Our project, SPARK (Signal-Programmable Activation of Regulated seCretion), began with a central question:

What prevents current designer cell therapies from becoming effective clinical tools?

We discovered that most gene-switch systems rely on transcriptional control, which requires new protein synthesis before any therapeutic effect occurs—a process that can take many hours. For time-critical treatments such as insulin release or cytokine delivery, such delays are unacceptable.

To overcome this limitation, we aimed to design a fast, modular, and safe post-translational switch capable of rapidly releasing pre-formed secretory vesicles in response to external or internal cues.

From the beginning, Human Practices was not a side effort but an integral part of the design-build-test-learn (DBTL) cycle. Each stage of development—concept generation, system optimization, hardware design, and future planning—was shaped by stakeholder feedback, expert consultation, and public engagement to ensure SPARK remains responsible, relevant, and beneficial to society.

Figure 1. Integrated Human Practices framework guiding SPARK development.

Context

Cell therapy holds tremendous promise, yet clinical translation is limited by inadequate timing control. To understand clinical needs, we interviewed Dr. Jie Zhou, Director of Endocrinology at Xijing Hospital.

"Even when gene switches work, their response time is too slow for physiological regulation. In diabetes control, for example, insulin should be expressed minutes after the increase of glycemia. Every minute matters."

Impact on Design

This conversation defined our engineering objective: build a rapid-response secretion switch operating on the timescale of minutes. Her feedback directly inspired the concept of a secretion-level regulatory system that bypasses the transcription-translation delay.

Figure 2. Consultation with Dr. Jie Zhou identified the unmet clinical need for faster therapeutic response.

Rationale

To achieve rapid control, we envisioned a mechanism in which pre-synthesized proteins are stored in vesicles and released upon receiving a signal. This concept evolved into the modular SPARK architecture, consisting of:

• a Release Module (signal peptide + target protein + furin + TMD),
• a Response Module (signal-sensitive dissociation pair), and
• a Retention Module (cytoskeletal anchor LifeAct).

Outcome

This tripartite design allowed us to separate biological functions into interchangeable modules—making SPARK compatible with different input signals and cell types (see the Design page for details).

Challenge

Our first prototype employed mMaple3, a violet-light photolytic protein. While conceptually successful, violet illumination produced strong cytotoxicity in mammalian cells.

Stakeholder Engagement

At the CCiC 2025 Conference, we discussed this issue with Dr. Haotian Guo (CEO, Ailurus Bio). He cautioned that violet-light photolysis was unsuitable for therapeutic use due to limited tissue penetration and phototoxic effects.

Action Taken

We pivoted from photolytic cleavage to signal-induced dissociation, leading to two complementary systems:

• OptoSPARK: activated by red or near-infrared light, ensuring deep tissue penetration and spatiotemporal control.
• chemSPARK: activated by the small molecule Grazoprevir via NS3a–ANR dissociation, enabling pharmacological precision using an FDA-approved drug.

Result

This design transition, motivated by expert and clinical feedback, made SPARK more biocompatible, translatable, and modular, aligning engineering feasibility with medical relevance.

Figure 3. Feedback from industry experts at CCiC 2025 drove the transition to dissociation-based regulation.

Context

Precise optical control required custom hardware for reproducible illumination.

Engagement and Feedback

We collaborated with Dr. Shaowei Zhang, a synthetic biologist at National University of Defense Technology, who advised on practical improvements:

• Prevent light leakage to ensure experimental consistency,
• Add ventilation and heat dissipation to avoid thermal stress, and
• Integrate adjustable power output for light-intensity optimization.

Implementation

We adopted all recommendations, producing a compact, modular, and cost-efficient illumination device. Later, we shared the prototype with Prof. Jiawei Shao (Zhejiang University), who confirmed its research utility and encouraged further cost reduction for community adoption.

Outcome

The result was not only a reliable research instrument for our team but also a hardware contribution that can support future synthetic-biology projects requiring light-controlled regulation.

Figure 4. Expert-guided improvement of SPARK illumination hardware.

Refining the Dissociation Module

Further consultation with Dr. Shaowei Zhang revealed that CarH (green light) and BphP1–PpsR2 (NIR) tend to form homodimers, reducing anchoring efficiency. He recommended exploring heterodimeric dissociation systems to enhance modularity and minimize crosstalk. This feedback defined our next engineering goal: build orthogonal heteromeric pairs for future versions of SPARK.

Moving Toward Closed-Loop Therapeutics

After validating chemSPARK for rapid, tunable secretion, we discussed future directions with Prof. Jiawei Shao (Zhejiang University School of Medicine). He suggested:

"Move beyond external triggers toward closed-loop regulation—for instance, calcium-responsive secretion systems that mirror natural endocrine feedback."

This conversation redirected our long-term vision toward endogenous signal-responsive SPARK, capable of autonomous homeostatic regulation—an important step toward smart cell therapies for diabetes and beyond.

Expanding Applications Through Public Dialogue

At the Yuelu Mountain Science Fair (Changsha, 2025), we presented SPARK to scientists and the public. Discussions inspired new cross-disciplinary ideas:

• Plant hormone-responsive secretion systems (Dr. Ruozhong Wang, Hunan Agricultural University)
• Deep-learning-enabled real-time monitoring (Dr. Zhihong Xiao, Hunan Academy of Forestry)

These dialogues broadened our perspective, showing that SPARK's modular logic could extend beyond medicine into agriculture, biosensing, and biomanufacturing.

Figure 5. Cross-disciplinary discussion at the Yuelu Mountain Science Fair inspired new directions for SPARK.

To ensure that others can learn from and build upon our approach, we documented our experience through collaborative initiatives:

These resources demonstrate our belief that responsible innovation requires collaboration, transparency, and continuous reflection.

At every stage, Human Practices shaped SPARK's evolution—from identifying unmet clinical needs to optimizing experimental design and expanding societal relevance.

Through sustained engagement with clinicians, synthetic biologists, engineers, and the public, we created a system that is:

• Clinically inspired: addressing real therapeutic challenges;
• Scientifically optimized: modular, orthogonal, and fast;
• Socially responsible: open, educational, and community-oriented.

By converting dialogue into design, SPARK exemplifies how engineering biology can be both innovative and ethical. We hope our integrated Human Practices workflow can serve as a model for future teams seeking to design technologies that are good for science and good for the world.