從阿司匹林到突破KRAS等不可成藥靶點：百年共價藥物背后，這些技術正延伸藥物發(fā)現(xiàn)邊界 | Bilingual

編者按：從阿司匹林的“意外誕生”，到如今不斷攻克“不可成藥”靶點，共價藥物走過了一條跨越百年的進化之路。截至2025年，已有超過50款共價藥物獲批上市，在腫瘤、感染以及神經(jīng)系統(tǒng)疾病等領域展現(xiàn)出獨特價值。如何更高效、精準地發(fā)現(xiàn)這類藥物，仍是擺在行業(yè)面前的一道關鍵難題。依托一體化、端到端的CRDMO賦能平臺，藥明康德致力為全球合作伙伴提供覆蓋共價藥物發(fā)現(xiàn)與開發(fā)的一體化解決方案。本文將重點介紹藥明康德在共價藥物發(fā)現(xiàn)領域的能力布局與實踐。

共價藥物的百年進化

1899年，一種名為乙酰水楊酸的化合物以阿司匹林之名正式上市，用于緩解疼痛與炎癥。這一現(xiàn)代藥物史上的里程碑事件，也在無意間揭開了共價藥物的序幕。

傳統(tǒng)小分子藥物與靶蛋白之間，通常依賴非共價相互作用。這種結合更像一次短暫的“握手”，藥物一旦被代謝清除，療效便隨之消失。

而共價藥物，則提供了另一種可能。

它們通過親電基團，與靶蛋白的特定氨基酸形成共價鍵。這個過程就如同用鑰匙插入目標蛋白的“鎖孔”，以“先結合、再鎖定”的方式延長藥效持續(xù)時間。

盡管阿司匹林早已惠及無數(shù)患者，但其作用機制卻在數(shù)十年后才揭曉。20世紀70年代，科學家發(fā)現(xiàn)阿司匹林通過乙酰基親電基團與環(huán)氧化酶的絲氨酸殘基結合，形成共價鍵。這一發(fā)現(xiàn)不僅解釋了其藥理機制，也讓“共價抑制”的概念逐漸清晰。

類似的故事并不罕見。青霉素、頭孢菌素、磷霉素等經(jīng)典抗生素，都是在臨床成功多年之后，人們才真正理解其共價作用機制。

由于對共價藥物可能引發(fā)脫靶效應的擔憂，這類分子在藥物研發(fā)中曾長期被邊緣化。隨著這些“意外發(fā)現(xiàn)”的分子不斷在臨床上證明其價值，局面開始改變。

進入21世紀，通過藥物理性設計，BTK抑制劑伊布替尼（ibrutinib）與EGFR抑制劑阿法替尼（afatinib）等共價抑制劑相繼問世。近年來，共價藥物的研發(fā)進入新階段。通過系統(tǒng)性篩選與特定殘基共價結合的配體，共價藥物叩開了曾經(jīng)“不可成藥”靶點的大門，促成了sotorasib、adagrasib等KRAS G12C抑制劑的誕生。

如今，已有超過50款共價藥物獲FDA批準上市。共價機制所帶來的高效力、長作用時間以及對新靶點空間的拓展，使其成為當下藥物研發(fā)的重要方向之一。

共價藥物的快速發(fā)展也對篩選技術提出了更高要求。從共價化合物庫高通量篩選，到片段篩選以及DNA編碼化合物庫（DEL）等新興技術，這些方法正不斷開拓共價藥物發(fā)現(xiàn)的新路徑，也為這一跨越百年的藥物故事開啟新的篇章。

共價DEL打開共價篩選新空間

在共價藥物的發(fā)現(xiàn)過程中，藥明康德生物學業(yè)務平臺團隊就曾面臨著一項挑戰(zhàn)。他們需要為具有促癌效應的糖酵解酶PGAM1設計共價抑制劑，但棘手之處在于，這一靶點的活性口袋中，并沒有共價藥物最常用的反應位點半胱氨酸。

在共價藥物發(fā)現(xiàn)中，半胱氨酸通常是首選“攻擊目標”。其攜帶的巰基反應活性強，易于與親電基團形成穩(wěn)定的共價鍵。但這一策略也有明顯的局限性。在人體的蛋白質組中，半胱氨酸僅占約2%。這意味著，大量潛在靶點缺乏可利用的反應位點。PGAM1正是其中的典型。

破局的關鍵，來自于藥明康德持續(xù)建設的DEL技術平臺。

在DEL體系中，每個化合物都連接著一段獨特的DNA標簽。當數(shù)百萬甚至數(shù)十億個攜帶標簽的化合物與靶蛋白共同孵育，研究人員只需讀取標簽，便能快速鎖定高親合力的苗頭化合物，大幅提升了篩選效率。

早在2018年，當DEL仍屬于一項少數(shù)實驗室掌握的復雜技術時，為了讓這項技術成為每一位科研人員都能輕松使用的藥物發(fā)現(xiàn)工具，藥明康德生物學業(yè)務平臺啟動了DEL平臺建設。從零起步，僅用數(shù)月首個DEL產品初步成型；一年后，平臺迎來了首位客戶。

如今，隨著DELopen、DELight、DELpro等產品的陸續(xù)推出，DEL平臺已成為藥明康德新藥發(fā)現(xiàn)體系的基石之一，每年為客戶執(zhí)行數(shù)百次篩選，覆蓋數(shù)十億級別的化合物空間。

在共價藥物發(fā)現(xiàn)領域，藥明康德生物學業(yè)務平臺多年前也已建立起共價DEL（cDEL）能力。2023年，隨著共價化合物庫的升級，cDEL平臺持續(xù)擴展，形成覆蓋可逆與不可逆共價藥物發(fā)現(xiàn)的完整體系。

例如，針對不可逆共價藥物發(fā)現(xiàn)，DELink Pro平臺涵蓋了16億化合物和184種專門設計的共價親電基團，并提供可定制的一站式解決方案；而在可逆共價領域，平臺也構建了超過4億化合物、涵蓋39類親電基團的篩選能力。

值得一提的是，這一平臺不再局限于半胱氨酸位點，而是將篩選范圍拓展至賴氨酸、絲氨酸、酪氨酸等多種氨基酸殘基。此外，新型親電基團的引入，也讓篩選具有更高的特異性與更低的脫靶風險。

在PGAM1項目中，基于這一綜合性能力，藥明康德團隊利用不可逆和可逆共價DEL技術，鑒定并驗證了一類靶向酪氨酸的新型抑制劑。得益于cDEL庫中多樣化的親電基團，這類抑制劑引入了未曾使用過的非經(jīng)典親電基團結構。

正如這一案例所展示的，依托cDEL平臺，共價藥物的“可探索空間”正被系統(tǒng)性地打開。

新進展：系統(tǒng)評估共價DEL親電基團

cDEL平臺能力的提升，不僅依賴于規(guī)模擴展，還需要方法的完善。在一項發(fā)表于Helvetica Chimica Acta的近期研究中，藥明康德生物學業(yè)務平臺團隊與蘇黎世聯(lián)邦理工學院合作，對共價DEL親電基團的適用性進行了系統(tǒng)性評估。

這項研究聚焦于cDEL面臨的反應性與穩(wěn)定性問題：高反應性的親電基團可能會導致選擇性下降，此外不同親電基團的穩(wěn)定性差異顯著。若缺乏系統(tǒng)評估，可能直接影響篩選結果的可靠性。

為此，研究團隊選取了59種具有代表性的親電基團，覆蓋8大類結構類型。其中，6類用于形成不可逆共價鍵，包括芳基鹵化物、烯烴、炔烴等；另外2類（醛和腈類）則傾向于形成可逆共價鍵。

▲該研究對59種親電基團的DEL適用性進行了系統(tǒng)評估（圖片來源：參考資料[1]）

在模擬真實DEL構建流程的條件下，研究團隊首先對各類親電基團的偶聯(lián)效率進行了系統(tǒng)評估。在59種候選親電基團中，有45種能夠以超過50%的效率成功偶聯(lián)，具備良好的DEL兼容性。

在此基礎上，研究團隊進一步考察了這45種親電基團對半胱氨酸和賴氨酸這兩類氨基酸殘基的反應活性、選擇性，以及候選親電基團的反應速度、穩(wěn)定性等指標。

最終，研究團隊篩選出21種具有高偶聯(lián)效率以及反應活性的親電基團，這些親電基團在后續(xù)檢驗中均表現(xiàn)出了良好的穩(wěn)定性。值得一提的是，醛和腈類可逆親電基團在此次實驗中均未能穩(wěn)定存在，表明后續(xù)研究中，需要額外的化學步驟將其轉化為穩(wěn)定形式。

這項研究系統(tǒng)評估了化學性質各異的親電基團在共價DEL中的適用性，有望為未來共價DEL的理性設計提供參考。

構建一體化共價藥物發(fā)現(xiàn)引擎

cDEL并不是孤立存在的技術。在藥明康德生物學業(yè)務平臺的共價藥物發(fā)現(xiàn)平臺中，cDEL與共價片段藥物發(fā)現(xiàn)（cFBDD）、共價高通量篩選（cHTS）技術深度融合，共同構成了協(xié)同互補的綜合性發(fā)現(xiàn)引擎。

其中，cHTS平臺以大規(guī)模共價化合物庫為基礎，構建了高效的苗頭化合物發(fā)現(xiàn)引擎。cHTS共價庫包含約6.9萬個分子，涵蓋50多種不同類型的共價親電基團，能夠作用于9類氨基酸殘基，將篩選范圍從傳統(tǒng)單一位點拓展至更廣泛的蛋白質空間。cHTS平臺還提供了“direct to biology”（D2B）解決方案，為苗頭化合物的發(fā)現(xiàn)，以及從苗頭到先導提供全方位支持。

藥明康德的cFBDD平臺則從“更小”的分子片段出發(fā)，提供另一條高效的共價藥物發(fā)現(xiàn)路徑。該平臺基于一個包含2600余個結構多樣化片段的化合物庫，并通過引入帶有親電基團的片段，快速識別可作為優(yōu)化起點的共價結合物。這一策略尤其適用于結構復雜、“不可成藥”的靶點。

一百多年間，共價藥物完成了一場從偶然發(fā)現(xiàn)到理性設計的跨越。而這場跨越的背后，是篩選技術的一次次突破與迭代。

如今，藥明康德cDEL、cHTS、cFBDD等技術平臺協(xié)同，為客戶提供系統(tǒng)性的解決方案，也讓共價藥物發(fā)現(xiàn)進入更加系統(tǒng)化的新階段。

依托端到端、一體化CRDMO賦能平臺，藥明康德致力于加速突破性療法的開發(fā)，幫助合作伙伴將創(chuàng)新成果高效轉化為造福全球患者的解決方案，以踐行“讓天下沒有難做的藥，難治的病”的愿景。

了解WuXi Biology如何賦能藥物研發(fā)，請長按掃描上方二維碼，與藥明康德生物學業(yè)務平臺聯(lián)系

From Aspirin to Targeting KRAS: These Technologies Are Redrawing the Frontiers of Drug Discovery

From the “accidental birth” of aspirin to recent advances against previously “undruggable” targets, covalent drugs have evolved for over a century. As of 2025, over 50 covalent drugs have been approved, demonstrating unique value in oncology, infectious diseases, neurological disorders and beyond. It has remained a central challenge to discover covalent molecules more efficiently and precisely. Leveraging its integrated, end-to-end CRDMO platform, WuXi AppTec is committed to providing global partners with integrated solutions for covalent drug discovery and development. This article highlights WuXi AppTec’s capabilities and practices in covalent drug discovery.

A Century-Long Evolution of Covalent Drugs

In 1899, a compound known as acetylsalicylic acid was officially launched under the name aspirin to relieve pain and inflammation. This milestone in modern pharmaceutical history also inadvertently marked the beginning of covalent drugs.

Traditional small-molecule drugs rely on non-covalent interactions with target proteins. Such binding resembles a transient “handshake”, once the drug is metabolized and cleared, its therapeutic effect dissipates.

Covalent drugs, however, offer a different paradigm.

They contain electrophiles that form covalent bonds with specific amino acid residues on target proteins. This process can be likened to inserting a key into a lock, enabling a “docking & locking” mechanism that prolongs the duration of action.

While aspirin has benefited countless patients, its mechanism of action was not elucidated until decades later. In the 1970s, scientists discovered that aspirin forms a covalent bond with a serine residue in cyclooxygenase via its acetyl electrophile. This finding not only explained its pharmacological activity but also helped define the concept of covalent inhibition.

Such stories are far from rare. Classic antibiotics such as penicillins, cephalosporins and fosfomycin achieved clinical success long before their covalent mechanisms were fully understood.

Due to concerns about potential off-target effects, covalent drugs were once sidelined in drug development. However, as these “accidentally discovered” molecules continued to prove their clinical value, perceptions began to shift.

Entering the 21st century, rational drug design enabled the emergence of covalent inhibitors such as the BTK inhibitor ibrutinib and the EGFR inhibitor afatinib. In recent years, covalent drug discovery has entered a new phase. Through systematic screening of ligands that form covalent bonds with specific residues, researchers have unlocked previously “undruggable” targets, leading to the development of KRAS G12C inhibitors such as sotorasib and adagrasib.

To date, more than 50 covalent drugs have been approved by the FDA. Their high potency, prolonged duration of action, and ability to expand the target landscape have made covalent mechanisms a key direction in modern drug discovery.

The rapid progress of covalent drugs has also raised the bar for screening technologies. From high-throughput screening of covalent libraries to fragment-based approaches and DNA-encoded libraries (DEL), emerging technologies are continuously opening new avenues for covalent drug discovery, ushering in a new chapter in this century-long journey.

Covalent DEL Expands the Landscape of Covalent Screening

In the course of covalent drug discovery, WuXi Biology (a business unit of WuXi AppTec) once encountered a notable challenge: designing covalent inhibitors for the pro-oncogenic glycolytic enzyme PGAM1. The difficulty lay in the absence of cysteine within the enzyme’s active site.

Cysteine is the most commonly targeted reactive residue for covalent drugs due to the high reactivity of its thiol group, which readily forms stable covalent bonds with electrophiles. However, this strategy has clear limitations. Cysteine accounts for only about 2% of residues in the human proteome, leaving many potential targets without accessible reactive sites. PGAM1 is a representative example.

The progress came from WuXi Biology’s continuously evolving DEL platform.

In a DEL system, each compound is tagged with a unique DNA tag. When millions—or even billions—of tagged compounds are incubated with a target protein, researchers can rapidly identify high-affinity hits by decoding the tags, dramatically improving screening efficiency.

As early as 2018, when DEL remained a technically demanding approach mastered by only a few laboratories, WuXi Biology initiated the construction of its DEL platform with the goal of making it broadly accessible for all researchers. Starting from scratch, the first DEL offering took shape within months, and the platform welcomed its first client after a year.

Today, with the launch of products such as DELopen, DELight and DELpro, the DEL platform has become a cornerstone of WuXi AppTec’s drug discovery engine, enabling hundreds of screening campaigns annually across chemical spaces encompassing billions of compounds.

In the field of covalent drug discovery, WuXi Biology established its covalent DEL (cDEL) capabilities years ago. In 2023, with the expansion of its covalent compound libraries, the cDEL platform further evolved into a comprehensive system, supporting both reversible and irreversible covalent drug discovery.

For example,in irreversible covalent discovery, the DELink Pro platform covers 1.6 billion compounds and 184 specifically designed covalent electrophiles, offering customizable, one-stop solutions.In the reversible covalent space, the platform supports screening of over 400 million compounds incorporating 39 classes of electrophiles.

Notably, the platform is no longer limited to cysteine-targeting strategies.It has expanded screening to multiple amino acid residues, including lysine, serine and tyrosine.The introduction of novel electrophiles also enhances specificity while reducing off-target risks.

In the PGAM1 project, leveraging this integrated capability, WuXi Biology utilized both irreversible and reversible cDEL technologies to identify and validate a novel type of tyrosine-targeting inhibitors. Thanks to the diversity of electrophiles in the cDEL library, these inhibitors incorporate previously unexplored, non-classical electrophile structures.

As this case illustrates, the “explorable space” for covalent drug discovery is being systematically expanded through cDEL.

New Advances: Systematic Evaluation of cDEL Electrophiles

Enhancing cDEL capabilities requires not only scale expansion but also methodological refinement. In a recent study published in Helvetica Chimica Acta, WuXi Biology’s team collaborated with ETH Zurich to systematically evaluate the applicability of covalent DEL electrophiles.

The study addressed key challenges in cDEL related to reactivity and stability. Highly reactive electrophiles may compromise selectivity, while stability can vary significantly across different electrophile types. Without systematic evaluation, these factors may constrain the reliability of screening outcomes.

To this end, the research team selected 59 representative electrophiles spanning eight structural classes. Six classes were designed to form irreversible covalent bonds, including aryl halides, alkenes, alkynes and beyond, while two classes—aldehydes and nitriles—tended to form reversible covalent bonds.

▲The study systematically evaluated the suitability of 59 electrophilic warheads for DEL applications.

Under conditions simulating real DEL construction workflows, the team first evaluated the coupling efficiency of each electrophile. Of the 59 candidates, 45 achieved a coupling yield above 50%, demonstrating ideal DEL compatibility.

Building on this, the team further assessed the reactivity and selectivity of these 45 electrophiles toward cysteine and lysine residues, as well as key parameters such as reaction kinetics and stability.

Ultimately, 21 electrophiles were identified that combined high coupling yield with strong reactivity, all of which also demonstrated high stability in subsequent evaluations. Notably, reversible electrophiles failed to maintain stability under the tested conditions, indicating that additional chemical steps may be required to improve their stability for future applications.

This study provides a systematic assessment of chemically diverse electrophiles within the cDEL framework and offers valuable guidance for the rational design of future covalent DEL libraries.

Building an Integrated Engine for Covalent Drug Discovery

cDEL does not operate in isolation. Within WuXi Biology’s covalent drug discovery platform, cDEL is deeply integrated with covalent fragment-based drug discovery (cFBDD) and covalent high-throughput screening (cHTS), forming a synergistic and complementary discovery engine.

The cHTS platform, built on large-scale covalent compound libraries, serves as an efficient hit identification engine. Its library comprises approximately 69,000 molecules featuring more than 50 distinct electrophile chemotypes.The library is purpose-built to target nine different amino acid residues, expanding screening beyond single-site approaches to broader protein space. The platform also offers “direct to biology” (D2B) solutions, providing end-to-end support from hit discovery to hit-to-lead optimization.

Meanwhile,WuXi AppTec’s cFBDD platform starts from smaller molecular fragments, offering another efficient pathway for covalent drug discovery.Based on a library of over 2,600 structurally diverse fragments, the platform incorporates electrophiles to rapidly identify covalent binders that serve as starting points for optimization. This strategy is particularly well-suited for structurally complex or traditionally “undruggable” targets.

Over more than a century, covalent drugs have evolved from serendipitous discoveries to rationally designed therapeutics, driven by continuous advances in screening technologies.

Nowadays, WuXi AppTec’s integrated platforms, including cDEL, cHTS, and cFBDD, work in concert to deliver systematic solutions for clients, ushering covalent drug discovery into a more systematic era.

Leveraging its end-to-end, integrated CRDMO platform, WuXi AppTec is committed to accelerating the development of breakthrough therapies, enabling partners translate innovation into solutions that benefit patients worldwide, and advancing its vision of “every drug can be made, and every disease can be treated.”

參考資料：

[1] Puff, Johanna, et al. "Systematic Evaluation of Electrophile Reactivity and Stability for Covalent DNA‐Encoded Libraries." Helvetica Chimica Acta (2026): e70064. https://doi.org/10.1002/hlca.70064

免責聲明：本文僅作信息交流之目的，文中觀點不代表藥明康德立場，亦不代表藥明康德支持或反對文中觀點。本文也不是治療方案推薦。如需獲得治療方案指導，請前往正規(guī)醫(yī)院就診。

版權說明：歡迎個人轉發(fā)至朋友圈，謝絕媒體或機構未經(jīng)授權以任何形式轉載至其他平臺。轉載授權請在「藥明康德」微信公眾號回復“轉載”，獲取轉載須知。