Immature neutrophils are elevated in human PGD and linked to G-CSF-driven injury in a murine model of lung ischemia–reperfusion

Abstract

Background:

Primary Graft Dysfunction (PGD) is an early post-lung transplant (LTx) inflammatory condition primarily driven by lung ischemia-reperfusion injury (LIRI). Neutrophils are key mediators of LIRI, but their phenotypic diversity and maturation state remain poorly characterized. In other inflammatory settings, early expansion of immature neutrophils has been linked to increased tissue injury and worse clinical outcomes. Whether immature neutrophils increase following LTx and contribute to PGD severity remains unclear.

Methods:

Circulating neutrophil heterogeneity was analyzed by flow cytometry in 20 LTx candidates with advanced lung disease and 30 LTx recipients. Matched plasma samples were used for cytokine profiling. The role of immature neutrophils in LIRI was studied using a murine left pulmonary hilar clamp model with or without anti-G-CSF treatment. The differentiation and effector functions of immature neutrophils derived from murine hematopoietic progenitors were studied in vitro.

Results:

LTx recipients exhibited an early rise in circulating immature neutrophils, correlated with higher G-CSF levels and PGD severity. In mice, LIRI was linked to increased G-CSF levels, significant mobilization, and lung infiltration of immature neutrophils with an activated, ROS-producing phenotype. These cells showed prolonged survival, strong ROS activity, but impaired phagocytosis. Preoperative anti-G-CSF treatment decreased lung injury while reducing immature neutrophil mobilization and recruitment to the lung.

Conclusions:

Our findings underscore the clinical significance of neutrophil heterogeneity in the early perioperative setting following LTx. Targeting the G-CSF-immature neutrophil axis may offer a novel strategy to improve early lung allograft outcomes.

INTRODUCTON

Primary Graft Dysfunction (PGD) is a form of early acute lung injury (ALI) following lung transplantation (LTx) and remains a major cause of post-transplant morbidity and mortality^1,2. Lung ischemia-reperfusion injury (LIRI) is the principal driver of PGD severity, triggering a sterile inflammatory response that rapidly recruits leukocytes to the graft^3,4. Experimental PGD models have demonstrated increased neutrophil mobilization and infiltration into the reperfused lung, where they exacerbate tissue damage by releasing reactive oxygen species (ROS) and other pro-inflammatory mediators^5–10. However, these studies often treat neutrophils as a homogeneous population of terminally differentiated granulocytes, potentially overlooking significant functional heterogeneity. Recent evidence challenges this conventional view, showing that acute inflammation promotes neutrophil heterogeneity by mobilizing immature neutrophils into the circulation^11–15. Although the functions of immature neutrophils remain incompletely understood, there is growing consensus that they can contribute to tissue injury in both infectious and “sterile” conditions^16–20. In this context, granulocyte colony-stimulating factor (G-CSF), a key regulator of granulopoiesis and neutrophil mobilization^21,22, is elevated in patients with ALI, and its association with worse clinical outcomes^23,24 suggests a link between G-CSF, immature neutrophils, and ALI severity. However, whether the mobilization of immature neutrophils early after LTx contributes to the severity of PGD remains unclear.

Here, we address this question by characterizing neutrophil heterogeneity in pre- and post-LTx settings, as well as in a murine model of LIRI. We found that the early accumulation of immature neutrophils in the peripheral blood of LTx recipients is linked to PGD severity and identify a G-CSF–immature neutrophil axis as a potential therapeutic target for PGD.

MATERIALS AND METHODS (detailed material and methods are described in the supplement)

Human study approval and design

This study was approved by the Institutional Review Board at St. Joseph’s Hospital (IRB# PHXB16-0027-10-18) and is compliant with ISHLT statement on transplant ethics. Study participants were enrolled between July 2023 and April 2024 at St. Joseph’s Hospital and Medical Center, Phoenix, (AZ). All subjects provided written informed consent. Whole Blood was collected at the time of the transplant evaluation for the pre-LTx group (n=20) and within 72 hours after reperfusion for the post-LTx group (n=30). Clinical and demographic characteristics were abstracted from electronic records (Table 1). PGD diagnosis and grades were determined in accordance with the latest ISHLT consensus¹.

Table 1.

Demographic and Clinical Characteristics

Characteristics	pre-LTX (n=20)	PGD 0-1 (n=20)	PGD 2-3 (n=10)	p-value a
Age (years) c	65.6 (63.8–73.2)	67.6 (62–72.7)	66.9 (64.2–70.5)	>0.9
Female (n, %) d	6 (30%)	5 (25%)	4 (40%)	0.7
Race d				0.08
Black or African American	0	1 (5%)	0
White	20 (100%)	15 (75%)	9 (90%)
Others	0	4 (20%)	1 (10%)
BMI (Kg/m2) e	29.2 ± 4.2	27.8 ± 4.7	29 ± 5.8	0.6
Primary Lung Disease d				0.8
ILD	12 (60%)	12 (60%)	5 (50%)
COPD	7 (35%)	5 (25%)	3 (30%)
CF	0	0	0
Others	1 (5%)	3 (15%)	2 (20%)
Comorbidities d
Coronary Artery Disease	7 (30%)	8 (40%)	2 (20%)	0.7
High Blood Pressure	9 (45%)	9 (45%)	4 (40%)	>0.9
Diabetes	3 (15%)	5 (25%)	1 (10%)	0.7
Chronic kidney Disease	2 (10%)	2 (10%)	1 (10%)	>0.9
History of Malignancy	3 (15%)	4 (20%)	2 (20%)	>0.9
History of Smoke	15 (75%)	13 (65%)	6 (60%)	0.7
Operative Characteristics		PGD 0-1 (n=20)	PGD 2-3 (n=10)	p-valueb
Intraoperative support ECLSd		2 (10%)	2 (20%)	0.6
Total Ischemic timef		421 ± 181	357 ± 137	0.3
Surgical procedure (BLT)d		20 (100%)	10 (100%)
Basiliximab + (TAC-MMF-steroids)d		20 (100%)	10 (100%)

Abbreviations: BMI, Body Mass Index; ILD, Interstitial Lung Disease; COPD, Chronic Obstructive Pulmonary Disease; CF, Cystic Fibrosis; ECLS, Extracorporeal Life Support; BLT, Bilateral Lung Transplant; TAC, Tacrolimus; MMF, Mycophenolate Mofetil

^ap value comparing pre-LTX vs PGD 01 vs PGD 2–3.

^bp value comparing PGD 01 vs PGD 2–3.

^cMedian (IQR) interquartile range; Kruskal-Wallis.

^dn(%); Fisher test.

^eMean ± (SD) standard deviation; one-Way ANOVA.

^fMean ± (SD) standard deviation; student`s t-test

Human flow cytometry

Surface staining was performed on 100 uL of whole blood within 4 hours from the collection. All samples were analyzed on the same Cytek^® Aurora station with QC program for ensuring control on potential longitudinal variability. The SpectroFlo unmixing algorithm, trained on a set of single-reference controls, was used to define positive and negative spectral signatures.

Human cytokines

Plasma cytokine concentration was assessed using the Cytokine Human Magnetic 25-Plex Panel (Invitrogen) on a Luminex 100/200 platform. G-CSF levels were measured using the Human G-CSF Quantikine ELISA Kit (R&D Systems) on a Biotek Epoch Microplate Spectrophotometer.

Animals and LIRI model

This study was performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals and was approved by the Institutional Animal Care and Use Committee at St. Joseph’s Hospital and Medical Center. An in vivo hilar clamp model^7,25,26 was used for inducing LIRI in 8 to 12 weeks old male C57BL/6J mice (Jackson laboratory). Mice underwent 1 hour of ischemia followed by 24 hours of warm reperfusion before euthanasia. In a series of experiments, 10 μg/g body weight of monoclonal Rat anti-mouse G-CSF (9B4CSF, eBioscience) or monoclonal Rat IgG2 Isotype control (eBR2a, eBioscience) was injected IV 1 hour before the surgery.

Mouse tissue processing and flow cytometry

Lung, blood, and bone marrow (femur and tibia) were collected and processed for flow cytometry following standard protocols.

Mouse cytokines

Plasma levels of G-CSF and IL-6 were measured using ELISA Kit-Quantikine (R&D System) on a Biotek Epoch Microplate Spectrophotometer.

Measures of ALI

Lung was collected and processed for histopathological analysis (H&E staining) and for assessment of pulmonary edema (wet-to-dry weight ratio).

Neutrophil differentiation and functional characterization

Bone marrow–derived hematopoietic progenitors were purified by immunomagnetic negative selection²⁷, and immature neutrophils were differentiated and functionally characterized in vitro.

Statistical analysis

Descriptive statistics are presented as mean ± standard deviation (SD) for normally distributed continuous variables and as median and interquartile range (IQR) for non-normally distributed continuous variables. Normality was assessed using the Shapiro-Wilk test. Fisher’s exact test was used for the analysis of contingency tables.

Differences between two groups were assessed using a Student’s t-test or the Mann-Whitney test as appropriate based on data distribution. For comparison involving more than two groups, one-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis with Dunn’s post hoc test were used as appropriate. Correlations between continuous variables were evaluated with Pearson’s correlation for normally distributed, linear data and Spearman’s rank correlation otherwise. Receiver operating characteristic (ROC) curve analysis was performed to assess sensitivity and specificity; AUC was reported and tested against the null hypothesis H0: AUC = 0.5. All the statistical analyses and figures were calculated and prepared using GraphPad Prism 10.4.1. For all the analyses, a P value less than 0.05 was considered statistically significant.

RESULTS

CD10^neg immature neutrophils increase early after LTx

To assess early post-LTx neutrophil phenotypic changes, we performed flow cytometry on fresh whole blood collected from 20 LTx candidates with advanced lung disease (pre-LTx) and 30 LTx recipients within 72 hours of reperfusion (post-LTx). As expected, post-LTx subjects had a higher number of white blood cells (WBC) (Fig.1A) and neutrophils (Fig.1B). Neutrophils were next characterized for markers of maturity and activation (Fig.1C). Relative to pre-LTx, post-LTx neutrophils had reduced levels of late-stage differentiation markers, including CD10 (p<0.0001), CD11b (p=0.0002), CXCR4 (p= 0.005), and PD-L1 (p = 0.02)²⁸. Post-LTx neutrophils had high expression of integrin alpha 4 (CD49d) (p = 0.02)²⁸, consistent with an immature granulocyte phenotype. Additionally, receptors that regulate migratory behavior, such as CXCR1 (p = 0.001)²⁹ and CD177 (p = 0.007)³⁰, were also elevated. CD10 is a cell surface zinc-dependent metalloprotease that is specifically expressed at later stages of neutrophil differentiation, and is widely used as a reliable marker to distinguish mature (CD10⁺) from immature (CD10^neg) neutrophils^{16,18,19,31–33}. Consistently with previous studies^16,19, CD10⁺ neutrophils from LTx recipients had a segmented, mature nuclear morphology, while CD10^neg neutrophils displayed a non-segmented immature nuclear morphology (Fig. 1D). Additionally, lack of CD10 expression identified circulating neutrophils with reduced apoptosis and phagocytosis but intact ROS activity, a functional phenotype we independently verified and found consistent with previous reports on immature neutrophils^16,19 (Fig.S1). Given this phenotypic and functional distinction, we utilized CD10 expression to assess levels of immature neutrophils in pre- and post-LTx recipients (Fig. 1E). Relative to pre-LTx, immature neutrophils were significantly more abundant in post-LTx recipients (3.1% [0.6–11.7] vs 42.4% [20.9–60.9]; p<0.0001). This significant increase persisted when the analysis was restricted to a small subset of matched pre- and post-LTx samples (n=10; p<0.01) (Fig. S2).

Figure 1. CD10^neg immature neutrophils increase early after LTx.

Whole blood was analyzed by flow cytometry in subjects with advanced lung disease at the time of LTx evaluation (pre-LTx, n=20) and from LTx recipients within 72 hours post-transplant (post-LTx, n=30). (A) Total white blood cell (WBC) counts in pre and post-LTx groups; (mean ± SD; t-test). (B) Circulating neutrophil (PMN) frequency (mean ± SD; t-test) and absolute counts (box and whisker plot; Mann-Whitney test) in pre and post-LTx groups. (C) MFI of neutrophil surface markers in pre and post-LTx groups; (Heatmap of Z-score; Mann-Whitney test). (D) Representative contour plots distinguishing mature CD10⁺ and immature CD10^neg neutrophils alongside Giemsa-stained images of sorted populations highlighting segmented nuclei in CD10⁺ and band-shaped nuclei in CD10^neg cells with corresponding quantification (mean ± SD). (E) Frequency of CD10^neg immature neutrophils in pre and post-LTx groups; (box and whisker plot; Mann-Whitney test).

For all analysis, *p<0.05, **p< 0.01, *** p<0.001, ****p< 0.0001.

Noting previous work which showed that immature neutrophil expansion could be due to infection³⁴, we examined whether positive microbiological cultures from the donor or recipient were linked to levels of CD10^neg neutrophils (Fig. S3). We found that immature neutrophil levels did not differ based on the presence or absence of positive cultures in either the donor or the recipient. Taken together, these findings reveal that early post-LTx neutrophil heterogeneity is characterized by the accumulation of immature neutrophils in the peripheral blood.

The accumulation of peripheral blood immature neutrophils is linked to inflammatory cytokine release

The expansion of immature neutrophils has been reported in different human acute inflammatory conditions^16–18. To assess the inflammatory environment associated with LTx and immature neutrophil expansion, we measured the levels of 26 cytokines in plasma samples matched to the whole blood specimens used for flow cytometric analysis, obtained from 10 pre-LTx and 28 post-LTx subjects (Table S1). Compared to pre-LTx, the post-LTx group exhibited significantly increased plasma levels of IL-6 (p<0.0001), G-CSF (p=0.0001), IL-2R (p<0.0001), IL-13 (p<0.0001) and IL-10 (p=0.01), along with a significant decrease in Eotaxin (p<0.0001) and MCP-1 (p=0.03).

Among the cytokines elevated post-LTx, the expansion of immature neutrophils significantly correlated with IL-6 [(p=0.02; r= 0.35), (Fig. 2A)], G-CSF [(p=0.03; r=0.34), (Fig. 2B)], IL-2R [(p=0.0008; r=0.52), (Fig. 2C)], and IL-13 [(p=0.0016; r=0.49), (Fig. 2D)]. In contrast, no significant correlation was observed with IL-10 (Fig. 2E). A negative, but not statistically significant correlation was noted with Eotaxin (Fig. 2F) and MCP-1 (Fig. 2G). These findings indicate that early expansion of CD10^neg immature neutrophils after LTx correlates with distinct post-transplant cytokine changes, consistent with stress-induced mobilization and inflammatory signaling.

Figure 2. Plasma cytokine levels and their correlation with immature neutrophil frequency.

Plasma from the same whole blood samples used for flow cytometry was analyzed for cytokine quantification in 10 pre-LTx and 28 post-LTx subjects. Cytokines shown were significantly altered following LTx and include: (A) IL-6, (B) G-CSF, (C) IL-2R, (D) IL-13, (E) IL-10, (F) Eotaxin, and (G) MCP-1. For each cytokine, the box and whisker plot shows the plasma concentrations in Pre vs Post-LTx groups (Mann-Whitney test), while the scatter plot shows the correlation with immature neutrophil frequency (Spearman’s correlation). For all analysis *p < 0.05; **p < 0.001; ***p < 0.0001.

Immature neutrophil expansion is associated with PGD severity

Given the variable increase in immature neutrophils observed across our LTx recipients and their positive correlation with inflammatory markers, we next asked whether this increase was associated with the severity of PGD. PGD was assessed at 72 hours after ICU admission and graded based on the presence of pulmonary edema on chest X-ray and the degree of hypoxemia, as measured by the PaO2/FiO2 ratio¹. LTx recipients were further stratified into mild (PGD 0–1; n=20) and severe (PGD 2–3; n=10) groups^35–37. While both groups showed a significant increase in immature neutrophils compared to pre-LTx levels, LTx recipients in the severe group had significantly higher frequencies than those in the mild group (Mild = 33.8% [14.4–43.9] vs severe 63.6% [43.5–72.2]; p=0.04) (Fig. 3A).

Figure 3. Higher circulating immature neutrophil levels are associated with PGD severity and ICU stay duration.

Post-LTx subjects were stratified based on their T72-PGD score into a mild (PGD 0–1; n=20) and a severe (PGD 2–3; n=10) group (A) Circulating immature neutrophil frequency in pre-LTx vs PGD 0–1 vs PGD 2–3; (box and whisker plots; Kruskal-Wallis with Dunn’s post hoc test). (B) ROC curve analysis evaluating the performance of immature neutrophil frequency in discriminating PGD severity. (C) Correlation between immature neutrophil frequencies with ICU stay. The correlation included subjects with ICU stay <14 days (n=26) and the p value was calculated using Pearson correlation. For all analysis *p<0.05, *** p<0.001, ****p<0.0001.

To further evaluate the ability of early immature neutrophil expansion to differentiate between mild and severe PGD, we conducted a receiver operating characteristic (ROC) analysis. The area under the curve was 0.83 (95% CI= 0.66–0.99) with p = 0.003 indicating that a high proportion of immature neutrophils is a prescient early indicator of severe PGD development (Fig. 3B). The early rise in immature neutrophils also positively correlated with prolonged ICU days suggesting that this early increase may contribute to a more complicated clinical course in LTx recipients; this association was statistically significant when we analyzed patients with ICU stays <14 days (p=0.03; r=0.4; n=26) (Fig. 3C) and remained consistent when patients with extreme ICU durations were also included (p=0.02; r=0.4; n=30) (Fig. S4).

Activated ROS-producing immature neutrophils accumulate in the lung during IRI

In mice, immature neutrophils are distinct by the lack of the maturation marker CD101 and, like in humans, are mobilized into the blood in response to lung infections^16,20,38,39. To further understand the role of immature neutrophils in PGD, we utilized a murine left lung hilar-clamp model, which reproduces many of the histological, immunological, and functional aspects of severe PGD^7,25,26. C57Bl/6 (B6) underwent 1 hour of left lung warm ischemia followed by 24 hours of reperfusion (Fig. 4A). In agreement with previous studies, neutrophils accounted for approximately 10% of CD45⁺ leukocytes in naïve lung⁴⁰. Phenotypic analysis revealed that only a small fraction of these neutrophils exhibited an immature CD101^neg phenotype (Fig. 4B), resulting in immature neutrophils comprising less than 0.5% of the total pulmonary leukocyte population under homeostatic conditions. However, at 24 hours after reperfusion, the lungs were densely infiltrated by immature neutrophils, which significantly outnumbered their mature counterparts (Fig. 4C). Compared to circulating immature neutrophils, those recruited to the lung after reperfusion exhibited more phenotypic signs of activation, including increased CD11b expression (Fig. 4D) and shedding of CD62L (Fig. 4E). Functionally, lung-infiltrating immature neutrophils exhibited higher ROS production (Fig. 4F), indicating a heightened ability to injure tissue. This prominent accumulation of activated, ROS-producing immature neutrophils was associated with signs of ALI, including inflammatory infiltrates, alveolar disruption, and interstitial thickening (Fig. 4G). Additionally, lung vascular permeability was significantly increased, as reflected by a higher wet-to-dry weight ratio (Fig. 4H). While immature neutrophils were also detected in the lungs of sham-operated mice, likely reflecting inflammatory responses to surgical stress and mechanical ventilation^41,42, their abundance was significantly lower, and their localization was mainly intravascular; their presence did not coincide with signs of ALI. (Figure S5).

Figure 4. Immature neutrophil expansion and functional impairment in a murine model of LIRI.

Phenotypic and functional characteristics of neutrophils were studied in a murine model of LIRI induced by left pulmonary hilar clamp. (A) Mice underwent 1 hour of lung ischemia followed by 24 hours of reperfusion. (B) Representative contour plot showing pulmonary neutrophils (Ly6G⁺CD11b⁺ within the Live CD45⁺ cells) from Naïve and LIRI mice. Neutrophils were further stratified into immature (CD101^neg) and mature (CD101⁺) subsets. (C) Quantification of lung immature and mature neutrophils is shown as both percentage of CD45⁺ cells and absolute number (Naïve in white, n=6; LIRI in orange, n=6; mean ± SD; One-Way ANOVA with Tukey’s post hoc test). (D-E) Adhesion molecules CD11b (D) and CD62L (E) were compared in immature neutrophils from the blood and the lung as a sign of activation during LIRI; representative histograms show MFI in the blood (red, n=6) and the lung (blue, n=6); the quantification is presented in the corresponding bar graph (mean ± SD; t-test); (F) ROS production in immature (white) vs mature (gray) lung neutrophils from LIRI mice. Data presented as DHR123 fold MFI (mean ± SD; t-test). (G) Representative H&E-stained lung sections. Lower panels: 200x (scale bar, 100 μm); upper panels from boxed regions 400x (scale bar, 50 μm). ALI was scored in Naïve (white; n = 5) and LIRI (orange; n = 5) mice using a standardized quantification system; (median/IQR; Mann-Whitney U test). (H) Lung edema was quantified by wet/dry ratio in Naïve (white; n=5) and LIRI (orange; n=5) mice; (median/IQR; Mann-Whitney test). For all analysis *p<0.05, ** p<0.01, ****p<0.0001.

These findings demonstrate that the extensive recruitment of activated, ROS-producing immature neutrophils plays a role in exacerbating pulmonary damage following LIRI.

LIRI drives accelerated granulopoiesis and mobilization of immature neutrophils from the bone marrow

IL-6 and G-CSF promote the mobilization of neutrophils from the bone marrow^22,43. Given the elevated IL-6 and G-CSF plasma levels correlating with immature neutrophil expansion observed in human LTx recipients, we asked whether these observations could be reproduced in the mouse hilar clamp model (Figs. 5A, B). Mouse LIRI resulted in high plasma levels of IL-6 (p = 0.0005) and G-CSF (p = 0.0003). Additionally, more than half of the circulating neutrophils exhibited an immature phenotype one day after reperfusion (Figs. 5C, D).

Figure 5. LIRI increases G-CSF levels and drives emergency granulopoiesis.

(A-B) Plasma levels of IL-6 (A) and G-CSF (B) in naïve (white; n=6) and LIRI (orange; n=5) mice. (C) Representative contour plots of circulating neutrophils (Ly6G⁺CD11b⁺ within live CD45⁺ cells) in blood stratified into immature (CD101^neg) and mature (CD101⁺) subsets. (D) Quantification of blood mature and immature neutrophils shown as percentage of CD45⁺ cells and absolute counts in Naïve (white; n=7) and LIRI (orange; n=7) mice. (E) Gating strategy for bone marrow neutrophil maturation in Naïve (white; n=7) and LIRI (orange; n=7) mice. After excluding Lin⁺ cells, eosinophils (SSC^high), and megakaryocyte-erythroid progenitors (c-Kit⁺-CD34^low), distinct neutrophil maturation stages were identified as shown, based on patterns of c-Kit and Ly6G expression. Defined stages include: Myeloblasts (c-Kit^high,Ly6G^neg); Pro-Myelocytes (c-Kit^int,Ly6G^neg); Myelocytes (c-Kit^neg/low,Ly6G^low); Meta-Myelocytes (c-Kit^neg/low,Ly6G^int); Neutrophils (c-Kit^neg,Ly6G⁺). Neutrophils were further stratified into Immature (Ly6G⁺,CD101^neg) and Mature (Ly6G⁺,CD101⁺). Dot plots show percentage of each stage among live bone marrow cells. (F) Representative histogram and quantification of CXCR2 expression on bone marrow immature neutrophils of Naïve (white; n=7) and LIRI (orange; n=7) mice. Data are presented as mean ± SEM. Multiple group comparisons were performed using one-way ANOVA with Tukey’s post hoc test. Two-group comparisons were analyzed by t-test.

For all analysis *p<0.05, **p< 0.01, *** p<0.001, ****p< 0.0001.

To determine if alterations in bone marrow granulopoiesis were linked to the expansion of immature neutrophils, we applied a flow cytometry-based gating strategy previously validated for dissecting granulopoiesis in mice (Fig. 5E)^39,44. We found that bone marrow from mice that underwent LIRI had a sharp reduction in neutrophil precursors at the myeloblast and pro-myelocyte stage of development. This was accompanied by a substantial accumulation of more differentiated precursors including myelocytes and metamyelocytes, possibly indicative of accelerated differentiation. In contrast, neutrophils, identified by their Ly6G positivity, were sharply reduced, consistent with the enhanced mobilization observed in the blood. Further stratification based on the maturation marker CD101 revealed that both immature (CD101^neg) and mature (CD101⁺) neutrophil subsets were substantially reduced following LIRI. CXCR2 promotes the egress of mature neutrophils from the bone marrow⁴⁵, but its expression on immature neutrophils is low or absent under homeostatic condition⁴⁶. Unlike immature neutrophils from naive mice, CXCR2 levels were markedly elevated in immature neutrophils from mice that underwent LIRI (Fig. 5F). Overall, our data indicate that LIRI promotes accelerated maturation of myeloid precursors and enhanced mobilization of immature neutrophils from the bone marrow.

In Vitro differentiation and functional properties of immature neutrophils

To study the phenotype and function of immature neutrophils across maturation stages, we induced neutrophil differentiation ex vivo of bone marrow–derived c-Kit⁺ hematopoietic progenitors isolated from naive B6 mice²⁷. After an initial expansion phase (EP) of hematopoietic progenitors driven by Stem Cell Factor (SCF) and IL-3 stimulation, neutrophil differentiation was induced by G-CSF (Fig. 6A). By day 4 (D4), c-Kit expression was lost and levels of neutrophil differentiation markers such as Ly6G, increased (Fig. 6B). Among differentiated neutrophils, the large majority displayed phenotypic and morphological characteristics of immaturity, including absent or low CD101 expression, and a characteristic ring-shaped nucleus (Fig. 6C). This immature population bound less Annexin V compared to mature neutrophils, indicative of reduced spontaneous apoptosis (Fig 6D). Analysis of ROS production showed comparable levels in immature and mature neutrophils (Fig. 6E). However, the phagocytic capacity of immature neutrophils was markedly compromised relative to mature neutrophils, suggesting an inability of immature neutrophils to remove tissue debris (Fig. 6F). By day 7, most neutrophils completed their maturation and displayed segmented nuclei, CD101 expression and high CD11b and CXCR2 levels (Figs. 6B, C).

Figure 6. In vitro differentiation and functional characterization of immature neutrophils.

(A) Lineage-negative (Lin^neg) c-Kit⁺ hematopoietic progenitor cells were isolated from naïve wild-type mice and cultured with SCF and IL-3 to promote early progenitor expansion (EP). G-CSF was added on day 0 (D0) to initiate neutrophil differentiation, and cells were maintained in G-CSF alone from day 2 to day 7 (D7) to support maturation. (B) Representative histograms demonstrating progressive phenotypic maturation from D0 to D7. (C) Representative Giemsa staining at D0, D4, and D7 showing morphological changes, alongside flow cytometry plots indicating CD101⁺ cell frequency within the neutrophil population. (D) Apoptosis assay: representative dot plot and quantification of Annexin V⁺ cells in immature vs mature neutrophils at D4. (E) ROS assay: DHR123 histograms and quantification of MFI fold change in immature vs mature neutrophils following PMA stimulation at D4. (F) Phagocytosis assay: dot plots and quantification of opsonized Fluoresbrite bead uptake by immature and mature neutrophils incubated at 4°C or 37°C at D4. All data were obtained from triplicate measurements and are representative of at least 2 independent experiments. Results are presented as mean ± SD. Multiple group comparisons were analyzed using one-way ANOVA with Tukey’s post hoc test. Two-group comparisons were analyzed by unpaired t-tests. For all analysis (*p<0.05, ** p<0.01, ****p< 0.0001).

Collectively, these findings show that CD101^neg immature neutrophils derived from bone marrow–resident hematopoietic progenitors have reduced spontaneous apoptosis, limited scavenging activity, but preserved oxidative burst capacity.

G-CSF-driven mobilization and lung accumulation of immature neutrophils are linked to severe LIRI

Prior studies show that G-CSF blockade reduces neutrophilia and mitigates ALI^5,47,48. However, it remains unclear whether protection from lung injury results from limiting the mobilization of distinct subsets. To address this question, we treated mice with a G-CSF neutralizing antibody or isotype control antibody and profiled neutrophil subsets in the lung, blood, and bone marrow compartments. Given the early increase in G-CSF and immature neutrophils levels observed in time course experiments of LIRI (Figure S6), anti-GCSF was administered one hour prior to LIRI surgery. Consistent with previous reports^5,48, pre-operative G-CSF blockade significantly inhibited lung neutrophilia (Fig. 7A). However, subset analysis showed that this reduction was primarily due to the loss of immature neutrophils, with no significant effect on mature neutrophil numbers (Fig. 7B). Additionally, G-CSF blockade attenuated histopathologic signs of injury (Fig. 7C) and improved pulmonary edema (Fig. 7D), supporting the notion that immature neutrophils promote LIRI.

Figure 7. G-CSF blockage improves LIRI while selectively blocking immature neutrophil mobilization and lung recruitment.

Mice received anti-G-CSF (a-G-CSF) or isotype control antibody (Isotype) 1 hour before pulmonary hilar clamp. At 24 hours post-reperfusion, lungs, blood, and bone marrow were collected for analysis. (A) Representative contour plots of lung neutrophils (Ly6G⁺CD11b⁺) gated from live CD45⁺ cells in naïve and LIRI mice. Neutrophils were divided into immature (CD101^neg) and mature (CD101⁺) subsets. (B) Quantification of immature and mature neutrophils is shown as percentage of CD45⁺ cells and absolute number per mg of lung tissue in Isotype (brown, n=6) and a-G-CSF (beige, n=6) groups. (C) Representative H&E-stained lung sections at 200x (scale bar = 100 μm) and 400x (scale bar = 50 μm, boxed regions). ALI was scored using a standardized system; results shown for Isotype (brown, n=5) and a-G-CSF (beige, n=5). (D) Lung edema quantified by wet/dry weight ratio in Isotype (brown, n=5) and a-G-CSF (beige, n=5) mice. (E) Representative plots of circulating neutrophils (Ly6G⁺CD11b⁺) within live CD45⁺ blood leukocytes, stratified into immature and mature subsets. (F) Quantification of immature and mature neutrophil as percentage of CD45⁺ cells and absolute counts in Isotype (brown, n=7) and a-G-CSF (beige, n=7) groups. (G) Bone marrow neutrophil maturation stages quantified as percentage of live marrow cells using c-Kit and Ly6G expression: Myeloblasts* (c-Kit^high,Ly6G^neg); Pro-Myelocytes (c-Kit^int, Ly6G^neg); Myelocytes (c-Kit^neg/low,Ly6G^low); Meta-Myelocytes (c-Kit^neg/low,Ly6G^int); Neutrophils (c-Kit^neg,Ly6G⁺). Neutrophils were further stratified into Immature (Ly6G⁺,CD101^neg) and Mature (Ly6G⁺,CD101⁺). Data shown for Isotype (brown, n=7) and a-G-CSF (beige, n=7) groups; (H) Representative histogram and quantification of CXCR2 MFI on bone marrow immature neutrophils from Isotype (brown, n=7) and a-G-CSF (beige, n=7) mice. Data are presented as mean ± SD. Multiple group comparisons were analyzed using one-way ANOVA with Tukey’s post hoc test. Two-group comparisons were analyzed using t-tests. *Myeloblast comparison is shown as median/IQR and analyzed using Mann-Whitney test.

For all analysis *p < 0.05, **p < 0.01, ***p < 0.001.

Consistent with findings in the lung, G-CSF blockade prevented peripheral blood neutrophilia following LIRI primarily by preventing the accumulation of immature neutrophils (Figs. 7E, F). Interestingly, bone marrow analysis revealed that anti-G-CSF did not substantially alter the frequency of early neutrophil precursors, but in the absence of effective mobilization, immature neutrophils significantly accumulated in the marrow (Fig 7G). This selective impairment of immature neutrophil egress was accompanied by a reduction in their CXCR2 expression (Fig. 7H). Collectively, our findings indicate that, in the context of LIRI, G-CSF-dependent lung injury is associated with enhanced mobilization and lung recruitment of immature neutrophils.

DISCUSSION

Excessive neutrophils play a crucial role in worsening PGD^5,7, yet their phenotypic diversity early after LTx remains poorly understood. While earlier studies have found immature neutrophils at later stages of LTx, particularly in cases of chronic lung allograft dysfunction (CLAD) or infection^49,50, our study uniquely shows a rapid post-LTx increase in immature neutrophils that strongly link to the severity of PGD.

CD10 has emerged as a reliable marker for distinguishing between immature and mature neutrophils, correlating well with nuclear morphology in both microbial and “sterile” inflammatory settings^16,18,31,32. In our study, CD10^neg neutrophils predominantly displayed immature non segmented nuclei, confirming CD10 as a reliable marker to identify immature neutrophils in the setting of LTx. Immature neutrophils have been extensively studied in infectious diseases, where their expansion has been linked to exaggerated immune responses and poorer clinical outcomes^{17,18,20,39,51}. However, more recent findings indicate that immature neutrophil expansion may represent a conserved immunological response to systemic inflammation rather than a specific response to infection. For example, Marini et al. showed that G-CSF stimulation in healthy human donors was sufficient to mobilize CD10^neg neutrophils with immature nuclear morphology and pro-inflammatory properties³¹. Moreover, the early accumulation of CD10^neg immature neutrophils was recently demonstrated in patients with acute myocardial infarction, a condition exacerbated by ischemia-reperfusion injury¹⁶.

Building on these observations, we examined whether LTx leads to an expansion of immature neutrophils. In our cohort, immature neutrophils increased markedly after transplantation. Although not all pre- and post-LTx samples were matched, the magnitude and consistency of our observations support a genuine post-transplant expansion unlikely to be explained by baseline variability alone. This increase in immature neutrophils was especially pronounced in patients with elevated G-CSF and IL-6, whose high post-LTx levels appeared consistent with emergency granulopoiesis. The concurrent post-LTx reductions in eosinophil- and monocyte-attracting chemokines further suggested a shift towards neutrophilic responses underscoring their role as early cellular effectors. In this study we therefore focused on G-CSF to investigate mechanisms of immature neutrophil mobilization and to evaluate the potential of G-CSF blockade as a therapeutic strategy. However, given the growing number of clinical trials targeting IL-6 signaling in solid organ transplant⁵², dedicated future studies are warranted to define the specific contribution of IL-6 to immature neutrophil differentiation and mobilization after LTx. Given the early post-LTx increase in immature neutrophils and their association with an inflammatory environment, we then asked whether their expansion was related to clinical outcomes. CD10^neg neutrophils were significantly higher in recipients who developed severe PGD and had longer ICU stays. These differences occurred despite all recipients undergoing bilateral LTx with uniform induction and immunosuppressive protocols, minimizing potential confounding effects from perioperative therapeutic interventions. The increase in immature neutrophils was also independent of infection status, as high levels were observed in patients with negative microbiological cultures. However, due to the high risk of infection, LTx recipients usually receive a potent antibiotic prophylaxis that may lead to gut dysbiosis. Short Chain Fatty Acids (SCFA) produced by the gut microbiota have been recently shown to regulate the balance between immature and mature neutrophils in homeostatic conditions⁵³. Moreover, gut microbiota appear to control ALI/ARDS severity⁵⁴ as well as lung allograft stability⁵⁵. Unraveling the intricate connections among gut microbiota, immature neutrophils and early lung graft injury may offer, in the future, novel strategies for improving clinical outcomes.

Overall, our data identify both a robust post-LTx expansion of immature neutrophils and its association with PGD severity, implicating these cells as potential contributors to post-transplant ischemia–reperfusion injury.

Prompted by our observations in human lung recipients, we examined the role of immature neutrophils in a murine model of LIRI using the well-established model of left pulmonary hilar clamp^7,25,26. Although this model lacks cold ischemia and alloimmunity, it reliably reproduces key immunological and histopathological features of LIRI observed in LTx and is therefore widely used to study innate immune responses relevant to PGD pathogenesis⁵⁶. Consistently, we found that elevated plasma G-CSF levels and neutrophil trafficking to the lung in our model were similar to those reported in murine orthotopic LTx models^5,48, further supporting the notion that immune responses to LIRI are conserved across different surgical contexts. Using CD101 as a marker to identify murine immature neutrophils^16,20,38,39, we found that most of the lung-infiltrating neutrophils had an immature phenotype. While this has been noted in infection models^20,39, our study is the first to implicate immature neutrophils in ALI secondary to LIRI. Interestingly, immature neutrophils infiltrating the perfused lung displayed higher ROS responses than mature neutrophils, suggesting metabolic hyperactivity within the inflammatory lung microenvironment and a strong potential for tissue damage. In line with our findings, previous work shows that ROS-producing immature neutrophils increase in systemic vasculitis where they promote vascular damage and increased permeability of the endothelial barrier¹⁹. However, despite their robust oxidative capacity, immature neutrophils displayed impaired phagocytosis. This observation also aligns with previous reports^16,19,57 and suggests that immature neutrophils may contribute to lung injury not only through ROS generation but also by failing to effectively clear cellular debris, which is critical to help resolve inflammation and promote tissue repair⁵⁸. Immature neutrophils released after acute myocardial infarction are resistant to apoptosis¹⁶. Similarly, we found that, immature neutrophils have reduced apoptosis compared with their mature counterpart. Delayed apoptosis amplifies neutrophil-mediated tissue damage prolonging the local release of cytotoxic mediators including ROS⁵⁹. Notably, we observed that the prolonged survival of immature neutrophils was accompanied by their retained capacity to differentiate in vitro into mature neutrophils, suggesting that immature neutrophils may function as a local reservoir of activated cells capable of sustaining ongoing inflammation.

CXCR2 is a key driver of mature neutrophil egress from the bone marrow⁴⁵; however, its role in the mobilization of immature neutrophils remains less defined. Under homeostatic conditions, immature neutrophils typically express low or absent levels of CXCR2, which contributes to their retention in the bone marrow⁴⁶. Our data show that LIRI induces G-CSF-dependent CXCR2 upregulation in immature neutrophils, which is associated with their premature mobilization. The relevance of the CXCR2 axis in neutrophil mobilization is supported by previous work demonstrating that G-CSF fails to induce neutrophil release in CXCR2-deficient mice⁴⁵. Moreover, circulating immature neutrophils displaying abnormally high CXCR2 levels have been recently reported in the context of severe COVID-19¹⁸, reinforcing the concept that inflammatory cues can override normal maturation-dependent regulation of CXCR2 and promote the accelerated trafficking of immature cells. Blocking G-CSF activity retained immature neutrophils in the marrow and reduced ALI, supporting their contribution to lung damage. This aligns with clinical data linking high G-CSF to poor outcomes in ARDS^23,24, implicating immature neutrophils as key effectors in G-CSF–driven injury.

Altogether, our preclinical in vivo and in vitro findings demonstrate that ROS producing immature neutrophils accumulate in the lung during pulmonary ischemia-reperfusion because of G-CSF–driven mobilization from the bone marrow. G-CSF blockade reduced their CXCR2 expression, impaired their egress and lung infiltration, and ultimately attenuated ALI.

Although our study offers new insights into early post-LTx neutrophil heterogeneity and the role of immature neutrophils in LIRI, it has some limitations. These include using a mouse model that may not fully reflect the complexity of human LTx⁵⁶. We chose to focus on early pulmonary injury without examining the long-term fate of recruited immature neutrophils and their potential effects on lung survival. To this end, G-CSF administration to human lung recipients with mild neutropenia has been reported to increase the risk of CLAD⁶⁰ and prevent the induction of tolerance in the mouse orthotopic lung transplant⁴⁸. Long-term clinical studies will be necessary to validate immature neutrophil accumulation as an early biomarker for poor post-LTx outcomes and to assess whether therapeutic targeting of the G-CSF can improve graft function and patient recovery.

Abbreviations

ALI

Acute Lung Injury

BLT

Bilateral Lung Transplant

CF

Cystic Fibrosis

COPD

Chronic Obstructive Pulmonary Disease

ECLS

Extracorporeal Life Support

G-CSF

Granulocyte Colony-Stimulating Factor

ICU

Intensive Care Unit

ILD

Interstitial Lung Disease

IQR

Interquartile Range

LIRI

Lung Ischemia-Reperfusion Injury

LTx

Lung Transplantation

MFI

Median Fluorescence Intensity

MMF

Mycophenolate Mofetil

PGD

Primary Graft Dysfunction

PMN

Neutrophils

ROS

Reactive Oxygen Species

SCF

Stem Cell Factor

SCFA

Short Chain Fatty Acid

SD

Standard Deviation

TAC

Tacrolimus

WBC

White Blood Cell