Organic Cosolvent-Free Antibody–Drug Conjugation Enabled by Hydrophilic Linker–Payloads

Abstract

Bioconjugation of linker–payloads (LPs) to antibodies has historically relied on organic cosolvents or surfactants to overcome the poor aqueous solubility of cytotoxic payloadsan assumption that has remained largely unquestioned in antibody–drug conjugate (ADC) manufacturing. Here, we systematically re-evaluate this paradigm by characterizing the aqueous solubility of 13 representative LPs at operationally relevant concentrations (5–50 mM) across water and acidic buffers. Classical hydrophobic LPs, including vedotin, mafodotin, and deruxtecan, were confirmed to be insoluble under all aqueous conditions tested, validating their established dependence on cosolvents. However, LPs incorporating hydrophilic elements, such as PEG chains or ionizable groups, dissolved readily in water at concentrations sufficient for bioconjugation. Most notably, the linker–payload corresponding to sacituzumab govitecan (CL2A-SN-38) achieved concentrations of at least 50 mM in pure water, and two different conjugation reactions were successfully executed in fully aqueous buffered media without any organic solvents or surfactants. Physicochemical analysis revealed that solubility is not governed solely by global hydrophobicity (A log P) but is codetermined by PEG length, charge distribution, pH-dependent ionization, and the continuity of hydrophobic surface topology. Conjugation under fully aqueous conditions yielded ADCs with average DAR values of 1.6–1.9 under target-DAR2 conditions and 6.1–7.0 under target-DAR8 conditions, with minimal aggregation and physicochemical properties comparable to those produced via conventional cosolvent-based workflows. These findings demonstrate that rational hydrophilicity-driven LP design enables cosolvent-free conjugation, thereby simplifying ADC manufacturing, eliminating flammability hazards, removing residual solvent testing requirements, and reducing regulatory complexity.

Introduction

Antibody–drug conjugates (ADCs) have emerged as a powerful modality in oncology, offering the potential to selectively deliver cytotoxic agents to tumor cells while minimizing systemic toxicity. − By combining the targeting precision of monoclonal antibodies with the potency of small-molecule drugs, ADCs represent a sophisticated therapeutic platform capable of addressing unmet medical needs across a range of cancer types.

A critical component of ADCs is the linker–payload, which connects the antibody to the cytotoxic drug. This modular element plays a central role in determining the conjugation efficiency, stability, pharmacokinetics, and overall therapeutic index of the final ADC product. , However, despite the growing sophistication of ADC design, bioconjugation of the linker–payload remains a bottleneck in process development due to its hydrophobic nature. , Most linker–payloads are poorly soluble in aqueous buffers, necessitating the use of organic cosolvents (e.g., DMSO, DMF, DMAc) or surfactants (e.g., polysorbates) to facilitate solubilization and conjugation. − While these additives enable conjugation, they introduce several challenges that complicate downstream processing and regulatory compliance with residual solvent levels. Surfactants are notoriously difficult to remove during purification, and their presence can interfere with analytical methods, requiring sensitive trace-level detection. , Moreover, surfactants have been implicated in promoting protein degradation and aggregation, which can compromise the stability and safety of the final ADC formulation. ,

Organic solvents, despite being widely used, present a different set of concerns. They are flammable, pose safety risks during handling and scale-up, and require residual solvent testing (e.g., gas chromatography) to ensure compliance with regulatory standards. These factors collectively increase the complexity, cost, and risk profile of ADC manufacturing. To address these limitations, recent efforts have focused on hydrophobicity tuning of linker–payloads. Strategies include the incorporation of hydrophilic spacers such as poly­(ethylene glycol) (PEG) chains or sulfonate groups, as well as structure–activity relationship (SAR) studies aimed at balancing solubility with cytotoxic potency. Despite these innovations, the use of cosolvents remains widespread, often driven by legacy practices or the assumption that aqueous solubility is inherently unachievable for most linker–payloads.

In this manuscript, we challenge that assumption by presenting a series of linker–payloads that exhibit unexpectedly high water solubility, enabling fully aqueous conjugation without the need for organic solvents or surfactants. Through systematic solubility screening and conjugation studies, we demonstrate that several linker–payloads can be solubilized and conjugated in aqueous buffers with high efficiency and minimal aggregation.

These findings suggest a shift in ADC process development. By eliminating the reliance on organic solvents and surfactants, we can simplify conjugation workflows, reduce safety and regulatory burdens, and enhance the robustness of ADC manufacturing.

Experimental Procedure

Materials

Human IgG1 trastuzumab (Herceptin) was purchased from the Roche Pharmaceutical Company (Switzerland). APL-1091 and APL-1092 were purchased from Ajinomoto Co, Inc. All other linker–payloads were purchased from MedChemExpress (Monmouth Junction, NJ) and the detailed information is in Table . All other chemical reagents were purchased from Sigma-Aldrich.

1. Detailed Characteristics of Linker–Payloads as Reported in the Vendors' COA.

	drug linker	CAS number	purity	molecular weight	salt form
1	vedotin (MC-VC-PAB-MMAE)	646502-53-6	99.9%	1316.63	free base
2	zovodotin (MT-VC-ZD2044)	1800462-96-7	99.8%	1292.58	free base
3	APL-1091 (MC-EXO-EEVC-MMAE)	2921735-87-5	99.2%	1631.91	free base
4	APL-1092 (MC-EXO-EEVC-exatecan)	2921735-91-1	99.6%	1349.37	free base
5	mafodotin (MC-MMAF)	863971-19-1	99.9%	925.16	free base
6	deruxtecan	1599440-13-7	99.9%	1034.05	free base
7	MC-β-glucuronide-MMAE	1703778-92-0	99.2%	1323.53	free base
8	MC-EVC-PAB-MMAE	2873452-49-2	99.0%	1445.74	free base
9	Mal-PEG2-Gly-Gly-Phe-Gly-exatecan	1599439-54-9	99.1%	1149.18	free base
10	Mal-PEG8-Val-Cit-PAB-MMAE	2353409-69-3	97.1%	1626.97	free base
11	MC-PEG-SN-38 (CL2A-SN-38)	1279680-68-0	99.1%	1480.61	free base
12	MC-β-glucuronide-exatecan	2763252-25-9	99.8%	984.89	free base
13	MC-Val-Cit-PAB-MMAF	863971-17-9	98.2%	1330.61	free base

Solubility Assessment

The aqueous solubility of representative linker–payloads (LPs) was evaluated at practical reaction concentrations of 5–50 mM depending on the linker–payload (Figure and Table ) to determine whether a cosolvent- and surfactant-free conjugation process could be implemented. Each LP was weighed accurately and evaluated in ultrapure water and 20 mM acetate buffers adjusted to acidic (approximately pH 5.5) conditions. No organic cosolvents or surfactants were used during this process. After initial mixing, the solutions were inspected visually for clarity. If precipitation was observed, sonication was applied for 10 min, and the effect of pH adjustment was also assessed. To quantify solubility, subvisible particles were removed by 0.2 μm filtration, and linker–payload concentrations were determined by UV–Vis spectroscopy by comparing absorbance values before and after filtration. Recovery percentages were calculated to assess whether the LP was fully dissolved or partially present as subvisible aggregates.

1.

Overview of the linker–payloads tested in this study.

2. Study of Water Solubility.

	drug linker	A log P	solubility in water	solubility in other buffer
1	vedotin (MC-VC-PAB-MMAE)	4.79	no	no (PBS, acetate)
2	zovodotin (MT-VC-ZD2044)	1.38	yes (50 mM)	yes (PBS, acetate)
3	APL-1091 (MC-EXO-EEVC-MMAE)	2.22	no	no (PBS, acetate)
4	APL-1092 (MC-EXO-EEVC-exatecan)	1.10	no	yes (30 mM acetate)
5	mafodotin (MC-MMAF)	4.31	no	no (PBS, acetate)
6	deruxtecan	0.29	no	no (PBS, acetate)
7	MC-β-glucuronide-MMAE	2.85	no	yes (30 mM acetate)
8	MC-EVC-PAB-MMAE	4.03	no	no (PBS, acetate)
9	Mal-PEG2-Gly-Gly-Phe-Gly-exatecan	–1.17	yes (50 mM)	yes (PBS, acetate)
10	Mal-PEG8-Val-Cit-PAB-MMAE	2.51	yes (50 mM)	yes (PBS, acetate)
11	MC-PEG-SN-38 (CL2A-SN-38)	2.09	yes (50 mM)	yes (PBS, acetate)
12	MC-β-glucuronide-exatecan	0.23	no	yes (30 mM acetate)
13	MC-Val-Cit-PAB-MMAF	4.86	no	no (PBS, acetate)

Stability Assessment

For three linker–payloads exhibiting acetate buffer-specific solubility (APL-1092, MC-β-glucuronide-MMAE, and MC-β-glucuronide-exatecan), chemical stability in aqueous solution was evaluated to determine whether dissolved LPs remained intact over operationally relevant timeframes. After incubation at 37 °C (time points: 0, 4, 24, and 72 h), N-acetylcysteine (NAC) was added to cap unreacted maleimide groups, followed by RP-HPLC analysis to assess degradation. As a control, DMSO solutions of the same linker–payloads were subjected to identical incubation conditions and NAC quenching followed by RP-HPLC analysis.

Instruments and Analytical Methods

The linker–payload concentration was measured using UV–Vis spectroscopy with a Lunatic system.

The ADC concentration was measured using the slope spectroscopy method and a Solo-VPE system.

Hydrophobic interaction chromatography (HIC), RP-HPLC analysis, and LC-MS were performed as previously reported. A single naked-trastuzumab reference chromatogram was used for each analytical method and reused as a common comparator across the target-DAR8 figures; fresh naked-antibody controls were not injected alongside each individual conjugation sample.

Size-exclusion chromatography (SEC)-HPLC analyses were performed as previously reported.

The theoretical partition coefficient (A log P) was calculated using BIOVIA Draw software.

Experimental Procedure for DAR = 2 ADC Production

ADC production was performed based on the reported procedures, which are briefly described here.

Trastuzumab at 10 mg/mL in 20 mM acetate buffer (pH 5.5) was reacted with 2.5 equiv of AJICAP peptide reagent prepared as a 20 mM solution in DMF. The mixture was incubated at 20 °C for 1 h. After this initial conjugation step, excess hydroxylamine hydrochloride (NH₂OH·HCl) was added directly to the reaction mixture, and the solution was stirred for an additional 1 h to cleave the linker in a one-pot process. The reaction mixture was then purified using a Zeba desalting column and eluted with PBS buffer (pH 7.4).

Following linker cleavage, the resulting antibody–thiol intermediate was conjugated to the payload using an organic solvent-free thiol–maleimide coupling procedure. The linker-payload was dissolved at the appropriate concentration in water and added to the antibody–thiol under controlled conditions. After completion of the reaction, excess linker-payload was removed using a Zeba desalting column, and the final ADC was formulated in the desired buffer. The drug-to-antibody ratio (DAR) was confirmed by HIC. Aggregation percentage was analyzed by SEC.

Experimental Procedure for DAR = 8 ADC Production

Interchain conjugation targeting full reduction (DAR = 8) was performed as previously described, with adaptation for cosolvent-free payload coupling. Trastuzumab at 10 mg/mL in 20 mM PBS (pH 7.4) was reacted with 10 equiv of TCEP·HCl prepared as a 50 mM solution in PBS. The mixture was incubated at 37 °C for 1.5 h to fully reduce interchain disulfide bonds. Following reduction, the resulting antibody–thiol intermediate was conjugated to the linker–payload using the same aqueous thiol–maleimide coupling procedure described above. The linker–payload was dissolved at the appropriate concentration in water and added to the antibody–thiol under controlled conditions. After completion of the reaction, the excess linker–payload was removed via a Zeba desalting column, and the final ADC was formulated in the desired buffer. The average DAR and chain-level chromatographic distributions were assessed by HIC and reduced RP-HPLC, and aggregation was determined by SEC. Reduced Q-TOF LC–MS was used to confirm the predominant modified light- and heavy-chain species for six of the seven linker–payloads. Parallel aqueous and DMSO-containing conjugation samples were characterized by HIC, reduced RP-HPLC, and SEC (Figures S36–S77), and the corresponding reduced Q-TOF LC–MS data are shown separately in Figures S78–S89.

Experimental Procedure for Freeze–Thaw Stability Study

ADC samples were subjected to freeze–thaw stress to evaluate their susceptibility to aggregation. Aliquots of ADC, prepared in the final formulation buffer, were frozen at −80 °C for 30 min and subsequently thawed at room temperature for 1 h. This process constituted one freeze–thaw cycle and was repeated for a total of three cycles. Following completion of the stress protocol, samples were directly analyzed without further treatment. Aggregation was assessed by SEC.

Thermal Stability Study

ADC samples were incubated under thermal stress to evaluate their stability. Aliquots of ADC, prepared in the final formulation buffer, were incubated at 37 °C for 14 days. Following incubation, samples were analyzed without further treatment. Aggregation was assessed by SEC.

Results and Discussion

The central hypothesis of this work is that a subset of structurally advanced LPs, particularly those incorporating hydrophilic design elements developed to improve ADC pharmacokinetics, possess sufficient intrinsic aqueous solubility to support bioconjugation without organic cosolvents. To test this hypothesis, the aqueous solubility of representative linker–payloads (LPs) at practical reaction concentrations (up to 50 mM depending on the linker–payload, Table ) was re-evaluated, and based on these data, the feasibility of implementing a cosolvent- and surfactant-free conjugation process was examined.

First, traditional hydrophobic LPs, including vedotin (MC-VC-PAB-MMAE), were assessed. Vedotin was completely insoluble at 5 mM in ultrapure water, PBS, and acidic buffers; sonication and pH adjustment were ineffective. Literature descriptions of vedotin-derived species being readily removed during postreaction TFF in acidic buffers were confirmed to pertain to diluted, postconjugation streams quenched with N-acetylcysteine, a context distinct from dissolving the raw LP in aqueous media. Mafodotin (MC-MMAF) and deruxtecan (MC-GGFG-DXd) were likewise insoluble, supporting the practical view that hydrophobic LPs generally require organic cosolvents.

Next, LPs incorporating hydrophilizing elements (e.g., PEG) were evaluated. Notably, zovodotin (MT-VC-ZD2044), used in the clinical-stage ADCs XB002 and ZW49, formed a clear solution in water to at least 50 mM. Commercial PEG-containing LPs, including Mal-PEG2-Gly-Gly-Phe-Gly-exatecan and Mal-PEG8-Val-Cit-PAB-MMAE, exhibited similarly robust aqueous solubility (up to 50 mM). Comparable solubility was also observed for hydrazino-iso-Pictet-Spengler (HIPS)-handle LPs (i.e., not limited to maleimide), suggesting that hydrophilicity-oriented design extends across conjugation chemistries (Figure S90). The presence of mobile hydrophilic chains such as PEG was found to improve practical handling across weighing, dissolution, mixing, and purification. To provide a semiquantitative assessment of solubility, samples were filtered (0.2 μm), and linker–payload concentrations in aqueous solution were measured by UV–Vis spectroscopy by comparison before and after filtration. The absorbance before and after filtration remained essentially unchanged for almost all linker–payloads, indicating minimal loss upon filtration (Figure ). Supporting data for the linker–payload appearance, filtration recovery, and N-acetylcysteine-quenching stability are provided in Figures S1–S14.

2.

UV–Vis spectra of linker–payloads before and after 0.2 μm filtration, demonstrating minimal changes in absorbance under aqueous conditions. Before- and after-filtration spectra are overlaid; the higher-absorbance trace represents the prefiltration sample. (a) Zovodotin (MT-VC-ZD2044) in water (50 mM); (b) APL-1092 in acetate buffer; (c) MC-β-glucuronide-MMAE in acetate buffer; (d) Mal-PEG2-Gly-Gly-Phe-Gly-exatecan in water (50 mM); (e) Mal-PEG8-Val-Cit-PAB-MMAE in water; (f) MC-PEG-SN-38 (CL2A-SN-38) in water; and (g) MC-β-glucuronide-exatecan in acetate buffer. Absorbance profiles before and after filtration remained largely unchanged for most LPs, indicating minimal loss of the soluble material; a modest decrease was observed for CL2A-SN-38.

CL2A-SN-38 (MC-PEG-SN-38), corresponding to the LP of Trodelvy, was then examined in detail. Although publicly available protocols commonly employ 5–10% DMSO, concentrations of at least 50 mM were achieved in pure water, enabling stable aqueous stock solutions in nonorganic buffers. Because thiol–maleimide coupling proceeds quantitatively in water, adequate payload solubility removes the theoretical need for a cosolvent. In subsequent site-specific conjugations (AJICAP; target DAR = 2), reactions proceeded smoothly without organic solvents or surfactants (Figure and Table ). The ability to handle the CL2A-SN-38 linker–payload used in sacituzumab govitecan in fully aqueous media challenges the implicit assumption that such LPs require organic cosolvents, with implications for process safety, residual solvent management, and analytical simplification. Notably, CL2A-SN-38 showed an approximately 17% decrease in apparent concentration after filtration; however, this reduction was not considered significant, and further investigation may be warranted.

3.

Schematic representation of the conjugation process for generating trastuzumab-based ADCs. The linker-payload is first prepared in the solid form and then dissolved in aqueous solution (2.7 equiv relative to the antibody). Conjugation to trastuzumab–thiol occurs in PBS buffer (pH 7.4) at room temperature for 1 h, resulting in the formation of trastuzumab–ADC with the linker-payload attached.

By contrast, LPs bearing sugar modifications or acidic functionalities displayed strong pH dependence. MC-β-glucuronide-MMAE and MC-β-glucuronide-exatecan were insoluble in water and PBS but dissolved rapidly in 20 mM sodium acetate at pH 5.5. APL-1092 (MC-EXO-EEVC-exatecan) dissolved readily at 30 mM in 20 mM sodium acetate buffer at pH 5.5. These behaviors are consistent with cationization and improved dispersion under mildly acidic conditions. In contrast, introducing glutamate at the P1 position of vedotin (MC-EVC-PAB-MMAE) did not confer solubility even in acidic buffer; the exo-EEVC analogue APL-1091 (MC-EXO-EEVC-MMAE) also remained insoluble. A direct comparison between APL-1092 and APL-1091 indicates that, on a matched linker scaffold, the exatecan payload confers superior aqueous solubility relative to MMAE.

Importantly, to address solution stability under acidic conditions, we assessed whether LPs (APL-1092, MC-β-glucuronide-MMAE, and MC-β-glucuronide-exatecan) that dissolve in acetate buffer remain chemically intact over operationally relevant timeframes (Figures S9–S13). These data indicate that the maleimide functionality remains chemically stable for at least 24 h under these conditions, based on comparison with DMSO solutions, supporting their suitability for conjugation operations.

To provide a broader physicochemical context, solubility was mapped against calculated hydrophobicity (A log P). Although higher A log P generally correlated with insolubility, a strict linear relationship was not observed. LPs with moderate A log P remained soluble when PEG segments disrupted contiguous hydrophobic surfaces, whereas certain LPs with only modest A log P remained insoluble when extended hydrophobic patches were present. Consequently, charge distribution, exposure of polar functionalities, PEG length and flexibility, and the continuity of hydrophobic surface are indicated as codeterminants of practical solubility.

Aqueous-soluble LPs were advanced to conjugation to assess process compatibility. To simplify product composition, a site-specific strategy (AJICAP; target DAR = 2) was employed. Reactions proceeded cleanly in fully aqueous buffers, yielding ADCs with DARs near the target. Initial SEC analyses showed no increase in high-molecular-weight species, indicating that solvent-free processing did not exacerbate aggregation through early purification. For comparison, ADCs prepared from LPs requiring organic solvent were evaluated by HIC; Vedotin-derived ADCs displayed slower elution, consistent with higher apparent surface hydrophobicity. However, the aqueous solubility of the LP alone did not map one-to-one onto ADC HIC retention (Figure ), which was influenced by additional factors including DAR, conjugation site, spacer architecture, PEG placement, and the local antibody surface environment. Given empirical links between HIC retention and in vivo half-life, selection of hydrophilized LPs should be considered part of a broader optimization of overall surface properties, size, and charge rather than a binary “soluble/insoluble” decision.

4.

Comparison of drug-linker hydrophobicity (A log P), HIC retention time of the resulting ADC (DAR = 2 peak), and water solubility of the linker-payload. Each point represents a distinct ADC, categorized by linker-payload solubility: orange circles indicate water-soluble, blue triangles acetate-soluble, and red crosses insoluble linker-payloads.

To further evaluate process robustness under more demanding conditions, interchain conjugation targeting higher DAR (target DAR = 8) was also performed using these water-soluble LPs. In this format, both fully aqueous conditions and conventional cosolvent conditions (DMSO-containing systems) were examined in parallel (Figure ). Conjugation efficiency (DAR), aggregation (SEC), and physicochemical properties including freeze–thaw stability and thermal stability of the resulting ADCs were evaluated to compare aqueous-derived ADCs with those prepared using traditional organic cosolvent processes (Table S1). Overall, the aqueous and DMSO-containing methods produced comparable average DAR values and initial aggregation levels, although construct-dependent differences were observed after thermal stress (Table S1). Under the target-DAR8 conditions, the reduced RP-HPLC-derived average DAR values ranged from 6.1 to 7.0 across the seven linker–payloads under aqueous and DMSO-containing conjugation conditions (Table ). HIC and reduced RP-HPLC were used to assess overall conjugation extent and chromatographic heterogeneity rather than to establish definitive structural identities for all product species. Reduced Q-TOF LC–MS confirmed predominant light- and heavy-chain masses consistent with the L1 and H3 conjugation states for six of the seven linker–payloads under both aqueous and DMSO-containing conditions (Figures S78–S89). The LC–MS analysis was not used to assign each individual RP-HPLC peak. No interpretable reduced LC–MS spectrum was obtained for the CL2A-SN-38 ADC; therefore, its interpretation is based on HIC, reduced RP-HPLC, and SEC results. Collectively, these analytical data support comparable overall conjugation extent and aggregation levels between the two conjugation conditions, while acknowledging the limitations of chromatographic peak assignment.

5.

Schematic representation of interchain ADC with DAR = 8. Native antibody interchain disulfide bonds are first reduced to generate reactive thiol groups, followed by conjugation with maleimide-functionalized linker–payloads to produce ADCs with up to eight linker–payload attachments.

3. Reduced RP-HPLC-Derived Average DAR Values for Trastuzumab ADCs Prepared Under Target-DAR8 Conditions Using Aqueous and DMSO-Containing Conjugation Methods.

linker–payload	aqueous	DMSO
zovodotin	6.1	6.1
APL-1092	7.0	7.0
MC-β-glucuronide-MMAE	6.8	6.5
Mal-PEG2-Gly-Gly-Phe-Gly-exatecan	6.6	6.4
Mal-PEG8-Val-Cit-PAB-MMAE	6.7	6.5
CL2A-SN-38	6.9	6.8
MC-β-glucuronide-exatecan	6.8	6.7

Taken together, several design principles are supported. PEG incorporation exerts a dual benefit by solubilizing the LP and masking hydrophobic patches after conjugation, thereby mitigating HIC retention and aggregation. Even with canonical triggers such as Val-Cit-PAB, insertion of hydrophilic spacers (e.g., EVC or EXO-EEVC) can facilitate both solubility and operational robustness. For payloads with polar functionalities (e.g., SN-38, exatecan), uninterrupted hydrophobic surface can still dominate; thus, conformational freedom and pH-dependent ionization that increase exposure of polarity are advantageous. For sugar- or acid-functionalized LPs, preparation of mildly acidic stock solutions followed by neutralization at dosing expands the operable space for fully aqueous processing.

Literature indicates that the dissolution rate of the drug-linker often limits reaction progress in slurry-based conjugations. By optimizing mixing, researchers achieved a 90% reduction in reaction time while simultaneously improving product quality through reduced overconjugation. In our study, reaction kinetics were not directly evaluated. Under the site-specific target-DAR2 conditions (2.7 equiv of linker–payload and a 1 h reaction time), the resulting ADCs showed hydrophobic interaction chromatography-derived average drug-to-antibody ratio values of 1.6–1.9. These results demonstrate that substantial conjugation was achieved under fully aqueous conditions.

From a process and CMC perspective, a fully aqueous conjugation workflow reduces residual-solvent and surfactant control burdens, mitigates flammability risks, and simplifies analytical assays. Use of aqueous LP stock solutions enables controlled dosing and improves reproducibility during scale-up. Potential trade-offs include increased molecular size associated with hydrophilicity (often via PEG), which may influence postrelease diffusion and membrane transit; accordingly, balanced development should incorporate a solubility map (water/PBS/acidic buffer), operational metrics (cosolvent need, maximum practical concentration, viscosity), HIC/SEC profiles (DAR distribution, HMWS), and serum stability with initial PK.

Conclusion

This study challenges the prevailing assumption that linker–payloads for ADCs typically require organic cosolvents or surfactants for conjugation. Through systematic solubility screening and conjugation experiments, we demonstrate that several linker–payloads, particularly those incorporating hydrophilic elements such as PEG or ionizable groups, achieve sufficient aqueous solubility to enable fully aqueous conjugation workflows. These findings indicate that hydrophilicity-driven design can overcome solubility limitations without compromising conjugation efficiency, allowing the elimination of organic solvents and surfactants. This shift simplifies process development, reduces safety and regulatory burdens, and enhances scalability and reproducibility. Importantly, solubility is influenced by multiple factors beyond global hydrophobicity, including PEG length, charge distribution, and the continuity of hydrophobic surfaces. While aqueous solubility improves handling and reduces aggregation risk, overall ADC performance depends on integrated design considerations such as hydrophobic interaction chromatography retention, DAR, and pharmacokinetics. Collectively, these results support a shift toward organic solvent-free ADC manufacturing and highlight the need for holistic optimization strategies that balance solubility, structural attributes, and biopharmaceutical properties.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c01543.

• Figures showing linker–payload solubility and stability assessments; HIC, reduced RP-HPLC, SEC, and reduced Q-TOF LC–MS characterization of trastuzumab ADCs; freeze–thaw and thermal-stability data; and site-specific HIPS conjugation under organic cosolvent-free conditions. (PDF)

†.

Astellas, West Coast Innovation Center, 480 Forbes Boulevard, South San Francisco, California 94080, United States

‡.

MBC Consulting LLC, San Diego, California 92129, United States.

The authors declare no competing financial interest.