Natural cytotoxicity, mediated by natural killer (NK) cells plays an important role in the inhibition and elimination of malignant tumor cells. To investigate the immunoregulatory role of NK cells and their potential as diagnostic markers, NK cell activity (NKA) was analyzed in prostate cancer (PCa) patients with particular focus on NK cell subset distribution.
Prospective data of NKA and NK cell subset distribution patterns were measured from 51 patients initially diagnosed with PCa and 54 healthy controls. NKA was represented by IFN-γ levels after stimulation of the peripheral blood with Promoca®. To determine the distribution of NK cell subsets, PBMCs were stained with fluorochrome-conjugated monoclonal antibodies. Then, CD16+CD56dim and CD16−CD56bright cells gated on CD56+CD3− cells were analyzed using a flow-cytometer.
NKA and the proportion of CD56bright cells were significantly lower in PCa patients compared to controls (430.9 pg/ml vs. 975.2 pg/ml and 2.3% vs. 3.8%, respectively; p<0.001). Both tended to gradually decrease according to cancer stage progression (p for trend = 0.001). A significantly higher CD56dim-to-CD56bright cell ratio was observed in PCa patients (41.8 vs. 30.3; p<0.001) along with a gradual increase according to cancer stage progression (p for trend = 0.001), implying a significant reduction of CD56bright cells in relation to the alteration of CD56dim cells. The sensitivity and the specificity of NKA regarding PCa detection were 72% and 74%, respectively (best cut-off value at 530.9 pg/ml, AUC = 0.786).
Reduction of CD56bright cells may precede NK cell dysfunction, leading to impaired cytotoxicity against PCa cells. These observations may explain one of the mechanisms behind NK cell dysfunction observed in PCa microenvironment and lend support to the development of future cancer immunotherapeutic strategies.
Citation: Koo KC, Shim DH, Yang CM, Lee S-B, Kim SM, Shin TY, et al. (2013) Reduction of the CD16−CD56bright NK Cell Subset Precedes NK Cell Dysfunction in Prostate Cancer. PLoS ONE 8(11): e78049. https://doi.org/10.1371/journal.pone.0078049
Editor: Stephanie Filleur, Texas Tech University Health Sciences Center, United States of America
Received: June 12, 2013; Accepted: September 8, 2013; Published: November 4, 2013
Copyright: © 2013 Koo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (2012-0000807). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: An author, SMK, is an employee at ATgen who helped with the experimental process. However, this does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.
Natural killer (NK) cells serve a major role in the innate and adaptive immune responses against tumor transformation or pathogen-infected cells . NK cells exert natural cytotoxicity to eliminate malignant cells without prior sensitization or class I MHC restriction , . Furthermore, NK cells stimulate the adaptive immune response by secreting proinflammatory cytokines to counteract the escape mechanisms promoted by tumor cells . Progress has been made in understanding the biology of NK cells; nonetheless, further clarification remains regarding anti-tumor effects of NK cell activity (NKA) and patterns of subset distribution in PCa.
NK cells are defined phenotypically by their expression of CD56 and lack of CD3 expression . According to membrane densities of CD56 and CD16, NK cells are classified into CD16+CD56dim and CD16−CD56bright subsets . The majority are CD56dim cells that mainly exert potent cytotoxicity . In contrast, CD56bright cells mediate low cytotoxicity but acquire greater cytolytic activity than CD56dim cells upon activation due to release of proinflammatory cytokines such as IFN-γ . The level of IFN-γ, i.e., NKA, is generally associated with oncological prognosis, which implies the essential role of differential NK cell subset expression in the immune regulation of tumor cells . NKA has shown to serve an important role in surveillance and in the elimination of tumor cells . Studies have shown that low NKA leads to high levels of tumor occurrence and metastasis, and that its degree correlates with invasiveness of malignancy . On the contrary, high NKA has been shown to correlate with lower incidence of tumors, and their infiltration in certain tumors, i.e., melanoma, head and neck squamous cell carcinomas, is an indicator for a better oncological outcome , .
There is accumulating evidence that an impaired immune response is a crucial factor in the pathogenesis of prostate cancer (PCa) , . NK cell dysfunction has been implicated in PCa along with a variety of tumors , . Despite several proposed mechanisms including reduced number, immunosuppressive cytokines, and receptor repertoire imbalance, the pathophysiology of NK cell dysfunction in PCa is not fully understood . Regarding the role of NKA in tumor suppression, harnessing the mechanisms of NK cells could clearly be an important component for successful immunotherapy against PCa. Prostate-specific antigen (PSA) is the most widely used serum marker that has revolutionized the early detection and management of PCa. However, the relative lack of cancer-specificity and lack of an upper or lower threshold value are major drawbacks.
To address these issues, NKA and the distributions of CD56dim and CD56bright subsets were analyzed between PCa patients and controls. Our findings indicate that immunoregulation in PCa is impaired due to a reduction in NKA preceded by redistribution of NK cell subsets. Moreover, evaluation of the diagnostic performance of NKA revealed that it may be applied as a supportive marker in addition to PSA.
Materials and Methods
1. Patients and Controls
This prospective cross-sectional analysis involved 51 patients with newly diagnosed biopsy-proven PCa due to a PSA elevation noted on health examinations from March to December, 2012. 54 age-matched controls were self-volunteered healthy individuals whose prostate volume, PSA, and DRE were within normal accepted ranges. None of the patients had received prior treatment for PCa, were known to have immunological or other malignant conditions, and were all free of active infection or inflammation as assessed by white blood cell count <10,000 cells/µl and C-reactive protein <1.0 mg/L (Table 1). All controls were free from inflammatory conditions without prior exposure to immunosuppressive agents. Independent approval was obtained from Yonsei University Ethics Committee (4-2011-0660), with all blood samples collected after obtaining informed consent prior to radical prostatectomy. All participants provided written consent to participate in the current study.
2. NK Cell Activity
Cytotoxic activity of NK cells was determined using the NK Vue-Kit® (ATgen, Sungnam, Korea). Whole blood was collected using BD Vacutainer® heparinN1 tubes. 1 ml of whole blood was incubated for 24 hrs, at 37°C, under 5% CO2 with indicated dose of Promoca® and 1 ml of RPMI 1640 media. Cell-free supernatants were harvested, and IFN-γ levels were determined according to manufacturer’s protocols.
3. NK Cell Subset Distribution
3.1. Preparation of PBMCs.
3 ml of heparinized venous blood was obtained and analyzed within 4 h of collection. PBMCs were isolated by density gradient centrifugation using CPT® cell preparation tubes (BD Vacutainer®) at 1600 g for 20 min at 20°C. The collected PBMCs (1−2×106 cells/ml) were washed and resuspended in 5% fetal bovine serum (FBS)+phosphate buffered saline (PBS).
3.2. Antibody staining.
For the expression of CD3, CD16, and CD56 on NK cells, PBMCs were stained with Alexa-anti-CD3, PE-anti-CD16, and FITC-anti-CD56 fluorochrome-conjugated monoclonal antibodies (BD Biosciences). After staining for 30 min at 4°C, cells were washed extensively and fixed in 1% paraformaldehyde-PBS until assessment.
3.3. Flow cytometry.
To determine the total percentage of NK cells gated in the CD3−CD56+ cell population, at least 10,000 target cells were acquired by LSRII flow-cytometry (BD Biosciences). The distribution of CD16+CD56dim NK cells and CD16−CD56bright NK cells gated from the CD3−CD56+ cell population is presented as the percentage of total NK cells. For each sample, the data were further analyzed by FlowJo 18.104.22.168 (Tree Star, Inc., Ashland, OR, USA).
4. Cancer Stage Classification
PCa staging was determined according to the 7th American Joint Committee on Cancer (AJCC) TNM system. Stage distribution and pathological characteristics are shown in Table 2. Pathology was confirmed by a single pathologist.
5. Statistical Analysis
Statistical analyses were performed using Mann-Whitney U tests when comparing unpaired two-group data and Kruskal-Wallis tests with Bonferroni post-hoc correction when comparing more than two groups. The accuracy of NKA and the CD56dim-to-CD56bright ratio in detecting PCa was determined by receiver operating characteristics-derived area under the curve (AUC). Correlation analysis was used to evaluate associations among NKA, CD56dim-to-CD56bright ratio, and clinicopathological variables. Statistical analyses were performed using SPSS (v.18.0).
1. Demographic Data
All patients and controls were clinically and pathologically investigated with respect to factors shown in Table 1. Factors that may influence one’s immune status manifested no differences between groups.
2. Frequency of NK Cells and Distribution of CD56dim and CD56bright NK Cell Subsets
Representative flow cytometric data shows the distribution of total NK cell population represented as CD3−CD56+ cells (Fig. 1A) and two major subsets, CD16+CD56dim and CD16−CD56bright, expressed as a percentage of total NK cells (Fig. 1B). Total NK circulating frequencies did not differ between patients and controls or between cancer stage groups (Fig. 2A; Table 2). However, a preferential decrease in frequency of CD56bright cells was noted in patients. Moreover, CD56bright cells tended to gradually decrease according to cancer stage progression, i.e., extracapsular extension, LN or adjacent organ metastasis (Fig. 2B; Table 2). A significantly higher CD56dim-to-CD56bright NK cell ratio was observed in patients compared to controls, with a tendency to increase according to stage progression (p for trend = 0.001) (Table 2).
(B) Representative flow cytometric data of two major NK cell subsets detected in peripheral blood. Left, upper box: CD16+CD56dim NK cell subset. Right, lower box: CD16−CD56bright NK cell subset.
No significant differences of total NK population were noted between patients and controls, and within stage groups. (B) Boxplot diagrams showing flow cytometric distribution results of CD56dim and CD56bright subset distributions within total NK cells between controls and patients, grouped according to cancer stage. *p<0.05, **p<0.01 in relation to controls.
3. NK Cell Activity
Results obtained are presented in Fig. 3 and Table 2. As indicated, patients showed significantly lower NKA. According to stage progression, those with higher stages showed a greater reduction of NKA (p for trend <0.001).
4. Analysis by ROC Curves
ROC curves and best cut-off values were used to calculate the sensitivity and specificity of NK cell-related parameters (Table 3). The sensitivity and specificity of NKA with respect to PCa detection were 72% and 74%, respectively, whereas the CD56dim-to-CD56bright cell ratio showed a sensitivity of 66% and a specificity of 71% (Fig. 4A). In further analysis, the sensitivity and specificity of NKA were determined according to two PSA values grouped as 4 to 10 ng/ml, which is the diagnostic grey-zone, and levels greater than 10 ng/ml. At a set specificity of 74%, NKA for PSA values within the grey-zone showed higher sensitivity (73% vs. 70%) and AUC (0.82±0.06 vs. 0.76±0.07) relative to PSA values greater than 10 ng/ml (Fig. 4B).
(AUC = Area Under the Curve). (B) ROC curves comparing the performances of NK cell activity measurement according to PSA grouped as; 4 to 10 ng/ml and greater than 10 ng/ml. (AUC = Area Under the Curve).
5. NK Cell Activity and CD56dim-to-CD56bright Cell Ratio According to Clinicopathological Variables
NKA showed negative correlations with PSA, cancer stage, and the CD56dim-to-CD56bright ratio. On the other hand, CD56dim-to-CD56bright cell ratio showed positive correlations with PSA and cancer stage (Table 4). NKA and CD56dim-to-CD56bright ratio was compared between controls and patients grouped according to clinicopathological variables (Table 5). Although CD56dim-to-CD56bright ratio failed to discriminate patients with Gleason scores <7 and those without extracapsular extension from controls, all other subgroups were distinguishable from controls by NKA and CD56dim-to-CD56bright ratio. Analysis in-between patient subgroups revealed significantly higher CD56dim-to-CD56bright ratio in patients with pathologically confirmed LN metastasis (p = 0.043; data not shown).
The aim of the present study was to clarify the role of NK cells in the immune response against PCa. Several mechanisms of PCa development and progression have been proposed, including hormonal, metabolic alterations, and immune response , . There is accumulating evidence that different lymphocyte populations are involved in cell-mediated immunosuppression that leads to occurrence and progression of PCa , , . However, there is limited information regarding the functional role of NK cells in the immune response to PCa. To address this issue, we investigated NKA as a marker for IFN-γ levels and the distribution of NK cell subsets in PCa patients. The results of our study indicate that impaired NKA is presumably preceded by a reduction in CD56bright cells, and that the level of NKA could be utilized as a supportive diagnostic marker for PSA.
1. Preferential Reduction of CD56bright NK Cells in PCa Patients
NK cells are functionally classified into CD56dim and CD56bright subsets. CD16+CD56dim cells are effector cells with high quantities of cytolytic granules that express potent cytotoxicity against tumor cells . CD16−CD56bright cells release proinflammatory cytokines such as IFN-γ which drives inflammatory mechanisms that regulate tumor initiation, immunoevasion, survival, and outgrowth , . Recent discoveries have revealed that CD56bright cells constitute the majority of NK cells in lymphoid tissues and that they are not just a minor subpopulation among NK cells but are immature precursors of CD56dim cells . This work focuses on this particular subset, regarding its importance in the regulation of NK cell-mediated response against tumor cells.
Investigation on distributional patterns of NK cell subsets revealed a significant decrease of CD56bright cells without alteration of CD56dim cells. Previous studies on various tumor-bearing hosts have reported rather distinct interrelationships between CD56dim and CD56bright subsets. In contrast to our results, a reduction in CD56dim cells without alteration of CD56bright cells was noted in gastric and esophageal cancers . On the other hand, a reduction in CD56bright cells was observed in breast, head and neck cancers, and an equal distribution of CD56dim cells in PCa; results that are consistent with the present study , .
2. Alteration of CD56bright NK Cells as a Response Mechanism to Tumor Microenvironment
Our study primarily observed a preferential reduction of CD56bright cells without alteration of CD56dim cells. Although the underlying cause has not yet been clearly defined, two possible explanations can be raised; maturation process and recruitment process.
As mentioned, CD56bright cells are accepted as precursors to CD56dim cells, with each subset representing a distinct maturation stage . Possibly, an excessive demand for effector cells in response to tumor may have provoked a transition of immature CD56bright cells into CD56dim cells. A similar explanation has been proposed for the reduction of CD56bright cells in patients with head and neck cancers . However, this presupposes a concomitant increase of CD56dim cells, which was not observed in the present study.
An alternative explanation without demonstration is that peripheral CD56bright cells may have been recruited to lymphoid tissue sites as metastatic LNs to acquire cytotoxicity. This idea was based on previous observations that CD56bright cells preferentially accumulate in the T cell area of LNs until being activated to produce proinflammatory cytokines , [22,22]. Moreover, the observation that CD56bright cells isolated from human LNs become strongly cytotoxic upon stimulation by IL-2 suggests that NK cells recruited to LNs might represent an immature pool of effector cells . A significantly higher CD56dim-to-CD56bright cell ratio in patients with pathologically confirmed LN metastasis was observed in our study, implying that these circulating cells may have been recruited to pathologic or secondary LNs in response to tumor. This is of relevance because LNs are usually the primary metastatic sites and CD56bright cells are the primary subset found in LNs that counteract the metastatic cells . To confirm this issue, it would be interesting to examine whether CD56bright cells are accumulated in metastatic LNs following LN dissection.
3. NK Cell Dysfunction as a Consequence of Reduction of CD56bright NK Cells
NKA was investigated to determine the influence of reduction of CD56bright cells on cytolytic activity against tumor cells. Parallel to observations with CD56bright cells, NKA was observed to be lower in PCa patients, along with a tendency to gradually decrease according to cancer stage progression. These findings are consistent with previous reports that showed NKA is compromised in a broad spectrum of hematological and solid tumors , , . Several mechanisms of compromised NKA have been proposed, such as decreased number of tumor-infiltrating NK cells , increased surface receptors for immune suppressor factors , and inactivation of effector cells .
Correlations observed between NKA and CD56dim-to-CD56bright cell ratio may be of direct relevance to suggest an additional mechanism that weak NKA is a consequence of reduced CD56bright cells. CD56bright cells are known to be major sources for IFN-γ , as observed in in vitro studies where CD56bright cells were shown to preferentially proliferate in co-culture with immature dendritic cells and lipopolysaccharides to produce IFN-γ . Also, stimulation of CD56bright cells with transduced carcinoma cells resulted in an enhanced ability to produce IFN-γ and impart high cytotoxicity . Further, in vivo studies have shown that tonsillar CD56bright cells produce IFN-γ before maturation into effector cells . Conversely, a reduction of CD56bright cells was observed to induce impaired secretion of IFN-γ in patients with allergic rhinitis . Considering these supportive findings that secretion of IFN-γ directly depends on CD56bright cells, we suggest that the reduction of CD56bright cells is a potential mechanism involved in low NKA, which leads to impaired cytotoxicity against PCa cells.
4. NK Cell Activity, a Supportive Diagnostic Marker for PSA
ROC curves revealed that NKA may serve as a supportive marker for PSA in diagnosing PCa. Although it is clear that PSA provides the highest diagnostic value for PCa, a major limitation is its lack of cancer-specificity which causes unnecessary risks and costs, especially in the diagnostic grey-zone . Although ongoing challenges strive to develop novel methods of PCa detection, none have clearly outweighed the benefits against drawbacks . This investigation raises the possibility that NKA may be utilized in combination with PSA to provide additional diagnostic value, especially for those within the diagnostic grey-zone.
This study was based on controls versus patients diagnosed with PCa due to elevated PSA on a routine health examination. Therefore, the absence of PCa patients with normal PSA (<4 ng/ml) was the major limitation of this study, which hampered a direct comparison of diagnostic yield between PSA and NKA. An extended population study is needed to confirm our preliminary findings and to assess cost-effectiveness.
This observational study provided novel findings that CD56bright cells serve an important role in adaptive response against PCa cells. This notion lends further support that longitudinal studies regarding NK cell immunosurveillance clearly deserve additional research to potentially lead to novel immunotherapeutic strategies for enhancing oncological outcomes of PCa.
Conceived and designed the experiments: KCK HGY KHR JML SJH. Performed the experiments: KCK DHS CMY SL SMK. Analyzed the data: KCK TYS KHK. Contributed reagents/materials/analysis tools: SMK HGY KHR JML. Wrote the paper: KCK. Dual corresponding contribution: JML SJH.
- 1. Vivier E, Raulet DH, Moretta A, Caligiuri MA (2011) Innate or Adaptive Immunity? The Example of Natural Killer Cells. Science 331: 44–49.
- 2. Koch J, Steinle A, Watzl C, Mandelboim O (2013) Activating natural cytotoxicity receptors of natural killer cells in cancer and infection. Trends Immunol 34: 182–191.
- 3. Takahashi E, Kuranaga N, Satoh K, Habu Y, Shinomiya N, et al. (2007) Induction of CD16+ CD56bright NK cells with antitumour cytotoxicity not only from CD16- CD56bright NK Cells but also from CD16- CD56dim NK cells. Scand J Immunol 65: 126–138.
- 4. Cooper MA, Fehniger TA, Caligiuri MA (2001) The biology of human natural killer-cell subsets. Trends Immunol 22: 633–640.
- 5. Watanabe M, Kono K, Kawaguchi Y, Mizukami Y, Mimura K, et al. (2010) NK cell dysfunction with down-regulated CD16 and up-regulated CD56 molecules in patients with esophageal squamous cell carcinoma. Dis Esophagus 23: 675–681.
- 6. Dowell AC, Oldham KA, Bhatt RI, Lee SP, Searle PF (2012) Long-term proliferation of functional human NK cells, with conversion of CD56dim NK cells to a CD56bright phenotype, induced by carcinoma cells co-expressing 4-1BBL and IL-12. Cancer Immunol Immunother 61: 615–628.
- 7. Bauernhofer T, Kuss I, Henderson B, Baum AS, Whiteside TL (2003) Preferential apoptosis of CD56dim natural killer cell subset in patients with cancer. Eur J Immunol 33: 119–24.
- 8. Lin CT, Yu MT, Li C, Ho YC, Shen CH, et al. (2010) Dysfunction of natural killer cells in patients with transitional cell carcinoma. Cancer Lett 291: 39–45.
- 9. Kim S, Iizuka K, Aguila HL, Weissman IL, Yokoyama WM (2000) In vivo natural killer cell activities revealed by natural killer cell-deficient mice. Proc Natl Acad Sci U S A 97: 2731–2736.
- 10. Takeuchi H, Maehara Y, Tokunaga E, Koga T, Kakeji Y, et al. (2001) Prognostic significance of natural killer cell activity in patients with gastric carcinoma: a multivariate analysis. Am J Gastroenterol 96: 574–578.
- 11. Schleypen JS, Baur N, Kammerer R, Nelson PJ, Rohrmann K, et al. (2006) Cytotoxic markers and frequency predict functional capacity of natural killer cells infiltrating renal cell carcinoma. Clin Cancer Res 12: 718–725.
- 12. Vivier E, Ugolini S, Blaise D, Chabannon C, Brossay L (2012) Targeting natural killer cells and natural killer T cells in cancer. Nat Rev Immunol 12: 239–252.
- 13. Klein EA, Silverman R (2008) Inflammation, infection, and prostate cancer. Curr Opin Urol 18: 315–319.
- 14. Basch E, Oliver TK, Vickers A, Thompson I, Kantoff P, et al. (2012) Screening for prostate cancer with prostate-specific antigen testing: American Society of Clinical Oncology Provisional Clinical Opinion. J Clin Oncol 30: 3020–3025.
- 15. Oikawa T, Kawai K, Ishiwata I, Ohno T, Akaza H (2003) Induction of potent antitumour natural-killer cells from peripheral blood of patients with advanced prostate cancer. BJU Int 92: 1009–1015.
- 16. Kastelan M, Kraljic I, Tarle M (1992) NK cell activity in treated prostate cancer patients as a probe for circulating tumor cells: hormone regulatory effects in vivo. Prostate 21: 111–120.
- 17. Sotosek S, Sotosek Tokmadzic V, Mrakovcic-Sutic I, Thomas MI, Dominovic M, et al. (2011) Comparative study of frequency of different lymphocytes subpopulation in peripheral blood of patients with prostate cancer and benign prostatic hyperplasia. Wien Klin Wochenschr 123: 718–725.
- 18. Ebelt K, Babaryka G, Figel AM, Pohla H, Buchner A, et al. (2008) Dominance of CD4+ lymphocytic infiltrates with disturbed effector cell characteristics in the tumor microenvironment of prostate carcinoma. Prostate 68: 1–10.
- 19. Wagenlehner FM, Elkahwaji JE, Algaba F, Bjerklund-Johansen T, Naber KG (2007) Hartung R, Weidner W. The role of inflammation and infection in the pathogenesis of prostate carcinoma. BJU Int 100: 733–737.
- 20. Harlin H, Hanson M, Johansson CC, Sakurai D, Poschke I, et al. (2007) The CD16- CD56bright NK cell subset is resistant to reactive oxygen species produced by activated granulocytes and has higher antioxidative capacity than the CD16+ CD56dim subset. J Immunol 179: 4513–4519.
- 21. Zaidi MR, Merlino G (2011) The two faces of interferon-gamma in cancer. Clin Cancer Res 17: 6118–6124.
- 22. Poli A, Michel T, Theresine M, Andres E, Hentges F, et al. (2009) CD56bright natural killer (NK) cells: an important NK cell subset. Immunology 126: 458–465.
- 23. Izawa S, Kono K, Mimura K, Kawaguchi Y, Watanabe M, et al. H2O2 production within tumor microenvironment inversely correlated with infiltration of CD56dim NK cells in gastric and esophageal cancer: possible mechanisms of NK cell dysfunction. Cancer Immunol Immunother 60: 1801–1810.
- 24. Chan A, Hong DL, Atzberger A, Kollnberger S, Filer AD, et al. (2007) CD56bright human NK cells differentiate into CD56dim cells: role of contact with peripheral fibroblasts. J Immunol 179: 89–94.
- 25. Ferlazzo G, Munz C (2004) NK cell compartments and their activation by dendritic cells. J Immunol 172: 1333–1339.
- 26. Sheikhi A, Saadati K, Salmani R, Yahaghi N, Siemens DR (2011) In vitro modulation of natural killer activity of human peripheral blood mononuclear cells against prostate tumor cell line. Immunopharmacol Immunotoxicol 33: 700–708.
- 27. Rabinovich GA, Gabrilovich D, Sotomayor EM (2007) Immunosuppressive strategies that are mediated by tumor cells. Annu Rev Immunol 25: 267–296.
- 28. Yamaguchi Y, Takashima I, Funakoshi M, Kawami H, Toge T (1994) Defective natural killer activity in gastric cancer patients: possible involvement of suppressor factor receptor. In Vivo 8: 279–283.
- 29. Vitale M, Della Chiesa M, Carlomagno S, Romagnani C, Thiel A, et al. (2004) The small subset of CD56brightCD16- natural killer cells is selectively responsible for both cell proliferation and interferon-gamma production upon interaction with dendritic cells. Eur J Immunol 34: 1715–1722.
- 30. Scordamaglia F, Balsamo M, Scordamaglia A, Moretta A, Mingari MC, et al. (2008) Perturbations of natural killer cell regulatory functions in respiratory allergic diseases. J Allergy Clin Immunol 121: 479–485.
- 31. Schroder FH (2012) Landmarks in prostate cancer screening. BJU Int 110 Suppl 13–7.