Reconstitution and functional analyses of neutrophils and distinct subsets of monocytes after allogeneic stem cell transplantation

Abstract

Purpose

The aim of the study was to investigate the recovery of the innate immune system within the first 100 days after allogeneic peripheral blood stem cell transplantation (PBSCT) and to elucidate a potential correlation with such important events as severe infectious complications or graft-versus-host disease (GvHD).

Methods

In 30 consecutive patients who underwent allogeneic PBSCT, absolute numbers of neutrophils and monocytes were determined and different functional analyses performed at different time points (day +30, +60 and +90, respectively). The capacity to phagocyte Escherichia coli (E. coli) as well as the induction of oxidative burst after incubation with different stimuli (Phorbol-12-myristate-13-acetate; PMA, the chemotactic peptide N-formyl-Met-Leu-Phe; f-MLP or opsonized E. coli) were analysed after engraftment.

Results

There was a rapid reconstitution concerning the capability of both neutrophils and monocytes to phagocyte E. coli without a significant increase between day +30 and +90. In contrast, a twofold increase of monocyte oxidative burst after incubation with PMA at day +90 was observed (P = 0.017). Furthermore, the ability of neutrophils to induce oxidative burst after ingestion with E. coli was impaired on day +30 with a significant functional reconstitution on day +60 (P = 0.01). The oxidative burst activity following incubation with f-MLP did not show significant changes after stem cell engraftment. Analysis of numeric reconstitution of CD14+CD16+ monocytes demonstrated a potential correlation with a decreased incidence of chronic GvHD.

Conclusion

The functional recovery of neutrophils and monocytes in the early period after allogeneic PBSCT differs not only concerning phagocytosis and oxidative burst but also with respect to the stimulus and the cell population that was analysed for oxidative burst activity. The subset of CD16+CD14+ monocytes might be a predictor for a reduced risk of chronic GvHD.

Introduction

The innate immune system is of fundamental importance for the defence against bacterial or fungal infections. Among the cellular components of the innate immunity, neutrophils and monocytes play a pivotal role not only because it represents the majority of cells as compared with natural killer (NK) cells. The defence mechanisms of the innate immune system can be observed immediately (0–4 h) while a response of adaptive immunity is much slower even in case of existing memory cells (4–96 h) (Kobayashi and DeLeo 2009; Yona and Jung 2010; Vivier et al. 2011). Nevertheless, there is a tight and complex interaction between the innate and the adaptive immune system. In detail, antigen-specific responses of adaptive immunity can be activated by distinct mechanisms of pathogen-specific innate immune recognition (Iwasaki and Medzhitov 2010; Diacovich and Gorvel 2010).

Several studies investigated the recovery of adaptive immunity after allogeneic stem cell transplantation focussing on the reconstitution of different lymphocyte subsets. Such data are available for patients who underwent either allogeneic bone marrow transplantation (BMT) or allogeneic PBSCT resulting in a detailed knowledge of several factors that have an impact on lymphocyte repopulation following transplantation. Some of the most important factors beside the stem cell source are the conditioning regimen, the immunosuppression after transplantation, the reactivation of cytomegalovirus (CMV) and the occurrence of graft-versus-host disease (GvHD) (Ottinger et al. 1996; Roberts et al. 1993).

In contrast to allogeneic PBSCT, there are some reports of functional recovery on the innate immune system especially in patients who received allogeneic BMT or autologous PBSCT (Volk et al. 2000; Dayyani et al. 2004). In detail, Krause and colleagues reported on a rapid numeral engraftment of neutrophils and monocytes after autologous PBSCT that were associated with an impaired ability to secrete proinflammatory cytokines (Krause et al. 2003).

Furthermore, van den Broek could demonstrate a reduced intracellular killing of bacteria while phagocytosis itself was not impaired shortly after allogeneic BMT (Van den Broek et al. 1981). In accordance with this observation, Miyagawa reported on an unaffected phagocytosis while they found a diminished oxidative burst in patients who underwent allogeneic BMT (Miyagawa and Klingemann 1997).

Peripheral blood monocytes represent a highly variable cell population regarding size, morphology and expression of surface antigens (Passlick et al. 1989; Ziegler-Heitbrock et al. 1993; Ziegler-Heitbrock 1996). By means of two-colour flow cytometry staining, the CD14 and CD16 antigens, two major populations of monocytes can be distinguished. The surface antigen CD16 represents a low-affinity Fc receptor that mediates not only phagocytosis but also antibody-dependent cytotoxicity (ADCC) of monocytes, neutrophils and NK cells (Delves and Roitt 2000; Ziegler-Heitbrock 1996). Flow cytometry analysis of monocytes in healthy volunteers can demonstrate a majority of CD14++CD16− cells in contrast to about ten per cent of CD14+CD16+ cells (Ancuta et al. 2000; Ziegler-Heitbrock et al. 1993; Ziegler-Heitbrock 1996).

The subpopulation of CD14+CD16+ monocytes are assumed to differentiate from CD14++CD16− cells representing circulating precursors of macrophages. In detail, CD14+CD16+ monocytes demonstrate a higher expression of such antigens as CD11a, CD11c and VLA (very late antigen)-4 that are important for monocyte migration in case of inflammation (Ziegler-Heitbrock et al. 1993; Ziegler-Heitbrock 1996).

Previously, we could show that the ability of neutrophils to induce oxidative burst after ex vivo stimulation with E. coli is highly impaired at day +30 after allogeneic PBSCT. Unexpectedly, there was no impairment regarding the migratory capacity of neutrophils shortly after transplantation (Scholl et al. 2007).

Therefore, the aim of our study presented as to investigate the capacity of oxidative burst as well as the ability of neutrophils and monocytes to phagocyte E. coli at three consecutive time points before day +100 after allogeneic PBSCT. Furthermore, we sought to determine the reconstitution of CD14+CD16+ monocytes in peripheral blood after allogeneic transplantation.

Patients and methods

Patient characteristics

This monocentric analysis was approved by the institutional review board of the Universitätsklinikum Jena, Germany. Thirty consecutive patients suffering from various hematologic malignancies were investigated after informed consent. Detailed patients′ characteristics are summarised in Table 1. Most patients who were treated with a myeloablative regimen received hyperfractionated total body irradiation (TBI, 12 Gy) and cyclophosphamide (60 mg/kg at day −3 and −2). In addition, the protocols for patients treated with a reduced TBI (8 Gy) regimen contained either cyclophosphamide or fludarabine and were supplemented with antithymocyte globulin (ATG) in case of an unrelated donor. In contrast, most patients who received a dose-reduced conditioning therapy were treated according to the protocol published by Slavin (Slavin et al. 1998). In detail, patients received a conditioning regimen consisting of fludarabine (30 mg/m², day −10 to −5), busulfan (4 mg/kg at day −6 and −5) and ATG (10 mg/kg, day −4 to −1). All patients received a GvHD prophylaxis containing cyclosporine A (CSA), starting with 3 mg/kg at day −1 and consecutive adaptation according to serum level. Additional GvHD prophylaxis were given dependent on the applied conditioning protocol and additional risk factors for GvHD, e.g. patients receiving 12 Gy TBI were treated with methotrexate (MTX) on day 1 (15 mg/m²), followed by 10 mg/m² on days 3, 6 and 11 while patients with a HLA mismatch or an unrelated donor were additionally given mycophenolate mofetil (MMF), beginning on day +10. All patients were treated with granulocyte colony-stimulating factor (G-CSF) subcutaneously starting from day +1 with 5 μg/kg body weight until stable neutrophil engraftment. CMV risk groups as indicated in Table 1 were defined according to recently published recommendations (George et al. 2010). In detail, in case of seronegativity of both donor and recipient, a patient was classified as ‘low risk’ while any constellation including seropositivity of the patient defined a ‘high-risk’ situation. The severity of acute and chronic GvHD was classified according to previously published consensus statements (Przepiorka et al. 1995; Filipovich et al. 2005).

Table 1.

Patients’ characteristics (n = 30)

Patients’ characteristics	N = 30
Sex (female/male)	15/15
Disease, cases
AML/MDS/CML	18
ALL/NHL/CLL	11
AA	1
Median age at transplantation (range, years)	45 (20–63)
Donor, cases
MRD	6
MUD	12
mMUD	12
Median CD34 content (×106 per kg, range)	6.0 (3.1–9.9)
Engraftment (WBC > 1 × 109/l, range, days)	13 (9–25)
Conditioning, cases
Myeloablative regimen	19
Dose-reduced regimen	11
ABO compatibility, cases
ABO ident	10
Minor incompatibility	8
Major incompatibility	12
CMV risk group, cases
Low risk	11
Intermediate risk	4
High risk	15
Acute GvHD
None and grade I GvHD	17
GvHD grade II-IV	13

AML indicates acute myeloid leukaemia, AA aplastic anaemia, MDS myelodysplastic syndrome, CLL chronic lymphatic leukaemia, CML chronic myeloid leukaemia, ALL acute lymphoid leukaemia, NHL non-Hodgkin’s lymphoma, HLA human leucocyte antigen, MRD matched related donor, MUD matched unrelated donor, mMUD mismatched unrelated donor, GvHD graft-versus-host disease, CMV cytomegalovirus

Quantification of cell numbers

Absolute numbers of neutrophils and monocytes in peripheral blood were calculated using total WBC counts and the proportion of each cell type identified in the differential hemogram. Subsequently, the distinct subsets of monocytes were determined in consideration of the percentage CD14++/CD16− and CD14+/CD16+ in a separate two-colour flow cytometry analysis of whole blood after lysis of erythrocytes (Fig. 1B). Monoclonal antibodies (CD14-APC and CD16-PE) as well as the corresponding isotype controls were obtained from Becton–Dickinson (Heidelberg, Germany).

Fig. 1.

Overview of methods—a dot-plot presentation (SSC vs. CD14) identifying monocytes (G1, gate 1) and neutrophils (G2, gate 2). Functional assays (Phagotest^® and Phagoburst^®) of leucocyte subpopulations were performed by separate analysis according to immunological gating (c and d, respectively). b. Double-staining of leucocytes analysing CD14++CD16− (UL, upper left) and CD14+CD16+ (UR, upper right) monocyte subpopulations. For further details, see also section ‘Patients and methods’

Measurement of E. coli phagocytosis

The phagocytic capacity of neutrophils and monocytes was determined using the Phagotest^® assay (ORPEGEN Pharma, Heidelberg, Germany) according to the manufacturer’s instructions with some modifications. Briefly, 200 μl of freshly drawn heparinized whole blood was aliquoted for each test on the bottom of a 14-ml falcon tube and cooled down to 0°C in an ice bath for 10 min. Subsequently, 20 μl of opsonized FITC-labelled E. coli was added to each sample and incubated at 37°C for 10 min. Phagocytosis was stopped by adding 200 μl of quenching solution followed by two washing steps. After this, erythrocytes were lysed at room temperature for 20 min and each sample was aliquoted and stained with APC-conjugated isotype control or CD14-APC, respectively. All analyses included control samples that were not incubated with FITC-labelled E. coli in order to assess the autofluorescence defining negative cells of the test samples. Multiparameter flow cytometry was used for measurement of phagocytosis (Fig. 1A). The monocyte population was identified by immunological gating (CD14-APC versus SSC) while neutrophils were gated according to its scatter characteristics. The uptake of E. coli was quantified by green fluorescence (FITC). All samples were analysed as duplicates.

Analysis of oxidative burst

Measurement of oxidative burst was performed using the Phagoburst^® assay kit (ORPEGEN Pharma, Heidelberg, Germany) that enables the evaluation of phagocytosis and oxidative burst. In detail, 200 μl of heparinized whole blood was pipetted for each sample and incubated with 20 μl washing solution (negative control), PMA, f-MLP or a suspension of opsonized E. coli at 37°C for 10 min. Subsequently, 40 μl of freshly solved nonfluorescent substrate DHR (dihydrorhodamine-123) was added to each samples and incubated at 37°C for additional 10 min. All reactions were stopped with 2 ml lysing solution and were washed with 3 ml lysing solution. Staining of CD14 was performed as described above. Oxidative burst was analysed by flow cytometry of green fluorescence resulting from oxidative conversion of DHR to a fluorescent product. All determinations were performed as duplicates.

Data description and statistical analyses

Flow cytometry analyses defined two subgroups of FITC-positive cells. In detail, all leucocytes with a fluorescence intensity higher than one log-range above the corresponding control sample were defined as ‘high positive’. In contrast, ‘total positive’ cells represent all leucocytes that were FITC-positive in consideration of samples that were not incubated with E. coli or DHR, respectively. Statistical analyses were performed with the Statistical Program for Social Science (SPSS, Chicago, IL, USA). Differences between groups were assessed using Student′s t-test. Results were considered statistically significant in case of P ≤ 0.05.

Results

Time-dependent phagocytosis of E. coli

The capability of neutrophils and monocytes was analysed at three different time points: day +30, day +60 and day +90, respectively. The mean values of the percentage of positive cells defined by E. coli uptake are indicated in Table 2. It can be demonstrated that especially neutrophils undergo a rapid functional reconstitution measured by phagocytosis of bacteria. In detail, about ninety per cent of neutrophils were able to phagocyte E. coli on day +30. Furthermore, almost all neutrophils were high positive reflecting a strong uptake of opsonized bacteria.

Table 2.

Time course of phagocytic activity of neutrophils and monocytes

	Day +30		Day +60		Day +90
	Neutrophils	Monocytes	Neutrophils	Monocytes	Neutrophils	Monocytes
	n = 30	n = 23	n = 20	n = 15	n = 14	n = 11
E. coli
Total positive (%, range)	90.0 (1–99)	68.4 (0–96)	93.5 (2–98)	78.7 (0–96)	91.8 (1–96)	84.3 (2–97)	n.s.
High positive (%, range)	83.7 (1–96)	53.5 (0–94)	86.3 (1–98)	61.2 (5–94)	82.5 (1–95)	69.0 (2–94)	n.s.

Data represent median values of total and high positive cells

Analysis of monocytes revealed a similar time-dependent pattern after allogeneic PBSCT. Thus, the phagocytic capacity of monocytes did also recovered quickly as demonstrated by a high proportion of E. coli-positive cells. Nevertheless, there was a slight increase of total positive cells between day +30 and day +90: 68% vs. 84%, respectively (not significant). The observation that a majority of cells were even high positive in this assay correlates with the results for neutrophils as described above.

The data presented in Table 2 also demonstrate a broad range of positive cells identified in our patients that reflects a high variety of phagocytosis after allogeneic stem cell transplantation.

Recovery of oxidative burst activity

Table 3 summarises the quantitative analyses of oxidative burst activity of neutrophils and monocytes after allogeneic PBSCT in a time-dependent manner. It can be shown that the responsiveness of monocytes following stimulation with PMA is still quite impaired until day +60 while a twofold increase of oxidative burst positive cells can be observed on day +90 (35% vs. 77%, P = 0.017).

Table 3.

Time-dependent oxidative burst of neutrophils and monocytes

	Day +30		Day +60		Day +90
	Neutrophils	Monocytes	Neutrophils	Monocytes	Neutrophils	Monocytes
	n = 30	n = 23	n = 20	n = 15	n = 14	n = 11
PMA
Total positive (%)	53.4	31.1	76.5	35.2	48.6	76.8	day +60 vs. +90 P = 0.017
High positive (%)	29.3	2.5	31.4	2.4	25.6	6.7	n.s.
fMLP
Total positive (%)	1.3	14.4	1.6	15.4	3.2	25.5	n.s.
High positive (%)	0.03	0.04	0.1	0.2	0.1	0.3	n.s.
E. coli
Total positive (%)	38.7	43.8	61.9	54.6	66.3	67.1	day +30 vs. +60 P = 0.01
High positive (%)	13.3	8.7	16.8	13.8	5.4	6.7	day +60 vs. +90 P = 0.056

Data represent median values of total and high positive cells

In contrast, the ability of both neutrophils and monocytes to induce oxidative burst after stimulation with the peptide f-MLP did not change significantly between day +30 and +90.

In addition, a reduced functional capacity of neutrophils concerning oxidative burst induction after E. coli uptake early after engraftment can be shown. Furthermore, this lack of neutrophil function is no longer apparent on day +60 as demonstrated by consecutive analyses (39% vs. 62%, P = 0.01). Interestingly, the comparison of high positive cells (day +60 vs. day +90) suggests a decrease with respect to neutrophils after oxidative burst stimulation with E. coli (17% vs. 5%, P = 0.056). The variety of the data describing the range of each analysis was comparable with that presented in Table 2.

Reconstitution of distinct monocyte subpopulations

In addition to the absolute numbers of peripheral blood monocytes at different time points following allogeneic PBSCT, the distribution of CD14++CD16− and CD14+CD16+ cells representing distinct monocyte subpopulations were assessed by means of two-colour flow cytometry (Fig. 2). There was a rapid reconstitution of overall monocytes already at day +30 without any significant change compared to both time points that were analysed later. In addition, the same observation was made with respect to the monocyte subpopulations CD14++CD16− and CD14+CD16+, respectively. The percentage of CD14+CD16+ was much higher than in healthy volunteers. In detail, at day +60 and day +90, the ratio of CD14+CD16+ and CD14++CD16− monocytes was >1.

Fig. 2.

Reconstitution of monocyte subpopulations after allogeneic PBSCT (black bars indicate total monocytes; grey bars CD14++CD16−; white bars CD14+CD16+ monocytes)

Association between monocyte reconstitution and GvHD

The next question to be addressed was if there is a correlation between the reconstitution of monocyte subpopulations and the occurrence of either severe acute GvHD (grades II–IV) or of chronic GvHD. Neither the absolute numbers of monocyte subpopulations nor any ratio of these monocyte subsets calculated on day +30 or day +60, respectively, could demonstrate a correlation with the incidence or severity of acute GvHD (data not shown).

A descriptive analysis of absolute CD14+CD16+ counts was performed with respect to the clinical follow-up of chronic GvHD in these patients. In detail, a serial ranking of absolute CD14+CD16+ cells as measured on day +90 was assessed and patients in the lower third were compared with those patients within the upper third of absolute CD14+CD16+ counts (Table 4). Interestingly, all patients with low absolute CD14+CD16+ counts developed chronic GvHD including three patients with extended GvHD. Surprisingly, all patients with extended chronic GvHD had a donor without a HLA mismatch constellation. In contrast, within the small subgroup of patients with the highest CD14+CD16+ values, four out of six patients did not demonstrate a chronic GvHD in the clinical follow-up. Furthermore, no extensive GvHD was observed in this subgroup while one of the patients with limited chronic GvHD had a donor with a double HLA mismatch.

Table 4.

Association between absolute CD14+CD16+ counts on day +90 and the occurrence of chronic GvHD (n = 18)

Patient ID	Rank	Absolute counts of CD14+CD16+ (per μl)	Chronic GvHD	Comment
27	1	0	Limited	Relapse
16	2	11	Limited	MUD
26	3	51	Extensive	MUD
21	4	70	Extensive	MUD
2	5	72	Limited	MUD
8	6	113	Extensive	MUD
23	13	279	No	MRD
12	14	284	No	MUD
11	15	291	No	MUD
5	16	318	No	MUD
19	17	365	Limited	MRD
9	18	505	Limited	mMUD (A + C)

MUD matched unrelated donor, MRD matched related donor, mMUD mismatched unrelated donor

Discussion

The aim of this study was to analyse the time-dependent reconstitution of distinct functional aspects of neutrophils and monocytes in patients who underwent allogeneic PBSCT. These data reflect one of the first comprehensive investigations on the functional recovery of the innate immune system including an assessment of different functions after allogeneic PBSCT. The study presented here analysed distinct functional aspects of neutrophils and monocytes with the focus on the uptake of E. coli and subsequent capacity of oxidative burst at different time points following stem cell transplantation. Furthermore, the recovery of different subsets of monocytes in patients receiving allogeneic PBSCT was investigated.

It can be demonstrated that the functional reconstitution of E. coli phagocytosis occurs rapidly for both neutrophils and monocytes after engraftment of peripheral blood stem cells. Thus, a normal capacity to phagocyte bacteria can be documented on day +30 as well as at the consecutive time points investigated. These data agree with results previously published by Miyagawa and Klingemann in patients after allogeneic BMT. Thus, a rapid reconstitution of the phagocytic capacity can be confirmed in the setting of allogeneic PBSCT.

In accordance with previously published observations, an impaired capability of neutrophils to induce oxidative burst after ingestion of E. coli shortly after engraftment can be shown as measured on day +30. In addition, there is a significant increase of the percentage of neutrophils that are positive in the oxidative burst assay on day +60. This reflects a normalization of neutrophil function within the second month after allogeneic PBSCT.

Furthermore, a distinct functional reconstitution of monocytes can be demonstrated. In detail, there was a significant increase of monocytes tested positive for oxidative burst activity following incubation with PMA. In contrast, analyses of monocytes showed a faster functional reconstitution with respect to its capability of oxidative burst after exposure with E. coli. Thus, the full responsiveness towards the high stimulus PMA seems to be delayed. The biological important oxidative burst induced by E. coli has been rapidly and fully reconstituted on day +30.

The data evaluating the induction of oxidative burst following incubation with f-MLP did not reveal a time-dependent change of this functional capacity. Similar results were obtained for neutrophils and monocytes in patients after allogeneic PBSCT. Nevertheless, absolute values suggest an increased responsiveness of monocytes on day +90 after stem cell transplantation.

A potential correlation between high absolute numbers of CD14+CD16+ monocytes and the absence of chronic GvHD was observed. Of course, the small numbers of patients in this study have to be considered and therefore, these data are preliminary. Nevertheless, this observation allows to hypothesise about a preventive character of CD14+CD16+ monocytes for severe chronic GvHD. The interpretation of this observation seems to be rather difficult. The CD14+CD16+ monocytes are attributed to reflect the monocyte subpopulation with the highest production of proinflammatory cytokines like tumour necrosis factor (Belge et al. 2002). An increase of CD14+CD16+ monocytes had been described in severe infections, autoimmune disease or carcinomas (Ziegler-Heitbrock 1996; Kawanaka et al. 2002; Saleh et al. 1995). So far, profound data on a potential regulatory role of CD14+CD16+ cells that might explain a negative impact on such complex immune processes as GvHD have not been published.

With respect to the small cohort of patients in this study, the distribution of CD14+CD16+ cells and the occurrence of severe chronic GvHD might be a result of coincidence. Furthermore, it cannot be excluded that the low absolute CD14+CD16+ counts reflect only an epiphenomenon as a result of a higher rate of prolonged treatment with glucocorticoids (Fingerle-Rowson et al. 1998). Thus, the reconstitution of distinct subsets of monocytes following allogeneic stem cell transplantation should be evaluated in a prospective manner.

In conclusion, the reconstitution of neutrophil and monocyte functions in the early period after allogeneic PBSCT differs in terms of both phagocytosis and oxidative burst. In addition, the velocity of functional recovery depends on the stimulus used for oxidative burst analysis. Furthermore, it might be hypothesised that high numbers of CD16+CD14+ monocytes correlate with a reduced risk of chronic GvHD.