CML-specific CTL response. PR1 is a peptide derived from myeloblastin (MBN), also known as proteinase 3. MBN is abundantly expressed in azurophil granules of normal myeloid cells and is substantially overexpressed in certain immature myeloid leukemia cells where it may be important for the maintenance of a leukemic phenotype. PR1-specific CTLs (PR1-CTLs) lyses and inhibits the proliferation of CML cells but not of normal myeloid precursors. The association between the emergence of PR1-CTLs and the response to IFN-a in vivo strongly implied that IFN-a can induce remissions via induction of a CML-specific PR1-CTLs response. Therefore IFN-a, given to CML patient with undetectable disease achieved with IM, might enable its discontinuation.Ten consecutive CML patients, who were in stable CMR for at least one year after IM therapy, discontinued IM and received IFN-a. This drug was given at the dose of 3M/U/m2 weekly. The median duration of IM therapy before discontinuation was 60 (range, 31-76) months. At a median follow-up of 40 months (range, 26-58 months), five patients still have undetectable level of BCR-ABL transcript. Five patients relapsed at a median of 8 months (range, 3 – 11 months) but all except one, achieved a new CMR after reintroduction of IM. Grade 2 intolerance (muscle pain, fever) to IFN-a developed in 6/10 patients. The details of patients characteristics will be discussed during the presentation. These preliminary results demonstrate that treatment with low dose IFN-a offered to CML patients with undetectable disease is feasible and this approach may become an attractive alternative to TKI discontinuation alone. Disclosures:No relevant conflicts of interest to declare.

Non-Expanded PR1-Specific CTL Sort-Purified Directly From Cord Blood Lymphocytes Deplete Human AML In Vivo

Relapse remains a significant problem following allogeneic stem cell transplantation (SCT) for AML. Adoptive transfer of leukemia-specific cytotoxic T lymphocytes (CTL), such as PR1-specific CTL, might be used to treat persistent leukemia after SCT by enhancing graft versus leukemia (GVL) while minimizing graft versus host disease (GVHD). A limitation of this approach is the limited persistence of adoptively transferred T cells in the recipient, due in part to the lengthy ex vivo expansion of low frequency cells necessary to obtain a sufficient cell number.

Adoptive transfer of PR1 cytotoxic T lymphocytes associated with reduced leukemia burden in a mouse acute myeloid leukemia xenograft model.

Tumor antigen-specific cytotoxic T lymphocytes (CTL) have been used in the treatment of human cancer, including leukemia. Several studies have established PR1 peptide, an HLA-A2.1-restricted peptide derived from proteinase 3 (P3), as a human leukemia-associated antigen. PR1-specific CTL elicited in vitro from healthy donors have been shown to lyse P3-expressing AML cells from patients. We investigated whether PR1-CTL can be adoptively transferred into NOD/SCID mice to eliminate human leukemia cells. PR1-CTL were generated in bulk culture from peripheral blood mononuclear cells (PBMC) stimulated with autologous dendritic cells. Human acute myeloid leukemia (AML) patient samples were injected and engrafted in murine bone marrow at 2 weeks post-transfer. Following adoptive transfer, bone marrow aspirate from mice that received AML alone had 72-88% blasts in a hypercellular marrow, whereas mice that received AML plus PR1-CTL co-infusion had normal hematopoietic elements and only 3-18% blasts in a hypocellular marrow. The PR1-CTL persisted in the bone marrow and liver and maintained a CD45RA⁻CD28+ effector phenotype. We found that adoptive transfer of PR1-CTL generated in vitro is associated with reduced AML cells in NOD/SCID mice. PR1-CTL can migrate to the sites of disease and maintain their capacity to kill the AML cells. The surface phenotype of PR1-CTL was consistent with their trafficking pattern in both vascular and end-organ tissues.

Peripheral Tolerance to PR1 Is Maintained After Uptake of Soluble Proteinas-3 and Neutrophil Elastase

AbstractAbstract 4301The leukemia associated antigens (LAA) proteinase-3 (P3) and neutrophil elastase (NE) are serine proteases found within neutrophil azurophil granules and are aberrantly expressed by leukemia. The human leukocyte antigen (HLA)-A2 restricted nonomeric peptide PR1, derived from P3 and NE, has been detected on leukemia cell surface in association with HLA-A2. PR1-specific cytotoxic T lymphocytes (CTL) were detected in patients following hematopoietic stem cell transplantation and were associated with clinical remission. Furthermore, in a phase I/II clinical trial, PR1-vaccine showed immunologic and clinical responses in patients with MDS and AML. Antigen cross presentation is a mechanism by which exogenous antigens are taken up by antigen presenting cells (APC) and presented in association with major histocompatibility (MHC)-I molecule. Cross presentation is thought to be the primary mechanism by which tumor antigens are presented to elicit tumor specific immune responses. Cross presentation by dendritic cells (DC), which express co-stimulatory molecules, leads to immune priming, while antigen presentation by B cells is associated with immune tolerance. Since cross presentation of leukemia antigens has been associated with anti-leukemia immune responses, we investigated whether P3 and NE are cross presented by DCs and B cells, and whether the source of P3 and NE (soluble vs. leukemia-cell associated) determines priming of the anti-PR1 immune response. We have previously reported that serum P3 levels are significantly higher in leukemia patients, and we now also report higher serum NE levels in patients with AML (median±standard deviation (SD)= 1236 ±572 ng/ml; n=15) and CML (median ± SD = 941 ± 494 ng/ml; n=15), in comparison with healthy donor blood (median ± SD = 65 ± 177 ng/ml; n=5) (p<0.05). Using flow cytometry and confocal microscopy, we demonstrate uptake of soluble P3 and NE into lysosomes by B cells as well as normal and leukemia-derived DCs. Using western blots (WB), we show that leukemia-cell associated P3 and NE are found in nuclear, membrane, cytosolic and skeletal compartments, unlike neutrophils where they are primarily limited to the granule-containing compartment. Moreover, immunoprecipitation and WB assays showed that intracellular P3 and NE are ubiquitinated, suggesting that they are processed by proteasomes for subsequent PR1 presentation. Using BrDU uptake assay, we show that cross presentation of PR1 by HMY.CIR-A2 APC cell line leads to a 2-fold increase in proliferation of in vitro generated PR1-specific CTLs through a proteasomal dependent pathway. Using the monoclonal antibody 8F4 (anti-PR1-HLA-A2), we show that PR1 is cross presented by normal B cells from soluble P3 and NE (1.6 fold increase in PR1 expression vs. unpulsed B cells). Furthermore, we show that B cells cross present PR1 from cell-associated NE and P3 from irradiated HLA-A2+ and HLA-A2- primary leukemia and leukemia cell lines (1.7 fold increase in PR1 expression vs. unpulsed B cells). In contrast to B cells, normal DCs cross presented cell-associated, but not soluble, P3 and NE (1.6-2.0 fold increase in PR1 expression vs. unpulsed or soluble NE/P3 pulsed DCs). Furthermore, cell-associated NE and P3 increased the co-stimulatory receptors CD83, CD86, and HLA-DR in normal DCs, while soluble NE and P3 did not affect, and in some cases decreased, expression of co-stimulatory DC receptors. In conclusion, this study shows that 1) soluble and cell-associated P3 and NE are cross presented by B cells, 2) only cell-associated P3 and NE are cross presented by DCs, and 3) cell-associated, but not soluble, P3 and NE increased co-stimulatory receptors in DCs, all suggesting that cross presentation of soluble P3 and NE leads to tolerance to PR1. Since serum P3 and NE are increased in leukemia, this may be a mechanism by which leukemia evades the anti-PR1 immune response. Our findings may also be applicable not only to the anti-leukemia immune response, but also to autoimmune diseases where serum P3 and NE were shown to be increased and to inflammatory states that are common to many cancers. Disclosures:No relevant conflicts of interest to declare.

Anti-PR1/HLA-A2 Antibody (8F4, A TCR-Mimic) Mediates Specific Lysis of AML Stem Cells but Not Normal Hematopoietic Stem Cells.

Abstract 15PR1 (VLQELNVTV) is an HLA-A2-restricted peptide derived from endogenous myeloid leukemia-associated antigens (LAA) proteinase 3 and neutrophil elastase. PR1/HLA-A2 is targeted by PR1-specific cytotoxic T lymphocytes (PR1-CTL) that contribute to cytogenetic remission in patients with CML after allogeneic stem cell transplantation or interferon therapy, and PR1 peptide vaccination induces specific CD8 immunity in AML, MDS, and CML patients. Because clinical effects after PR1 vaccination are associated with a low amount of leukemia and because the effects on leukemia stem cells (LSCs) are unknown, we sought to develop a PR1/HLA-A2-specific MAB that could be more effective in high disease burden cases and that might also target LSC. We previously reported a mouse IgG2A antibody (8F4) with PR1/HLA-A2 specificity. We now used 8F4 to study PR1 expression by FACS and confocal mircroscopy on leukemia subsets and normal peripheral blood mononuclear cells (PBMC). Circulating blasts from peripheral blood (PB) or leukapheresis products (LP) from AML patients (5 HLA-A2+, 2 control HLA-A2-,) were studied (see Table 1). PR1 expression (MFI±SEM) was higher in AML (23.7±5.18) compared to control (4.0±2.42) and normal PBMC (13.6±0.23; p = 0.046). Confocal microscopy confirmed bright heterogeneous surface expression of PR1 on AML, but was nearly absent on healthy donor PB. Interestingly, PR1 expression was higher on LSC (CD34+CD38-lin-) (76±44) compared to healthy donor HSC (CD34+CD38-lin-) (13.4), which suggests that only differentiated myeloid cells might be more susceptible to 8F4 toxicity. To study whether 8F4 mediated specific lysis of PR1+ cells, we determined cell-mediated cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC) of 8F4 compared to isotype control. 8F4 induced 51% specific lysis by CDC and 15% specific lysis by ADCC of PR1-pulsed T2 cells, which was dose-dependent on both 8F4 and PR1, with no lysis of target cells by isotype control. Moreover, 8F4 induced CDC-mediated lysis of blasts from 5/5 HLA-A2+ AML patients, but not blasts from an HLA-A2- AML patient (see Table 1), or mononuclear cells from HLA-A2+ healthy donors. To study the in vivo effects of 8F4 on AML, we used an established xenogeneic model of human AML in NOD/SCID mice. 1 × 106 blasts from 4 HLA-A2+ AML patients (including the HLA-A2+ patients susceptible to 8F4-mediated lysis in vitro) and 1 HLA-A2- patient were mixed for one hour under serum-free conditions with 8F4 or isotype control antibody and adoptively transferred to NOD/SCID mice. Engraftment of AML was compared on day 14 and AML growth was confirmed on day 28 by comparing FACS analysis and histopathology of BM from sacrificed animals. 8F4 treatment of AML cells, but not isotype treatment, abrogated engraftment of 4/4 HLA-A2+ AML patients, but not of 1 HLA-A2- AML patient on d14. By day 28, AML expanded by 150% in one surviving isotype-treated animal. Thus, 8F4-mediated specific lysis of AML and LSC by CDC and ADCC mechanisms in vitro were confirmed by in vivo results of AML engraftment. Taken together, these results show that 8F4, a novel TCR-mimic antibody that targets only PR1/HLA-A2+ cells, mediates lysis of AML blasts and LSC, but not total normal BM cells or HSC. Therefore, 8F4 may be useful as an immunotherapeutic agent in the treatment of HLA-A2+ myeloid leukemia, and as the first antibody therapy that may target and eliminate LSC.Table 1Patients' characteristics and 8F4 effectsUPNDiagnosis+Disease statusHLA-A2SexAgeSample*Blast8F4 FCMIn vitro CDC lysisIn vivo Engraftmentiso8F4AML1AML-M1 secondary MyelofibrosisUntreated+M79LP87%++++−AML2AML-M2Relapsed+M61PB66%++++Low*−AML3AML-M5Refractory+F73PB64%+++++−AML4AML-M1Untreated+F55LP92%+++N/DN/DAML5AML-M2Refractory relapsed+M33PB98%N/DN/D+−AML6AML-M5Untreated+M53LP96%++N/DN/DN/DAML7AML-M1Untreated+M24LP85%++−−AML8AML-M1Untreated−F41PB91%−−LowLow+French-American-British AML subtypeFCM flowcytometry staining intensity (+ indicates staining intensity close to those of normal PBMC, ++ and +++ indicate increased intensity)N/D The experiment was not performed due to low viability of sample (less than 50%) or sample was not availableLow* - indicates engraftment <0.5% of viable mCD45- cells, effect of treatment is inaccessible.*Peripheral blood (PB) or leukopheresis (LP) Disclosures:No relevant conflicts of interest to declare.

Leukemia-Associated Primary Granule Proteins (PGPs) Elastase-2 and Proteinase-3 Are Aberrantly Expressed in Solid Tumors: A Potential Therapeutic Target for PR1-Directed Immunotherapy

Neutrophil elastase (ELA2) and proteinase-3 (PRTN3) are the source proteins for the nonomeric HLA-A2-restricted peptide PR1 (VLQELNVTV), a recognized leukemia-associated antigen (LAA). In myeloid leukemia, high expression of ELA2 and PRTN3 has been correlated with improved clinical outcomes following hematopoietic stem cell transplant (HSCT). Vaccination with PR-1 peptide showed efficacy in myeloid leukemia and myelodysplastic syndrome, and was associated with immunologic response (≥ 2 fold increase in PR1-specific cytotoxic T lymphocytes (PR1-CTL)) and long-lasting clinical remissions. PR1-CTLwere shown to mediate immunity elicited by PR-1 vaccine, as they preferentially recognize and kill myeloid leukemia cells. Since ELA2 and PRTN3 have been reported in breast cancer biopsies, we hypothesized that ELA2 and PRTN3 may be mislocalized in solid tumors, leading to MHC-I presentation, therefore providing a therapeutic target for PR1 vaccine. To study the subcellular localization of ELA2 and PRTN3, the cytoplasmic, nuclear, membrane, and golgi/ER fractions were obtained from inflammatory breast cancer (IBC) (MDA-IBC1 and SUM149) and melanoma (526, 624, 888 and 926) cell lines. ELA2 was preferentially expressed in the cytoplasmic fraction in the IBC cell line SUM149, while PRTN3 was expressed in the membrane and cytoskeletal fractions of the IBC cell line MDA-IBC1. In the melanoma cell lines, ELA2 was expressed in the cytoplasmic fractions, while PRTN3 was noted in the nuclear, cytoplasmic and cytoskeletal fractions. Furthermore, the HLA-A2+ melanoma cell line 888 was susceptible to lysis by PR1-CTL (approximately 30% lysis; Effector:Target ratio=1:2). Together, these results demonstrate that ELA2 and PRTN3 are aberrantly localized in non-hematopoietic cell lines, and that this can be associated with susceptibility to PR1- CTL lysis. In particular, PR1 is a potential immunotherapeutic target for patients with melanoma and breast cancer.

T-Cell Autoimmunity May Contribute to Neutropenia in a Patient with Cyclic Neutropenia (CN) and Double De-Novo Mutations in Gfi-1.

CN is a rare autosomal dominant disease characterized by predictable cycling of patients' neutrophils from normal levels to a nadir of ≤100 cells / μL. While heterozygous mutations in the neutrophil elastase gene (ELA2) occur in the majority of patients and altered subcellular trafficking of neutrophil elastase (NE) has also been demonstrated, it remains unknown how these defects lead to the CN phenotype. Similarly, mutations in the gene for the beta subunit of adapter protein 3 (AP3), which binds NE and shuttles it to endosomal vesicles, are present in severely neutropenic patients with Hermansky-Pudlak syndrome (type 2). Interestingly, separate single base-pair mutations in the zinc finger domains of Gfi-1, a gene that maintains the long-term potential of hematopoeitic stem cells and that negatively regulates ELA2 expression, were reported in two different patients with severe congenital neutropenia (SCN). Decreased immunity against recall antigens, and an increase in the proportion of T-cells bearing a mature memory phenotype have also been described in patients with Gfi-1 mutations. As yet, there is no unifying hypothesis linking these various abnormalities. We hypothesized that aberrantly expressed NE, resulting from mutations in ELA2 or associated regulatory genes, could predispose autologous T-cells to target myeloid precursors that aberrantly express NE. Such a reversal of immune tolerance could contribute to neutropenia in CN and SCN patients. We report in a 25 year-old man with CN from birth the simultaneous occurrence of heterozygous mutations in Gfi-1 (N382S and K403R), which were previously identified in separate patients with SCN. Neither mutation was found in this patient's parents however, and sequencing of mRNA encoding PRTN3, ELA2 and AP3 showed no mutations. Because Gfi-1 serves as a transcriptional repressor of ELA2, we hypothesized that NE may be overexpressed and / or mis-trafficked in neutrophils. To address this hypothesis, we analyzed granulocytes by FACS staining and confocal microscopy. NE was disproportionately expressed in the plasma membrane of 75% of the patient's granulocytes compared to only 3% of granulocytes from healthy donors (HD). We have previously demonstrated that cytotoxic T-lymphocytes (CTLs) that recognize proteinase 3 (P3) and NE-derived peptides, such as the HLA-A2 restricted peptide PR1 (VLQELNVTV), can kill leukemia but not normal myeloid cells due to aberrant expression of P3 and NE in the leukemia and that these PR1 specific CTLs (PR1-CTL) are not detected in healthy donors. Because this patient expressed the HLA-A2 allele, we used fluorescently labeled PR1/HLA-A2 tetramers to study whether PR1-CTL were present. Strikingly, 0.5% to 1% of total CD8+ cells from the patient were specific for PR1, as measured over 2 years, whereas no PR1-CTL were present in control HD or the patient's mother, who is also HLA-A2+. The PR1-CTLs co-expressed CCR7 and CD45, consistent with a naïve phenotype, and cytokine secretion after crosslinking the T-cell receptors was similar to controls. These data suggest that aberrant NE expression can lead to a reversal of immune tolerance and a corresponding increase in PR1-CTL. The mechanism of neutrophil cycling in CN patients is not known; however, an autoimmune response against peptides derived from aberrantly expressed granule proteins in myeloid precursors could lead to a CTL-mediated feedback loop that would also be consistent with the chalone hypothesis of feedback regulation already proposed for this disease.

Aberrant Subcellular Localization of Azurophil Granule Proteins in Myeloid Leukemia Favors Peptide Antigen Presentation on MHC-I and Susceptibility to Killing by Cytotoxic T Lymphocytes.

The recognition of leukemia-associated antigens (LAA) by immune cells is essential for generating anti-leukemia responses. Whether elicited by vaccine or hematopoietic stem cell transplantation (HSCT), anti-leukemia immunity requires that cytotoxic T lymphocytes (CTL) recognize and mount an effective response against leukemia cells. PR1 is a nonomeric HLA-A2-restricted peptide (VLQELNVTV) derived from the myeloid-restricted serine proteases neutrophil elastase (ELA2) and proteinase-3 (PRTN3), recognized azurophil granule proteins. PR1-specific cytotoxic T lymphocytes (PR1-CTL) have been shown to preferentially recognize and kill myeloid leukemia cells. We hypothesized that aberrant protein expression in target leukemia cells predisposes to CTL killing. To determine the subcellular localization of ELA2 and PRTN3 in leukemia and in healthy donor (HD) neutrophils, high-speed ultracentrifugation of leukapheresed cells from patients with AML and CML in blast crises yielded 4 fractions: cytoplasmic, nuclear, membrane and golgi/endoplasmic reticulum (ER). Western blot analyses identified higher levels of ELA2 and PRTN3 in cytoplasmic, nuclear, and golgi/ER fractions in leukemia, compared to HD granulocytes. Confocal microscopy of leukapheresed myeloid blasts and HD granulocytes confirmed these findings. We further hypothesized that cross-presentation of PR1 is necessary for priming anti-leukemia immunity. Confocal microscopy of the human B-cell line HMy.CIR-A2, co-incubated with either purified or recombinant ELA2 and PRTN3 demonstrated co-localization of both proteins with LAMP-2, a marker of lysosomes and early endosomes, and small distribution in cytoplasm after 2 hours. Uptake of ELA2 and PRTN3 by HMy.CIR-A2 cells was additionally confirmed by Western blots. Priming of naive PR1-CTL by co-culture of HD lymphocytes with PRTN3 or ELA2-transfected HMy.CIR-A2 cells was confirmed by cell proliferation assays using BrDU. This could be abrogated by addition of lactacystin and cyclohexamide to cell culture. Together, these results demonstrate that ELA2 and PRTN3 are aberrantly localized preferentially in gogli/ER and cytoplasmic fractions in leukemia, rather than azurophil granules, which are often absent or deficient in myeloid leukemia. Similar forced subcellular localization of these proteins in B cells leads to their susceptibility to PR1-CTL killing. Therefore, we conclude that the natural mistrafficking of PRTN3 and ELA2 in myeloid leukemia is necessary to render cells susceptible to CTL attack and that cross-presentation of ELA2 and PRTN3 contributes to priming of a PR1-CTL response.

PR1 Peptide Vaccine-Induced Immune Response Is Associated with Better Event-Free Survival in Patients with Myeloid Leukemia.

Background: PR1 is a nonomeric HLA-A2-restricted peptide derived from the myeloid leukemia-associated antigens, proteinase 3 and neutrophil elastase. PR1-specific cytotoxic T lymphocytes (PR1-CTL) selectively kill MDS, AML and CML and contribute to complete cytogenetic remission. We conducted a phase I/randomized phase II trial to evaluate the safety and efficacy of a PR1 peptide vaccine to elicit PR1-CTL in leukemia patients.Methods: Sixty-six HLA-A2+ patients with acute myeloid leukemia (42), chronic myeloid leukemia (13) or myelodysplastic syndrome (11) were treated with PR1 peptide vaccine. The first 54 patient received three vaccinations, and the last 12 patients received 6 vaccinations. PR1 was injected with incomplete Freud's adjuvant (Montanide ISA) subcutaneously, at 3 week intervals at one of three dose levels: 0.25, 0.5 or 1.0 mg per vaccination. GM-CSF at a dose of 75 mg was injected subcutaneously into the same site. Immune response to the vaccine was defined as a ≥ 2-fold increase in peripheral blood PR1-CTL. Fifty-three patients had measurable disease and 13 were in complete remission at enrollment.Results: The vaccine was well tolerated, with toxicity limited to grade I and II injection site reactions. PR1-specific immune response was observed in 25/53 (47%) patients with measurable disease. Clinical responses were observed in 9/25 (36%) immune responders versus 3/28 (10%) immune non-responders (p=0.03). The PR1 vaccine-induced immune response was associated with a longer event-free survival, 8.7 months vs. 2.4 months (p = 0.03), and a trend towards longer overall survival. Among the 13 patients treated in complete remission, 4 have remained in remission for a median of 30.5 months (range 11–59 months). This includes continued complete remissions in 3/10 who were IRV+, and 1/3 who were IRV-. PR1-specific immune response lasted for up to 4 years in some patients. In a multivariate model, low bone marrow blast count (<10%) emerged as a significant predictor of an immune response and a longer event-free survival (p=0.001).Conclusion: PR1 peptide vaccine-induced immune response is associated with a clinical response and better event-free survival in patients with myeloid leukemia, who had measurable disease at vaccination. Patients with ≤10% bone marrow blasts are likely to obtain the maximum benefit. [Display omitted]

Interferon-alpha (IFN-a) Given Alone to CML Patients with Undetectable Disease after Imatinib (IM) Is Able To Maintain Their Complete Molecular Remission after IM Discontinuation.

Cytotoxic lymphocytes (CTLs) which recognize distinct antigenic peptides on CML cells, contributed greatly to the cure after allografting. Previously, BCR-ABL specific peptides and the PR1 non-apeptide, were found to elicit a CML-specific CTL response (Molldrem, 1996/1997; Bocchia 1996, Clark 2001). PR1 is a peptide derived from myeloblastin (MBN), also known as proteinase 3. MBN is abundantly expressed in azurophil granules of normal myeloid cells and is substantially overexpressed in certain immature myeloid leukemia cells where it may be important for the maintenance of a leukemic phenotype. PR1-specific CTLs (PR1-CTLs) lyses and inhibits the proliferation of CML cells but not of normal myeloid precursors. The association between the emergence of PR1-CTLs and the response to IFN-a in vivo strongly implied that IFN-a can induce remissions via induction of a CML-specific PR1-CTLs response (Molldrem et al, 2000). Recent data demonstrating a failure of IM to kill cell cycle-arrested CML precursors advice us to be attentive to arising escape mechanisms of even chronic phase CML; therefore the sequential combination therapy of IFN-a given alone to CML patient with undetectable disease achieved with IM, might be more effective than any current monotherapy. We have treated until now 4 CML patients who had undetectable residual disease for at least one year after IM therapy. The patients discontinued IM and started IFN-a at the dose of 3 M/U/m2 weekly. The median duration of IM therapy was 58 (range, 31–68) months before IM interruption. At a median follow-up of 5 months (range, 3–12 months) after IFN-a, all patients still have an undetectable level of BCR-ABL transcript. The details of patients characteristics will be discussed during the presentation. These preliminary results demonstrate that treatment with low dose IFN-a in CML patients with undetectable disease is feasible and did not lead to a molecular relapse. We are starting a large prospective randomized study to evaluate the role of IFN-a employed alone after 1 year of complete molecular remission achieved with IM (Arm A) vs contination of IM alone (Arm B). The main objective of this study is to verify the real effect of the immune modulation following IFN-a in CML patients with undetectable disease.

A Mouse Model of PR1-CTL Adoptive Cell Therapy Demonstrates Rapid Reduction of Leukemia Burden in Both Bone Marrow and Extramedullary Organs.

Cytotoxic T lymphocytes (CTL) specific for a nine amino acid HLA-A2-restricted peptide, PR1 (derived from proteinase 3, P3), are capable of killing leukemia cells and contribute to the elimination of CML. PR1-specific CTL (PR1-CTL) elicited in vitro from healthy donors also can lyse P3-expressing AML blasts. Here, we examined whether PR1-CTL can be adoptively transferred to NOD-SCID/HLA-A2 transgenic mice to eliminate blasts as a preclinical model for PR1-CTL cell therapy for AML. Human AML blasts isolated from bone marrow were transferred into irradiated (200 rad) NOD-SCID/HLA-A2 mice and engrafted in bone marrow at various doses (107, 106, 105) between 2–8 weeks. Human CD45 Ab and mouse CD45.1 Ab were used to identify cells of human origin by FACS and IHC, and human CD33 and CD34 Abs were used to identify AML cells. PR1-CTL were elicited from healthy donors that specifically lysed P3-expressing leukemia. These PR1-CTL were predominantly CD45RA−/CCR7−/CD28+ and maintained an activated effector phenotype following adoptive transfer. Transferred AML blasts were significantly decreased in the bone marrow of mice receiving co-transferred PR1-CTL compared to those that received only AML: 0.5×106 AML cells plus 1×104 PR1-CTL or 0.5×106 AML cells (control) were infused into irradiated mice respectively. Mice were sacrificed at 2–4 weeks post-transfer and tissues were analyzed by FACS and IHC. Bone marrow aspirate from mice that received AML alone had 72–88% blasts in hypercellular marrow, whereas mice that received AML plus PR1-CTL had normal hematopoietic elements and only 10–11% blasts in hypocellular marrow. PR1-CTL also migrated to the secondary lymphoid organs including spleen, peripheral and mesenteric lymph nodes, and in the liver. Leukemia cells also infiltrated extramedullary organs. In the liver, intrasinuosidal and periportal collections of leukemic blasts were associated with vascular damage, and PR-CTL infiltrated these same regions by 2 weeks post-transfer. The PR1-CTL in the bone marrow and liver maintained an effector phenotype similar to the pre-infusion CTL, whereas PR1-CTL in secondary lymph organs such as blood and spleen also contained some cells expressing CCR7 that either reverted to a naïve phenotype or had expanded from the precursor population. By 4 weeks post-transfer, leukemia was no longer evident in the liver, although PR1-CTL remained. In conclusion, we found that PR1-CTL generated in vitro can eliminate the AML cells in NOD-SCID/HLA-A2 mice. PR1-CTL can migrate to sites of disease and maintain their capacity to eliminate leukemic blasts. Surface phenotype of PR1-CTL were consistent with their trafficking pattern in both vascular and end-organ tissues. Our results justify a clinical trial with adoptively transferred PR1-CTL in patients with AML.

Elevated Serum Neutrophil Elastase but Not Proteinase 3 Is Present in Patients with Myeloid Leukemia.

Proteinase 3 (P3) and neutrophil elastase (NE) are abnormally expressed in the cytoplasm of myeloid leukemia, and we have shown that the HLA-A2-restricted peptide PR1 is processed and cross-presented from both proteins by antigen-presenting cells (APC) to cytotoxic T lymphocytes (CTL). PR1-specific CTL kills leukemia cells and contributes to cytogenetic remission in patients with chronic myelogenous leukemia. Both exogenous P3 and NE can bind to monocytes and P3 promotes DC maturation in vitro. Since high levels of circulating P3 were found in patients with the ANCA-associated systemic vasculitis Wegener's granulomatosis, modulation of DC by P3 has been implicated in reversal of tolerance. The role of extracellular P3 and NE in reversing tolerance in patients with myeloid leukemia is unknown. We used an ELISA assay to assess P3 and NE level in the sera of patients with AML and CML. Serum NE level was highly increased in all 15 AML patients studied (1045±149 ng/ml, n = 15) and in all 15 CML patients (1013±128, n = 15), compared to a control group of healthy donors (54±13 ng/ml, n = 4; p = 0.0038 and p = 0.0014, respectively). In contrast, serum P3 was increased in only 5 out of 18 (28%) AML patients (47 ± 18, n = 18 ng/ml) and in 8 out of 28 (29%) CML patients (95 ± 44 ng/ml, n = 28) compared to a group of 14 healthy donors (10 ± 2.2 ng/ml; p = 0.09 and p = 0.06, respectively). There was no relationship between elevated serum NE and P3 levels (r=0.16, n=35). P3 level in untreated patients and healthy donors did not change within at least 13 months, although P3 decreased after allogeneic BMT in 5 out of 5 patients (p < 0.05) with high serum levels of P3 pre-BMT. Because P3 is expressed by myeloid progenitor cells in the bone marrow, we studied the concentration of P3 in bone marrow-derived serum (BMS). To our surprise, BMS P3 concentration in patients with myeloid leukemia (23 ± 9.5 ng/ml, n = 9) was similar to that of a control group of patients with non-myeloid hematological malignancies (27 ± 11.8, n = 5) (p = 0.08), and BMS P3 concentration was similar to sera P3 concentration in the same patient (p = 0.52). Interestingly, BMS P3 level correlated with the level of BMS GM-CSF (quantified by ELISA) in a small group of patients with AML and CML (r = 0.94, n = 9), suggesting that P3 secretion by leukemia cells may be downstream of GM-CSF. We also analyzed membrane expression of P3 and NE in 5 AML patients using flow cytometry and found that P3, but not NE, was expressed on circulating cells in 3 of 5 patients. This unique expression of P3 and NE suggests that circulating NE might be the predominant source of cross-presented PR1 antigen by APC or myeloid leukemia cells.

Antigen Cross-Presentation Allows the PR1 Leukemia-Associated Antigen To Be Processed from Both Proteinase 3 and Neutrophil Elastase to Prime T Cells.

We have shown that the HLA-A2-restricted PR1 leukemia-associated peptide antigen is processed and presented endogenously to cytotoxic T lymphocytes (CTL) from both proteinase 3 (P3) and neutrophil elastase (NE). It has recently been shown that antigen cross-presentation of exogenous proteins through the endogenous antigen presentation pathway is required to elicit naïve CTL responses, however. Professional antigen presenting cells (APC), such as dendritic cells (DC), are especially able to cross-present antigens and prime CTL immunity. Since DC express both P3 and NE, however, they could not be used to study whether PR1 can be cross-presented. Therefore, we used our P3- and NE-transfected HMy.C1R B cells (which normally do not express either P3 or NE) to determine whether PR1 could be cross-presented and its mechanism. P3- and NE-transfected HMy.C1R B cells express P3 and NE in the cytoplasm, as determined by immunofluorescence microscopy. Concentrated supernatants derived from P3- and NE-transfected B cells also contain abundant amounts of both proteins by Western blotting. To define the pathway of P3 and NE trafficking, we measured PR1-CTL proliferation after co-incubation with P3- and NE-transfected B cells that were previously treated with the proteosome inhibitor lactacystin (LAC), or the trans-Golgi network (TGN) inhibitor brefeldin A (BFA). Co-immunoprecipitation experiments were performed to further map intracellular location of the proteins. Both P3- and NE-transfected B cells treated with LAC showed a dose-dependent decrease in PR1-CTL proliferation. PR1-CTL proliferation was also abrogated by BFA. Next, we immunoprecipitated ubiquitin-conjugated proteins derived from P3- and NE-transfected whole cell lysates and immunoblotted with anti-P3 and anti-NE, respectively. By Western blot, using a blocking peptide to ubiquitin and the non-transfected B cells as controls, we found that both P3 and NE were conjugated to ubiquitin, which facilitates processing through the proteosome. To determine how P3 and NE might be shunted into the endogenous pathway, we immunoprecipitated P3- and NE-transfected whole cell lysates with antibodies to adaptor protein 3 (AP3) and the endoplasmic reticulum chaperone GRP94. We found both P3 and NE bound to both of these intracellular proteins. In contrast, U937 leukemia cells, which also express P3 and NE assessed by intracellular flow cytometry, did not secrete either P3 or NE into culture supernatants, assessed by Western blotting. Together, these results suggest that both P3 and NE, which are the source proteins of the PR1 leukemia-associated antigen, can be cross-presented to PR1-specific CTL by hematopoietic-derived APC, such as B cells, through the endogenous pathway involving ubiquitin conjugation, proteosome cleavage and ER to TGN trafficking to the cell surface to prime T cells. Leukemia cells, while susceptible to lysis by PR1-CTL, are likely incapable of priming naïve CTL since they lack exogenous production of either P3 or NE. These findings have important implications in the design of future immunotherapy strategies against leukemia.

Cytotoxic T Lymphocytes Specific for a Nonpolymorphic Proteinase 3 Peptide Preferentially Inhibit Chronic Myeloid Leukemia Colony-Forming Units

AbstractWe previously showed that a peptide (PR1) derived from the primary granule enzyme proteinase 3 induced peptide specific cytotoxic T lymphocytes (CTL) in a normal HLA-A2.1+ individual. These CTL showed HLA-restricted cytotoxicity to myeloid leukemias (which overexpress proteinase 3). To further investigate their antileukemic potential, we studied the ability of PR1-specific CTL, derived from two HLA-A2.1+ normal individuals, to inhibit colony-forming unit granulocyte-macrophage (CFU-GM) from normal and leukemic individuals. CTL from 20 day PR1 peptide-pulsed lymphocyte cultures showed 89% to 98% HLA-A2.1–restricted colony inhibition of chronic myeloid leukemia targets. Colony formation in normal HLA-A2.1+ bone marrow or HLA-A2.1− CML cells was not inhibited. Sequencing of the exon encoding PR1 showed that colony inhibition was not caused by polymorphic differences in proteinase 3 between effectors and targets. Analysis by flow cytometry showed that proteinase 3 was overexpressed in the leukemia targets compared with normal marrow targets (median channel fluorescence 1,399 v 298, P = .009). These results show that PR1-specific allogeneic T cells preferentially inhibit leukemic CFU-GM based on overexpression of proteinase 3, and that proteinase 3-specific CTL could be used for leukemia-specific adoptive immunotherapy.

Cytotoxic T Lymphocytes Specific for a Nonpolymorphic Proteinase 3 Peptide Preferentially Inhibit Chronic Myeloid Leukemia Colony-Forming Units

We previously showed that a peptide (PR1) derived from the primary granule enzyme proteinase 3 induced peptide specific cytotoxic T lymphocytes (CTL) in a normal HLA-A2.1+ individual. These CTL showed HLA-restricted cytotoxicity to myeloid leukemias (which overexpress proteinase 3). To further investigate their antileukemic potential, we studied the ability of PR1-specific CTL, derived from two HLA-A2.1+ normal individuals, to inhibit colony-forming unit granulocyte-macrophage (CFU-GM) from normal and leukemic individuals. CTL from 20 day PR1 peptide-pulsed lymphocyte cultures showed 89% to 98% HLA-A2.1–restricted colony inhibition of chronic myeloid leukemia targets. Colony formation in normal HLA-A2.1+ bone marrow or HLA-A2.1− CML cells was not inhibited. Sequencing of the exon encoding PR1 showed that colony inhibition was not caused by polymorphic differences in proteinase 3 between effectors and targets. Analysis by flow cytometry showed that proteinase 3 was overexpressed in the leukemia targets compared with normal marrow targets (median channel fluorescence 1,399 v 298, P = .009). These results show that PR1-specific allogeneic T cells preferentially inhibit leukemic CFU-GM based on overexpression of proteinase 3, and that proteinase 3-specific CTL could be used for leukemia-specific adoptive immunotherapy.