Culture Dimensionality Regulates Protein Expression and Bioactivity in THP-1-Derived Macrophages

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study investigated how culture dimensionality (2D vs. 3D) and extracellular matrix composition (Matrigel vs. type I collagen) influence the phenotype, function, and post-translational modifications of THP-1-derived macrophages. Several concerns should addressed.

The manuscript did not clearly state the number of biological replicates for any experiment.

While the study identifies differences between Matrigel and collagen, it does not explore the underlying mechanisms.

The intriguing finding of altered serglycin glycosylation in 3D is presented without functional follow-up. Does this altered glycosylation affect serglycin's ability to bind and present cytokines?

The expression of markers like CD80 (high in M0) and CD206 (higher in M0 2D vs. 3D) did not strictly align with classical M1/M2 definitions. The authors acknowledge this but do not sufficiently discuss the implications.

 

Author Response

Comment 1: The manuscript did not clearly state the number of biological replicates for any experiment.
Response 1: We apologize for this error. The number of replicates has now been indicated in the figure legends for all figures that include statistical analysis (highlighted in yellow). 

Comment 2: While the study identifies differences between Matrigel and collagen, it does not explore the underlying mechanisms.
Response 2: We thank the reviewer for this insightful comment. It is true that, in the present study, we did not conduct additional experiments to investigate the underlying mechanisms responsible for the observed differences in the survival of THP-1-derived macrophages cultured in Matrigel versus collagen. However, we have thoroughly discussed the possible explanations for this observation in the Discussion section (Lines 387–409). We are currently planning further experiments to elucidate the underlying mechanisms. In the Discussion, we wrote:
Multiple factors likely contributed to this phenomenon. One possible explanation is that the composition of Matrigel is not macrophage-supportive. Matrigel is primarily composed of laminin, collagen IV, entactin, and growth factors, which mimic the basement membrane rather than interstitial tissue[21]. In vivo, macrophages typically reside and migrate within interstitial matrices rich in fibrillar type I collagen, not in basement membrane-like environments[22]. Moreover, the crosslinked basement membrane proteins presented in Matrigel could have resisted the degradation by macrophage-secreted proteases[23]. Hence, macrophages could not remodel the matrix; they would become physically constrained and metabolically suppressed. In contrast, collagen is a natural macrophage substrate and is readily remodeled by macrophage-secreted proteases such as MMPs and cathepsins. Additionally, macrophages rely on integrin-mediated adhesion signals (e.g., β1/β2 integrins) for survival and activation, and type I collagen is known to strongly support these interactions[24]. Thus, ECM ligands presented in Matrigel probably could not efficiently engage macrophage adhesion receptors and reduced adhesion signaling that triggers anoikis-like responses or quiescence[25]. These findings suggested that matrix biochemical composition and mechanical properties differentially influenced macrophage survival and cytoskeletal organization. Since collagen is a major structural component of interstitial tissues, our results suggested that it provided adhesive ligands and mechanical cues that better recapitulated native macrophage niches, thereby enabling phenotypic features that were absent or attenuated in Matrigel or planar systems. The improved growth of M1 and M2 macrophages in collagen further indicated that matrix-specific signaling supports subtype-dependent adaptation.

Comment 3: The intriguing finding of altered serglycin glycosylation in 3D is presented without functional follow-up. Does this altered glycosylation affect serglycin's ability to bind and present cytokines?
Response 3: We thank the reviewer for this insightful question. Based on our Western blot analysis in Figure 5, we observed that THP-1-derived macrophages cultured in collagen gels exhibited glycosaminoglycans (GAGs), including dermatan sulfate (DS), chondroitin sulfate (CS), and heparan sulfate (HS), at higher molecular weights compared to those cultured in 2D conditions. This suggests that proteoglycans in 3D-cultured cells are more extensively glycosylated and therefore more negatively charged, which may enhance their capacity to bind and retain cytokines.

Moreover, cytokine expression varied depending on macrophage subtype. As shown in Figure 4, M2 macrophages cultured in 3D collagen exhibited higher expression of proangiogenic factors such as FGF2 compared to 2D cultures, suggesting that proteoglycans in these cells may preferentially interact with proangiogenic cytokines. In contrast, TNF-α expression was significantly higher in M1 macrophages cultured in 3D, indicating that its retention and presentation may also be influenced by highly glycosylated proteoglycans such as serglycin.

In the Discussion section, we have added the following statements (Line 447-457, highlighted in yellow):
The higher molecular weight distribution of CS, DS, and HS in THP-1-derived macrophages cultured in 3D collagen suggests increased glycosylation, which may enhance cytokine-binding affinity and influence cellular bioactivity. The enhanced secretion of serglycin in 3D cultures, particularly in M1 and M2 macrophages, supports the concept that extracellular matrix architecture can regulate intracellular trafficking and secretory pathways. Thus, structural context may modulate immune signaling not only at the transcriptional level but also through post-translational modification of key regulatory molecules.

Comment 4: The expression of markers like CD80 (high in M0) and CD206 (higher in M0 2D vs. 3D) did not strictly align with classical M1/M2 definitions. The authors acknowledge this but do not sufficiently discuss the implications.

Response 4: We thank the reviewer for this important comment. We agree that the expression patterns of certain markers, such as CD80 and CD206, do not strictly conform to the classical M1/M2 framework. These observations suggest that macrophage phenotypes exist along a continuum rather than as discrete M1 or M2 states, and that marker expression is highly context-dependent. A recent publication has shown that primary monocytes cultured in 2D also exhibit CD80 and CD206 expression levels similar to those of M1 and M2 macrophages, respectively (Front. Immunol. 2025, 16:1589553)

Furthermore, in the present study, we demonstrated that CD80 expression was clearly higher in THP-1-derived M1 macrophages compared to M2 populations, particularly in 3D collagen cultures. Likewise, M2 macrophages displayed higher CD206 expression than M0 and M1 subtypes in both 2D and 3D conditions. Thus, while our results are broadly consistent with classical M1/M2 definitions, they also demonstrate that marker expression is strongly influenced by the microenvironment.

We have added the following statements to the Discussion section (Line 420-427, highlighted in yellow):

Interestingly, markers such as CD80 and CD206 are classical representatives of M1 and M2 macrophages, respectively; however, their expression was also detected in M0 macrophages under both 2D and 3D conditions. A recent study on the differentiation of peripheral monocytes in 2D culture similarly reported that M0 macrophages exhibited CD80 and CD206 expression levels comparable to those of M1 and M2 macrophages[27]. These observations support the concept that macrophage phenotypes exist along a continuum rather than as discrete states, and that marker expression is highly context-dependent.

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript investigates how culture dimensionality (2D vs. 3D) influences macrophage morphology, viability, phenotype, and function using THP-1-derived macrophages. The study is timely and relevant, as 3D culture systems are increasingly recognized as more physiologically relevant models. The work provides a reasonably comprehensive comparison across morphology, surface markers, cytokine production, and functional tumor co-culture assays. However, several aspects of experimental design, data interpretation, and clarity require improvement before the manuscript is suitable for publication. In particular, the biological interpretation of dimensionality effects, methodological rigor (especially for 3D cultures), and phenotype characterization need strengthening.

1. The study relies entirely on THP-1–derived macrophages, which may not fully recapitulate primary human macrophage biology. The limitations of this model should be discussed more explicitly. If feasible, key findings should be validated using primary human monocyte-derived macrophages or at least an additional macrophage model to enhance biological relevance.
2. The statement (line 205) that Matrigel “does not support robust cell expansion” appears overstated. The observed decline in viability (>50%; Fig. 1D) could be influenced by factors such as limited nutrient diffusion or matrix stiffness rather than an intrinsic inability of Matrigel to support growth. To clarify this point, the authors should consider including apoptosis assays (e.g., cleaved caspase-3) and/or proliferation markers (e.g., Ki-67) to better define the underlying cause of reduced cell numbers.
3. Although the study compares Matrigel and collagen, key physicochemical properties—such as stiffness, porosity, and ligand composition—are not characterized. Because these parameters strongly influence macrophage behavior, the observed differences may reflect matrix-specific effects rather than dimensionality per se. This limitation should be acknowledged, and conclusions should be tempered accordingly.
4. The characterization of macrophage polarization is based on a relatively limited panel of markers (CD11b, CD68, CD80, CD86, CD163, CD206, TNF-α, FGF2). A more comprehensive assessment would strengthen the conclusions. The authors should consider incorporating additional transcriptional markers (e.g., IL1B, IL10, ARG1, IL6) or performing RNA-seq to enable an unbiased comparison between 2D and 3D conditions.
5. Figures 4B and 4D show that collagen 3D culture induces TNFα expression even under M0 conditions, suggesting the presence of unintended inflammatory stimuli in the system. In a physiological context, TNFα production at this level would not be expected. The authors may need to optimize the culture conditions to minimize such basal activation.
6. The assessment of antitumor activity relies primarily on imaging and cell counts, which makes it difficult to distinguish between effects on tumor cell death versus proliferation. The authors should include more direct functional assays, such as apoptosis measurements (e.g., cleaved caspase-3), proliferation assays (e.g., Ki-67), or tumor cell–specific viability assays, to better support their conclusions.

Author Response

Comment 1: The study relies entirely on THP-1–derived macrophages, which may not fully recapitulate primary human macrophage biology. The limitations of this model should be discussed more explicitly. If feasible, key findings should be validated using primary human monocyte-derived macrophages or at least an additional macrophage model to enhance biological relevance.

Response 1: We thank the reviewer for this insightful comment. In lines 59–67 of the Introduction, we outline our rationale for focusing on THP-1 cells and their macrophage polarization in 2D and 3D conditions. We agree that the limitations of this model should be more explicitly stated. While validating these key findings using primary human monocyte-derived macrophages falls outside the scope of the current study, we have expanded our Discussion to address these constraints and highlight this as a critical next step. We have added the following text to the Discussion section (Line 497-509, highlighted in yellow):

Although THP-1–derived macrophages are a controlled and reproducible model, several limitations should be acknowledged. As an immortalized monocytic leukemia cell line, THP-1 cells do not fully recapitulate the heterogeneity, differentiation pathways, and functional complexity of primary human macrophages. Furthermore, PMA-induced differentiation may introduce baseline activation that does not reflect physiological conditions, and the resulting macrophages may exhibit altered gene expression, receptor profiles, and cytokine responses compared to primary cells. In addition, THP-1–derived macrophages lack donor variability and tissue-specific imprinting, which are important determinants of macrophage behavior in vivo. Consequently, while this model is well-suited for dissecting microenvironmental effects under controlled conditions, our findings should be interpreted with caution when extrapolating to primary human macrophages. Future studies incorporating primary cells and more complex tissue models will be essential to validate and extend these observations.

 

Comment 2: The statement (line 205) that Matrigel “does not support robust cell expansion” appears overstated. The observed decline in viability (>50%; Fig. 1D) could be influenced by factors such as limited nutrient diffusion or matrix stiffness rather than an intrinsic inability of Matrigel to support growth. To clarify this point, the authors should consider including apoptosis assays (e.g., cleaved caspase-3) and/or proliferation markers (e.g., Ki-67) to better define the underlying cause of reduced cell numbers.

Response 2: We thank the reviewer for this insightful comment. In this study, both Matrigel and collagen gels were formulated at a concentration of 3 mg/mL (please see lines 116–122 of the Materials and Methods section). We agree with the reviewer that physical factors, such as limited nutrient diffusion or matrix stiffness, could play a significant role in the observed decline in viability in Matrigel. Because our primary objective was to empirically determine which matrix component better enabled cell growth to establish our 3D model, defining the exact underlying cause of reduced cell numbers (e.g., via apoptosis or proliferation assays) falls outside the scope of the current study. However, we completely agree that our original phrasing was overstated. To avoid making intrinsic mechanistic claims about Matrigel, we have revised the text in Line 205 (highlighted in yellow) to read:

Since the Matrigel culture did not support sufficient cell growth under these conditions, subsequent 3D experiments were performed using collagen gels.

 

Comment 3: Although the study compares Matrigel and collagen, key physicochemical properties—such as stiffness, porosity, and ligand composition—are not characterized. Because these parameters strongly influence macrophage behavior, the observed differences may reflect matrix-specific effects rather than dimensionality per se. This limitation should be acknowledged, and conclusions should be tempered accordingly.

Response 3: We thank the reviewer for bringing up this important issue. We have included the following sentences in the Discussion section (Line 511-519, highlighted in yellow):
Another important limitation of this study is that key physicochemical properties of the Matrigel and collagen matrices, such as stiffness, porosity, and ligand composition, were not explicitly characterized. Macrophage behavior is highly responsive to both mechanical forces and biochemical cues. Therefore, we cannot rule out that the phenotypic differences observed between our culture conditions reflect matrix-specific biochemical and biophysical effects rather than dimensionality per se. As such, conclusions regarding the pure effect of 3D spatial geometry should be tempered. To fully decouple the influence of matrix architecture from structural mechanics and ligand presentation, future studies utilizing tunable engineered hydrogels are warranted.

 

Comment 4: The characterization of macrophage polarization is based on a relatively limited panel of markers (CD11b, CD68, CD80, CD86, CD163, CD206, TNF-α, FGF2). A more comprehensive assessment would strengthen the conclusions. The authors should consider incorporating additional transcriptional markers (e.g., IL1B, IL10, ARG1, IL6) or performing RNA-seq to enable an unbiased comparison between 2D and 3D conditions.

Response 4: We thank the reviewer for this valuable suggestion. We agree that the panel of markers examined in the present study is relatively limited and that a more comprehensive transcriptional analysis would further strengthen the conclusions. In this work, we intentionally selected a focused set of well-established surface markers and functional readouts to systematically evaluate the effects of culture dimensionality and matrix composition under controlled conditions. Importantly, these markers were complemented by functional assays, including cytokine secretion and tumor cell coculture, which support the phenotypic interpretations.

We acknowledge that additional transcriptional markers (e.g., IL1B, IL10, ARG1, IL6) or unbiased approaches such as RNA sequencing would provide deeper insight into macrophage polarization states. However, such analyses are beyond the scope of the current study, which is primarily designed to establish the impact of extracellular dimensionality on macrophage behavior.

We are currently planning follow-up studies to perform comprehensive transcriptomic and proteomic profiling to further characterize these effects, including comparisons between THP-1–derived macrophages and primary human monocyte-derived macrophages. We have now clarified this point and expanded the Discussion to acknowledge this limitation and outline future directions.

 

Comment 5: Figures 4B and 4D show that collagen 3D culture induces TNFα expression even under M0 conditions, suggesting the presence of unintended inflammatory stimuli in the system. In a physiological context, TNFα production at this level would not be expected. The authors may need to optimize the culture conditions to minimize such basal activation.

Response 5: We thank the reviewer for this important comment. In the present study, THP-1 monocytes were first differentiated into M0 macrophages using PMA under both 2D and 3D conditions, followed by further polarization into M1 or M2 phenotypes. It is well established that PMA-induced differentiation can result in a basal level of activation, including the expression of pro-inflammatory cytokines such as TNFα, even in M0 macrophages. This phenomenon has been reported in previous studies (Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy, 2020, 248:119118; Medicina, 2024, 60:1009).

Therefore, the observed TNFα expression under M0 conditions does not necessarily reflect unintended inflammatory stimuli arising from the culture system itself, but rather an inherent feature of the PMA differentiation protocol. Importantly, all experimental groups were subjected to the same differentiation conditions, allowing for a controlled comparison of dimensionality-dependent effects.

To our knowledge, this study is the first to systematically compare TNFα expression and secretion across M0, M1, and M2 THP-1–derived macrophages in both 2D and 3D environments. While we acknowledge that the basal activation level may differ from that of primary macrophages in vivo, the relative differences observed between conditions remain valid and informative.

 

Comment 6: The assessment of antitumor activity relies primarily on imaging and cell counts, which makes it difficult to distinguish between effects on tumor cell death versus proliferation. The authors should include more direct functional assays, such as apoptosis measurements (e.g., cleaved caspase-3), proliferation assays (e.g., Ki-67), or tumor cell–specific viability assays, to better support their conclusions.

Response 6: We thank the reviewer for this important suggestion. We agree that distinguishing between effects on apoptosis and proliferation would provide additional mechanistic insight. However, the primary objective of this experiment was to determine whether the capacity of THP-1–derived M1 or M2 macrophages to modulate MDA-MB-231 cell numbers is influenced by 2D versus 3D culture conditions, rather than to dissect the specific mechanisms of tumor cell regulation.

To address this objective, we employed consistent imaging and cell quantification approaches across all conditions, which provide a reliable and comparative assessment of overall tumor cell outcomes. Importantly, all experimental groups were analyzed under identical conditions, ensuring that the observed differences reflect the impact of macrophage phenotype and culture dimensionality.

While additional assays such as apoptosis or proliferation markers (e.g., cleaved caspase-3 or Ki-67) would further refine mechanistic interpretation, these analyses are beyond the scope of the current study. We have clarified this point in the revised manuscript and highlighted it as an important direction for future investigation (Line 472-475, highlighted in yellow):

It would be of interest in the future to conduct additional assays, such as apoptosis or proliferation markers (e.g., cleaved caspase-3 or Ki-67), to further clarify whether dimensionality affects how THP-1–derived M1 macrophages restrict cancer cell growth.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

My concerns have been addressed.

Reviewer 2 Report

Comments and Suggestions for Authors

I have no more comments for this manuscript.