Hydrogel Chemistry in Expansion Microscopy: From Polyacrylamide to Next‐Generation Swellable Polymers

Abstract

Expansion microscopy (ExM) is a transformative biological imaging approach that achieves nanoscale resolution on conventional diffraction‐limited microscopes by physically expanding specimens embedded in swellable hydrogels. Since its introduction in 2015, the hydrogel chemistry underpinning ExM has evolved considerably, progressing from the original sodium polyacrylate/polyacrylamide (PAAm) formulations toward advanced polymer architectures that address persistent limitations in expansion factor, isotropy, and structural fidelity. This review comprehensively analyzes the hydrogel chemistry landscape in ExM, tracing the progression from classical polyacrylamide‐based gels through modified acrylamide formulations and high‐expansion dimethylacrylamide (DMAA) systems to next‐generation click‐chemistry‐based tetra‐gel platforms. We critically examine the polymerization mechanisms, monomer composition, crosslinker design, and anchoring strategies governing gel performance, and evaluate how these parameters influence expansion factor, mechanical integrity, and biomolecular spatial fidelity. Emerging approaches, including single‐shot 20‐fold expansion protocols and microwave‐accelerated workflows, are discussed in the context of their underlying polymer chemistry. By synthesizing findings from the primary literature through 2025, this review aims to guide researchers in selecting and optimizing hydrogel systems for specific imaging applications and to identify promising directions for future polymer innovation in ExM.

Expansion microscopy magnifies specimens embedded in a swellable hydrogel, enabling nanoscale imaging on conventional microscopes. The gel is the engine of the method, and its chemistry setting the expansion factor, isotropy, and biomolecular compatibility. From classic polyacrylamide to high‐expansion (20ExM) and rapid (BOOST) gels, chemical design pushes effective resolution from ∼70 nm to below 20 nm.

1. Introduction

The optical diffraction limit constrains the lateral resolving power of conventional fluorescence microscopy to approximately 200–300 nm, precluding direct visualization of many subcellular structures and molecular assemblies. Super‐resolution techniques such as stimulated emission depletion (STED), photoactivated localization microscopy (PALM), and stochastic optical reconstruction microscopy (STORM) circumvent this barrier through sophisticated optical and computational strategies but typically require expensive specialized instrumentation and prolonged acquisition times. Expansion microscopy (ExM), introduced by Chen, Tillberg, and Boyden in 2015, offers a conceptually distinct solution: rather than improving the optics, the specimen itself is physically magnified [1]. Preserved biological samples are embedded in a dense, swellable polyelectrolyte hydrogel and subsequently expanded through osmotic swelling in water, isotropically increasing the distances between labeled biomolecules and enabling nanoscale features to be resolved with standard confocal microscopes [1, 2] (Figure 1A).

FIGURE 1.

Principles and chemistry of expansion microscopy hydrogels. (A) Schematic of the original ExM concept: methacryloyl‐functionalized oligonucleotide probes anchor fluorescent labels into the polyelectrolyte gel network, enabling isotropic separation of labels upon expansion in water. Right: fluorescence images of a fixed mouse brain slice before and after expansion. Adapted with permission from Ref. [1]. Copyright 2015 American Association for the Advancement of Science. (B) Workflow of protein retention ExM (proExM): anchoring, gel polymerization, homogenization, and expansion, with molecular‐level schematics illustrating each step. Bottom: photographs of the expanded gel and fluorescence images of cells demonstrating preserved structural detail after expansion. The proExM workflow shown was originally reported in Refs. [3, 4]. Reproduced with permission from Ref. [5]. Copyright 2024 The Company of Biologists. (C) Summary of biomolecular anchoring workflows: three protein retention strategies (antibody staining with AcX anchoring, fluorescent protein expression with AcX anchoring, and gentle disruption for postexpansion antibody staining) and RNA anchoring using LabelX, with chemical structures of key anchoring reagents shown at the bottom. Adapted with permission from Ref. [6]. Copyright 2018 John Wiley and Sons.

The foundational chemistry of ExM relies on in situ polymerization of sodium acrylate and acrylamide monomers, crosslinked by N,N′‐methylenebisacrylamide (bis‐acrylamide), to form a swellable polyelectrolyte network within the fixed specimen [1]. Upon dialysis in water, the charged polyacrylate chains repel one another, driving isotropic expansion of approximately 4–4.5‐fold linearly. This elegant yet simple mechanism has since been adapted into numerous variants—protein retention ExM (proExM), ExFISH for RNA imaging, iterative ExM (iExM), and others—each refining different aspects of the workflow, including anchoring, homogenization, and labeling strategies [2, 3, 4, 7].

However, the hydrogel remains the central enabling element of all ExM protocols, and its chemical composition directly governs critical performance parameters: achievable expansion factor, mechanical robustness of the expanded gel, isotropy and fidelity of nanoscale expansion, and compatibility with diverse tissue types and biomolecular anchoring strategies. Recognizing this, considerable effort has been devoted to engineering improved gel formulations, ranging from incremental modifications of the original PAAm recipe—such as adjusting monomer ratios and crosslinker concentrations—to fundamentally different polymer architectures such as the tetra‐gel system based on click chemistry [6, 8].

This review systematically examines the hydrogel chemistry underlying ExM, organized around the evolution from classical polyacrylamide systems to next‐generation swellable polymers. We discuss the physicochemical principles governing gel swelling behavior, critically evaluate the advantages and limitations of each gel platform, and highlight emerging innovations that promise higher expansion factors with improved nanoscale accuracy. Our analysis draws primarily on the peer‐reviewed literature from 2015 through 2025, with emphasis on developments reported in the most recent 3 years. Whereas earlier surveys—most notably Truckenbrodt's 2023 account—already treat the classical ExM, proExM, iExM, and X10 chemistries in depth, the present review is deliberately weighted toward post‐2023 developments such as single‐shot 20‐fold expansion (20ExM), microwave‐accelerated workflows (BOOST), universal anchoring (Magnify), iterative ultrastructure protocols (iU‐ExM), single‐molecule‐scale expansion (ONE microscopy), and clinical FFPE applications, together with the underlying polymer chemistry—in particular the self‐crosslinking behavior of DMAA networks—that makes these advances possible. In concrete terms, relative to that 2023 review this article newly incorporates and discusses ExR, iU‐ExM, ExESRRF, GelMap, continuous membrane labeling, and the first reports of approximately 1000‐fold gels, most of which postdate it; the classical SA/AAm/Bis, proExM, iExM, and X10 sections are retained only as the concise mechanistic foundation needed to make the newer chemistry intelligible, so that the bulk of the analysis concerns developments from 2023 onward.

2. Fundamentals of Hydrogel Swelling in ExM

Polyelectrolyte hydrogel swelling in water is driven by three factors: osmotic pressure from mobile counterions, electrostatic repulsion between charged polymer segments, and the thermodynamic tendency of the network to mix with the solvent.

In ExM, sodium acrylate incorporation introduces negatively charged carboxylate groups along the polymer backbone. When the gel is placed in deionized water, the high counterion (Na⁺) concentration within the gel generates an osmotic pressure gradient that drives water influx, causing the network to swell. The extent of swelling is balanced by the elastic restoring force of the crosslinked network, which resists deformation [1, 9].

The equilibrium swelling ratio is determined by several controllable parameters: polymer charge density (governed by sodium acrylate content), crosslink density (governed by bisacrylamide concentration), total polymer concentration, and ionic strength of the surrounding medium. Reducing crosslink density or increasing the charge‐bearing monomer fraction leads to higher expansion factors at the cost of mechanical fragility. This fundamental trade‐off has been a central theme in the development of improved ExM gel formulations [10, 11].

An ideal ExM hydrogel must satisfy several stringent requirements simultaneously. First, it must swell isotropically—uniformly in all three spatial dimensions—to preserve the relative positions of anchored biomolecules. Second, the gel must be optically transparent after expansion to permit fluorescence imaging. Third, the polymerization reaction must proceed uniformly throughout the specimen, including within densely packed tissue volumes, to avoid local network heterogeneity. Fourth, the gel chemistry must be compatible with biomolecular anchoring strategies that enable covalent attachment of proteins, nucleic acids, lipids, or other target molecules to the polymer matrix. Finally, the expanded gel must possess sufficient mechanical integrity for handling, mounting, and imaging without tearing or collapse [2, 12].

3. Classical Polyacrylamide‐Based Hydrogels

The original ExM protocol employed a monomer solution of sodium acrylate (SA; the charge‐bearing monomer), acrylamide (AAm; a neutral comonomer contributing to network strength), and N,N′‐methylenebisacrylamide (bis‐acrylamide, Bis; the bifunctional crosslinker) [1]. Polymerization was initiated by ammonium persulfate (APS) with tetramethylethylenediamine (TEMED) as the accelerator, following the standard free‐radical chain‐growth mechanism used in polyacrylamide gel electrophoresis (PAGE). The polymerization inhibitor 4‐hydroxy‐TEMPO (4‐HT) was included to delay gelation onset, providing time for thorough monomer diffusion into the tissue before network formation [1, 13].

In the original formulation, the monomer solution typically contained 8.6 g/100 mL sodium acrylate, 2.5 g/100 mL acrylamide, and 0.15 g/100 mL bis‐acrylamide, yielding gels that expanded approximately 4.5‐fold linearly in deionized water. This corresponds to a roughly 100‐fold volumetric expansion and an effective lateral resolution of approximately 60–70 nm when combined with conventional confocal microscopy [1]. The method was first demonstrated on mouse brain tissue and cultured cells, achieving three‐color super‐resolution imaging of approximately 10⁷ μm³ of hippocampal tissue [1].

Free‐radical chain‐growth polymerization offers several practical advantages for ExM: reagents are inexpensive, widely available, and well‐characterized; the reaction proceeds at room temperature (or 37 °C for tissue samples); and polymerization kinetics can be tuned by adjusting initiator and inhibitor concentrations. However, this mechanism introduces inherent structural heterogeneity into the gel network [8, 14]. During chain growth, local fluctuations in monomer and crosslinker concentration produce regions of varying crosslink density, creating spatial inhomogeneity on the tens‐of‐nanometers scale—a long‐recognized and intrinsic feature of free‐radical copolymerization rather than a peculiarity of ExM gels. Because the bifunctional crosslinker (bis‐acrylamide) is generally more reactive and is consumed faster than the monoacrylate monomer, densely crosslinked nanoscale clusters form early in the reaction and become embedded within more loosely crosslinked regions; the resulting spatial fluctuations in crosslink density are effectively frozen into the network at the gel point and cannot subsequently relax. Such inhomogeneity was characterized quantitatively in synthetic hydrogels by scattering and related techniques well before its relevance to ExM was appreciated [15, 16]. In the ExM context it is a direct origin of locally nonuniform swelling: regions of high crosslink density expand less than their loosely crosslinked surroundings, displacing anchored molecules from their true relative positions and thereby limiting isotropy. Additionally, topological defects—including dangling chain ends, intramolecular loops, and trapped entanglements—further compromise network uniformity [8] (Figure 2A).

FIGURE 2.

Evolution of hydrogel architectures and expansion performance in ExM. (A) Gel network topology and tetra‐gel monomer design. Top: schematic of free‐radical chain‐growth polymerized PAAm network with inherent structural defects—(1) local density fluctuations, (2) dangling chain ends, and (3) intramolecular loops. Bottom: chemical structures of complementary tetrahedral monomers for tetra‐gel formation via click chemistry. Adapted with permission from Ref. [8]. Copyright 2021 Springer Nature. (B) Quantitative comparison of gel embedding accuracy using DNA origami nanorulers. Left: nearest‐neighbor distance histograms from STORM imaging before and after tetra‐gel expansion. Right: gel‐introduced distortion at 28 and 84 nm spacings across PA gel, tetra‐gel, and tetra‐gel after expansion, showing that tetra‐gel maintains <5 nm distortion versus ∼16 nm for PA gel. Reproduced with permission from Ref. [14]. Copyright 2021 Springer Nature. (C) Ultrastructural visualization enabled by 10‐fold TREx expansion. Left: expanded Caco‐2 cell gel labeled with mCLING, with electron microscopy images of brush border microvilli as reference. Right: comparison of mCLING fluorescence at ∼10‐fold versus ∼4.5‐fold expansion, with zoomed views showing improved resolution at higher expansion. Quantitative analysis yields a mean microvilli diameter of 1.08 ± 0.16 μm after 10‐fold expansion. Reproduced with permission from Ref. [10]. Copyright 2022 eLife Sciences Publications.

These structural imperfections manifest as local variations in swelling behavior, introducing nanoscale distortions during expansion. Empirical measurements using DNA‐origami nanorulers have shown that the act of gel embedding alone—measured before any expansion—displaces labeled sites in polyacrylamide gels with a standard deviation of approximately 16 nm, whereas tetra‐gel embedding preserves these positions to within less than 5 nm [14]. It should be emphasized that these are gel‐induced embedding distortions on pre‐expansion rulers and not a direct measure of distortion accrued during expansion; they nonetheless illustrate the superior structural homogeneity of the tetra‐gel network (Figure 2B). Although this distortion level is acceptable for many biological applications, it becomes a significant limitation when resolving structural features below 20 nm or when combining ExM with additional super‐resolution methods such as STORM or STED.

The original ExM protocol required custom DNA‐oligonucleotide‐conjugated antibodies, limiting its accessibility. Beyond the polymer network itself, a second chemical requirement of every ExM workflow is anchoring. However homogeneous and well‐formed a gel may be, it conveys no biological information unless the molecules of interest are themselves coupled to it; anchoring is precisely this covalent coupling of target biomolecules to the network, ensuring that their relative positions survive homogenization and expansion. The need for custom probes was first relieved by two independent, contemporaneous protein‐retention strategies reported in 2016. Tillberg et al. introduced protein‐retention ExM (proExM), which uses the small bifunctional molecule Acryloyl‐X SE (AcX) to anchor proteins to the growing polymer network [3]; concurrently and independently, Chozinski et al. demonstrated an equivalent strategy based on the methacrylic acid N‐hydroxysuccinimidyl ester (MA‐NHS) anchor [4]. In both, the NHS‐ester group reacts with primary amines on proteins, while the acryloyl or methacryloyl group participates in radical polymerization, becoming incorporated into the gel backbone. This approach enabled retention and imaging of genetically encoded fluorescent proteins and standard immunofluorescence labels after expansion [5] (Figure 1B).

Alternative anchoring strategies have since been developed for other biomolecule classes. LabelX, a linker based on the alkylating agent Label‐IT, anchors RNA molecules to the hydrogel for expansion fluorescence in situ hybridization (ExFISH) [7]. More recently, glutaraldehyde and tetrafluorophenyl (TFP) ester crosslinkers have proven effective as protein anchors, and click‐chemistry‐based approaches have extended anchoring compatibility to lipids, glycans, and small molecules [17, 18] (Figure 1C). Lipid retention in particular has continued to improve: Shin et al. recently introduced a continuous membrane‐labeling strategy that preserves and uniformly labels lipid membranes throughout the gel, enabling membrane ultrastructure to be traced across expanded specimens and extending ExM compatibility to the lipidome [19].

4. Modified Acrylamide Formulations for Enhanced Expansion

The most straightforward strategy for increasing expansion factor within the polyacrylamide/sodium acrylate framework is reducing crosslinker concentration. Lowering bisacrylamide content decreases crosslink density, allowing the network to swell further before the elastic restoring force balances the osmotic driving force. The original ExM report noted this possibility, and subsequent studies systematically explored the parameter space [1, 10].

M’Saad and Bewersdorf's pan‐ExM offers a complementary route to very high expansion factors (approximately 13‐ to 21‐fold): rather than optimizing a single‐step recipe, it performs two sequential rounds of gelation and expansion combined with N‐hydroxysuccinimide (NHS)‐ester pan‐proteome (total‐protein) labeling, an iterative strategy considered in detail in Section 5 [20]. Within the single‐step polyacrylamide/sodium acrylate framework, the most direct route to higher expansion is to push the same crosslinker‐reduction strategy to its limit—ultimately dispensing with a separately added crosslinker altogether. Truckenbrodt and colleagues developed the X10 protocol, which uses a monomer solution of N, N‐dimethylacrylamide (DMAA) and sodium acrylate without an externally added crosslinker [21]. In this system, the network forms not by classical bifunctional crosslinking but by chain transfer to the polymer backbone: a growing radical abstracts a backbone hydrogen at an α‐C—H or N‐methyl position of a previously incorporated DMAA unit, generating a mid‐chain radical that reinitiates a new DMAA chain and thereby creates a covalent branch point [21, 22].

At the high DMAA concentrations used in X10, these stochastically distributed branch points become frequent enough to gel the system without any added crosslinker, and—crucially—their uniform spatial distribution gives the network a more homogeneous crosslink density than chain‐growth SA/AAm/Bis gels, which is the physicochemical basis for the higher and more isotropic expansion achieved by DMAA‐based systems (X10, 20ExM, and the DMAA component of Magnify). This self‐crosslinking behavior was first characterized in polymer chemistry by Cipriano et al., who showed that DMAA with sodium acrylate alone yields superabsorbent gels (water uptake ∼3000‐fold, elongation at break ∼1350%) and attributed these properties to the uniform distribution of self‐crosslinks—the foundation on which X10 was subsequently built [22]. X10 provides up to ∼11.5‐fold linear expansion (typically ∼10‐fold). Although X10 achieved remarkable resolution (∼25 nm effective) on conventional microscopes, it required rigorous degassing with N₂ gas bubbling for at least 40 min to remove dissolved oxygen prior to gelation, and its application to thick tissue specimens was limited [21, 23].

Building on the X10 approach, Damstra et al. developed Ten‐fold Robust ExM (TREx) through systematic exploration of the ExM gel recipe space [10]. TREx retains the sodium acrylate/acrylamide/bisacrylamide chemistry of the original ExM protocol but with optimized ratios: sodium acrylate and acrylamide concentrations were increased while bisacrylamide crosslinker was substantially reduced. Crucially, TREx does not require the specialized degassing procedures of X10, making it significantly more practical for routine laboratory use [10].

TREx achieves robust 10‐fold expansion of both cultured human cells and thick (50–60 μm) mouse brain tissue sections in a single expansion step. When combined with small‐molecule total protein and membrane stains, 10‐fold TREx expansion sufficiently decrowds biological specimens to reveal ultrastructural features, including mitochondrial cristae, nuclear pore complexes, and microvillar morphology [10] (Figure 2C). TREx thus represents an important advance in making high‐expansion ExM accessible to laboratories already familiar with standard ExM procedures, without requiring fundamentally different chemistry or equipment [24].

Klimas et al. introduced Magnify, an ExM variant that addresses the long‐standing limitation of molecule‐specific anchoring [11]. Magnify employs a mechanically sturdy gel formulation that retains a broad spectrum of biomolecules—including proteins, nucleic acids, and lipids—without a dedicated, molecule‐class‐specific anchoring step (Figure 3B). Its universal retention does not arise from the absence of an anchor, however, but from methacrolein, a small‐molecule aldehyde–vinyl reagent added directly to the monomer solution: the aldehyde moiety reacts with nucleophilic groups present on diverse biomolecules, while the vinyl group copolymerizes into the growing network. The key innovation is therefore that anchoring is folded into the gelation step itself—a single reagent grafts essentially all biomolecule classes to the matrix as the gel forms—rather than requiring separate, class‐specific anchoring chemistries. This is combined with a relatively gentle homogenization protocol that preserves a wider range of molecular species than the aggressive proteinase K digestion used in standard ExM [11].

FIGURE 3.

Frontier gel technologies for expansion microscopy. (A) Single‐shot 20ExM workflow (AX anchoring, SA/DMAA gelation, N₂ glove‐bag deoxygenation, softening, decrowding to ∼4×, and postimmunostaining expansion to ∼20×; steps (a–d). Reproduced with permission from Ref. [25]. Copyright 2024 Springer Nature. (B) Schematic of universal biomolecule retention in Magnify (i, prepared tissue → ii, direct gelling with methacrolein anchoring of proteins, nucleic acids, and lipids → iii, homogenization → iv, expansion). Adapted with permission from Ref. [11]. Copyright 2023 Springer Nature. (C) Microwave‐accelerated BOOST workflow: the optimized protocols for three sample types, a processing‐time comparison across ExM methods (BOOST completing in ∼1.1–1.5 h), optimization of acrylamide and bis‐acrylamide concentrations for maximal expansion factor, the gel chemistry, and a deformation‐index analysis showing that BOOST gels are more mechanically robust than TREx. Reproduced with permission from Ref. [26]. Copyright 2025 Springer Nature.

Magnify achieves expansion factors of up to approximately 11‐fold in a single step, providing an effective resolution of ∼25 nm on a conventional confocal microscope with a 1.15 NA objective. It has been demonstrated to be compatible with multiple tissue types, including mechanically challenging specimens such as kidney, as well as formalin‐fixed paraffin‐embedded (FFPE) clinical tissue sections. It should be noted that the ∼25 nm figure reflects the gel (diffraction‐limited confocal readout) alone, whereas the ∼15 nm figure is obtained only when Magnify is paired with an additional super‐resolution method—super‐resolution optical fluctuation imaging (SOFI); the two values should not be conflated. With this combination, the Magnify–SOFI approach resolves features as small as ∼15 nm [11, 27].

5. Next‐Generation Polymer Architectures

A fundamentally different approach to ExM hydrogel design emerged with tetra‐gel, developed by Gao et al. [8] Rather than relying on free‐radical chain‐growth polymerization, tetra‐gel is assembled through nonradical terminal linking of two complementary tetrahedral monomers: a tetra‐arm polyacrylate bearing azide terminal groups and a tetra‐arm polyethylene glycol (PEG) bearing complementary alkyne groups. The two monomers join through copper‐free click chemistry (strain‐promoted azide–alkyne cycloaddition, SPAAC), forming a diamond lattice‐like polymer network [8].

This design principle directly addresses the structural heterogeneity inherent in radical‐polymerized gels. In the tetra‐PEG architecture, polymer chain lengths and crosslinking densities are highly uniform because each monomer is of consistent size and the stoichiometric terminal‐linking mechanism is self‐limiting. Topological defects such as loops and dangling ends are largely eliminated by design [8]. Small‐angle neutron scattering (SANS) and static light scattering (SLS) studies have confirmed that tetra‐PEG gels possess structural homogeneity approaching that of an ideal polymer network [8, 28].

When applied to ExM, the gel‐embedding step in tetra‐gel preserves labeled molecular positions to within a standard deviation of less than 5 nm, compared with approximately 16 nm for conventional polyacrylamide gels, as quantified by STORM imaging of DNA‐origami nanostructures before expansion [14]. As noted in Section 3, these figures quantify embedding fidelity on pre‐expansion rulers rather than distortion accrued over the full expansion process, and should not be read as implying that tetra‐gel expansion is itself threefold more isotropic; they do, however, reflect the markedly more uniform network architecture of the click‐chemistry gel. Iterative expansion using a cleavable tetra‐gel variant (incorporating disulfide bonds cleavable by TCEP) achieved approximately 10–16‐fold linear expansion of HeLa cells, revealing microtubule sidewall separations and viral envelope structures with a median spatial error of 9.2 nm [8]. Tetra‐gel thus represents a significant step toward structurally faithful expansion at the single‐digit nanometer scale.

Two further developments in this space deserve emphasis. First, ultrastructure ExM (U‐ExM), introduced by Gambarotto et al., reoptimized the gel and homogenization chemistry to preserve the molecular architecture of multiprotein complexes and, critically, established postexpansion immunostaining as a standard practice by exposing epitopes only after the specimen is anchored and expanded; U‐ExM has since become a reference method for imaging centrioles, cilia, and other macromolecular assemblies [29]. Second, the iterative‐expansion lineage initiated by iExM has continued to advance: expansion revealing (ExR), developed by Sarkar et al., applies repeated rounds of gelation and re‐embedding to approach the single‐protein scale and is the most significant iterative protocol to follow iExM [30], while iterative U‐ExM (iU‐ExM), reported by Louvel et al. [31], couples the structural preservation of U‐ExM with iterative re‐embedding and currently represents the state of the art among iterative protocols. Most recently, one‐step nanoscale expansion (ONE) microscopy, reported by Shaib et al., pairs a DMAA‐based gel with a fluctuation‐based super‐resolution readout (SRRF) to image individual proteins and even their conformational states, underscoring how far DMAA gel chemistry can be pushed toward true molecular resolution [32].

The most recent landmark in ExM gel chemistry is 20ExM, developed by Wang, Shin, and colleagues, which achieves approximately 20‐fold linear expansion in a single step—matching iterative expansion performance with the simplicity of a one‐step protocol [25] (Figure 3A). The 20ExM gel is based on DMAA and sodium acrylate, building on X10 and TREx chemistry but increasing the monomer‐to‐water ratio to 3:2 (w/v) and carefully optimizing polymerization conditions.

A critical innovation was the identification of dissolved oxygen concentration as a dominant source of batch‐to‐batch variability in DMAA–sodium acrylate gels. Rather than degassing by nitrogen bubbling (as in X10), the authors developed an improved physical oxygen exclusion strategy: the gelation solution is deoxygenated using a gas dispersion tube (50 s of N₂ flow), and the entire gelation process is carried out inside an N₂‐filled glove bag to reproducibly control dissolved oxygen levels. Furthermore, gelation time acts as a tunable parameter for controlling expansion factor: shorter gelation times produce less‐crosslinked networks with higher expansion, while longer times yield more‐crosslinked, lower‐expansion gels [25].

With optimized initiator concentration and gelation time (2 h for cell culture, 16–20 h for brain tissue), 20ExM yields gels that expand approximately 20‐fold linearly in deionized water, providing an effective resolution below 20 nm on a conventional confocal microscope. The microtubule sidewall‐to‐sidewall distance measured after 20‐fold expansion was consistent with the known outer diameter of microtubules, validating the nanoscale fidelity of the expansion [25]. Additionally, 20ExM supports postexpansion immunostaining, a workflow choice with substantive consequences rather than a minor procedural detail. Staining after expansion keeps antibodies and other labels out of the still‐crowded pre‐expansion specimen, so that epitopes are decrowded and made sterically accessible before labeling, and the signal is no longer diluted across the expanded volume. U‐ExM first established this strategy as standard for relatively sparse, well‐defined targets such as cilia and centrioles; 20ExM extends postexpansion labeling to the far denser and more heterogeneous setting of neural tissue, enabling biomolecular labeling of tightly packed neuropil structures that are difficult to resolve when stained beforehand [25].

While most advances in ExM gel chemistry have focused on expansion factor and structural fidelity, the overall workflow duration has received comparatively less attention. The BOOST protocol (reported in 2025) addresses this gap by introducing microwave‐accelerated chemistry to achieve rapid, robust 10‐fold expansion [26]. Microwave irradiation accelerates all major steps—anchoring, gelation, denaturation/hydrolysis, staining, and expansion—reducing the total processing time from the typical one to several days to under 2 h for most sample types, which is 6–40 times faster than comparable protocols. A key design choice of BOOST is the elimination of sodium acrylate (SA) from the gel formulation. Instead, BOOST uses only acrylamide (AAm) and N, N’‐methylenebisacrylamide (Bis) as gel monomers, with Bis serving as the covalent crosslinker under free‐radical polymerization. Since the resulting polyacrylamide gel is electrically neutral and thus nonswellable, a subsequent alkaline hydrolysis step (10% SDS, 0.05 M boric acid, pH 12, under microwave irradiation) converts amide side groups to charged carboxyl groups, transforming the neutral gel into a swellable polyelectrolyte hydrogel. These are two logically independent innovations and should not be conflated: the 6–40‐fold reduction in processing time derives from microwave acceleration alone, whereas removing sodium acrylate contributes nothing to the speedup and instead serves a separate purpose—improving reproducibility by eliminating the batch‐to‐batch quality variability and tissue shrinkage associated with sodium‐acrylate stocks [26].

BOOST was demonstrated on cultured cells, vibratome‐cut tissue sections, FFPE specimens, organoids, and millimeter‐sized whole lymph nodes, pioneering single‐step ∼10‐fold expansion of large three‐dimensional specimens [26] (Figure 3C). This broad compatibility makes BOOST relevant to both research and clinical pathology. The workflow innovation highlights that continued progress in ExM depends not only on novel polymer chemistry but also on optimizing the physical and kinetic parameters of established gel systems [26]. Consistent with this principle, the BOOST formulation was tuned by optimizing the acrylamide and bis‐acrylamide concentrations for maximal expansion, and the resulting gels deform less than the comparably expanding TREx gel—a lower deformation index—indicating greater mechanical robustness (Figure 3C).

6. Critical Parameters in Gel Design and Optimization

Table 1 summarizes the key characteristics of the major ExM gel platforms described in the literature. Several trends are evident. First, the achievable expansion factor has increased dramatically, from the original 4.5‐fold to 20‐fold in single‐shot protocols, driven primarily by modifications to monomer composition, crosslinker concentration, and polymerization control. Second, there is a clear trade‐off between expansion factor and structural fidelity: while high‐expansion radical‐polymerized gels provide remarkable magnification, the tetra‐gel system offers superior positional accuracy at the molecular scale. Third, practical considerations—including workflow simplicity, tissue compatibility, and the need for specialized equipment—have strongly influenced which gel platforms gain widespread adoption [1, 3, 8, 10, 11, 13, 21, 25, 26, 29].

TABLE 1.

Comparison of the major ExM gel platforms discussed in this review.

Gel system	Polymer basis	Expansion factor	Effective resolution	Crosslinking	Year	Key features	Reference
Original ExM	SA/AAm/Bis	∼4.5×	∼60–70 nm	Bis‐AAm (radical)	2015	Foundational protocol; widely adopted	[1]
ProExM	SA/AAm/Bis	∼4–5×	∼70 nm	Bis‐AAm (radical)	2016	Protein retention via AcX	[3]
U‐ExM	SA/AAm/Bis (+FA/AA)	∼4×	∼35 nm	Bis‐AAm (radical)	2019	Ultrastructure preservation; established postexpansion immunostaining	[29]
iExM	SA/AAm/cleavable‐Bis (two rounds)	∼20×	∼25 nm	Cleavable Bis – DHEBA (radical)	2017	DNA oligo linker; two‐round iterative expansion; cleavable DHEBA crosslinker enables re‐embedding	[13]
X10	DMAA/SA	∼10× (up to ∼11.5×)	∼25 nm	Self‐crosslinking (radical)	2018	No external crosslinker; N2 degassing required	[21]
TREx	SA/AAm/Bis (opt.)	∼10×	∼25–30 nm	Low Bis (radical)	2022	Robust; no degassing; thick tissue compatible	[10]
Tetra‐gel	PA/PEG tetra‐arm	∼3–3.5× (single); ∼10–16× (iter.)	∼25–30 nm (iter.)	Click chemistry	2021	High isotropy; ∼9 nm positional error	[8]
Magnify	DMAA + SA + AAm + Bis +methacrolein	∼11×	∼25 nm (gel, confocal); ∼15 nm (gel + SOFI)	Radical	2023	Universal anchoring; retains lipids	[11]
20ExM	DMAA/SA	∼20× (∼21.5× cell; ∼18.4× tissue)	∼18 nm (FRC)	Self‐crosslinking (radical)	2024	Single‐shot; O2 exclusion (N2 glove bag + gas dispersion); postexpansion staining for brain tissue	[25]
BOOST	AAm/Bis	∼10×	∼25–30 nm	Crosslinking: Bis‐AAm (radical). Charge introduced separately by alkaline hydrolysis (amide → carboxylate; not a crosslinking step)	2025	Microwave‐accelerated; rapid workflow (<2 hr); SA‐free gel; compatible with FFPE and large 3D specimens	[26]

Uniform gel formation within biological specimens is essential for isotropic expansion. The monomer solution must fully permeate the tissue before polymerization begins. For thin cultured cells, this is straightforward; for thick tissue sections (50–200 μm), extended incubation times and the use of polymerization inhibitors such as 4‐HT are necessary [1, 13].

The rate of gelation relative to monomer diffusion determines whether the gel forms homogeneously throughout the tissue or develops gradients in crosslink density from the surface inward. The 20ExM protocol specifically addresses this issue by using very low initiator concentrations (1.6 μM for brain tissue) and extended gelation times (16–20 h) to allow monomers to equilibrate before significant polymerization occurs [25]. More fundamentally, the gel forms by free‐radical chain‐growth polymerization: initiator‐derived radicals (from APS/TEMED) successively add monomer units, growing chains are tied together by the bifunctional crosslinker, and a system‐spanning network appears abruptly at the gel point. Two transport‐and‐kinetics factors largely determine whether this proceeds cleanly, and both have been studied directly in the polymer‐physics literature. First, diffusivity: monomers, initiator, and anchoring reagents must diffuse into the fixed specimen faster than the network closes; otherwise their local concentrations—and hence the resulting crosslink density—vary from the surface to the core, producing graded swelling. Measurements of small‐molecule and macromolecular diffusivity in swollen hydrogels quantify these transport limits and how steeply they scale with network mesh size [33]. Second, grafting: each anchored biomolecule, and each pendant vinyl group, must be covalently grafted onto a growing chain in order to be retained, and the efficiency of grafting onto a forming network is itself concentration‐ and kinetics‐dependent [34]. When grafting is incomplete, biomolecules are washed out during homogenization (signal loss) or remain only weakly tethered (positional error); when diffusion is rate‐limiting, the crosslink‐density gradients that develop translate directly into anisotropic expansion. These two effects, together with the intrinsic crosslink‐density inhomogeneity of radical gels discussed in Section 3 [15, 16] and the dissolved‐oxygen sensitivity discussed below, constitute the principal ways in which gelation can go wrong in practice; the corresponding chemical levers—slower initiation, longer equilibration, and oxygen exclusion—are used to control them.

Dissolved oxygen is a well‐known inhibitor of radical polymerization, as molecular oxygen reacts with carbon‐centered radicals to form peroxyl radicals that are less reactive and can terminate growing polymer chains. This sensitivity is particularly problematic for DMAA‐based gels, where the absence of a bifunctional crosslinker means that the network depends entirely on radical transfer reactions (self‐crosslinking via hydrogen abstraction and branching) for its mechanical coherence. Variations in oxygen concentration therefore produce large batch‐to‐batch differences in expansion factor and gel quality [21, 25]. The physical oxygen exclusion approach (N₂‐filled glove bag and gas dispersion tube) developed for 20ExM provides a more reproducible alternative to the simple nitrogen bubbling methods used in earlier protocols, and may prove broadly useful for other radical‐polymerized ExM gel systems [25].

The fidelity of ExM imaging depends not only on the gel itself but also on the efficiency and specificity of biomolecular anchoring. If anchoring is incomplete, biomolecules will be washed away during homogenization, resulting in signal loss. Conversely, nonspecific anchoring can introduce artifacts. The choice of anchoring chemistry must also be compatible with the polymerization mechanism: acryloyl and methacryloyl anchors are incorporated through radical polymerization, while azide anchors are used with click‐chemistry‐based tetra‐gels [3, 8]. The development of universal anchoring strategies, such as the one employed in Magnify, aims to simplify this requirement by grafting essentially all biomolecule classes to the matrix with a single small‐molecule aldehyde–vinyl anchor (methacrolein) that is incorporated during gelation [11].

A conceptually important question that is often overlooked is whether ExM merely separates biomolecules from one another after homogenization or whether the biomolecules are themselves physically enlarged. The two outcomes follow from different chemistries. In classical, digestion‐based ExM, proteinase K cleaves proteins into short peptide fragments before expansion, so the gel separates (decrowds) these fragments and increases the distance between labels without enlarging any individual protein. In DMAA‐based gels, by contrast, the network can be grafted densely along the polypeptide backbone, so that when the gel swells it stretches the unfolded chain itself and the protein is genuinely expanded. For most chemistries the question has not been resolved experimentally, but for DMAA gels the expansion of individual proteins has now been demonstrated directly by structural analyses (ONE microscopy; see Section 5) [32], a finding examined in detail in a recent review [35] and recently extended down to the level of individual amino acid residues by the 1000‐fold‐expansion architecture discussed in Section 7 [36]. Whether other gel chemistries also expand proteins remains an open question. Some protocols clearly do not: those that leave green fluorescent protein (GFP) functional after expansion, as in proExM, must preserve the folded protein largely intact rather than unfolding and enlarging it [3]. Critically, the outcome may depend as much on the homogenization step as on the gel chemistry itself, because only a fully denatured, unfolded polypeptide can be threaded and stretched by the swelling network, whereas a gently homogenized protein that retains its tertiary fold resists per‐molecule expansion. This distinction—mere separation versus true molecular magnification—has direct consequences for how quantitatively ExM images can be interpreted at the scale of single macromolecules.

7. Emerging Directions and Future Perspectives

Future ExM gel development may benefit from hybrid architectures that combine the high expansion factors achievable with polyelectrolyte gels and the structural homogeneity of click‐chemistry networks. For example, incorporating a tetra‐PEG backbone with charged polyacrylate arms could potentially yield gels that swell substantially while maintaining the positional accuracy associated with tetra‐gel topology. The development of stimuli‐responsive gels for ExM, including thermoresponsive and photoresponsive systems, could enable reversible or tunable expansion, opening new possibilities for correlative and multimodal imaging; Kang et al., for example, reported stimuli‐responsive hydrogel chemistry of direct relevance to this goal [37].

The application of ExM to clinical pathology specimens—particularly FFPE tissues, the standard format for archival pathology—has gained significant momentum [11, 38]. However, the chemical fixation and paraffin embedding inherent in FFPE processing introduces crosslinks and modifications that can impede gel infiltration and biomolecular anchoring. Optimized gel formulations and homogenization protocols specifically designed for FFPE compatibility, such as those in Magnify and BOOST, are essential for realizing the diagnostic potential of ExM. Combining expansion with computational super‐resolution offers a complementary route: Kylies et al. paired ExM with enhanced super‐resolution radial fluctuations (ExESRRF) to achieve high‐resolution imaging of FFPE and other archival kidney specimens on conventional microscopes, illustrating how gel‐based expansion and fluctuation‐based super‐resolution can be layered for clinical samples [39]. Future clinical applications may also require standardized, quality‐controlled gel kits to ensure reproducibility across different laboratories and tissue types.

Even with the most homogeneous gel systems, some degree of local distortion during expansion is inevitable. Computational methods for post‐hoc distortion correction—using fiducial markers, repeated labeling, or structural priors—are increasingly recognized as an important complement to gel chemistry optimization. A representative strategy is GelMap, introduced by Damstra et al., which polymerizes a fluorescent reference grid directly into the expansion gel; this embedded coordinate system allows local deformations to be measured and computationally corrected, and simultaneously provides a built‐in scale bar and quality‐control readout for the expansion [40]. These approaches can partially compensate for the structural imperfections of radical‐polymerized gels, potentially relaxing the requirements for gel perfection while still achieving high‐accuracy structural measurements [31, 41].

The combination of ExM with other super‐resolution techniques (ExM‐STED, ExM‐SIM, ExM‐STORM) and with spatial omics technologies (ExSeq, spatial proteomics) places additional demands on gel chemistry. These applications require gels that are compatible with the specific buffer conditions, imaging reagents, and multiround labeling strategies employed by each technique. The continued development of chemically versatile, optically transparent, and mechanically robust gel platforms will be essential for supporting these integrative workflows [2, 27, 42, 43].

Finally, the upper bound of achievable expansion continues to be pushed dramatically. A very recent preprint introduces 1000‐fold ExM (1000ExM), a four‐network interpenetrating hydrogel constructed by recursive “ionic‐in‐ionic” casting that expands a specimen successively from ∼18‐fold to more than 1000‐fold linearly—on the order of a billion‐fold in volume [36]. By anchoring biomolecules to the network, cleaving their backbone bonds, and then separating the resulting building blocks, the approach reportedly maintains the structures of test proteins and peptides (nanobodies, GFP, and the mCLING peptide) across expansion, and computational analysis indicates that it can separate adjacent amino acid residues, approaching subnanometer precision on a conventional light microscope. Although this is a nonpeer‐reviewed preprint, it is conceptually central to the field: it represents the logical endpoint of the high‐expansion, dense‐grafting DMAA lineage and, if independently validated, would extend ExM from resolving molecular positions toward reading molecular shape and sequence. Achieving such extreme expansion reproducibly will place stringent demands on network homogeneity, anchoring density, and mechanical stability across successive expansion rounds, defining a clear frontier for the gel‐chemistry strategies surveyed here.

8. Summary and Outlook

The evolution of hydrogel chemistry in ExM over the past decade reflects a broader trend in biological imaging: the recognition that sample preparation—rather than optical hardware alone—can be a powerful lever for improving imaging resolution and accessibility. From the original PAAm formulation that achieved 4.5‐fold expansion to the latest 20ExM protocol that delivers 20‐fold magnification in a single step, each advance in gel chemistry has expanded the resolution frontier of conventional microscopes.

Several key lessons emerge from this trajectory. First, the fundamental physics of polyelectrolyte swelling sets upper bounds on what can be achieved within a given gel chemistry, but creative manipulation of polymerization kinetics, monomer composition, and network topology can push these bounds further than initially anticipated. Second, practical usability matters: the widespread adoption of ExM by nonspecialist laboratories has been facilitated by protocols that use familiar reagents and standard equipment, suggesting that future gel innovations must balance performance gains against workflow complexity. Third, the development of tetra‐gel demonstrates that fundamentally different polymerization mechanisms can overcome limitations inherent to radical chemistry, pointing toward a rich design space for future polymer engineering.

Limitations of the current literature include the relatively sparse quantitative characterization of gel microstructure—most studies assess expansion quality through distortion measurements on cellular structures, with fewer investigations of the gel network itself using techniques such as SANS or dynamic light scattering. Additionally, the long‐term stability of expanded gels (relevant for archival or high‐throughput applications) has not been systematically studied. Finally, the field would benefit from standardized benchmarking protocols that allow direct comparison of different gel systems under identical conditions.

Hydrogel chemistry is the enabling foundation of ExM, and its continued innovation is essential for advancing the resolution, accuracy, and accessibility of this transformative imaging approach. The field has progressed from a single polyacrylamide‐based formulation to a diverse toolkit of gel platforms, each offering distinct advantages in terms of expansion factor, structural fidelity, tissue compatibility, and workflow simplicity. Classical PAAm gels remain the workhorse of standard ExM applications; modified formulations such as TREx and Magnify have democratized high‐expansion imaging; tetra‐gel has set a new standard for nanoscale positional accuracy; and single‐shot 20‐fold expansion protocols are pushing the resolution of conventional microscopes toward the sub‐20‐nm regime. As the ExM community continues to grow, we anticipate that advances in polymer science—including hybrid architectures, stimuli‐responsive systems, and computationally guided gel design—will further expand the boundaries of what can be visualized with a standard microscope, ultimately bridging the gap between optical and electron microscopy.

Funding

This work was supported by the Natural Science Foundation of Hunan Province Joint Fund for Universities (2026JJ90110).

Declaration on the Use of Generative AI

The graphical abstract (table‐of‐contents figure) was created with the assistance of a generative AI image‐generation tool (Nano Banana); the authors have reviewed and edited the graphic, verified its scientific accuracy, and accept full responsibility for its content. No generative AI tools were used to create, generate, alter, or enhance the scientific figures (Figures 1–3), which are either original schematics prepared by the authors or panels reproduced or adapted, with permission from the original publishers, from the cited primary literature, nor to generate the written content of this manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Biographies

Min Hou obtained her Ph.D. in 2022 from Hunan University. Currently, she works at Hunan First Normal University. Her research fields involve the application of DNA hydrogel materials and nucleic acid nanomaterials in tumor imaging.

Min Tang is currently studying chemistry at Hunan First Normal University and is expected to graduate in July 2027. Her research fields involve the application of DNA hydrogel materials in tumor imaging.

WeiJin Peng is currently studying chemistry at Hunan First Normal University and is expected to graduate in July 2027. Her research fields involve the application of DNA hydrogel materials in tumor imaging.

Ang Li received his bachelor's degree in applied chemistry in 2019 from Huazhong Agricultural University and obtained her master's degree in Environmental Material Science from Hokkaido University, Japan. His research focuses on imaging based on expansion microscopy.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.