Machine Learning-Based Single Cell and Integrative Analysis Reveals That Baseline mDC Predisposition Correlates With...

Area of Science:

• Immunology
• Vaccinology
• Genomics

Background:

• Vaccination is a key public health tool, but individual responses vary significantly.
• Understanding the mechanisms behind this variability is crucial for improving vaccine efficacy.
• Previous studies showed inconsistent antibody levels after hepatitis B virus (HBV) vaccination.

Purpose of the Study:

• To investigate the immune mechanisms underlying variable responses to the HBV vaccine.
• To identify baseline immune cell characteristics that predict vaccine response.
• To explore the potential of dendritic cell subsets as biomarkers for vaccine efficacy.

Main Methods:

• Single-cell RNA sequencing was used to analyze innate immune cell subsets.
• Two distinct myeloid dendritic cell subsets (mDC2 and mDC4) were identified and characterized.
• Machine learning models were developed using gene expression data from these subsets to predict antibody titers.

Main Results:

• The baseline ratio of NDRG1-expressing mDC2 and CDKN1C-expressing mDC4 cells correlated with the immune response to a single HBV vaccine dose.
• Participants were categorized into two distinct "dispositional endotypes" based on their baseline dendritic cell populations.
• Predictive models incorporating gene expression from these mDC subsets showed improved correlation with serum antibody titers after full vaccination.

Conclusions:

• Pre-vaccination levels of specific circulating dendritic cell subsets, termed "dispositional endotypes," can predict an individual's response to vaccination.
• This finding offers a potential mechanism for understanding and potentially improving vaccine design and efficacy.
• Targeting or understanding these dendritic cell states may enhance future vaccination strategies.