Dr. Annie De Groot is an invited speaker at the 6th Workshop on Monoclonal Antibodies in Basel, Switzerland, on September 12th, 2013. She will presenting information on CHO host cells proteins (HPC’s) and their effects on the immunogenicity of biologics. She will also present information regarding the CHOPPI website, a free online immunogenicity screening tool for CHO Host Cell Proteins (HCP’s), that is powered by EpiMatrix and produced in collaboration with Dartmouth College.

See below for the presentation slides, a link to the CHOPPI website, and a recent CHO Immunogenicity Poster.

CHOPPI Website

See CHO Poster Here

Immunogenicity Analysis of Chinese Hamster Ovary (CHO) Host Cell Protein Contaminants in Therapeutic Protein Formulations

Andres H. Gutierrez1, Leonard Moise1,2, Frances Terry2, Chris Bailey-Kellogg3, William Martin2, Anne S. De Groot1,2

1Institute for Immunology and Informatics, University of Rhode Island, Providence, RI;

2EpiVax, Inc., Providence, RI; and

3Department of Computer Science, Dartmouth College, Hanover, NH.

Purpose: Concerns over immune responses to contaminating host cell proteins (HCPs) in therapeutics have recently emerged. For example, the presence of Chinese hamster ovary (CHO)-derived HCPs contributed to the cancellation of two clinical trials in 2012. Publication of the CHO-K1 genome and transcriptome offers an opportunity to analyze and understand the immunogenic potential of CHO HCPs.

Methods: Four datasets were assembled: The CHO-K1 genome, the CHO-K1 transcriptome, published CHO proteins identified in biotherapeutics, and mouse secreted proteins downloaded from the LOCATE and UniProtKB/Swiss-Prot databases. Validated HCP contaminants, originally identified by sequence similarity with mouse homologs, were discovered by BLAST in the CHO genome. Their expression and secretion status were confirmed using BLAST against the CHO transcriptome and mouse secreted protein datasets, respectively; SignalP further confirmed secretion status. Secretion was assessed because exported proteins may contaminate any recombinant CHO-expressed product. Confirmed sequences from CHO-K1 were analyzed for potential CD4+ T cell immunogenicity using the epitope-mapping algorithm EpiMatrix and overall protein immunogenicity scores calculated using the EpiMatrix approach. Human homologs discovered by BLAST were similarly screened for T-cell epitope content; predicted immunogenicity was then adjusted for human homology.

Results: Predicted overall protein immunogenicity of 28 validated CHO HCP contaminants ranged from 51.8 (low immunogenic potential) to 35.8 (high immunogenic potential). Five proteins (18%) scored over the EpiMatrix cutoff score of 20 for high potential immunogenicity, suggesting these proteins could stimulate T cell responses. While most validated CHO HCPs scored below this threshold, a closer look at potential immunogenicity on the 9-mer (epitope) level revealed concerns for potential immunogenicity in lower scoring proteins.

Conclusions: Further study is required to assess the potential for autoimmune responses, as several widely publicized cases of severe adverse immune responses to autologous therapeutic proteins are published and the FDA and EMA take the potential for such effects very seriously. Finally, we built a website (CHOPPI; www.cs.dartmouth.edu/~cbk/choppi) to give protein manufacturers free access to information on potential immunogenicity of CHO-K1 sequences. We look forward to its further development in collaboration with protein therapeutics industry, as this and additional means of performing immunogenicity analysis continue to be refined.