FirstWord Pharma

Bristol Myers Squibb will be implementing additional job cuts in New Jersey as part of its ongoing cost-saving efforts, laying off 195 employees, according to a recent Worker Adjustment and Retraining Notification (WARN) for the state. The latest round of pink slips follows a decision to eliminate 938 jobs at the New Jersey site under a cost-cutting initiative unveiled earlier the year.A BMS spokesperson confirmed the cuts in an email to FirstWord, noting that many of the jobs referenced in the WARN notice were included in previous announcements. Affected employees were already notified and the dates listed on the WARN notice, spanning from February to December 2025, represent their final dates of employment."As we communicated earlier this year, we are focused on strengthening the company's long-term growth profile," the BMS spokesperson said. "We are optimising operations across the company while prioritising investments in innovative and transformational medicines where we can deliver the highest value for patients and shareholders."In April, the company announced a cost-saving programme designed to shave off $1.5 billion in expenses by the end of next year. The initiative aims to reduce management layers and consolidate sites, with two-thirds of savings expected to come from R&D restructuring.

As biopharma begins to scratch the surface of what AI-enabled protein design is capable of, Aizen Therapeutics is targeting the flipside of the coin. The San Diego-based startup, which emerged from stealth Thursday with $13 million in seed funding, aims to generate peptides that are the mirror image of naturally occurring molecules, and have the added bonus of improved stability and lower immunogenicity. Aizen's AI-powered DaX platform allows it to design D-amino acids, which are the exact opposite of nature's L-amino acids. The technology was spun out of the lab of California Institute of Technology professor David Van Valen, who is also the newco's scientific founder.With Thursday's financing, which is expected to provide more than two years of runway, Aizen is aiming to select an IND-enabling lead programme and secure a strategic partner that validates its mirror peptide technology. The company's backers include venture capital firms Wilson Hill, Madrona and Cercano.AI accelerationThe drug development space has been interested in D-amino acids for at least 20 years, Aizen CEO Ajay Kshatriya told FirstWord, given their therapeutically beneficial properties. When proteins composed of L-peptides are introduced to the body, they're quickly broken down, and the smaller pieces of amino acids are then clocked as foreign, triggering an immune response, he explained. However, proteases can't engage with mirror peptides, thus bypassing the breakdown process and avoiding alerting the immune system altogether. Several drugs that feature D-amino acids in their molecular makeup have been approved by the FDA, but designing entire mirror peptides has historically posed a time-consuming problem due to a lack of data. The entire process, if relying solely on wet lab work and a method called reverse mirror image phage display, could take one to three years, Kshatriya said. While Van Valen believed that the process could be sped up with generative AI (genAI), the next roadblock he faced was that the protein repositories assembled thus far relied completely on L-amino acids."AI needs data, and all the data in the Protein Data Bank, all the crystal structures and all the images, were all looking at L-proteins binding to L-receptors," Kshatriya said, adding that existing computational programmes such as AlphaFold and Rosetta were designed to work only for L-amino acids. "It's a pretty challenging transformation to leverage L-data and then design a D-model," he remarked. But, thanks to a "clever algorithmic design" and recent innovations in guided diffusion models, Van Valen was able to build and train a genAI programme to design D-peptides from scratch. More importantly, the Aizen team has shown that the algorithm works. The company's scientists synthesised D-peptides spit out by the model and then tested them against known receptors — and were able to demonstrate binding activity. With the scientific and technological groundwork laid for Aizen, the company has now begun to explore all the clinical applications of a compound that can now be designed in weeks, instead of years.  Best of both therapeutic worldsAccording to Kshatriya, the startup's mirror peptides are "ideal targeting vehicles" because they have the "benefits of size and tissue penetration of a small molecule, but the specificity of binding of an antibody."One application Aizen is pursuing is the development of peptides that target receptors at the interface of the blood-brain barrier, thus acting as a delivery vehicle to transport a linked payload into the brain. Another area ripe for innovation is radioligand therapeutics (RLTs), a modality that has seen big-ticket investments and deals from pharma this year (see – ViewPoints: Exploring the strategy behind Sanofi's recent investment in RLTs with global head of oncology Olivier Nataf and ViewPoints: RadioMedix CEO sees lead as the future of RLTs).Mirror peptides could grant cancer cell surface-specific targeting to RLTs, while also being able to penetrate deep into the tumour microenvironment, Kshatriya explained.Beyond those two applications, Aizen's DaX platform is capable of designing peptides for a variety of receptor classes, including GPCRs, transmembrane proteins and ion channels."We're in the early innings of the potential for computation and AI to impact the design of proteins," Kshatriya commented.

Valora Therapeutics has emerged as the new kid on the block in immunotherapy, securing $30 million in seed funding to develop a different approach to immune system modulation, inspired by research in glycobiology. The investment, led by a consortium of investors including Avalon BioVentures, Bregua Corporation and TigerGene, positions the San Diego-based startup to advance a platform that could reshape how researchers tackle oncology and autoimmune disorders.At the heart of Valora's approach is its antibody-lectin chimera (AbLec) platform, which allows it to engineer antibodies to interact with the glycocalyx — a complex layer of glycoproteins, proteoglycans, and glycolipids that surrounds cells and helps them communicate with the immune system. In certain diseases, this "sugar code" can get disrupted and trigger abnormal immune responses — such as persistent immunosuppression within tumours, allowing cancer to evade the immune system, or poor immune regulation and hyperactivity, as seen in autoimmune diseases.Manipulating glyco-immunologyAccording to the startup, AbLec represents a departure from traditional protein-based checkpoint approaches in that it offers a more nuanced way of manipulating immune responses by precisely targeting glycan structures that play a critical role in immune system signalling.AbLecs pairs two key components: an antibody domain for precise cell targeting and a lectin domain that interacts with specific glycan structures on cell surfaces. Valora says the combination allows AbLecs to target specific cells, adjust immune signals at those cells, and either boost or suppress immune responses depending on their design."The glyco-immune checkpoint is present in cancer, it drives persistent immune suppression, and it is orthogonal to the PD-1/PD-L1 axis," CEO Miguel Garcia-Guzman told FirstWord, elaborating on the platform's potential in oncology. "Therefore, the AbLec platform provides a new way to restore innate and adaptive anti-cancer immunity and could provide therapeutic opportunities for patients that are refractory or do not respond to PD-1/PD-L1 modulators."The technology builds on research by Carolyn Bertozzi, the 2022 Nobel Prize recipient for chemistry, and Jessica Stark, who brings expertise in glycobiology from her position at MIT. Both serve as scientific founders and advisors for Valora.Setting prioritiesLooking ahead, Garcia-Guzman says the company is pursuing "a data-driven approach" as to which therapeutic area to pursue first. "In oncology, we're exploring constructs for superior anti-cancer activity and T-cell modulation, including molecules that may serve as alternatives to IgGs or antibody-drug conjugates, as well as constructs showing superior T-cell enhancement and engagement capabilities," he said.For autoimmune diseases, the company is focusing on specifically targeting hyperreactive T- and B-cells to enhance the glyco-immune checkpoint in these cells. "While novel from a therapeutic perspective," Garcia-Guzman said, the approach "holds significant potential based on our understanding of glyco-immune biology."The company is mum on specific timelines, but says it is advancing multiple programmes through early research toward preclinical development.