402 - Kinetics of disease, immune responses, and plasma biomarkers in infant baboons challenged with Bordetella pertussis

Publication Number: 2391.402

Byron Brook, Harvard Medical School, Boston, MA, United States; Etsuro Nanishi, Boston Children's Hospital, Boston, MA, United States; Soumik Barman, Boston Children’s Hospital, Boston, MA, United States; Christopher Chen, PKD Foundation, Ponte Vedra Beach, FL, United States; Deborah Chavez, Texas Biomedical Research Institute, San Antonio, TX, United States; Fernando Varnador, Southwest National Primate Research Center, San Antonio, San Antonio, TX, United States; Bernadette Guerra, Texas Biomedical Research Institute, San Antonio, TX, United States; Hanno Steen, Boston Children's Hospital, Boston, MA, United States; Jacob A.. Koster, Boston Children's Hospital, Boston, MA, United States; Kevin Lyskawa, Boston Children's Hospital, Boston, MA, United States; Asimenia Angelidou, Harvard Medical School, Boston, MA, United States; Caitlin Syphurs, Boston Children's Hospital, Boston, MA, United States; Joann Diray-Arce, Boston Children's Hospital, Boston, MA, United States; Ofer Levy, Precision Vaccines Program, Boston Children's Hospital, Boston, MA, United States; David Dowling, Boston Children's Hospital, Boston, MA, United States

Background: Rising Bordetella pertussis infections, the cause of whooping cough, has resulted in high infant hospitalization rates and substantial disease burden in the United States, and globally. The B. pertussis infection trajectory is incomplete due to variation of bacterial strain, infectious dose, elapsed time between infection and healthcare assessment, different patient ages, sampling limitations, and population variability.

Objective: In a nonhuman primate model, assess pathogen and immunological kinetics, and identify biomarkers via metabolomics and proteomics.

Design/Methods: We employed an infant baboon model (N = 8), standardizing challenge strain, dose, and time of infection, aligning bacterial burden, pathogenicity, and immunological outcomes. B. pertussis burden was inferred by nasopharyngeal wash (NPW) PCR, while pathogenicity tracked percent weight change, coughs, and complete blood count over time. IgM and IgG specific to pertussis toxin (PT), filamentous hemagglutinin (FHA), and pertactin (PRN) were longitudinally assessed. Untargeted metabolomics and proteomics revealed potential acute infection biomarkers.

Results: Pathogen burden peaked on day 7 post-challenge. Subsequent notable antibody responses included FHA IgM and IgG, and PT IgG, and PRN IgG. The leucyl glycine dipeptide metabolite and hydroxyasparagine (plasma alanine and aspartate metabolism) were downregulated during infection. Proteomics identified upregulation of many proteins, including SERPINA3, SAA1, CPN1, and ITIH3. Notable upregulated responses included heme-associations (haptoglobin and hemopexin), serine protease inhibitors (SERPINA3 and SERPINA1), a lipid metabolism and bacterial clearance factor (serum amyloid A1), complement inactivator (CFI), granulocyte differentiation marker (leucine rich alpha 2 glycoprotein 1), and an acute phase protein (orosomucoid 1). Infant baboons classified as early clearers (D21+ full NPW clearance) and delayed clearers (burden on D21 or 28), N = 4 per group differed. Lower protein abundance was seen in early clearance versus delayed clearance for inflammation-associated alarmin (S100A8), immunoglobulins (IGHV2.70D, IGLV5.48, IGHA1), APO-AI, heme-binders (HPX), and a protein impacting vascular integrity (APOM).

Conclusion(s): A longitudinal infant baboon B. pertussis challenge model kinetic characterization of disease responses, pathogenicity, immune responses, and plasma metabolomic and proteomic biomarker modulation, potentially informing novel approaches to diagnose, prognose, and treat whooping cough in early life.

Figure 1. Kinetics of infant baboon Bordetella pertussis bacterial burden.
Fig 1.pdfWeanling baboons were challenged with Bordetella pertussis (B. pertussis), and nasopharyngeal washes (NPW) were collected weekly through day 42 post-challenge, with a terminal collection at day 88. NPW were assessed for B. pertussis burden via a BP485 PCR assay, aligned to a standard curve. Peak of disease burden was noted at day 7 post-challenge, persisting through day 14, with some animals fully clearing by day 21 (an ‘early clearance’ group), and others with detectable burden on d21 and/or d28 (a ‘delayed clearance’ group). N = 8, with significance determined by Shapiro-Wilk and Friedman tests, and annotated as * p < 0.05, ** p < 0.01.

Figure 2. LC/MS Proteomic assessment of Baboon samples over time.
Fig 2.pdf(A) Induction of Serum amyloid A1 (SAA1, left) and Haptoglobin (HP, right), respectively involved in bacterial clearance and binding of free hemoglobin, were significantly induced on day 3 and 7 post-challenge, and followed a similar trajectory of pathogen burden, reducing in intensity by day 14 and 28. (B) Infant baboons that had early clearance (undetectable NPW BP on or after day 21 post-challenge) had significantly different LC/MS proteomic profiles from baboons with delayed clearance (detectable NPW BP on or after day 21 post-challenge). Example proteins of interest that were induced to a higher degree in animals that cleared bacteria faster include an alarmin, S100 calcium-binding protein A8 (S100A8, left), and a vascular integrity factor, Apolipoprotein M (APOM, middle), and a protein involved in binding free heme, Hemopexin (HPX, right). (A) N = 8, with significance determined by Shapiro-Wilk and Friedman tests, and annotated as * p < 0.05, and (B) N = 4 / clearance group, with significance determined by GEEGLM assessment of clearance grouping impact on protein intensity over time.

Figure 3. Aligned kinetics of infant baboon B. pertussis burden, pathogenicity, and immunological outcomes.
Fig 3.pdfReadouts were calculated as a percentage of start, or a percentage of maximum observed throughout the experiment, enabling concomitant graphing of a best-fit loess curve. Nasopharyngeal wash (NPW, solid line) and direct swab (DS, long dash) sampling with PCR of B. pertussis inferring bacterial burden enabling pathogen burden kinetics, with alignment to pathogenicity readouts percent weight change (dot-dash), and percent coughs of maximum (long-short dash). Pathogenicity with an increase of coughs was noted in the days post-challenge, preceding the pathogen burden by NPW or DS. Immunological kinetics of pertussis toxin (PT), filamentous hemagglutinin (FHA), and pertactin (PRN) binding antibodies were quantified as a percent of log10-transformed maximum observed, assessed for IgG and IgM isotypes. IgM was slightly induced by 14-21 days post-challenge, while IgG was notably induced by day 28 post-challenge. PT and FHA specific IgG had notably sustained immunity while PRN IgG began to wane by day 88. Alignment of bacterial burden, pathogenicity, and immunological outcomes were made possible in this animal model, standardizing based on the day of challenge.