Journal: PLOS Computational Biology

Article Title: Quantitative mechanistic model reveals key determinants of placental IgG transfer and informs prenatal immunization strategies

doi: 10.1371/journal.pcbi.1011109

Figure Lengend Snippet: (A) The schematic depicts the ODE model representing transport of IgG through the placenta, mapped onto a graphical representation of the maternal-fetal interface. Reaction rate constants associated with each process are depicted on the corresponding arrow. Each box represents a molar concentration in each compartment. “C” denotes complexes of bound IgG and Fc receptors either in endosomes (FcRn) or at the cell surface (FcγRIIb). Δ F c R n S T B t o t a l , Δ F c γ R I I b E C t o t a l , and Δ F c R n S T B t o t a l represent synthesis of new FcRn and FcγRIIb molecules over time as the placenta grows. For simplicity, the schematic represents bulk IgG transport, whereas the model includes separate equations for each IgG subclass. (B) Simulated fetal IgG subclass levels vs. time (lines) are overlaid with the data used for parameter estimation. The overlaid data are the mean and standard deviation of IgG subclasses in the umbilical cord between 17–41 weeks gestation (n = 107) measured using cordocentesis by Malek et al . Gray lines represent 1000 perturbation simulations resulting from sampling parameters from the ranges determined by CaliPro. (C) The swarm charts show the fetal/maternal ratio (F:M ratio) for each IgG subclass at 40 weeks gestation resulting from 1000 simulations sampling from the optimized range of parameters (filled symbols) compared to the mean cord/maternal ratio for each IgG subclass compiled across cohort studies mined from the literature (hollow symbols). Each symbol represents a single simulation or cohort average. Sample means are indicated by black lines. The symbols outlined in black represent the mean of IgG subclass cord/maternal transfer ratios at parturition from the human study used during parameter optimization . Dashed line at F:M ratio = 1 indicates equal concentration in maternal and fetal blood at parturition.

Article Snippet: We performed parameter estimation for 11 parameters using CaliPro, a flexible optimization software implemented in MATLAB and described by Joslyn et al [ ].

Techniques: Concentration Assay, Standard Deviation, Sampling

Journal: PLOS Computational Biology

Article Title: Quantitative mechanistic model reveals key determinants of placental IgG transfer and informs prenatal immunization strategies

doi: 10.1371/journal.pcbi.1011109

Figure Lengend Snippet: (A-B) The OPLSR model predicts the optimal immunization time (t vax , weeks gestational age) from perturbed parameter sets. (A) The scores plot depicts the X scores on LV1 and LV2 of the optimal t vax for 1000 perturbed parameter sets. Each circle represents a single perturbation experiment colored proportionally to the predicted optimal t vax . The model was orthogonalized such that LV1 is in the direction of maximum variance. (B) The bar plot shows VIP scores that quantify the relative contribution of each parameter to the optimal t vax . VIP scores are artificially oriented in the direction of their loadings on LV1. The model performed with predictive power (Q 2 = 47%) and with a mean squared error lower than 1000 models based on randomly permutated labels (p < 0.001). (C) The value of F c R n S T B t o t a l , e n d (left), F c γ R I I b E C t o t a l , e n d (center), or F c R n E C t o t a l , e n d (right) was varied from 10-fold lower to 10 times higher than the mean of their optimal range determined by CaliPro. Vaccination was simulated at the optimal time for a term newborn (t vax = 25 weeks) for each FcR condition and the resulting percent changes in anti-pertussis toxin IgG (α-PT IgG) in a newborn delivered at 40 weeks is depicted. The dashed line represents the baseline simulation of an average patient using the optimized values of F c R n S T B t o t a l , e n d , F c γ R I I b E C t o t a l , e n d , or F c R n E C t o t a l , e n d , which are marked by vertical lines. (D) The chart shows the effect of increasing vaccine dosage on α-PT IgG transfer for patients with a 10-fold deficiency in F c R n S T B t o t a l , e n d or F c γ R I I b E C t o t a l , e n d expression. Dosages are shown as fold-changes from the baseline simulation dosage. (E) The effect of increasing vaccine-induced α-PT IgG affinity for FcRn, FcγRIIb, or FcRn and FcγRIIb on α-PT IgG transferred to a fetus delivered at term.

Article Snippet: We performed parameter estimation for 11 parameters using CaliPro, a flexible optimization software implemented in MATLAB and described by Joslyn et al [ ].

Techniques: Expressing