A team member mentioned that his father gradually gained weight after middle age, and his increasing weight was starting to affect his daily life. He realized that obesity raises the risk of many diseases and poses a serious threat to his health. As a result, he decided to start losing weight. He planned to jog for an hour every day, eat fewer carbohydrates, cut down on snacks, and limit meat intake to just two meals a day. However, he gave up after less than two days because it was simply too painful.

During a health check-up, the doctor recommended a weight-loss medication called semaglutide. After a week of treatment, the effects of semaglutide were quite significant. However, he was very resistant to the idea of needing to inject the medication each time. Additionally, the pharmacokinetics of semaglutide showed that the drug's concentration in the body rises initially and then significantly drops, resulting in strong efficacy in the first few days after injection, followed by a gradual decline in effectiveness.

This led us to wonder: if we could use a probiotic to continuously express the main active component of semaglutide, which is glucagon-like peptide-1 (GLP-1), would it help us avoid the inconveniences of injections and the issue of uneven drug concentration?

Obesity is a chronic disease characterized by the excessive accumulation of body fat, typically defined based on measurements of body mass index (BMI). According to international standards, a BMI of ≥ 30 is considered a hallmark of obesity[1]. The occurrence of obesity is closely related to various factors, including genetic susceptibility, environmental influences, lifestyle choices, and metabolic abnormalities[2][3].

Obesity has numerous negative impacts on health. Firstly, it increases the risk of cardiovascular diseases, such as hypertension, coronary heart disease, and stroke, as fat deposits on the inner walls of blood vessels can lead to atherosclerosis[4]. Secondly, it is significantly associated with an increased risk of developing type 2 diabetes, as excess fat interferes with the normal function of insulin[5]. Obesity can also put additional strain on joints, leading to osteoarthritis, and negatively affect the respiratory system, potentially triggering sleep apnea[6]. In addition to these physical health issues, obesity can impact mental health, resulting in decreased self-esteem, anxiety, and depression[6]. Therefore, controlling weight and maintaining a healthy lifestyle are crucial for overall health.

From 1990 to 2022, there has been a significant increase in the global obesity problem. The obesity prevalence has risen in 188 countries (94%) among women and in all but one country among men, with notable increases in countries like the United States and Romania. In terms of gender-specific obesity populations, the number of obese women reached 504 million, which represents a substantial increase of 377 million compared to 1990. Meanwhile, the number of obese men increased by 307 million, reaching 374 million[7]. The rapid rise in obesity rates is not limited to adults; many children and adolescents also suffer from obesity. Currently, it is estimated that over 159 million children worldwide are affected by obesity[7]. Therefore, addressing the issue of obesity to improve public health and quality of life is an urgent matter.

There are many weight loss methods, each with its drawbacks. Dietary control aims to reduce weight by adjusting dietary structure, including decreasing carbohydrate and fat intake while increasing high-quality protein and vegetables[8]. However, this approach requires long-term commitment and high self-discipline, making it prone to rebound.Exercise-based weight loss relies on increasing calorie expenditure through activities such as running, swimming, and fitness training[8]. However, excessive exercise or improper technique can lead to physical injuries and even strain the heart. Pharmaceutical weight loss options are abundant on the market, with significant variations in effectiveness and side effects. For example, orlistat may cause diarrhea and gastrointestinal discomfort, while naltrexone can lead to nausea and headaches; some medications may also not meet legal standards[9].Surgical weight loss is suitable for patients with severe obesity and includes methods like liposuction. Although the results can be significant, side effects are common, and there is a risk of rebound, along with inherent surgical risks[10].

Semaglutide was first approved for the treatment of type 2 diabetes in 2017[11], produced by Novo Nordisk. It was subsequently approved for weight management in 2021[12] (Fig 3). The introduction of semaglutide for weight loss marked an important expansion of its application in weight management, quickly garnering widespread attention. Notable public Figures such as former UK Prime Minister Boris Johnson, Hollywood star Rebel Wilson, and Tesla founder Elon Musk have all successfully used this medication for weight loss.

Semaglutide is a medication used for the treatment of type 2 diabetes and belongs to the class of glucagon-like peptide-1 (GLP-1) receptor agonists. GLP-1 hormones help lower blood sugar levels, promote insulin secretion, and inhibit glucagon secretion[13]. Additionally, semaglutide slows gastric emptying, which increases feelings of fullness and aids in weight reduction[14]. It is typically administered as an injection and has shown significant effectiveness in improving blood sugar control in diabetic patients while also assisting in weight management.

However, semaglutide has several drawbacks as a weight loss medication that can be difficult for people to accept:

It is administered via subcutaneous injection, which is an invasive method that can be inconvenient for patients and may lead to difficulties in long-term adherence. Additionally, injection methods carry risks of needle-related infections.

As a peptide, semaglutide exhibits a pharmacokinetic profile characterized by an initial sharp increase in drug concentration followed by a gradual decline. This means that after stopping the injections, the effectiveness diminishes over time, which can lead to weight regain.

The production process for semaglutide involves extracting GLP-1 from microbial hosts and then chemically modifying it to obtain the final product. This process is costly and yields low returns, contributing significantly to the high price of the medication.

Therefore, there is an urgent need to design a non-invasive GLP-1 treatment method that ensures stable drug concentration, effectively mitigates side effects, and is cost-effective.

To assist individuals in society who wish to lose weight in a healthier and more effective manner, while avoiding the inconveniences of traditional weight loss methods and the drawbacks of semaglutide injections, as well as the high cost of medications, our team has decided to create a probiotic that can aid in weight loss. This probiotic aims to help obese patients and individuals needing to lose weight achieve their goals in a more efficient, gentle, and effective way.

The probiotic is based on the microbial chassis Escherichia coli Nissle 1917, into which we have genetically engineered three systems: a fat-reduction system (GLP-1 secretion system), a muscle-gaining system (Bimagrumab secretion system), and a suicide system (Fig 4). Ultimately, this engineered probiotic can help patients lose weight while preventing muscle loss and effectively mitigating the potential side effects of GLP-1.

E. coli serves as an excellent chassis for genetic engineering due to its rapid growth, ease of cultivation, straightforward genetic manipulation, low cost, and the ability for precise genome modification. It is an ideal choice for the production of proteins and other biomolecules [15]. However, common strains like BL21 and DH5α can secrete intracellular toxins, which may pose potential risks to the host [16]. The secretion of these toxins can lead to a range of adverse reactions, including gastrointestinal discomfort, inflammatory responses, and even more serious health issues [16].

E. coli Nissle 1917 was isolated from the feces of a soldier during World War I who never contracted dysentery, making it unique in that it does not secrete intracellular toxins while retaining the manipulability of standard E. coli strains [17]. Therefore, when selecting a strain for developing a consumable engineered bacterium, we specifically chose E. coli Nissle 1917 as our chassis microorganism.

Glucagon-like peptide-1 (GLP-1) is a hormone primarily produced by intestinal L cells and belongs to the incretin family. GLP-1 enhances insulin secretion in a glucose-dependent manner by activating GLP-1 receptors, suppresses glucagon secretion [13], and delays gastric emptying, which reduces food intake through central appetite suppression [18]. These actions help lower blood glucose levels and promote weight loss (Fig 5).

However, GLP-1 is susceptible to degradation by dipeptidyl peptidase-4 (DPP-4) in the bloodstream, resulting in a half-life (t1/2) of only a few minutes [19]. Therefore, our team has decided to design a DPP-4 protein inhibitor to extend the half-life of GLP-1 in the body, allowing it to function more effectively.

After secretion from intestinal L cells, GLP-1 acts on pancreatic β cells via the bloodstream. It binds to GLP-1 receptors, activating adenylate cyclase and increasing cAMP levels. This increase in cAMP activates PKA and EPAC2, leading to the closure of ATP-sensitive potassium channels and subsequent depolarization of the cell membrane. This process opens voltage-gated calcium channels, raising intracellular calcium concentrations. These mechanisms promote the release of insulin granules and enhance insulin gene transcription, ultimately increasing insulin secretion. There is also evidence that GLP-1 can inhibit glucagon secretion [21].

Additionally, GLP-1 plays a crucial role in stabilizing blood glucose levels and managing weight by delaying gastric emptying. It does this by increasing gastric capacity and inhibiting cholinergic nerves, which slows the rate of gastric emptying [22]. Clinical studies indicate that the reduction in gastric motility induced by GLP-1 is primarily mediated through vagal nerve mechanisms, involving GLP-1 receptors on intestinal muscular layer neurons and related signaling pathways. Moreover, GLP-1 inhibits vagal nerve activity, further delaying gastric emptying and reducing gastric acid secretion [18].

Finally, to ensure the continuous secretion of GLP-1, we added the secretion signal peptide, pectate lyase B from Erwinia carotovora CE (PelB), to the N-terminus of GLP-1. PelB is a secretion tag composed of 22 N-terminal amino acid residues. This sequence allows for fusion with other proteins, facilitating the transfer of the fusion protein into the periplasmic space of Gram-negative bacteria, such as E. coli [36]. Additionally, we utilized a constitutive promoter to ensure the sustained expression of GLP-1, achieving the desired therapeutic effect.

With the rapid development of deep learning methods in biology in recent years, de novo protein design has become feasible. Significant advancements have been made in the field of protein structure prediction and design, particularly through the application of deep learning technologies. In protein structure prediction, DeepMind's AlphaFold[24] and the University of Washington's RoseTTAFold[25] have achieved breakthrough results, significantly enhancing prediction accuracy, especially in template-free structure prediction. In May 2024, DeepMind released AlphaFold3, which can predict the structures and interactions of all life molecules (including proteins, nucleic acids, small molecules, and ions) with unprecedented accuracy [24].

In terms of protein structure design, David Baker's ProteinMPNN[26] and RFdiffusion[27] algorithms have demonstrated powerful capabilities in sequence design and structural generation using deep learning. ProteinMPNN has shown a significantly higher recovery rate of sequences on natural protein scaffolds compared to traditional methods [26], while RFdiffusion, through diffusion model training, can more accurately generate protein structures with realistic folding properties, significantly improving design quality and efficiency [27]. The application of these technologies continues to push the frontiers of protein research and applications, showcasing the immense potential of deep learning in the field of biology.

Therefore, our team utilized existing advanced deep learning tools to design a comprehensive computational workflow for developing DPP-4 protein inhibitors. This computational framework has demonstrated favorable characteristics at the computational level, with the potential to subsequently help extend the half-life of GLP-1 and enhance its efficacy.

Weight loss often leads to muscle reduction, primarily due to caloric deficits that cause the body to not only burn fat but also break down muscle for energy. As body weight decreases, the basal metabolic rate drops, and hormonal changes further affect muscle maintenance. Therefore, an effective weight loss strategy should include adequate protein intake and strength training, which is where Bimagrumab comes in[29].

Bimagrumab is a humanized monoclonal antibody that binds to type II hormone receptors on muscle cells. It is primarily used to treat muscle-wasting diseases and has been evaluated in several clinical trials. Bimagrumab binds to ActRIIA and ActRIIB, which play critical roles in mediating the signaling of myostatin and activin. This binding prevents myostatin and activin from attaching to type IIB activin receptors on muscle membranes, thereby inhibiting the activation of the transmembrane kinases ALK4 and ALK5.

As a result, ALK4 remains inactive, preventing the phosphorylation of Smad2 and Smad3, which means Smad4 cannot be recruited to form the Smad complex, thereby failing to induce changes in gene transcription that would lead to muscle atrophy. Additionally, the inactivation of ALK5 keeps AKT in a phosphorylated state, which prevents the dephosphorylation of FOXO. Consequently, FOXO cannot enter the nucleus to activate specific muscle atrophy E3 ligases (such as MuRF1 and Atrogin-1/MAFbx) and other related atrophy genes. Thus, this process avoids the ubiquitination and degradation of muscle proteins via proteasomal or autophagic pathways[30].

In summary, Bimagrumab can bind to ActRIIA/ActRIIB, inhibiting downstream signaling pathways, preventing muscle wasting, and increasing skeletal muscle density.

Similar to GLP-1, we also chose to use the PelB secretion tag to facilitate the effective secretion of Bimagrumab into the extracellular space, allowing it to exert its effects.

To prevent gene leakage, we have designed a suicide system controlled by an arabinose operon and a low-temperature promoter.To prevent gene leakage, we designed an OR-gate suicide system. In this system, the arabinose-inducible element and the cold-inducible promoter control the expression of the suicide genes. When the engineered bacteria are excreted from the gut, the low temperature triggers bacterial suicide. If we want to eliminate the bacteria inside the gut, we can administer arabinose to activate the suicide genes. Additionally, this system provides dual control, both inside and outside the body, ensuring that gene leakage is effectively prevented during use and after environmental release.

In the absence of arabinose, the arabinose regulatory protein AraC binds to the pBAD promoter, inhibiting gene transcription. When L-arabinose is present, the arabinose operon is activated, leading to the rapid induction of the pBAD promoter, which can reach maximum expression levels within minutes (Fig 12) [31]. This system allows for the induction of lysis in the engineered probiotics by administering arabinose, effectively triggering the secretion of lysozyme and causing the probiotics to undergo apoptosis. This can be particularly useful if the patient experiences side effects from the probiotics or wishes to stop GLP-1 production.

To prevent gene leakage into the environment and avoid contamination, we have chosen the low-temperature inducible promoter pCspA to induce the expression of lytic proteins for bacterial suicide. Low temperatures enhance the activity of the CspA promoter, increasing the transcription level of the CspA gene. The expression of the CspA gene under cold shock conditions is primarily regulated at the post-transcriptional level, with the secondary structure of its mRNA exhibiting different stability and translation efficiency at 37°C and low temperatures (Fig 13). Under cold induction, the secondary structure of CspA mRNA undergoes temperature-dependent rearrangement, forming a "cold shock" structure that has higher translation efficiency and lower degradation rates compared to the structure at 37°C. These data indicate that CspA mRNA can directly sense temperature changes and respond to cold shock through structural changes, without the involvement of transcription factors [32]. Therefore, we have chosen a cold-inducible promoter to ensure that the engineered bacteria express the suicide gene under low-temperature conditions, preventing gene leakage.

Finally, the two promoters are connected through an "OR" gate, allowing the downstream expression of the T4 Holin and T4 Lysozyme genes upon activation. This design enables the engineered bacteria to express the suicide genes both inside and outside the gut, providing a safeguard against gene leakage, a feature not present in existing methods.

T4 Holin is a small membrane protein with two transmembrane domains, with its N-terminus and C-terminus located in the cytoplasm and extracellular space, respectively. This structure allows Holin to form pores in the cell membrane, leading to membrane disruption (Figure 14, left) [33].

T4 Lysozyme is a small globular protein composed of 164 amino acids (Figure 14, right). It initially binds to the bacterial cell wall, recognizing and cleaving the chemical bonds between N-acetylglucosamine and N-acetylmuramic acid in the cell wall. This action compromises the integrity of the cell wall, disrupting the osmotic balance inside the cell, ultimately leading to cytoplasmic leakage and cell lysis [34].

The synergistic action of these two proteins results in the destruction of both the bacterial cell membrane and wall, causing the leakage of cellular contents and ultimately leading to the lysis of the entire bacterial cell.

Our probiotic is designed for individuals with a BMI greater than 30, categorized as overweight. According to a report by the China Business Industry Research Institute, the market for overweight/obesity treatment drugs in China is projected to reach 1.7 billion yuan in 2023, with a year-on-year growth of 30.8%. Analysts predict this market will expand to 3.9 billion yuan in 2024. This highlights a significant and growing demand for our probiotic among the overweight population.

Initially, we will cultivate the engineered bacteria in fermentation tanks. After cultivation, we will utilize freeze-drying technology to extract the E. coli colonies. The lyophilized powder will then be mixed with auxiliary materials to create either a powder formulation or capsules. Throughout the production process, we will precisely control moisture, oxygen, and temperature to ensure the stability and viability of the probiotic strains are effectively maintained.

The probiotic freeze-dried powder will be made into capsules for easy portability. When consumers want to take it, they can simply swallow the capsule with warm water. If consumers prefer not to take the capsule, they can purchase commercial probiotic powder and dissolve the freeze-dried powder in warm water (note: absolutely do not use boiling water) or milk. It's important to avoid taking antibiotics during this process and for four hours after consuming the freeze-dried powder, as antibiotics can diminish its effectiveness. If consumers wish to quickly kill the engineered bacteria, they can consume an appropriate amount of arabinose to induce the E. coli to activate its internal suicide system.

Lastly, since our product essentially functions as a weight loss medication, it is not suitable for every consumer. According to the "Chinese Guidelines for Medical Nutrition Therapy for Overweight/Obesity (2021)," it is recommended that individuals with a BMI ≥ 30 kg/m² or those with comorbidities (such as hypertension or diabetes) with a BMI ≥ 28 kg/m² consider using medication for weight loss. Therefore, our product is designed to serve individuals who have a genuine need to lose weight, alleviating their challenges and inconveniences in the weight loss process.

Our project offers several advantages over traditional weight loss methods:

No injections: Our product does not require injections, alleviating patient fears associated with needle use.

Continuous GLP-1 secretion: Our probiotics continuously secrete GLP-1 in the gut, maintaining a consistent concentration of the drug in the body.

Fat loss and muscle preservation: Our product helps patients lose fat while preventing muscle loss.

Cost-effective production: The production process is simple and cost-effective, resulting in a low retail price.

Side effect mitigation: Traditional methods often lead to side effects that cannot be promptly addressed. If patients experience side effects from our probiotics, they can take arabinose to induce the engineered bacteria to self-destruct, mitigating adverse effects.

Efficient fat loss: Our product enables patients to achieve fat loss efficiently and with minimal time investment, providing convenience.

[1] World Health Organization. Surveillance of chronic disease risk factors: country level data and comparable estimates[J]. 2005.

[2] Bleich S N, Cutler D, Murray C, et al. Why is the developed world obese?[J]. Annu. Rev. Public Health, 2008, 29(1): 273-295.

[3] Kassotis C D, Vandenberg L N, Demeneix B A, et al. Endocrine-disrupting chemicals: economic, regulatory, and policy implications[J]. The lancet Diabetes & endocrinology, 2020, 8(8): 719-730.

[4] Blüher M. Obesity: global epidemiology and pathogenesis[J]. Nature Reviews Endocrinology, 2019, 15(5): 288-298.

[5] Grundy S M. Obesity, metabolic syndrome, and cardiovascular disease[J]. The Journal of Clinical Endocrinology & Metabolism, 2004, 89(6): 2595-2600.

[6] Haslam DW, James WP. Obesity. Lancet. 2005 Oct 1;366(9492):1197-209. doi: 10.1016/S0140-6736(05)67483-1. PMID: 16198769.

[7] Phelps N H, Singleton R K, Zhou B, et al. Worldwide trends in underweight and obesity from 1990 to 2022: a pooled analysis of 3663 population-representative studies with 222 million children, adolescents, and adults[J]. The Lancet, 2024, 403(10431): 1027-1050.

[8] Jensen M D, Ryan D H, Apovian C M, et al. 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society[J]. Circulation, 2014, 129(25_suppl_2): S102-S138.

[9] Heymsfield S B, Wadden T A. Mechanisms, pathophysiology, and management of obesity[J]. New England Journal of Medicine, 2017, 376(3): 254-266.

[10] Yoshino M, Kayser B D, Yoshino J, et al. Effects of diet versus gastric bypass on metabolic function in diabetes[J]. New England Journal of Medicine, 2020, 383(8): 721-732.

[11] "Ozempic- semaglutide injection, solution". DailyMed. Archived from the original on 5 June 2021. Retrieved 5 June 2021.

[12] Phillips A, Clements J N. Clinical review of subcutaneous semaglutide for obesity[J]. Journal of clinical pharmacy and therapeutics, 2022, 47(2): 184-193.

[13] Li Y, Hansotia T, Yusta B, et al. Glucagon-like peptide-1 receptor signaling modulates β cell apoptosis[J]. Journal of Biological Chemistry, 2003, 278(1): 471-478.

[14] Doggrell S A. Sgemaglutide in type 2 diabetes–is it the best glucagon-like peptide 1 receptor agonist (GLP-1R agonist)?[J]. Expert Opinion on Drug Metabolism & Toxicology, 2018, 14(3): 371-377.

[15] Lee S Y. High cell-density culture of E. coli [J]. Trends in biotechnology, 1996, 14(3): 98-105.

[16] Roth J, Blatteis C M. Mechanisms of fever production and lysis: lessons from experimental LPS fever[J]. Compr Physiol, 2014, 4(4): 1563-1604.

[17] Roth J, Blatteis C M. Mechanisms of fever production and lysis: lessons from experimental LPS fever[J]. Compr Physiol, 2014, 4(4): 1563-1604.

[18] Kim K S, Park J S, Hwang E, et al. GLP-1 increases preingestive satiation via hypothalamic circuits in mice and humans[J]. Science, 2024, 385(6707): 438-446.

[19] Andersen E S, Deacon C F, Holst J J. Do we know the true mechanism of action of the DPP‐4 inhibitors?[J]. Diabetes, Obesity and Metabolism, 2018, 20(1): 34-41.

[20] Wikipedia contributors. (2024, August 8). Glucagon-like peptide-1. In Wikipedia, The Free Encyclopedia. Retrieved 07:57, September 9, 2024, from https://en.wikipedia.org/w/index.php?title=Glucagon-like_peptide-1&oldid=1239355190

[21] Kobayati A, Haidar A, Tsoukas M A. Glucagon‐like peptide‐1 receptor agonists as adjunctive treatment for type 1 diabetes: Renewed opportunities through tailored approaches?[J]. Diabetes, Obesity and Metabolism, 2022, 24(5): 769-787.

[22] Schirra J, Nicolaus M, Woerle H J, et al. GLP‐1 regulates gastroduodenal motility involving cholinergic pathways[J]. Neurogastroenterology & Motility, 2009, 21(6): 609-e22.

[23] Drucker D J, Habener J F, Holst J J. Discovery, characterization, and clinical development of the glucagon-like peptides[J]. The Journal of clinical investigation, 2017, 127(12): 4217-4227.

[24] Abramson J, Adler J, Dunger J, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3[J]. Nature, 2024: 1-3.

[25] Baek M, DiMaio F, Anishchenko I, et al. Accurate prediction of protein structures and interactions using a three-track neural network[J]. Science, 2021, 373(6557): 871-876.

[26] Dauparas J, Anishchenko I, Bennett N, et al. Robust deep learning–based protein sequence design using ProteinMPNN[J]. Science, 2022, 378(6615): 49-56.

[27] Watson J L, Juergens D, Bennett N R, et al. De novo design of protein structure and function with RFdiffusion[J]. Nature, 2023, 620(7976): 1089-1100.

[28] Krishna R, Wang J, Ahern W, et al. Generalized biomolecular modeling and design with RoseTTAFold All-Atom[J]. Science, 2024, 384(6693): eadl2528.

[29] Morvan F, Rondeau J M, Zou C, et al. Blockade of activin type II receptors with a dual anti-ActRIIA/IIB antibody is critical to promote maximal skeletal muscle hypertrophy[J]. Proceedings of the National Academy of Sciences, 2017, 114(47): 12448-12453.

[30] Tsuchida K, Nakatani M, Hitachi K, et al. Activin signaling as an emerging target for therapeutic interventions[J]. Cell Communication and Signaling, 2009, 7: 1-11.

[31] Guzman L M, Belin D, Carson M J, et al. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter[J]. Journal of bacteriology, 1995, 177(14): 4121-4130.

[32] Giuliodori A M, Di Pietro F, Marzi S, et al. The cspA mRNA is a thermosensor that modulates translation of the cold-shock protein CspA[J]. Molecular cell, 2010, 37(1): 21-33.

[33] Mehner-Breitfeld D, Schwarzkopf J M F, Young R, et al. The phage T4 Antiholin RI has a cleavable signal peptide, not a SAR domain[J]. Frontiers in Microbiology, 2021, 12: 712460.

[34] Krieger I V, Kuznetsov V, Chang J Y, et al. The structural basis of T4 phage lysis control: DNA as the signal for lysis inhibition[J]. Journal of molecular biology, 2020, 432(16): 4623-4636.

[35] 孙铭遥, 陈伟.《中国超重/肥胖医学营养治疗指南 (2021)》解读[J]. 协和医学杂志,2022, 13(2): 255-262.

[36] Sockolosky J T, Szoka F C. Periplasmic production via the pET expression system of soluble, bioactive human growth hormone[J]. Protein expression and purification, 2013, 87(2): 129-135.