Human neonatal CITE-seq atlas identifies an immune transition at 32 weeks' gestation from CD15+ myeloid-dominated to interferon-primed immunity

This study presents a comprehensive human neonatal CITE-seq atlas that reveals a critical immune transition at 32 weeks' gestation, shifting from a CD15+ myeloid-dominated state in extremely preterm infants to an interferon-primed immunity in more mature neonates, thereby defining gestational age-specific immune programs and vulnerabilities.

The Gist

Imagine the human immune system as a highly sophisticated security team. For a baby growing inside the womb, this team has a very specific job: keep the peace. The womb is a safe, sterile environment, and the immune system's main goal is to make sure the baby doesn't accidentally attack its own mother or itself. It's like a security guard who is told, "Do not shoot; just stand there and watch."

This study is like a high-tech security audit of newborns. The researchers looked at the "security guards" (immune cells) in babies born at different times and compared them to adults. They discovered a critical turning point in how this security team is trained: 32 weeks of pregnancy.

Here is the story of what they found, broken down into simple concepts:

1. The "32-Week Switch"

Think of gestational age (how long a baby has been in the womb) as the "training level" of the immune system.

• Before 32 Weeks (The "Peacekeepers"): Babies born extremely early (before 32 weeks) have an immune system that is still stuck in "Womb Mode." Their security team is dominated by a specific type of cell called CD15+ myeloid cells. You can think of these as super-sleepy, peace-keeping guards. Their main job is to suppress any reaction. They are like security guards who have been told, "If you see a fire, just put it out immediately and don't let anyone get excited." They are very good at stopping inflammation, but they are slow to react to real threats.
• After 32 Weeks (The "Alert Team"): Babies born at 32 weeks or later have a different setup. Their immune system has flipped a switch. They start producing a "wake-up call" signal called Interferon. This is like a siren that says, "Alert! We are entering the real world! Get ready to fight viruses and bacteria!" Their security team becomes much more active and ready to defend.

2. The Problem with Being Born Too Early

When a baby is born before this 32-week switch flips, they enter the real world (which is full of germs) with a security team that is still in "Womb Mode."

• The Mismatch: The outside world is chaotic and full of bacteria. The baby's immune system is still trying to keep the peace and suppress reactions.
• The Result: Because their "peacekeeper" cells are so dominant, they struggle to mount a strong defense against infections. They are like a security team that is too busy trying to calm everyone down to notice a burglar breaking in. This explains why extremely premature babies are so vulnerable to infections.

3. The "Bystander" Effect

The researchers also tested what happens when they "tricked" the immune cells with a fake alarm (simulating a bacterial or viral infection).

• The Mature Babies (≥32 weeks): When the alarm went off, their immune cells shouted back loudly. They activated their "Interferon" siren, which helped other parts of the immune system (like T-cells and NK cells) jump into action. It was a coordinated, strong defense.
• The Immature Babies (<32 weeks): When the alarm went off, their cells didn't shout as loudly. Instead, they started talking about "metabolism" and "regulation." They were still trying to calm things down. Even when they tried to fight, they were using a different playbook that focused on stopping the fight rather than winning it.

4. The Communication Breakdown

The study also looked at how these cells talk to each other.

• In Mature Babies: The cells talk to each other like a well-oiled machine. The "guards" tell the "fighters" (T-cells and NK cells) to get ready. They say, "We see a virus; let's make some Interferon and attack!"
• In Premature Babies: The communication is mixed up. The guards are telling the fighters, "No, no, let's not get excited. Let's stay calm and suppress this." They are sending signals that tell the immune system to hold back, which leaves the baby defenseless against real infections.

The Big Takeaway

This study is a map. It tells us that 32 weeks is a magic number for immune development.

• If a baby is born after 32 weeks, their immune system is roughly "ready for the real world," even if they are still a bit small.
• If a baby is born before 32 weeks, they are essentially born with a "fetal" immune system that is designed for the womb, not the outside world.

Why does this matter?
Understanding this "32-week switch" helps doctors. Instead of treating all premature babies the same, they might be able to develop new treatments that help babies born before 32 weeks "flip the switch" earlier. Maybe they can give them a little nudge to wake up their immune system, helping them fight infections better without causing dangerous inflammation.

In short: The immune system isn't just "weak" in premature babies; it's just programmed for the wrong environment. This study helps us understand exactly how that programming works and where the switch is located.

Technical Summary

1. Problem Statement

The neonatal immune system is developmentally specialized to balance effective pathogen defense with the prevention of excessive inflammation during the perinatal transition. Prematurity disrupts this finely tuned balance, leading to heightened susceptibility to infections and long-term immune dysregulation. However, the specific cellular and molecular mechanisms governing the transition from fetal to postnatal immunity remain incompletely understood. Key knowledge gaps include:

• Whether prematurity represents a mere delay in maturation or the persistence of a distinct fetal immune program.
• The specific gestational age (GA) at which critical immune transitions occur.
• How GA influences baseline immune states and the responsiveness of immune cells to inflammatory challenges (e.g., TLR stimulation).
• The limitations of existing studies, which often rely on unimodal readouts, narrow GA windows, or baseline measurements without functional perturbation.

2. Methodology

The authors constructed a comprehensive multimodal CITE-seq (Cellular Indexing of Transcriptomes and Epitopes by Sequencing) atlas to profile circulating immune cells across a wide spectrum of gestational ages.

• Cohort:
  • Neonates: 25 individuals (82 samples) ranging from 24 to 42 weeks GA, categorized as:
    • Extremely preterm (<28+0 weeks, n=5)
    • Very preterm (28+0 to 31+6 weeks, n=2)
    • Late preterm (32+0 to 36+6 weeks, n=5)
    • Full-term (≥37+0 weeks, n=13)
  • Controls: 10 healthy adults.
  • Sampling: Peripheral blood mononuclear cells (PBMCs) and cord blood mononuclear cells (CBMCs) collected at birth and/or Day of Life 3 (DOL3).
• Experimental Design:
  • Stimulation: Samples were subjected to in vitro stimulation for 2 hours with:
    • LPS (TLR4 agonist).
    • Resiquimod (R848) (TLR7/8 agonist).
    • Baseline: Untreated controls.
  • Technology: 36-antibody CITE-seq panel integrating single-cell RNA sequencing (scRNA-seq) with antibody-derived tags (ADTs) for surface protein quantification.
  • Total Data: ~195,000 cells profiled.
• Analytical Approach:
  • Unsupervised clustering (UMAP) and dimensionality reduction (PCA).
  • Differential expression analysis (DEGs) and Gene Set Enrichment Analysis (GSEA).
  • Bayesian compositional analysis for cell type proportions.
  • Cell-cell interaction inference using CellPhoneDB.
  • Cytokine activity inference (huCIRA) to map signaling pathways.
  • Validation via flow cytometry on independent cohorts.

3. Key Contributions

• First Multimodal Neonatal Atlas: This is the first study to systematically compare human neonatal immune cell states across a broad GA spectrum (24–42 weeks) in the immediate postnatal period, integrating both baseline states and innate immune responses.
• Identification of a Critical Developmental Threshold: The study defines 32 weeks of gestation as a pivotal transition point separating two distinct immunobiological states.
• Multimodal Resolution of Rare Populations: By combining surface protein (CD15, CD16) and transcriptomic data, the authors successfully identified and characterized ARG1high CD15+ myeloid cells (gMDSC-like) that are often missed in unimodal studies.
• Mechanistic Insight into Prematurity: The study moves beyond descriptive differences to propose a mechanistic model where extremely preterm infants (<32 weeks) fail to initiate a critical "interferon spike," remaining in a regulatory, suppressive fetal state.

4. Key Results

A. The 32-Week Transition Point

Principal Component Analysis (PCA) revealed a clear stratification of neonatal samples based on GA, with a distinct separation between <32 weeks (extremely/very preterm) and ≥32 weeks (late preterm/term) neonates. This separation was evident at baseline and accentuated upon TLR stimulation.

B. Baseline Immune States (<32w vs. ≥32w)

• <32 Weeks (The "gMDSC-Dominated" State):
  • Myeloid: Enrichment of CD15+ myeloid cells with a gMDSC-like phenotype (high ARG1, ALPL, CCL4, high CD16 surface expression). These cells are immunosuppressive.
  • NK Cells: Enrichment of a CD8+ NK cell subset (Cluster I) expressing CX3CR1 and TXNIP, associated with reduced IFN-γ production and regulatory features.
  • Transcriptional State: Low interferon signaling; cells exhibit signs of basal preactivation but lack the coordinated antiviral program seen in older neonates.
• ≥32 Weeks (The "Interferon-Primed" State):
  • Myeloid: Shift toward pro-inflammatory CD15+ cells and classical monocytes.
  • Transcriptional Remodeling: By DOL3, these neonates exhibit a robust upregulation of Interferon-Stimulated Genes (ISGs) (e.g., GBP1, STAT1, OAS1) across monocytes, T cells, and B cells, establishing a tonic antiviral state.
  • Comparison to Adults: Neonates show widespread basal preactivation compared to adults, but the quality of this activation differs by GA.

C. Response to TLR Stimulation (LPS/R848)

Upon stimulation, both groups mount robust responses, but the qualitative nature diverges:

• <32 Weeks:
  • Myeloid: Preferential activation of KRAS and mTORC1 signaling pathways. Upregulation of regulatory genes like THBD (thrombomodulin) and TNFSF14, suggesting a bias toward differentiation and anti-inflammatory responses.
  • Lymphocyte Crosstalk: Predicted interactions are dominated by TGF-β signaling from NK cells to T cells, promoting a regulatory environment.
  • T Cell Response: Naïve CD4+ T cells adopt Treg-associated and Th2-biased transcriptional programs (high FOXP3, SNAI1, TIGIT).
• ≥32 Weeks:
  • Myeloid: Strong induction of Interferon-driven programs (Type I and II IFN) and chemokines like CXCL10.
  • Lymphocyte Crosstalk: Shift toward IFN-γ and TRAIL-mediated signaling.
  • T Cell Response: Naïve CD4+ T cells shift toward Th1-effector programs (high IFNG, TNFRSF4/OX40, BHLHE40).
  • NK Cells: CD16hi NK cells show enhanced IFN-γ expression and effector gene upregulation (GNLY, GZMB, PRF1).

D. Cell-Cell Communication

• <32 Weeks: Monocytes signal primarily to other myeloid cells (self-reinforcing regulatory loop). NK-T cell crosstalk is inhibitory (TGF-β).
• ≥32 Weeks: Monocytes signal to lymphoid populations (NK and T cells) via IL-15, IL-18, and CXCL10, facilitating coordinated innate-adaptive activation.

5. Significance and Implications

• Redefining Prematurity: The study suggests that the immune vulnerability of extremely preterm infants is not merely a "delay" in maturation but the persistence of a fetal regulatory program optimized for in utero tolerance. These infants are delivered before the critical "interferon spike" that triggers the transition to an adult-like, antiviral-ready state.
• Clinical Relevance: The 32-week threshold aligns with clinical improvements in neonatal outcomes. Understanding this transition provides a biological basis for the increased susceptibility of <32-week infants to sepsis and their altered response to vaccines or infections.
• Therapeutic Potential: The findings suggest that immune vulnerability in preterm infants could potentially be addressed by GA-adapted immunomodulatory strategies. For instance, therapeutically inducing interferon signaling or modulating the mTOR pathway might help "prime" the immune system of extremely preterm infants, though caution is required to avoid excessive inflammation.
• Methodological Advance: The study demonstrates the power of CITE-seq in resolving complex immune states (e.g., distinguishing gMDSCs from neutrophils via surface protein + transcriptome) that are invisible to standard scRNA-seq or flow cytometry alone.

In conclusion, this atlas provides a foundational framework for understanding how gestational age dictates immune architecture, identifying a specific developmental window (~32 weeks) where the immune system switches from a suppressive, myeloid-dominated state to an interferon-primed, effector-ready state.